jfa journal

AND option

OR option



B.W.J. Pang1,*, S.-L. Wee1,2,*, L.K. Lau1, K.A. Jabbar1, W.T. Seah1, D.H.M. Ng1, Q.L.L. Tan1, K.K. Chen1, M.U. Jagadish1,3, T.P. Ng1,4

1. Geriatric Education and Research Institute (GERI), Singapore; 2. Faculty of Health and Social Sciences, Singapore Institute of Technology, Singapore; 3. Geriatric Medicine, Khoo Teck Puat Hospital, Singapore; 4. Department of Psychological Medicine, National University of Singapore, Singapore.
Corresponding author: Shiou-Liang Wee, Geriatric Education and Research Institute (GERI), 2 Yishun Central 2, Tower E Level 4 GERI Admin, 768024, Singapore, Phone: +65 6807 8011, weeshiouliang@gmail.com; Benedict Wei Jun Pang, Geriatric Education and Research Institute (GERI), 2 Yishun Central 2, Tower E Level 4 GERI Admin, 768024, Singapore, Phone: +65 6807 8030, L3enanapang@gmail.com (B.W.J. Pang)

J Frailty Aging 2020;in press
Published online December 11, 2020, http://dx.doi.org/10.14283/jfa.2020.65



Objectives: Due to the lack of a uniform obesity definition, there is marked variability in reported sarcopenic obesity (SO) prevalence and associated health outcomes. We compare the association of SO with physical function using current Asian Working Group for Sarcopenia (AWGS) guidelines and different obesity measures to propose the most optimal SO diagnostic formulation according to functional impairment, and describe SO prevalence among community-dwelling young and old adults. Design: Obesity was defined according to waist circumference (WC), percentage body fat (PBF), fat mass index (FMI), fat mass/fat-free mass ratio (FM/FFM), or body mass index (BMI). SO was defined as the presence of both obesity and AWGS sarcopenia. Muscle function was compared among phenotypes and obesity definitions using ANOVA. Differences across obesity measures were further ascertained using multiple linear regressions to determine their associations with the Short Physical Performance Battery (SPPB). Setting: Community-dwelling adults 21 years old and above were recruited from a large urban residential town in Singapore. Participants: 535 community-dwelling Singaporeans were recruited (21-90 years old, 57.9% women), filling quotas of 20-40 participants in each sex- and age-group. Measurements: We took measurements of height, weight, BMI, waist and hip circumferences, body fat, muscle mass, muscle strength, and functional assessments. Questionnaire-based physical and cognitive factors were also assessed. Results: Overall prevalence of SO was 7.6% (WC-based), 5.1% (PBF-based), 2.7% (FMI-based), 1.5% (FM/FFM-based), and 0.4% (BMI-based). SO was significantly associated with SPPB only in the FMI model (p<0.05), and total variance explained by the different regression models was highest for the FMI model. Conclusions: Our findings suggest FMI as the most preferred measure for obesity and support its use as a diagnostic criteria for SO.

Key words: Sarcopenic obesity, sarcopenia, obesity, prevalence, Singapore.

Abbreviation: ALMI: Appendicular Lean Mass Index; AWGS: Asian Working Group for Sarcopenia; FM/FFM: Fat Mass to Fat-Free Mass ratio; FMI: Fat Mass Index; GPAQ: Global Physical Activity Questionnaire; GS: Gait Speed; HGS: Handgrip Strength; KES: Knee Extensor Strength; LASA: Longitudinal Aging Study Amsterdam; MNA: Mini Nutritional Assessment; PBF: Percentage Body Fat; RBANS: Repeatable Battery for the Assessment of Neuropsychological Status; SO: Sarcopenic Obesity; SPPB: Short Physical Performance Battery; TUG: Timed Up-and-Go; WC: Waist Circumference.



The rising tide of obesity prevalence is currently a global public health problem of epidemic proportions in all ages. At the same time, population ageing will double the number of persons aged 65 years and over worldwide in the next three decades, reaching 1.5 billion in 2050 (1), many of whom will be physically frail and disabled from the progressive loss of muscle mass and function (sarcopenia) (2). Obesity, the excessive accumulation of body fat, is an important factor in the development of metabolic syndrome and cardiovascular disease (3), and is also an important contributing cause of muscle loss. The two conditions have overlapping causes and feedback mechanisms that are interconnected and mutually-aggravating (3). Coexistence of both, a condition known as sarcopenic obesity (SO), has been shown to act synergistically to exacerbate metabolic impairment, disability, cardiovascular disease and mortality more so than either condition alone (3,4).
Although the diagnosis of sarcopenia has been increasingly harmonized by consensus working groups such as the Asian Working Group for Sarcopenia (AWGS) (2), the lack of a uniform obesity definition in the context of SO has led to great variation in the methods and cut-offs applied to define obesity, resulting in marked variability in reported SO prevalence as well as conflicting data on observed adverse health outcomes (5). Obesity is officially considered a disease that requires clinical treatment, however, there are currently no universally-accepted definitions for it (3). Commonly used measures include the body mass index (BMI), waist circumference (WC), percentage body fat (PBF), fat mass index (FMI) and fat mass to fat-free mass (FM/FFM) ratio. BMI provides a good indication of disease risk, but does not distinguish between fat and fat-free mass, thus making its clinical value questionable (3,5,6). WC indicates central obesity and serves as a surrogate measure of visceral adiposity (5-7), while PBF gives an objective indication of total body fat and its distribution (5). FM and FFM are the most frequently used adiposity indexes for SO classification (7), with the FM/FFM ratio deemed clinically-suitable in the diagnosis of SO (6). However, each of these measures assesses a different construct of obesity and are not interchangeable (5). To better understand its underlying physiological processes, and to determine disease prevalence and design clinical interventions, it is necessary to progress towards a unified criteria for the diagnosis and classification of obesity.
Aside from the wide heterogeneity of obesity measures used in studies, inconsistent observations of associations between SO and disease risk (7) may also result from the criteria used for defining sarcopenia in most studies, which other than muscle mass did not always consider muscle strength and physical function (3,5). Muscle strength and function have been shown to decline more rapidly with age and contribute more significantly to physical decline and frailty than muscle mass.
The primary aim of the present study was to propose the most optimal SO diagnostic formulation by comparing the association of SO with physical function using different obesity measures (WC, PBF, FMI, FM/FFM and BMI). We hypothesize that the most optimal diagnostic formulation would be one that is most significantly associated with physical functional impairment. The secondary aim was to compare and describe estimates of SO prevalence among community-dwelling younger and older adults in the Singapore population using AWGS guidelines for sarcopenia diagnosis and the different obesity definitions.




Community-dwelling adults (≥21 years) were recruited from the town of Yishun, one of the largest north-residential towns in Singapore, residential population of 220,320 (50.6% females), with 12.2% older adults (≥65 years), similar to the overall Singapore residential population of 4,026,210 (51.1% females), with 14.4% older adults (≥65 years) (8).


Random sampling was employed to obtain a representative sample of approximately 300 male and 300 female participants, filling quotas of 20-40 participants in each sex- and age-group (10-year age-groups between 21-60; 5-year age-groups after 60). Detailed recruitment methods and exclusion criteria have been reported previously (9). Ethics approval was obtained from the National Healthcare Group DSRB (2017/00212). All respondents signed informed consent before participating in the study.


Participants answered questionnaires pertaining to education level, housing type (a proxy for socio-economic status), living arrangement, marital status, smoking and drinking (more than four days a week), a health and medical questionnaire indicating medical conditions and comorbidities, a mini nutritional assessment (MNA) (10), a global physical activity questionnaire (GPAQ) (11) and the LASA physical activity questionnaire (12).


Body weight to the nearest 0.1 kg and height to nearest millimeter were measured using a digital balance and stadiometer (Seca, GmbH & Co. KG, Hamburg, Germany). Waist and hip circumferences were measured using a non-elastic, flexible measuring tape around the navel and widest part of the hips respectively. Body mass index (BMI) was calculated as weight (kg) divided by height (m) squared. ‘Obesity’ according to waist circumference (WC) was defined as ≥90 and ≥80 cm for men and women respectively (13). A BMI of ≥27.5 kg/m2 was used to define obesity as recommended by the World Health Organization for Asian populations (14).

Cognitive Assessment

Global cognition and cognitive domains including immediate and delayed memory, visuospatial, language and attention were assessed using the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) (15).

Body Composition

Bone mineral density, total percentage body fat (PBF), fat mass (FM), fat-free mass (FFM) and appendicular lean mass (ALM) were measured using DXA (Discovery WI, Hologic, Inc., Marlborough, USA). Fat mass index (FMI) and appendicular lean mass index (ALMI) were calculated as FM (kg) and ALM (kg) divided by height (m) squared, where FM equals to total body fat mass and ALM equals to the sum of lean mass in the upper and lower limbs. FM/FFM ratio was calculated by taking FM (kg) divided by FFM (kg). ‘Obesity’ according to PBF, FMI and FM/FFM were defined using the upper two quintiles of PBF and FMI, and a ratio of >0.80 for FM/FFM (6).

Functional Performance

Muscle function, as a gauge of functional deterioration and impairment, was measured using objective and validated assessments including handgrip strength (HGS) (16), knee extensor strength (KES) (17), usual gait speed (GS) (18), the Short Physical Performance Battery (SPPB) (19) and the Timed Up-and-Go (TUG) (20). HGS was assessed using the Jamar Plus+ Digital Hand Dynamometer (Patterson Medical, Evergreen Boulevard, Cedarburg, USA), and the highest of four readings (two trials per arm) recorded. KES was assessed using a spring gauge strapped 10 cm above the ankle joint, and the highest of four readings (two trials per leg) recorded. GS was measured using the 6 m GAITRite Walkway (CIR Systems Inc., Sparta. New Jersey, USA) with a 2 m lead in and out phase, and the average speed (three trials) recorded. TUG measures the progress of balance, sit-to-stand and walking. The average timing (two trials) was recorded. A composite score was calculated for SPPB, which comprises three components: balance, gait speed and repeated chair stands.

Sarcopenia and Sarcopenic Obesity

Sarcopenia was assessed using the AWGS criteria (2). Poor physical function was defined as GS <1.0 m/s, low muscle mass as ALMI <7.0 and <5.4 kg/m2, and muscle strength by HGS <28 and <18 kg for men and women respectively. Presence of low muscle mass and poor muscle strength and/or physical performance constitutes ‘sarcopenia’ (2). Participants with both sarcopenia and obesity were classified as ‘sarcopenic obese (SO)’. Those who had neither were classified as ‘normal’.

Statistical Analysis

SPSS version 22 (Chicago, Illinois, USA) was used for analysis. Prevalence of SO were extrapolated to the general population weights by age groups. In statistical analyses, the sarcopenia component was defined according to low muscle mass and strength only, and one-way analysis of variance (ANOVA) with Bonferroni correction for post-hoc comparisons were performed to compare the four phenotypes – ‘Normal’, ‘Obese’, ‘Sarcopenic’ and ‘Sarcopenic Obese’ – against muscle functions for those 50 years and older. To further ascertain the impact of obesity definitions on physical function, univariate and multiple linear regressions were performed to determine their associations with SPPB. Statistical significance was set at p<0.05.



A total of 542 participants (57.9% females) aged 21-90 years were recruited. Due to incomplete data from seven participants, data from 535 participants were analyzed (Figure 1.). Of these, 81.9% were Chinese, 8.6% Malays, 6.7% Indians, and 2.8% from other races. Mean age was 58.6 (18.8) years. Reference values and descriptive statistics are presented in Supplementary Tables S1. and S2.

Figure 1
Participant flowchart


Cut-off values for obesity using the sex-specific upper two quintiles of PBF and FMI were 31.0% and 7.63 kg/m2 for men, and 41.4% and 9.93 kg/m2 for women. Overall population-adjusted prevalence of sarcopenic obesity (SO) was 7.6% (WC-based: men 7.2%; women 7.9%), 5.1% (PBF-based: men 4.4%; women 5.7%), 2.7% (FMI-based: men 2.2%; women 3.2%), and 1.5% (FM/FFM-based; men 0%; women 2.9%). Population-adjusted prevalence of SO for older adults (≥60 years) was 21.6% (WC-based), 16.1% (PBF-based), 9.5% (FMI-based) and 3.7% (FM/FFM-based; Table 1.).

Table 1
Prevalence estimates in study sample and adjusted to the Singapore general population age groups weights

SO: Sarcopenic obese. Values are presented as percentages (%)


Participant Characteristics and Sarcopenic Obesity

Across all five obesity measures, the SO phenotypes had higher age compared to the overall sample (Table 2.). Individuals with the lowest education levels, smallest housing types, lived alone, were widowed, had diabetes, hypertension or high cholesterol, or had one or more medical conditions were more likely to have SO. Individuals who smoked were more likely to have SO using the PBF, FMI and BMI phenotypes, while individuals who drank were more likely to have SO using the WC, PBF, FMI and BMI phenotypes. The FM/FFM-based definition did not identify any males with SO (0%).

Table 2
Participant characteristics and sarcopenic obesity statuses

SO: Sarcopenic Obesity; WC: Waist Circumference; PBF: Percentage Body Fat; FMI: Fat Mass Index; FM: Fat Mass; FFM: Fat-Free Mass. BMI: Body Mass Index. Values are presented as mean (SD) or number (%)


Muscle Function (ANOVA)

The SO phenotype consistently performed poorer than the normal group in functional measures for the WC, PBF and FMI obesity definitions (p<0.05, Table 3.). The SO phenotype also persistently performed poorer than the obese group for the PBF definition (p<0.05), and in HGS, KES, GS and TUG for the WC and FMI definitions (p<0.05). Compared to the sarcopenic group, the SO phenotype performed poorer in HGS for the WC and FM/FFM definitions, and in TUG for the PBF definition (p<0.05).

Table 3
Comparison of muscle function among phenotypes and obesity definitions using ANOVA (≥50 years old)

P value (<0.05); 1 denotes a significant post-hoc Bonferroni test between Normal and Obese (P<0.05); 2 denotes a significant post-hoc Bonferroni test between Normal and Sarcopenic (P<0.05); 3 denotes a significant post-hoc Bonferroni test between Normal and SO (P<0.05); 4 denotes a significant post-hoc Bonferroni test between Obese and Sarcopenic (P<0.05); 5 denotes a significant post-hoc Bonferroni test between Obese and SO (P<0.05); 6 denotes a significant post-hoc Bonferroni test between Sarcopenic and SO (P<0.05)


Multiple Linear Regression for SPPB

We adjusted for age, gender, education level, housing type, diabetes, GPAQ activity level and RBANS global (Table 4.). Small sample sizes for SO as defined by FM/FFM (n=9) and BMI (n=2) precluded multiple linear regression analysis using these definitions. Using the WC-based, PBF-based and FMI-based definitions, the total variance explained by the regression models were 22.6% [F(8, 352)=14.147, p<0.001], 23.1% [F(8, 352)=14.534, p<0.001] and 23.6% [F(8, 352)=14.867, p<0.001] respectively. SO was significantly associated with SPPB only in the FMI model (p<0.05).

Table 4
Multiple linear regression analysis for Short Physical Performance Battery (≥50 years old)

β: Standardized Coefficient; B: Unstandardized Coefficient; SE: Standard Error; GPAQ: Global Physical Activity Questionnaire; MET: Metabolic Equivalent of Task; RBANS: Repeatable Battery for the Assessment of Neuropsychological Status SO: Sarcopenia Obesity; * denotes a significant P value (<0.05)



In this study, we present reference values for waist circumference (WC), percentage body fat (PBF), fat mass index (FMI), fat mass to fat-free mass (FM/FFM) ratio, and body mass index (BMI), as well as sex-specific cut-off values for PBF and FMI, to define alternative phenotypic representations of sarcopenic obesity (SO). We describe the corresponding SO prevalences using AWGS guidelines for sarcopenia, across the age groups of healthy adults. Our estimated prevalence of SO for adults aged ≥21 years and ≥60 years were 7.6% and 21.6% (WC-based), 5.1% and 16.1% (PBF-based), 2.7% and 9.5% (FMI-based), 1.5% and 3.7% (FM/FFM-based), and 0.4% and 1.6% (BMI-based) respectively.
Comparatively, a recent study on 200 cognitively-intact and functionally-independent community-dwelling adults in Singapore (≥50 years) reported a prevalence of 10.5% and 10.0% based on WC-based and PBF-based definitions of SO respectively (5), slightly lower than the 15.4% and 10.7% in the present study (≥50 years). Notably, the authors used the original AWGS criteria to define sarcopenia (21), with lower cut-offs for muscle strength and gait speed and thus lower detection rates compared to the updated AWGS criteria (2). Another study involving 591 healthy volunteers in Korea found a prevalence of 10.9% (40-59 years) and 18.0% (≥60 years) using the PBF-based definition (4), higher than the 2.1% (40-59 years) and 16.1% (≥60 years) in the present study. However, their criteria did not include the functional components of sarcopenia (2), and their population-derived cut-offs for low muscle mass were much higher at 8.81 and 7.36 kg/m2 compared to AWGS’ 7.0 and 5.4 kg/m2 used in the present study (2, 21).
Using one-way analysis of variance (ANOVA) with Bonferroni correction for post-hoc comparisons, the SO phenotype consistently performed poorer than the normal group in functional measures for the WC, PBF and FMI obesity definitions. The SO phenotype also persistently performed poorer than the obese group for the PBF definition, and in HGS, KES, GS and TUG for the WC and FMI definitions. Compared to the sarcopenic group, the SO phenotype performed poorer in HGS for the WC and FM/FFM definitions, and in TUG for the PBF definition.
Multiple linear regression results for SPPB revealed that only the FMI-based definition of SO was significantly associated with poorer SPPB scores, and the total variance explained by the different regression models was highest for the FMI definition, followed by PBF and WC. Physical function impairment in the absence of disability likely represents the shared core of sarcopenia and physical frailty. Such functional deterioration with deficits in gait speed, balance, and muscle strength, can be objectively assessed through the SPPB (22). Given that the SPPB is considered one of the most reliable and valid assessments for functional performance (5,19,22), our findings suggest FMI to be the most preferred obesity measure for defining SO.
Although PBF gives an objective indication of total body fat, it does not discern between visceral and subcutaneous fat (5). While WC provides an estimate of visceral adiposity which is associated with higher morbidity than its subcutaneous counterpart (23), it is not adjusted for height and is thus insensitive to body size (5-7). BMI gives a good indication of disease risk, but does not differentiate fat from fat-free mass (3, 5, 6). In addition, the BMI definition led to a noticeably much lower SO detection rate (0.4%) compared to the other definitions, similar to what was reported in a previous study.5 In corroboration with the literature, fat mass was previously reported to be the most frequently used adiposity index for the classification of SO, and its adjustment to height squared (FMI) has been the preferred method to account for differences in body size across age and between the sexes (6). In terms of physical performance, FMI is also considered an accurate indicator of total body adiposity that could improve the predictive value of SO in functional deterioration (6, 7). In addition, FMI was found to be a better screening tool in the prediction of metabolic syndrome in Chinese men and women (24) and more accurately assessed obesity in Mexican Americans (25) compared to BMI or PBF.
The FM/FFM definition did not identify any men with SO. This is similar to the findings of previous studies, where using the FM/FFM definition led to markedly disproportionate low numbers of men identified with SO (6, 7, 26). Women inherently have much higher relative fat mass than men (27), and conversely, men have higher relative fat-free mass (total body water, muscle and bone mass) than women at all ages (27,28). This is primarily due to the hormonal differences between men and women; men have higher testosterone levels which exhibits anabolic effects on muscle and bone (29), while higher estrogen levels in women promote subcutaneous fat deposition especially in the hips, thighs and chest (30). In addition, approximately 75% of skeletal muscle tissue is composed of water (31). Thus, with higher muscle mass, men inadvertently hold more total body water, further contributing to the discrepancy in fat mass and fat-free mass between men and women. To address the underlying gender-bias of the FM/FFM criteria and improve its accuracy in identifying gender-specific obesity and SO prevalence, different cut-off values for men and women (lower cut-off values for men) should be explored.
A recent study on 1235 adults with type 2 diabetes (T2D) in Singapore (≥45 years) reported a SO prevalence of 19.4% using the FM/FFM-based definition, higher than the 2.3% reported in this study, although the criteria for diagnosing SO in that study did not include the AWGS functional components for sarcopenia, which could possibly have inflated the proportions identified with SO (6). Furthermore, previous studies have shown a close link between sarcopenia and obesity through insulin resistance (3). Visceral fat accumulation (which promotes secretion of pro-inflammatory cytokines) is a contributing factor to the loss of skeletal muscle (which is the largest insulin-responsive tissue). Obesity and sarcopenia have a synergistic effect on promoting insulin resistance which could exacerbate T2D (4). In addition, patients with T2D exhibit insulin resistance, systemic inflammation and metabolic complications that could in turn perpetuate excess adiposity accumulation and loss of muscle mass (3), leading to a vicious cycle of worsening insulin resistance, T2D, sarcopenia and obesity (4).
The strengths of this study are its population-based nature, thoroughness of data collection and application of up-to-date and evidence-based consensus. It also has a few limitations. It presents cross-sectional data on obesity, muscular health and function of Singaporeans, which precludes inferences on causality. Agreement amongst the obesity definitions was also not investigated, though it has previously been established that different obesity definitions intrinsically measure different constructs and are therefore not interchangeable (5). While the AWGS criteria is for older adults, we also applied the same criteria to estimate prevalence of SO in younger adults, and so this might have been an underestimate, though we only included those 50 years and older in our statistical analyses. Finally, the participants were community-dwelling adults; thus, the findings may not be generalizable to hospitalized, institutionalized or disabled individuals.



This study presents new and much-needed data that help to better define and document sarcopenic obesity across age groups of healthy, community-dwelling Asian adults. To address the variability in sarcopenic obesity prevalence and conflicting data on its associations with adverse health consequences, a universally-accepted obesity definition is of utmost importance. Our findings suggest that FMI is the most preferred method for measuring obesity, and support its use as a diagnostic criteria for sarcopenic obesity.


Disclosure statement: The authors declare no conflict of interest. The research work conducted for this study comply with the current laws of the country in which they were performed.
Acknowledgements: This research was supported as part of a core funding from the Ministry of Health of Singapore to GERI. The authors gratefully acknowledge the strong support of Prof. Pang Weng Sun in making this Yishun Study possible, and the support of Dr. Lilian Chye, Sylvia Ngu Siew Ching, Aizuriah Mohamed Ali, Mary Ng Pei Ern, Chua Xing Ying and Shermaine Thein in this study. BWJP, SLW, MUJ, TPN contributed to the research design. BWJP, LKL, KAJ, WTS, DHMN, QLLT, KKC conducted the research. BWJP, LKL, KAJ, WTS, DHMN, QLLT, KKC analyzed data and performed statistical analysis. BWJP, SLW, TPN wrote the paper. BWJP, SLW, TPN had primary responsibility for final content. All authors have read and approved the final version.




1. United Nations, Department of Economic and Social Affairs, Population Division. World Population Prospects 2019: Highlights. New York: United Nations. 2019.
2. Chen L, Woo J, Assantachai P, Auyeung T, Chou M, Iijima K, Jang H, Kang L, Kim M, Kim S et al. Asian Working Group for Sarcopenia: 2019 Consensus Update on Sarcopenia Diagnosis and Treatment. Journal of the American Medical Directors Association. 2020;21(3):300-307.e2. https://doi.org/10.1016/j.jamda.2019.12.012.
3. Kalinkovich A, Livshits G. Sarcopenic obesity or obese sarcopenia: A cross talk between age-associated adipose tissue and skeletal muscle inflammation as a main mechanism of the pathogenesis. Ageing Research Reviews. 2017;35:200-221. https://doi.org/10.1016/j.arr.2016.09.008.
4. Kim T, Yang S, Yoo H, Lim K, Kang H, Song W, Seo J, Kim S, Kim N, Baik S et al. Prevalence of sarcopenia and sarcopenic obesity in Korean adults: the Korean sarcopenic obesity study. International Journal of Obesity. 2009;33(8):885-892. https://doi.org/10.1038/ijo.2009.130.
5. Khor E, Lim J, Tay L, Yeo A, Yew S, Ding Y, Lim W. Obesity Definitions in Sarcopenic Obesity: Differences in Prevalence, Agreement and Association with Muscle Function. The Journal of Frailty of Aging. 2019;000-000. https://doi.org/10.14283/jfa.2019.28.
6. Low S, Goh K, Ng T, Ang S, Moh A, Wang J, Ang K, Subramaniam T, Sum C, Lim S. The prevalence of sarcopenic obesity and its association with cognitive performance in type 2 diabetes in Singapore. Clinical Nutrition. 2019; 000-000. https://doi.org/10.1016/j.clnu.2019.10.019.
7. Prado C, Wells J, Smith S, Stephan B, Siervo M. Sarcopenic obesity: A Critical appraisal of the current evidence. Clinical Nutrition. 2012;31(5):583-601. https://doi.org/10.1016/j.clnu.2012.06.010.
8. Singapore Department of Statistics (DOS) [Internet]. Base. 2020 [cited 5 May 2020]. Available from: https://www.singstat.gov.sg.
9. Pang B, Wee S, Lau L, Jabbar K, Seah W, Ng D et al. Prevalence and Associated Factors of Sarcopenia in Singaporean Adults—The Yishun Study. Journal of the American Medical Directors Association. 2020;000-000. https://doi.org/10.1016/j.jamda.2020.05.029.
10. Kaiser M, Bauer J, Ramsch C, Uter W, Guigoz Y, Cederholm T, Thomas D, Anthony P, Charlton K, Maggio M et al. Validation of the Mini Nutritional Assessment short-form (MNA®-SF): A practical tool for identification of nutritional status. The Journal of Nutrition, Health and Aging. 2009;13(9):782-788. https://doi.org/10.1007/s12603-009-0214-7.
11. Armstrong T, Bull F. Development of the World Health Organization Global Physical Activity Questionnaire (GPAQ). Journal of Public Health. 2006;14(2):66-70. https://doi.org/10.1007/s10389-006-0024-x.
12. Stel V, Smit J, Pluijm S, Visser M, Deeg D, Lips P. Comparison of the LASA Physical Activity Questionnaire with a 7-day diary and pedometer. Journal of Clinical Epidemiology. 2004;57(3):252-258. https://doi.org/10.1016/j.jclinepi.2003.07.008.
13. Alberti K, Zimmet P, Shaw J. International Diabetes Federation: a consensus on Type 2 diabetes prevention. Diabetic Medicine. 2007;24(5):451-463. https://doi.org/10.1111/j.1464-5491.2007.02157.x.
14. WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. The Lancet. 2004;363(9403):157-163. https://doi.org/10.1016/s0140-6736(03)15268-3.
15. Collinson S, Fang S, Lim M, Feng L, Ng T. Normative Data for the Repeatable Battery for the Assessment of Neuropsychological Status in Elderly Chinese. Archives of Clinical Neuropsychology. 2014;29(5):442-455. https://doi.org/10.1093/arclin/acu023.
16. Rantanen T. Midlife Hand Grip Strength as a Predictor of Old Age Disability. JAMA. 1999;281(6):558. https://doi.org/10.1001/jama.281.6.558.
17. Guralnik J, Ferrucci L, Simonsick E, Salive M, Wallace R. Lower-Extremity Function in Persons over the Age of 70 Years as a Predictor of Subsequent Disability. New England Journal of Medicine. 1995;332(9):556-562. https://doi.org/10.1056/NEJM199503023320902.
18. Studenski S. Gait Speed and Survival in Older Adults. JAMA. 2011;305(1):50. https://doi.org/10.1001/jama.2010.1923.
19. Puthoff M. Research Corner Outcome Measures in Cardiopulmonary Physical Therapy: Short Physical Performance Battery. Cardiopulmonary Physical Therapy Journal. 2008;19(1):17-22. https://doi.org/10.1097/01823246-200819010-00005.
20. Shumway-Cook A, Brauer S, Woollacott M. Predicting the Probability for Falls in Community-Dwelling Older Adults Using the Timed Up & Go Test. Physical Therapy. 2000;80(9):896-903. https://doi.org/10.1093/ptj/80.9.896.
21. Chen L, Liu L, Woo J, Assantachai P, Auyeung T, Bahyah K, Chou M, Chen L, Hsu P, Krairit O et al. Sarcopenia in Asia: Consensus Report of the Asian Working Group for Sarcopenia. Journal of the American Medical Directors Association. 2014;15(2):95-101. https://doi.org/10.1016/j.jamda.2013.11.025.
22. Calvani R, Marini F, Cesari M, Tosato M, Anker S, von Haehling S, Miller R, Bernabei R, Landi F, Marzetti E et al. Biomarkers for physical frailty and sarcopenia: state of the science and future developments. Journal of Cachexia, Sarcopenia and Muscle. 2015;6(4):278-286. https:doi.org/10.1002/jcsm.12051.
23. Lee Y, Biddle S, Chan M, Cheng A, Cheong M, Chong Y, Foo L, Lee C, Lim S, Ong W et al. Health Promotion Board–Ministry of Health Clinical Practice Guidelines: Obesity. Singapore Medical Journal. 2015;57(06):292-300. https://doi.org/10.11622/smedj.2016103.
24. Liu P, Ma F, Lou H, Liu Y. The utility of fat mass index vs. body mass index and percentage of body fat in the screening of metabolic syndrome. BMC Public Health. 2013;13(1). https://doi.org/10.1186/1471-2458-13-629.
25. Peltz G, Aguirre M, Sanderson M, Fadden M. The role of fat mass index in determining obesity. American Journal of Human Biology. 2010;22(5):639-647. https://doi.org/10.1002/ajhb.21056.
26. Biolo G, Di Girolamo F, Breglia A, Chiuc M, Baglio V, Vinci P, Toigo G, Lucchin L, Jurdana M, Praznikar Z et al. Inverse relationship between “a body shape index” (ABSI) and fat-free mass in women and men: Insights into mechanisms of sarcopenic obesity. Clinical Nutrition. 2015;34(2):323-327. https://doi.org/10.1016/j.clnu.2014.03.015.
27. Cheng Q, Zhu X, Zhang X, Li H, Du Y, Hong W, Xue S, Zhu H. A cross-sectional study of loss of muscle mass corresponding to sarcopenia in healthy Chinese men and women: reference values, prevalence, and association with bone mass. Journal of Bone and Mineral Metabolism. 2013;32(1):78-88. https://doi.org/10.1007/s00774-013-0468-3.
28. Roberts S, Williamson D. Causes of Adult Weight Gain. The Journal of Nutrition. 2002;132(12):3824S-3825S. https://doi.org/10.1093/jn/132.12.3824S.
29. Flöter A, Nathorst-böös J, Carlström K, Ohlsson C, Ringertz H, von Schoultz B. Effects of combined estrogen/testosterone therapy on bone and body composition in oophorectomized women. Gynecological Endocrinology. 2005;20(3):155-160. https://doi.org/10.1080/09513590400021193.
30. Lizcano F, Guzmán G. Estrogen Deficiency and the Origin of Obesity during Menopause. BioMed Research International. 2014;2014:1-11. https://doi.org/10.1155/2014/757461.
31. Listrat A, Lebret B, Louveau I, Astruc T, Bonnet M, Lefaucheur L, Picard B, Bugeon J. How Muscle Structure and Composition Influence Meat and Flesh Quality. The Scientific World Journal. 2016;2016:1-14. https://doi.org/10.1155/2016/3182746.



R.S. Crow1,2, C.L. Petersen3,4, S.B. Cook5, C.J. Stevens1, A.J. Titus6,4, T.A. Mackenzie1,2,3, J.A. Batsis1,2,3

1. Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, United States; 2. Department of Medicine, Dartmouth Hitchcock Medical Center, Lebanon, New Hampshire, United States; 3. The Dartmouth Institute for Health Policy, Lebanon, New Hampshire, United States; 4 . Quantitative Biomedical Sciences Program, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, United States; 5. Department of Kinesiology, University of New Hampshire, Durham, NH, United States; 6. Department of Epidemiology, Geisel School of Medicine at Dartmouth, Lebanon NH, United States.
Corresponding author: Rebecca Crow DO, Section of General Internal Medicine, Dartmouth-Hitchcock Medical Center, 1 Medical Center Drive, Lebanon, NH 03756, Telephone: (603) 653-9500, Facsimile: (603) 650-0915, E-mail: Rebecca.s.crow@hitchcock.org

J Frailty Aging 2020;9(2)74-81
Published online January 6, 2020, http://dx.doi.org/10.14283/jfa.2019.44



Objective: A 5% change in weight is a significant predictor for frailty and obesity. We ascertained how self-reported weight change over the lifespan impacts rates of frailty in older adults. Methods: We identified 4,984 subjects ≥60 years with body composition measures from the National Health and Nutrition Examination Survey. An adapted version of Fried’s frailty criteria was used as the primary outcome. Self-reported weight was assessed at time current,1 and 10 years earlier and at age 25. Weight changes between each time point were categorized as ≥ 5%, ≤5% or neutral. Logistic regression assessed the impact of weight change on the outcome of frailty. Results: Among 4,984 participants, 56.5% were female, mean age was 71.1 years, and mean BMI was 28.2kg/m2. A weight loss of ≥ 5% had a higher association with frailty compared to current weight, age 25 (OR 2.94 [1.72,5.02]), 10 years ago (OR 1.68 [1.05,2.69]), and 1 year ago (OR 1.55 [1.02,2.36]). Weight gain in the last year was associated with increased rate of frailty (1.59 [1.09,2.32]). Conclusion: There is an association between frailty and reported weight loss over time while only weight gain in the last year has an association with frailty.

Key words: Obesity, frailty, sarcopenia, pre-frailty.

Abbreviations: BMI: body mass index; NHANES: National Health and Nutrition Examination Survey.



Frailty is the result of physical, psychological and social factors that contribute to a decline in the body’s physiological reserve and its reduced ability to maintain homeostasis among life’s stressors (1, 2). While a standardized pragmatic definition of frailty is still debated, Fried’s landmark study operationalized frailty as a phenotype defined by a set of variables: unintentional weight loss of ≥10 lbs, self-reported exhaustion, slow gait speed, low energy expenditure and weak grip strength (frail ≥3, pre-frail 1 or 2, robust =0) (3). Frailty is strongly associated with functional losses, disability, increased healthcare utilization and higher cost of healthcare (1, 4-8).
Weight status is commonly assessed in healthcare settings; however, it may be underutilized as a metric to indicate current and future adverse health outcomes (9, 10). Changes in body composition occur with each decade of life and include a peaking of fat mass in the 7th decade followed by subsequent decline in both skeletal muscle and fat (11, 12) making the most commonly used metric, body mass index (BMI), a highly insensitive measure of weight status in older adults(13). Some older adults are at risk for what is known as sarcopenic obesity or a disproportionate loss of lean mass to gain of fat mass which is shown to be associated with higher rates of morbidity and mortality outside of both sarcopenia or obesity alone (14, 15).  Long-term changes of loss or gain in body weight are associated with the highest mortality rates among persons in the general population while mortality is the lowest among those with modest weight changes(16). Weight cycling, or gaining/losing a similar amount of weight repeatedly, is known to be associated with higher disability and mortality rates(17). This harmful cycle emphasizes the importance of obtaining a weight history in clinical practice as even lower percent changes may be significant in those with frailty (18-22).
Gaining an understanding of longitudinal weight measures over a lifespan could be helpful for predicting future risk of disease and functional status. Such measures are easily captured using most outpatient electronic medical records. As both frailty and obesity are associated with similar adverse outcomes (4-8, 23-25), weight change trajectories may help clinicians assess for the development of frailty and other adverse health outcomes in older adulthood. The purpose of this study was to evaluate the relationship between self-reported weight change over a lifespan and frailty in a representative sample of US older adults.



Study Design and Participants

Participants included in the analysis were community dwelling older adults identified from 1999-2004 National Health and Nutrition Survey (NHANES) data. NHANES is a multistage probability survey conducted by the National Center for Health Statistics designed to assess the health and nutritional status of adults and children in the United States. The survey oversamples Non-Hispanic blacks, Mexican Americans, persons greater than 60 years of age. Results are therefore generalizable to the United States population. All manuals, procedures and data files are publicly available at http://www.cdc.gov/nchs/nhanes.html.
NHANES screened 38 077 individuals, interviewed 31 125, and then examined 29 402 in a mobile examination unit, with exams conducted by trained medical personnel. For this secondary analysis of data, we included participants aged 60 years older with body composition measures and frailty variables for a final analytical cohort of N =4 984. The local Institutional Review Board at Dartmouth College exempted this study from review due to the de-identified nature of all NHANES data in the database.

Baseline Characteristics

Self-reported sociodemographic characteristics including age, race, sex, physical activity levels, smoking status and co-morbid conditions were obtained from questionnaires completed by participants or their primary caregivers. Age was stratified into three categories as performed in our previous analyses: 60-69, 70-79, 80+ years (13, 26, 27). Race was reported as non-Hispanic White, non-Hispanic Black, and Hispanic American. Co-morbid conditions were self-reported using the question, “Has a doctor or other health professional ever told you that you have [medical condition]?” Smoking status was classified as current smoker, former smoker or never smoker. Physical activity was categorized as sitting, walking, performing light loads or heavy work using the question, “Please tell me which of these four sentences best describes your usual daily activities?”

Study Variables

Frailty: We defined frailty according to the phenotypic model (3), using participant self-reported and objectively measured data. This phenotypic definition consists of five criteria derived from the Cardiovascular Health Study (3, 28) as follows: unintentional weight loss of 10 pounds or more in a year; self-reported exhaustion; weakness defined by grip strength; slow walking speed; and low physical activity. We adapted the criteria to define each variable(26), respectively, using data available in NHANES: low body mass index (BMI)<18.5kg/m2 [59 (1.3%)]; difficulty walking between rooms [586 (10%)]; difficulty lifting or carrying 10 pounds [1,455 [27.1%)], gait speed <0.8 m/s [1,865 (31.13%)]; and self-reported perception of reduced physical activity compared to others [735 (14.1%)]. Frailty was defined as meeting three or more of the five following criteria and pre-frailty was defined as meeting 1 or 2 criteria. Individuals not meeting any criteria were classified as robust.
Anthropometric Measures: Weight was assessed using a self-reported questionnaire. Participants were asked to report their current weight, weight one year ago, weight 10 years ago, and weight at age 25. Participants were asked to provide their specific weight in kg and if not known to give their best guess. Additional questions included “During the past 12 months, have you tried to lose weight?” and “During the past 12 months have you done anything to keep from gaining weight?”  Objective weight was measured on the right side of the body to the nearest tenth of a centimeter on an electronic digital scale (calibrated in kilograms), and height was measured using a stadiometer.

Statistical Analysis

All data were downloaded in September 2015 into a single dataset. Weight history data was combined in November 2017 following NHANES standard operating procedures, accounting for weighting, strata, primary sampling unit, and cluster. Descriptive statistics are presented as means ± standard errors, and counts (weighted percentages). Comparisons between groups were conducted using t-tests and chi-square tests of independence. We calculated self-reported percent weight change as the quotient of the difference between baseline year and year in question (1 year prior, 10 years prior or at age 25 years old). Meaningful weight loss/gain is categorized as ± a change of 5% or more (29). We created three categories: ≥5% weight loss; ≥5% weight gain; or no change in weight (-5 to +5% weight change). The latter category is represented as the referent in our models. A 5% change in weight loss or gain was used since it has been used as a significant predictor for both frailty and obesity in past studies (30, 31). Slope for each individual change was calculated as the participant’s age at each of the three time points (quotient of Weight Time1.- Weight Time2 and Δ Age) and is represented as the change in weight per year. Multiple models were constructed to evaluate the effect of weight change (primary predictor – gain/loss of 5%) on the presence of frailty (primary outcome). Gait speed was not assessed in NHANES 2003-2004 therefore imputation by mean was used conditional on covariates to account for missing values using R (v 3.3.2) and the package mice for 3,645 participants(www.r-project.org). The package creates plausible data values from a distribution specifically designed for each data point; five imputed data sets are generated using predictive mean matching. The correction variables used were age, sex, education, race, diabetes, arthritis, congestive heart failure, cancer, and lean mass percent. The five data sets were averaged, resulting in a final imputed data set used for analysis. Analyses were run on the full imputed data set as well as a subset excluding the imputed variables to test the quality. The data presented in our results is based on full imputed data alone; data excluding imputed variables is not shown and presented elsewhere (26).
We constructed three incremental logistic regression models adjusting for co-variates: age, gender (Model 1); Model 1 co-variates plus race, education, smoking (Model 2); Model 2 co-variates plus diabetes, arthritis, coronary artery disease, and cancer (Model 3). Data are not shown for Model 1 and Model 2 but results did not change as we adjusted for additional variables. All logistic regression models assessed the impact of weight change (gain, loss, no change) for two different frailty outcomes (Frailty vs. Pre-Frailty/Robust, and Pre-Frailty vs. Robust). Visual representation of change in weight over time was plotted per individual with a LOESS local regression line (span = 0.7). All analyses were conducted using STATA v.14 (College Station, Texas) and R v3.5 (www.r-project.org). P values were considered statistically significant if they were less than the criterion level of 0.05.



Table 1 presents the baseline characteristics of participants. Of 4,984 participants, 56.5% were female, mean age was 71.1 years and BMI was 28.2 kg/m2. Prevalence of pre-frailty and frailty was 40.1% and 9.1%, respectively. Robust participants were more likely to be non-Hispanic white and have a higher education level (p=<0.001). Frail patients were more likely to have comorbidities such as arthritis, diabetes and coronary artery disease (p=<0.001) but not more likely to have cancer (p=0.48). Nearly all frail patients met the criterion of weakness (96.9%) and the most common criterion identified for pre-frailty was slowness of gait (59.1%).

Table 1 Baseline Characteristics of Participants

Table 1
Baseline Characteristics of Participants

Values represented are means± standard errors or counts (weighted percentages).


Table 2 outlines participants’ self-reported weights and weight changes over time. Weight increased in all groups over time, with individuals with frailty losing weight, as compared to the pre-frail or robust groups within the past year or 10 years. Weight gain of ≥5% in the past year was associated with higher rates of frailty among those classified in other categories. When comparing each participant’s weight at age 25 years old to his/her current weight, fewer individuals with frailty gained clinically significant weight and most had a greater than 5% weight loss. Changes in weight and rates of combination of frailty, pre-frailty and robust percentages can be found in Supplementary Table 1.

Table 2 Weight Change and Rates of Classification Along Frailty Spectrum

Table 2
Weight Change and Rates of Classification Along Frailty Spectrum

This was an ANOVA test and all values represented are mean ± standard error or counts (weighted percentage).

Multivariable logistic regression models evaluating the relationship between weight change during a specific time interval, and presence of frailty or pre-frailty are presented in Table 3.  Weight loss of ≥5% was strongly associated with presence of frailty compared to pre-frailty/robust and the relationship strengthens with longer time intervals. A ≥5% weight gain only was associated with frailty if the gain occurred within the last year. Similar associations were seen with pre-frailty when comparing pre-frailty to robust status, albeit weaker than the association seen with frailty.

Table 3 Association of Weight Change and Frailty Status

Table 3
Association of Weight Change and Frailty Status

Data are represented as odds ratio [95% Confidence interval]. Multivariable logistic regression models (referent category: no change in weight) are represented as odds ratios (95% confidence intervals). The primary predictor was weight change (gain ≥5%, loss of 5%, or no change in weight) during the time period. Separate multivariable models were created for the outcomes of Frailty vs. Pre-Frailty/Robust, and Pre-Frailty vs. Robust (each yes/no). Models presented were adjusted for: age (years as continuous variable), sex (female=1 male=0), race (non-Hispanic white (ref), non-Hispanic black, Hispanic and Other) education status >12 years (yes=1, no=0), smoking status (former, current=1, never=0) self-reported diabetes (yes=1, no=0) , arthritis (yes=1, no=0), coronary artery disease (yes=1, no=0), cancer (yes=1, no=0).


Figure 1 represents the change in weight as a function of age plotted with a LOESS smoothed line by frailty status. Graphically, a period of increased rate of weight gain is noted prior to a steep decline in weight for individuals who were frail at time of study inclusion; conversely, this was not observed in those with pre-frail or robust status. Additionally, frail individuals had a higher peak weight (87.3 kg), occurring earlier in life (58.4 years), relative to both pre-frail (peak weight 79.9 kg at age 59.9 years) and robust (79.0 kg peak weight at 63.0 years) individuals. This suggests that the earlier one reaches their peak weight in life, the higher incident frailty.

Figure 1 Change in Weight by Frailty Status. This figure represents a LOESS smoothed line demonstrating the change in weight as a function of age by frailty status

Figure 1
Change in Weight by Frailty Status. This figure represents a LOESS smoothed line demonstrating the change in weight as a function of age by frailty status



Our findings demonstrate the importance of dynamic weight changes over a lifetime in the future development of frailty. While frailty is classically defined by weight loss (3), here we show that significant changes in weight gain or loss over a lifetime were associated with frailty progression making this a more complicated relationship than was previously thought. We created two separate analyses looking at frailty versus pre-frail and robust participants and pre-frail vs robust participants alone. Pre-frailty has individually been associated with an increase in overall mortality and cardiovascular mortality (26). Here we demonstrate that significant weight loss or weight gain is strongly associated with frailty and has a weaker, but statistically significant, association with pre-frailty supporting the significance of each step along the frailty spectrum.
While weight loss’ association with frailty is well accepted, the relationship of weight gain with frailty is less understood. A 22 year follow-up study demonstrated obesity was associated with higher rates of pre-frailty and frailty compared to robust individuals, suggesting obesity could be a contributing factor to progression along the frailty spectrum (32, 33). Unsurprisingly, we showed that older adults with a ≥5% weight loss compared to 1 year ago, 10 years ago, and age 25 have significantly increased odds of having frailty suggesting that weight loss is a strong indicator of potential frailty development. Yet, a ≥5% weight gain only increased odds of frailty when this occurred within the last year, but not compared to 10 years ago or age 25. This seems to suggest that a sudden increase in weight may not be marker of health, but a weight trajectory trending more toward the concept of sarcopenic obesity (34).
Prior work has shown that when evaluating weight gain and loss, more lean mass is lost than is gained over repeated fluctuations suggesting weight cycling could accelerate sarcopenia in older adults and contribute to sarcopenic obesity (35). Weight cycling leads to greater central body fat and increased mortality (22). Some theorize that the trajectory of pathological aging is to move from robust status to sarcopenic obesity to frailty then to disability and mortality (36). Our previous data suggest that late adulthood weight changes over a 10 year period are predictive of sarcopenia development and that there is a healthy, natural propensity to gain weight over a life course (37). Our current findings demonstrate a similar trend as robust participants were more likely than those pre-frail or frail to have gained ≥5% weight over the course of the prior 10 years.
A natural, healthy trend toward weight gain could explain why the classification of overweight status in older adults, in part, has been noted as somewhat “protective” (38). Overweight status could be beneficial for reducing disability and functional loss with reduced osteoporosis and injuries from falls (39). An ability to gain or maintain weight demonstrates a reduced vulnerability to stressors from comorbid health conditions (40).
This study is not without limitations. First, the analysis relied on self-reported weights which may be impacted by recall bias but studies looking at self-reported weight accuracy and actual weight are reasonably comparable (41), however we were not able to assess the accuracy of recall in those with cognitive impairment included in the study. This was not done due to concurrent limitations in how those with cognitive impairment would be identified from NHANES variables (41). In clinical practice, weight 10 years prior and at age 25 would most likely be ascertained by recall, and therefore, use of self-report is ecologically valid. Second, the sample was comprised of community dwelling adults. Individuals living in facilities were not included, which limits the ability to generalize these findings to the older adult population as a whole. To operationalize Fried’s frailty criteria in NHANES some of the original definitions were modified. Those labeled frail could be considered those at “high likelihood of frailty” by our modified metrics. However the accuracy of our results are further supported by the fact the prevalence of each component is comparable to those observed in other studies who did not require modification (3, 42, 43). As walking speed was missing for 3,645 patients, multiple imputations needed to be performed to maximize the number of participants with appropriate data. Multivariate imputation by chained equations, a robust method that generates multiple predictions for each missing value, taking the uncertainty of the imputations into account and yielding accurate standard errors (44), was used to handle missing data. The reasons for unintentional weight loss in participants was unknown therefore the implications of these factors behind their weight loss are also unknown. As with all cross-sectional studies, we are not able to make causal inferences. Lastly, the true relationship between weight change and incident to frailty may not necessarily be linear.



The results demonstrate an association between frailty and weight change over time. Weight loss over a lifespan is strongly associated with frailty while weight gain in the last year also is association with higher rates of frailty and can therefore be a marker of declining health. A natural trend toward weight gain and overweight status in a lifetime could actually be a sign of metabolic health and longevity. These findings demonstrate the clinical value in weight trends obtained at most clinical visits and highlight potential trends that may warrant closer evaluation for syndromes like frailty that require additional intervention to help curb poorer health outcomes.


Acknowledgements: The authors would like to acknowledge Meaghan A. Kennedy MD, MPH who helped with editing and reviewing the transcript as well as support from the National Institute on Aging of the National Institutes of Health, Burroughs-Wellcome/Dartmouth Big Data in the Life Sciences Training Program, National Institute of Mental Health, Dartmouth Health Promotion and Disease Prevention Research Center, The Dartmouth Center for Health and Aging and the Dartmouth Hitchcock Medical Center Department of Medicine.
Disclosures: There are no conflicts of interest pertaining to this manuscript
Funding: Dr. Batsis receives funding from the National Institute on Aging of the National Institutes of Health under Award Number K23AG051681. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Dr. Batsis has also received honoraria from the Royal College of Physicians of Ireland, Endocrine Society, and Dinse, Knapp, McAndrew LLC, legal firm. Dr. Mackenzie:  none. Dr. Cook: none. Alexander Titus’ research reported in this publication was supported in part by the National Institutes of Health under Award Number T32LM012204. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Mr. Petersen’s research reported in this publication was supported by the Burroughs-Wellcome/Dartmouth Big Data in the Life Sciences Training Program. Dr. Crow’s research reported in this publication was supported by The Dartmouth Center for Health and Aging and the Department of Medicine. Dr. Stevens’ research is supported by National Institute of Mental Health (T32 MH073553, PI: Bruce, Fellow: Stevens). Support was also provided by the Dartmouth Health Promotion and Disease Prevention Research Center supported by Cooperative Agreement Number U48DP005018 from the Centers for Disease Control and Prevention. The findings and conclusions in this journal article are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. The authors acknowledge Friends of the Norris Cotton Cancer Center at Dartmouth and NCI Cancer Center Support Grant 5P30 CA023108-37 Developmental Funds.




1.    Chong E, Ho E, Baldevarona-Llego J, Chan M, Wu L, Tay L. Frailty and Risk of Adverse Outcomes in Hospitalized Older Adults: A Comparison of Different Frailty Measures. J Am Med Dir Assoc. 2017;18(7):638.e7-.e11.
2.    Clegg A, Young J, Iliffe S, Rikkert MO, Rockwood K. Frailty in elderly people. Lancet. 2013;381(9868):752-62.
3.    Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56(3):M146-56.
4.    Robinson TN, Wu DS, Stiegmann GV, Moss M. Frailty predicts increased hospital and six-month healthcare cost following colorectal surgery in older adults. Am J Surg. 2011;202(5):511-4.
5.    Bock JO, Konig HH, Brenner H, Haefeli WE, Quinzler R, Matschinger H, et al. Associations of frailty with health care costs–results of the ESTHER cohort study. BMC Health Serv Res. 2016;16:128.
6.    Makizako H, Shimada H, Doi T, Tsutsumimoto K, Suzuki T. Impact of physical frailty on disability in community-dwelling older adults: a prospective cohort study. BMJ Open. 2015;5(9):e008462.
7.    Comans TA, Peel NM, Hubbard RE, Mulligan AD, Gray LC, Scuffham PA. The increase in healthcare costs associated with frailty in older people discharged to a post-acute transition care program. Age Ageing. 2016;45(2):317-20.
8.    Avila-Funes JA, Amieva H, Barberger-Gateau P, Le Goff M, Raoux N, Ritchie K, et al. Cognitive impairment improves the predictive validity of the phenotype of frailty for adverse health outcomes: the three-city study. J Am Geriatr Soc. 2009;57(3):453-61.
9.    Meltzer AA, Everhart JE. Unintentional weight loss in the United States. Am J Epidemiol. 1995;142(10):1039-46.
10.    Willett WC. Weight loss in the elderly: cause or effect of poor health? Am J Clin Nutr. 1997;66(4):737-8.
11.    Beaufrere B, Morio B. Fat and protein redistribution with aging: metabolic considerations. Eur J Clin Nutr. 2000;54 Suppl 3:S48-53.
12.    Gallagher D, Visser M, De Meersman RE, Sepulveda D, Baumgartner RN, Pierson RN, et al. Appendicular skeletal muscle mass: effects of age, gender, and ethnicity. J Appl Physiol (1985). 1997;83(1):229-39.
13.    Batsis JA, Mackenzie TA, Bartels SJ, Sahakyan KR, Somers VK, Lopez-Jimenez F. Diagnostic accuracy of body mass index to identify obesity in older adults: NHANES 1999-2004. Int J Obes (Lond). 2016;40(5):761-7.
14.    Choi KM. Sarcopenia and sarcopenic obesity. Korean J Intern Med. 2016;31(6):1054-60.
15.    Cauley JA. An Overview of Sarcopenic Obesity. J Clin Densitom. 2015;18(4):499-505.
16.    Andres R, Muller DC, Sorkin JD. Long-term effects of change in body weight on all-cause mortality. A review. Ann Intern Med. 1993;119(7 Pt 2):737-43.
17.    Oh TJ, Moon JH, Choi SH, Lim S, Park KS, Cho NH, et al. Body-Weight Fluctuation and Incident Diabetes Mellitus, Cardiovascular Disease, and Mortality: A 16-Year Prospective Cohort Study. J Clin Endocrinol Metab. 2019;104(3):639-46.
18.    Payette H, Coulombe C, Boutier V, Gray-Donald K. Weight loss and mortality among free-living frail elders: a prospective study. J Gerontol A Biol Sci Med Sci. 1999;54(9):M440-5.
19.    Somes GW, Kritchevsky SB, Shorr RI, Pahor M, Applegate WB. Body mass index, weight change, and death in older adults: the systolic hypertension in the elderly program. Am J Epidemiol. 2002;156(2):132-8.
20.    Tully CL, Snowdon DA. Weight change and physical function in older women: findings from the Nun Study. J Am Geriatr Soc. 1995;43(12):1394-7.
21.    Alibhai SM, Greenwood C, Payette H. An approach to the management of unintentional weight loss in elderly people. CMAJ. 2005;172(6):773-80.
22.    Murphy RA, Patel KV, Kritchevsky SB, Houston DK, Newman AB, Koster A, et al. Weight change, body composition, and risk of mobility disability and mortality in older adults: a population-based cohort study. J Am Geriatr Soc. 2014;62(8):1476-83.
23.    Dee A, Kearns K, O’Neill C, Sharp L, Staines A, O’Dwyer V, et al. The direct and indirect costs of both overweight and obesity: a systematic review. BMC Res Notes. 2014;7:242.
24.    Visscher TL, Seidell JC. The public health impact of obesity. Annu Rev Public Health. 2001;22:355-75.
25.    Hirani V, Naganathan V, Blyth F, Le Couteur DG, Seibel MJ, Waite LM, et al. Longitudinal associations between body composition, sarcopenic obesity and outcomes of frailty, disability, institutionalisation and mortality in community-dwelling older men: The Concord Health and Ageing in Men Project. Age Ageing. 2016.
26.    Crow RS, Lohman MC, Titus AJ, Bruce ML, Mackenzie TA, Bartels SJ, et al. Mortality Risk Along the Frailty Spectrum: Data from the National Health and Nutrition Examination Survey 1999 to 2004. J Am Geriatr Soc. 2018;66(3):496-502.
27.    Rippberger PL, Emeny RT, Mackenzie TA, Bartels SJ, Batsis JA. The association of sarcopenia, telomere length, and mortality: data from the NHANES 1999-2002. Eur J Clin Nutr. 2018;72(2):255-63.
28.    Kulminski AM, Ukraintseva SV, Kulminskaya IV, Arbeev KG, Land K, Yashin AI. Cumulative deficits better characterize susceptibility to death in elderly people than phenotypic frailty: lessons from the Cardiovascular Health Study. J Am Geriatr Soc. 2008;56(5):898-903.
29.    Jensen MD, Ryan DH, Apovian CM, Ard JD, Comuzzie AG, Donato KA, et al. 2013 AHA/ACC/TOS guideline for the management of overweight and obesity in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and The Obesity Society. J Am Coll Cardiol. 2014;63(25 Pt B):2985-3023.
30.    Painter SL, Ahmed R, Hill JO, Kushner RF, Lindquist R, Brunning S, et al. What Matters in Weight Loss? An In-Depth Analysis of Self-Monitoring. J Med Internet Res. 2017;19(5):e160.
31.    Bieniek J, Wilczynski K, Szewieczek J. Fried frailty phenotype assessment components as applied to geriatric inpatients. Clin Interv Aging. 2016;11:453-9.
32.    Stenholm S, Strandberg TE, Pitkala K, Sainio P, Heliovaara M, Koskinen S. Midlife obesity and risk of frailty in old age during a 22-year follow-up in men and women: the Mini-Finland Follow-up Survey. J Gerontol A Biol Sci Med Sci. 2014;69(1):73-8.
33.    Strandberg TE, Sirola J, Pitkala KH, Tilvis RS, Strandberg AY, Stenholm S. Association of midlife obesity and cardiovascular risk with old age frailty: a 26-year follow-up of initially healthy men. Int J Obes (Lond). 2012;36(9):1153-7.
34.    Batsis JA, Mackenzie TA, Lopez-Jimenez F, Bartels SJ. Sarcopenia, sarcopenic obesity, and functional impairments in older adults: National Health and Nutrition Examination Surveys 1999-2004. Nutr Res. 2015;35(12):1031-9.
35.    Newman AB, Lee JS, Visser M, Goodpaster BH, Kritchevsky SB, Tylavsky FA, et al. Weight change and the conservation of lean mass in old age: the Health, Aging and Body Composition Study. Am J Clin Nutr. 2005;82(4):872-8; quiz 915-6.
36.    Cederholm T. Overlaps between Frailty and Sarcopenia Definitions. Nestle Nutr Inst Workshop Ser. 2015;83:65-9.
37.    Batsis JA, Petersen CL, Crow RS, Cook SB, Stevens CJ, Lillian SM, et al. Weight Change and Risk of Sarcopenia: Data from the National Health and Nutrition Examination Surveys 1999-2004 [Abstract]. 2018 International Frailty & Sarcopenia Conference, Miami Beach, FL, https://frailty-sarcopeniacom/docs/abstracts-2018pdf [accessed 18 September 2018]. 2018.
38.    Flegal KM, Kit BK, Orpana H, Graubard BI. Association of all-cause mortality with overweight and obesity using standard body mass index categories: a systematic review and meta-analysis. JAMA. 2013;309(1):71-82.
39.    Bowen ME. The relationship between body weight, frailty, and the disablement process. J Gerontol B Psychol Sci Soc Sci. 2012;67(5):618-26.
40.    Villareal DT, Banks M, Siener C, Sinacore DR, Klein S. Physical frailty and body composition in obese elderly men and women. Obes Res. 2004;12(6):913-20.
41.    Luo J, Thomson CA, Hendryx M, Tinker LF, Manson JE, Li Y, et al. Accuracy of self-reported weight in the Women’s Health Initiative. Public Health Nutr. 2019;22(6):1019-28.
42.    Shamliyan T, Talley KM, Ramakrishnan R, Kane RL. Association of frailty with survival: a systematic literature review. Ageing Res Rev. 2013;12(2):719-36.
43.    Fernandez-Garrido J, Ruiz-Ros V, Buigues C, Navarro-Martinez R, Cauli O. Clinical features of prefrail older individuals and emerging peripheral biomarkers: a systematic review. Arch Gerontol Geriatr. 2014;59(1):7-17.
44.    Azur MJ, Stuart EA, Frangakis C, Leaf PJ. Multiple imputation by chained equations: what is it and how does it work? Int J Methods Psychiatr Res. 2011;20(1):40-9.



N. Buchmann1,2, D. Spira1, M. König1,3, I. Demuth1,4, E. Steinhagen-Thiessen1,2


1. Charité – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Germany, Lipid Clinic at the Interdisciplinary Metabolism Center; 2. Universität Greifswald; Department of Internal Medicine B; 3. Charité – Universitätsmedizin Berlin Department of Nephrology and Internal Intensive Care Medicine; 4. Berlin-Brandenburg Center for Regenerative Medicine (BCRT), Charité University Medicine Berlin, Germany.
Corresponding author: Nikolaus Buchmann, Universität Greifswald Bereich Geriatrie der Universitätsmedizin Greifswald, Fleischmannstraße 8 17475 Greifswald, Email: Nikolaus.buchmann@uni-greifswald.de, Phone: ++49 3836 257 591, FAX: ++49 3836 257 202

J Frailty Aging 2019;8(4)169-175
Published online May 28, 2019, http://dx.doi.org/10.14283/jfa.2019.15



Background: Frailty and the metabolic Syndrome (MetS) are frequently found in old subjects and have been associated with increased risk of functional decline and dependency. Moreover, central characteristics of the MetS like inflammation, obesity and insulin resistance have been associated with the frailty syndrome. However, the relationship between MetS and frailty has not yet been studied in detail. Aim of the current analysis within the Berlin Aging Study II (BASE-II) was to explore associations between MetS and frailty taking important co-variables such as nutrition (total energy intake, dietary vitamin D intake), physical activity and vitamin D-status into account. Methods: Complete cross-sectional data of 1,486 old participants (50.2% women, 68.7 (65.8-71.3) years) of BASE-II were analyzed. MetS was defined following the joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity in 2009. Frailty was defined according to the Fried criteria. Limitations in physical performance were assessed via questionnaire, muscle mass was measured using dual energy X-ray absorptiometry (DXA) and grip strength using a Smedley dynamometer. Adjusted regression models were calculated to assess the association between MetS and Frailty. Results: MetS was prevalent in 37.6% of the study population and 31.9% were frail or prefrail according to the here calculated frailty index. In adjusted models the odds of being frail/prefrail were increased about 50% with presence of the MetS (OR1.5; 95% CI 1.2,1.9; p= 0.002). Moreover the odds of being prefrail/frail were significantly increased with low HDL-C (OR: 1.5 (95%CI: 1.0-2.3); p = 0.037); and elevated waist circumference (OR: 1.65 (95%CI: 1.1-2.3); p = 0.008). Conclusion: The current analysis supports an association between MetS and frailty. There are various metabolic, immune and endocrine alterations in MetS that also play a role in mechanisms underlying the frailty syndrome. To what extent cytokine alterations, inflammatory processes, vitamin D supply and hormonal changes in age and in special metabolic states as MetS influence the development of frailty should be subject of further research.

Key words: Frailty, obesity, metabolic syndrome, elderly.




The rising life expectancy in the western industrialized countries leads to a gain of relatively healthy lifetime upon the largest part of the population, however, an increase of chronic diseases and health limitations in advanced aged often spoils the life time won (1).  Quite a considerable proportion of the older population will face increased vulnerability for functional decline and dependency (2). Chronic diseases and their symptoms drive functional decline, even in absence of acute illness. Moreover, chronic disease is associated with unintentional weight loss, exhaustion, muscle weakness, slow walking speed, and low levels of activity, which has been summarized as the frailty syndrome (3). Although the understanding of frailty has grown in recent years, mechanisms involved in the development of this syndrome are still under discussion and different approaches have been suggested to identify frail elderly (4). There is evidence that chronic inflammation, immune activation and changes in the musculoskeletal and endocrine systems are centrally involved the development of frailty (5, 6) Moreover, genetic, epigenetic and metabolic factors, environmental and lifestyle stressors as well as acute or chronic diseases have been suggested to contribute to the development of frailty (5-7).
Another common disease-cluster in the elderly population is the metabolic syndrome (MetS). As with frailty, the MetS has been associated with increased risk of functional decline and dependency (8-10). The MetS has already been linked with sarcopenia, another important geriatric syndrome which is closely related to the frailty syndrome (11, 12). However, the relationship between MetS and frailty has not yet been studied in detail. It has already been shown that MetS parameters such as high blood pressure, obesity, lipids or diabetes are associated with frailty (13-18). Moreover, MetS has been associated with frailty parameters such as hand grip strength or low physical activity (11, 19). From a pathophysiological point of view, a connection between frailty and MetS seems plausible and few studies confirmed an association (20-24). Pérez-Tasigchana et al. demonstrated that subjects with MetS are at increased risk of frailty (22). However, it was shown that frailty does not add significantly to the prediction of mortality in subjects with MetS, other than the presence of chronic diseases (25). Notably, in diabetic subjects, which often also have Mets, an increased mortality risk has been detected (15, 26).
The Berlin Aging Study II (BASE-II) is a prospective epidemiological study to investigate factors associated with “healthy” or “unhealthy” aging. In this analysis of BASE-II baseline data we aimed to explore associations between MetS and Frailty in order to generate hypotheses about the nature of the supposed link between these two syndromes, nutritional aspects have been suggested to play important roles in both MetS and frailty. One major strength of BASE-II is the detailed coverage of nutritional information (total energy intake, dietary vitamin D load, among others), which allowed us to put emphasis on these issues in this analysis (27, 28).



Study population

In this cross-sectional analysis, we included 1,486 participants (50.2% women, 68.7 (65.8-71.3) years) of the Berlin Aging Study II (BASE-II). BASE-II is a prospective epidemiological study which investigates factors associated with “healthy” or “unhealthy” aging, and has been described previously in detail elsewhere. (29, 30) Briefly, participants were community-dwelling, comparably well-functioning older adults aged between 60 and 84 years. An important aspect here is the investigation of disease development. All subjects gave written informed consent to participate in the study. The study was conducted according to the declaration of Helsinki. The study was approved by the ethics committee of Charité – University Medicine Berlin (project number: EA2/029/09).


We calculated a frailty index based on the validated definition proposed by Fried and colleagues.(3) As described previously some minor modifications compared with the original methodology had to be made since not all assessment methods used by Fried and colleagues were available in the BASE-II study.(31) The five Frailty-criteria were defined as:
–    unintentional loss of at least 5% of the body weight in the least year (“weight loss”).
–    “Self-reported exhaustion” (two questions from the “Center for Epidemiological Studies depression” scale).(32)
–    “Weakness” (assessed by measuring hand grip strength with a Smedley Dynamometer (Scandidact, Denmark). The cut-off values stratified by gender and BMI as suggested by Fried and colleagues were used to identify reduced grip strength reflecting weakness).
–    “Slow walking speed” (assessed by timed “Up&Go” test). (33)
–    “Low physical activity” (based on the question “Are you seldom or never physically active?” If answered “Yes” the criterion was fulfilled).

According to how many criteria were met, participants were ranked as frail (3–5), prefrail (1–2), or not frail (0).

Metabolic Syndrome (MetS)

MetS was defined as suggested by the International Diabetes Federation/American Heart Association/ National Heart, Lung and Blood Institute (IDF/AHA/NHLBI 2009).(8). Blood pressure was measured with an electronic sphygmomanometer (boso-medicus memory, Jung Willingen, Germany), waist circumference was assessed using a non-elastic tape measure at the level of the umbilicus and elevated waist circumference was classified as ≥94 cm in men and ≥80 cm in women (8) .Triglycerides and high-density lipoprotein (HDL-C) cholesterol were measured with enzymatic color tests (Roche/Hitachi Modular; device: ACN 435 und ACN 781). Measurement of glucose level (pre- and post-load) was carried out by photometric concentration determination, insulin levels were analysed by chemiluminescence immunoassays, HbA1c by ion exchange high performance liquid chromatography. An oral glucose tolerance test (OGTT) (34) was performed in BASE-II in participants without previously known diabetes and T2D. Parameters of the MetS were defined as:
–    Elevated waist circumference (waist circumference ≥94 cm in men and ≥80 cm in women)
–    Elevated triglycerides (≥150 mg/dl)
–    Reduced HDL-C (<40 mg/dl in men; <50 mg/dl in women),
–    Elevated blood pressure (Systolic blood pressure ≥130 and/or diastolic blood pressure ≥85 mmHg or prevalent arterial hypertension or use of antihypertensive medication in subjects with arterial hypertension)
–    Elevated fasting glucose (fasting glucose ≥100 mg/dL, prevalent T2D or use of antidiabetic medication in subjects with T2D).

MetS was defined as fulfilling ≥ 3 of the above mentioned MetS parameters.


C-reactive protein level was determined using an immunoturbidimetric assay. Thyroid-stimulating hormone was measured using an electro-chemiluminescence immunoassay (ECLIA). 25[OH]D3 concentration was measured using automated chemiluminescence immunoassays (IDS-iSYS 25-Hydroxy Vitamin D Immunoassay [IDS] and LIAISON 25 OH Vitamin D TOTAL Assay [DiaSorin]). Vitamin D insufficiency was defined as 25[OH]D3-concentrations lower 50 nmol/l (35). To estimate usual nutrient intake, participants completed a validated, self-administered 146-item EPIC-FFQ Potsdam Germany (European Prospective Investigation into Cancer and Nutrition) (36-38). Regular alcohol intake (yes/no) and current smoking status (yes/no) were assessed by standardized questions. As part of the medical examination, diagnoses were obtained through participant reports, with select diagnosis (e.g. diabetes mellitus) being verified by additional blood-laboratory tests. Diagnoses were used to compute a morbidity index largely based on the categories of the Charlson index, which is a weighted sum of moderate to severe, mostly chronic physical illnesses, including cardiovascular (e.g., congestive heart failure), cancer (e.g., lymphoma), and metabolic diseases (e.g., diabetes mellitus) (39, 40). We used the Rapid Assessment of Physical Activity (RAPA) questionaire to assess physical activity of the study population (41). Dual-energy X-ray absorptiometry (DXA): Body composition was assessed with DXA Hologic Discovery Wi (software APEX version 3.0.1). A trained technician performed the DXA measurement protocol. Appendicular lean mass (ALM) in kilograms was calculated as the sum of the non-bone lean mass in arms and legs.


Statistical analyses were carried out using the software package IBM Statistics SPSS 24. Data are given as percentages or as median and interquartile range (IQR). Mann-U-test (not normally distributed variables) or t-test (normally distributed variables) were performed to assess differences between continuous data. The Chi2 test was used to compare proportions. Binary logistic regression models, stepwise adjusted for potential confounders were performed to calculate estimates (odds ratio and 95% confidence interval) for being prefrail/frail (combined variable). The final model (model 3) was adjusted for sex, age, comorbidities, current smoking and alcohol intake as well as C-reactive protein (CRP), thyroid stimulating hormone (TSH), HbA1c, appendicular lean mass (ALM), vitamin D supplementation, vitamin D insufficiency, nutritional intake of vitamin D and total energy intake. We moreover calculated models with the same covariates as in model 3 but replaced MetS by one of its constituents each (model 3B “low HDL-C”; model 3C “elevated triglycerides”; model 3D “elevated waist”; model 3E “insulin resistance; model 3F “elevated blood pressure). An acceptable level of statistical significance was established a priori at p < 0.05.



Basic characteristics

Complete cross-sectional data for metabolic syndrome (MetS) and frailty were available for 1,486 older BASE-II participants (50.2% women, 68.7 [65.8-71.3] years). Baseline characteristics of the study population are given in Supplementary Table 1. MetS was prevalent in 37.6% of the study population and 31.9% were frail or prefrail according to the here calculated frailty index. Elevated waist circumference was the most common of the single components of MetS, followed by elevated blood pressure and elevated fasting glucose. Regarding the single frailty criteria, slow walking speed was detected most frequently and 9% reported low physical activity or exhaustion, respectively.

Table 1 Clinical characteristics of the study population according to frailty

Table 1
Clinical characteristics of the study population according to frailty

*n (proportions) or Median (IQR)


Clinical characteristics of the study population according to frailty

Characteristics of the study population according to the presence of frailty are displayed in Table 1. The prevalence of MetS was higher in frail than in non-frail subjects. Except for high blood pressure, all of the other MetS criteria were more prevalent in frail/prefrail subjects than in non-frail subjects. Moreover, it was noted that on average frail/prefrail participants were older, had higher CRP and HbA1c, and had a lower appendicular lean mass. Also, a higher proportion was classified as vitamin D-deficient.  No differences were observed with respect to alcohol intake, smoking status or total energy intake.

Clinical characteristics of the study population according to MetS

Conversely, when comparing subjects with and without MetS, we found that among those with MetS higher proportions of subjects reported low physical activity or weight loss and showed slow walking speed (Table 2). Notably, as with frailty, participants with MetS were more likely to have vitamin D insufficiency compared to subjects without MetS (52.6% vs. 44.4%; p < 0.001).

Table 2 Parameters of frailty and frailty status according to metabolic syndrome (MetS)

Table 2
Parameters of frailty and frailty status according to metabolic syndrome (MetS)

*Cochran-Armitage test for trend


Logistic regression models assessing the association of MetS and Frailty

We calculated different logistic regression models to estimate the association between MetS and frailty, controlling for potential confounding factors. Stepwise adjustment for an increasing number of covariates was computed (Table 3). Model 1 shows the association between MetS and frailty, adjusted for sex, age, morbidities, smoking, alcohol intake. In model 2 in addition to the covariates of model 1, CRP, TSH, HbA1c and ALM were included. Finally, in model 3A1, we adjusted for vitamin D supplementation, vitamin D insufficiency, nutritional intake of vitamin D and total energy intake. Adjusting for CRP; TSH, HbA1c, ALM, and vitamin D did not significantly change the strength of the association between MetS and frailty, i.e. there was no confounding by these factors. The odds of being frail were increased about 50% with presence of the MetS (OR 1.5; 95% CI 1.2,1.9; p= 0.002). To consider the role of physical inactivity in the MetS-frailty association, we additionally included physical inactivity as a co-variable (model 3A1). However, no significant association between MetS with frailty could be observed (OR 1.2; 95% CI 0.8,1.6; p=0.397).

Table 3 Logistic regression models assessing the association of MetS and frailty

Table 3
Logistic regression models assessing the association of MetS and frailty

Model 0: unadjusted; Model 1: adjusted for sex. age. morbidities. smoking. alcohol intake; Model 2: Model 1 + CRP. TSH. HbA1c. ALM; Model 3 A1: Model 2 + Vitamin D supplementation. vitamin D insufficiency. nutritional intake of Vitamin D. total energy intake; Model 3 A2: Model 2 + Vitamin D supplementation. vitamin D insufficiency. nutritional intake of Vitamin D. total energy intake. Physical activity


Logistic regression models assessing the association of parameters of MetS and frailty

Considering the association between frailty and each of the constituents of MetS separately, the odds of being prefrail/frail were significantly increased with low HDL-C (OR: 1.5 (95%CI: 1.0-2.3); p = 0.037); model 3B) and elevated waist circumference (OR: 1.65 (95%CI: 1.1-2.3); p = 0.008; model 3D), whereas there was no association between elevated triglycerides (modes 3C, elevated fasting glucose (model 3D) and elevated blood pressure (model 3E) with frailty (Table 3).



To date, there are only few studies which have investigated the association between frailty and MetS (20-24). This is surprising, given that subjects with MetS have an increased risk for poor outcomes and are faced with early physical limitations and higher vulnerability (8-10). In a large analysis of US National Health and Nutrition Examination Survey (NHANES) data, frailty and MetS were positively associated in younger subjects (20–65 years), but not among elderly adults (65+ years) (42). The authors of this study concluded that the frailty syndrome was more suitable to predict mortality compared to MetS. In another survey focusing on subjects > 90 years, Hao Q et al. analyzed data of frail participants and found that prevalent MetS did not increase the risk of near-term death in these frail and old subjects (43). Notably, when analyzing data of subjects > 90 years, the aspect of a selection must be considered. Pérez-Tasigchana and colleagues followed 1,499 community-dwelling individuals aged ≥60 and during the follow-up period of 3.5 years, there were 84 incident cases of frailty and the incidence risk was increased in subjects with (44).  This finding was in line with results of Viscogliosi et al (23).
Notably, there is not only consistency as to findings of a link between MetS and frailty, but such an (bilateral) association would also be plausible. Central characteristics of the MetS like inflammation, obesity, diabetes and insulin resistance have been associated with the frailty syndrome (17, 45-48). Intervention strategies like e.g. resistance training are both effective in subjects with MetS and frailty (49).
Our data likewise suggest an association between MetS and frailty. MetS was more prevalent in prefrail/frail subjects. This association was consistent, even after including a larger number of potential confounders into our analyses. A relationship between muscle mass and muscle function with parameters of the frailty syndrome appears to be obvious. Sarcopenia, the age-related loss of muscle mass and frailty overlap in some aspects and with respect to outcomes (50).  The role of MetS in sarcopenia has already been addressed by our group (11). In particular, the interplay between insulin resistance/diabetes and muscle mass/function is a key point. On the one hand, insulin resistance can promote muscle degradation, since glucose serves as a source of energy in muscle cells; on the other hand, reduced muscle mass enhances insulin resistance, since the main breakdown of glucose occurs here (51). Notably, elevated concentrations of glucose and insulin activate the renin-angiotensin-system, which may contribute to the development of hypertension in patients with insulin resistance and may deduce a link between hypertension and muscle mass (52).
Obesity is a main feature of the MetS and favors the development of limitations in physical functions (53). Although in obese subjects, muscle mass may not be reduced, muscle function is impaired, this is particularly true in the elderly (11). This aspect is taken into account with respect to the identification of persons who show an increased risk of loss of function (low grip strength and low walking speed) (54). Endocrine effects of fat, fat infiltration in muscle cells and inflammatory capacities of obesity are putative mechanisms promoting loss of muscle mass or muscle function (9, 55). These mechanisms are being discussed in the context of MetS but are also suggested to play a role in the development of frailty.
Circulating free-fatty-acids (FFA) are commonly found in abdominal obesity and are associated with loss of muscle mass and muscle function (9). On the one hand, FFA can inhibit insulin-mediated glucose uptake in muscle cells. On the other hand FFA affect pancreatic β-cell function (9). Thus, an association between MetS with frailty appears to be biologically plausible, since obesity and insulin resistance are central factors of MetS and are related to key-features of frailty.
The role of lipoprotein patterns is more difficult to discuss in this context. In the current analysis, we found that frail/prefrail participants had more often low HDL-C and elevated triglycerides. Moreover, in adjusted models, higher odds of being frail/prefrail in subjects with low HDL-C was observed. MetS is a state of chronic low-grade inflammation, which has been shown to be associated with frailty and prefrail/frail participants had higher CRP-concentrations in the current analysis as well (5, 45). One purpose of the endocrine active adipose tissue is to store and release lipids. Adiponectin, which is secreted by adipocytes is a tissue hormone that is associated with the metabolism of high-density lipoprotein (HDL-C) and triglycerides. The release of adiponectin is down regulated in insulin resistance and a down-regulation results in higher triglyceride and lower HDL-C concentrations (56). This provides a link to the declaration approach of the joint appearance of parameters of the MetS but also states a link to frailty. An association between obesity and lipid patterns was found in the current analysis, however in adjusted models there was no significant association between insulin resistance (as a parameter of MetS) and frailty status as described in other publications (17, 45). Next to previously described associations between lipids and MetS, physical activity is a way to raise HDL-C concentrations. On the other hand, high concentrations of triglycerides are associated with low HDL concentrations, as activation of the enzyme cholesterol ester transfer protein (CETP) lowers HDL through an exchange of HDL cholesteryl esters with triglycerides. This mechanism in combination with low physical activity in frail/prefrail subjects might explain results regarding the HDL-C-frailty association found here (model 3B).
Notably, participants with MetS as well as subjects with frailty had statistically more often vitamin D insufficiency in our analysis. Vitamin D insufficiency has been linked to wide range of physical limitations, muscle weakness, falls or decline in muscle function (57-60) and receptors for vitamin D were found in many tissues, including muscle and fat (61-63). This adds an additional mechanism, through which an explanation of the relationship between MetS and frailty could be suggested, but the results of the here calculated regression analysis were not affected by vitamin D insufficiency. Nevertheless, a mediating role of vitamin D might be possible.
Low physical activity is both common in MetS and also a definition criteria of frailty (3, 64, 65). Thus we recalculated model 3A1 including physical inactivity in the regression model. Here we found that the association between MetS and frailty vanished. This suggests that physical activity might act as a strong link between MetS and frailty. Low physical activity was a main definition criteria of frailty in our cohort, 33% of the subjects defined as prefrail/frail fulfilled this criteria and due to low physical activity the prevalence of MetS might increase (64, 65). However, the variable for physical inactivity included in model 3A2 was equal to the one we used for the definition of frailty. As no other or more objective measurements for physical activity was available for our analysis, this point cannot sufficiently be clarified with the current analysis. However, insulin resistance, obesity, lipid concentrations and hypertension are positively influenced by sports, so it can be suspected that physical more active subjects face a lower risk for both frailty and MetS.
Our results are subject to limitations. First, the results arebased on cross-sectional information. There are various confounders associated both MetS and frailty. The cause-effect relationship cannot be concluded from a cross-sectional design. Furthermore, although the questionnaires used to assess physical activity, unintentional weight loss and exhaustion are well established instruments (RAPA questionnaire and items from the “Center for Epidemiological Studies depression” scale (32, 41), the data were self-reported and thus subjective. Finally, the BASE-II cohort is a typical convenience sample. Diseases such as chronic obstructive pulmonary disease (COPD) and coronary heart disease are under-represented, and therefore the results from this study cohort cannot simply be extrapolated to the general population. Strength of the study is the detailed information on medication, parameters of the MetS and potential confounders that we adjusted for.
In conclusion, our current analysis of cross-sectional BASE-II baseline data supports an association between MetS and frailty. There are various metabolic, immune and endocrine alterations in MetS that also play a role in mechanisms underlying the frailty syndrome. To what extent physical inactivity, cytokine alterations, inflammatory processes, vitamin D supply and hormonal changes in age and in special metabolic states as MetS influence the development of frailty should be subject of further research. The metabolic syndrome is a state of low-grade inflammation, which is also generally observed in aging. The role of such inflammatory processes in the development of frailty represents an interesting field to investigate possible pathomechanisms and therapeutic approaches, particularly in MetS subjects. Also, the current analysis indicates that nutritional intervention studies with i.e. Vitamin D substitution or intensified physical activity in subjects with MetS and Frailty could potentially show positive effects on these syndromes. It seems to be particularly worthwhile to focus on older cohorts, as both syndromes are highly prevalent in the old.


Conflicts of Interest: None declared by the authors.



1.    Perenboom RJM, Van Herten LM, Boshuizen HC, Van Den Bos GAM. Trends in disability-free life expectancy. Disability and Rehabilitation. 2004;26:377-386.
2.    Harwood RH, Sayer AA, Hirschfeld M. Current and future worldwide prevalence of dependency, its relationship to total population, and dependency ratios. Bulletin of the World Health Organization. 2004;82:251-258.
3.    Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, et al. Frailty in older adults: evidence for a phenotype. The journals of gerontology Series A, Biological sciences and medical sciences. 2001;56:M146-156.
4.    Walston JD, Bandeen-Roche K. Frailty: a tale of two concepts. BMC Medicine. 2015;13:185.
5.    Chen X, Mao G, Leng SX. Frailty syndrome: an overview. Clinical Interventions in Aging. 2014;9:433-441.
6.    Fougère B, Vellas B, van Kan GA, Cesari M. Identification of biological markers for better characterization of older subjects with physical frailty and sarcopenia. Translational Neuroscience. 2015;6:103-110.
7.    Clegg A, Young J, Iliffe S, Rikkert MO, Rockwood K. Frailty in elderly people. Lancet (London, England). 2013;381:752-762.
8.    Alberti KG, Eckel RH, Grundy SM, Zimmet PZ, Cleeman JI, Donato KA, et al. Harmonizing the metabolic syndrome: a joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation. 2009;120:1640-1645.
9.    Kaur J. A Comprehensive Review on Metabolic Syndrome. Cardiology Research and Practice. 2014;2014:21.
10.    Penninx BWJH, Nicklas BJ, Newman AB, Harris TB, Goodpaster BH, Satterfield S, et al. Metabolic Syndrome and Physical Decline in Older Persons: Results from the Health, Aging and Body Composition Study. The Journals of Gerontology: Series A. 2009;64A:96-102.
11.    Buchmann N, Nikolov J, Spira D, Demuth I, Steinhagen-Thiessen E, Eckardt R, et al. Identifying sarcopenia in metabolic syndrome: data from the Berlin Aging Study II. Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences. 2015;71:265-272.
12.    Scott D, Park MS, Kim TN, Ryu JY, Hong HC, Yoo HJ, et al. Associations of low muscle mass and the metabolic syndrome in Caucasian and Asian middle-aged and older adults. The journal of nutrition, health & aging. 2016;20:248-255.
13.    Castrejón-Pérez RC, Gutiérrez-Robledo LM, Cesari M, Pérez-Zepeda MU. Diabetes mellitus, hypertension and frailty: A population-based, cross-sectional study of Mexican older adults. Geriatrics & gerontology international. 2017;17:925-930.
14.    Lee JSW, Auyeung TW, Leung J, Kwok T, Leung PC, Woo J. Physical frailty in older adults is associated with metabolic and atherosclerotic risk factors and cognitive impairment independent of muscle mass. The journal of nutrition, health & aging. 2011;15:857-862.
15.    Morley JE, Malmstrom TK, Rodriguez-Mañas L, Sinclair AJ. Frailty, Sarcopenia and Diabetes. Journal of the American Medical Directors Association. 2014;15:853-859.
16.    Blaum CS, Xue QL, Michelon E, Semba RD, Fried LP. The Association Between Obesity and the Frailty Syndrome in Older Women: The Women’s Health and Aging Studies. Journal of the American Geriatrics Society. 2005;53:927-934.
17.    Hubbard RE, Lang IA, Llewellyn DJ, Rockwood K. Frailty, Body Mass Index, and Abdominal Obesity in Older People. The Journals of Gerontology: Series A. 2010;65A:377-381.
18.    Fulop T, Dupuis G, Fortin C, Douziech N, Larbi A. T cell response in aging: influence of cellular cholesterol modulation.  Lymphocyte Signal Transduction. Springer; 2006:157-169.
19.    Laaksonen DE, Lakka H-M, Salonen JT, Niskanen LK, Rauramaa R, Lakka TA. Low Levels of Leisure-Time Physical Activity and Cardiorespiratory Fitness Predict Development of the Metabolic Syndrome. Diabetes Care. 2002;25:1612-1618.
20.    Fulop T, Larbi A, Witkowski JM, McElhaney J, Loeb M, Mitnitski A, et al. Aging, frailty and age-related diseases. Biogerontology. 2010;11:547-563.
21.    Lin F, Roiland R, Chen D-G, Qiu C. Linking cognition and frailty in middle and old age: metabolic syndrome matters. International Journal of Geriatric Psychiatry. 2015;30:64-71.
22.    Perez-Tasigchana RF, Leon-Munoz LM, Lopez-Garcia E, Gutierrez-Fisac JL, Laclaustra M, Rodriguez-Artalejo F, et al. Metabolic syndrome and insulin resistance are associated with frailty in older adults: a prospective cohort study. Age Ageing. 2017:1-6.
23.    Viscogliosi G. The Metabolic Syndrome: A Risk Factor for the Frailty Syndrome? Journal of the American Medical Directors Association.17:364-366.
24.    Ishii S, Tanaka T, Akishita M, Ouchi Y, Tuji T, Iijima K. Metabolic syndrome, sarcopenia and role of sex and age: cross-sectional analysis of Kashiwa cohort study. PloS one. 2014;9:e112718.
25.    Hoogendijk EO, Huisman M, van Ballegooijen AJ. The role of frailty in explaining the association between the metabolic syndrome and mortality in older adults. Experimental Gerontology. 2017;91:5-8.
26.    Cacciatore F, Testa G, Galizia G, Della-Morte D, Mazzella F, Langellotto A, et al. Clinical frailty and long-term mortality in elderly subjects with diabetes. Acta diabetologica. 2013;50:251-260.
27.    Manal B, Suzana S, Singh DK. Nutrition and Frailty: A Review of Clinical Intervention Studies. The Journal of frailty & aging. 2015;4:100-106.
28.    Halfon M, Phan O, Teta D. Vitamin D: a review on its effects on muscle strength, the risk of fall, and frailty. Biomed Res Int. 2015;2015:953241.
29.    Gerstorf D, Bertram L, Lindenberger U, Pawelec G, Demuth I, Steinhagen-Thiessen E, et al. Editorial. Gerontology. 2016;62:311-315.
30.    Bertram L, Böckenhoff A, Demuth I, Düzel S, Eckardt R, Li S-C, et al. Cohort profile: the Berlin aging study II (BASE-II). International journal of epidemiology. 2013:dyt018.
31.    Spira D, Buchmann N, Nikolov J, Demuth I, Steinhagen-Thiessen E, Eckardt R, et al. Association of low lean mass with frailty and physical performance: a comparison between two operational definitions of sarcopenia—data from the Berlin Aging Study II (BASE-II). Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences. 2015;70:779-784.
32.    Radloff LS. The CES-D Scale. Applied Psychological Measurement. 1977;1:385-401.
33.    Podsiadlo D, Richardson S. The timed “Up & Go”: a test of basic functional mobility for frail elderly persons. Journal of the American geriatrics Society. 1991;39:142-148.
34.    Organization WH. Screening for type 2 diabetes: report of a World Health Organization and International Diabetes Federation meeting. 2003.
35.    Del Valle HB, Yaktine AL, Taylor CL, Ross AC. Dietary reference intakes for calcium and vitamin D. National Academies Press; 2011.
36.    Kroke A, Klipstein-Grobusch K, Voss S, Möseneder J, Thielecke F, Noack R, et al. Validation of a self-administered food-frequency questionnaire administered in the European Prospective Investigation into Cancer and Nutrition (EPIC) Study: comparison of energy, protein, and macronutrient intakes estimated with the doubly labeled water, urinary nitrogen, and repeated 24-h dietary recall methods. The American Journal of Clinical Nutrition. 1999;70:439-447.
37.    Fielding RA, Vellas B, Evans WJ, Bhasin S, Morley JE, Newman AB, et al. Sarcopenia: an undiagnosed condition in older adults. Current consensus definition: prevalence, etiology, and consequences. International working group on sarcopenia. Journal of the American Medical Directors Association. 2011;12:249-256.
38.    Bohlscheid-Thomas S, Hoting I, Boeing H, Wahrendorf J. Reproducibility and relative validity of energy and macronutrient intake of a food frequency questionnaire developed for the German part of the EPIC project. European Prospective Investigation into Cancer and Nutrition. International Journal of Epidemiology. 1997;26:S71.
39.    Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. Journal of chronic diseases. 1987;40:373-383.
40.    Meyer A, Salewsky B, Spira D, Steinhagen-Thiessen E, Norman K, Demuth I. Leukocyte telomere length is related to appendicular lean mass: cross-sectional data from the Berlin Aging Study II (BASE-II). The American journal of clinical nutrition. 2016;103:178-183.
41.    Topolski TD, LoGerfo J, Patrick DL, Williams B, Walwick J, Patrick MMB. Peer reviewed: the Rapid Assessment of Physical Activity (RAPA) among older adults. Preventing chronic disease. 2006;3.
42.    Kane AE, Gregson E, Theou O, Rockwood K, Howlett SE. The association between frailty, the metabolic syndrome, and mortality over the lifespan. GeroScience. 2017;39:221-229.
43.    Hao Q, Song X, Yang M, Dong B, Rockwood K. Understanding Risk in the Oldest Old: Frailty and the Metabolic Syndrome in a Chinese Community Sample Aged 90+ Years. J Nutr Health Aging. 2016;20:82-88.
44.    Pérez-Tasigchana RF, León-Muñoz LM, Lopez-Garcia E, Gutierrez-Fisac JL, Laclaustra M, Rodríguez-Artalejo F, et al. Metabolic syndrome and insulin resistance are associated with frailty in older adults: a prospective cohort study. Age and Ageing. 2017:1-6.
45.    Barzilay JI, Blaum C, Moore T, et al. Insulin resistance and inflammation as precursors of frailty: The cardiovascular health study. Archives of Internal Medicine. 2007;167:635-641.
46.    Morley JE, Malmstrom TK, Rodriguez-Mañas L, Sinclair AJ. Frailty, Sarcopenia and Diabetes. Journal of the American Medical Directors Association.15:853-859.
47.    Atiénzar P, Abizanda P, Guppy A, Sinclair A. Diabetes and frailty: An emerging issue. Part 2: Linking factors. 2012.
48.    Cobo A, Vazquez LA, Reviriego J, Rodriguez-Manas L. Impact of frailty in older patients with diabetes mellitus: An overview. Endocrinologia y nutricion : organo de la Sociedad Espanola de Endocrinologia y Nutricion. 2016;63:291-303.
49.    Sundell J. Resistance Training Is an Effective Tool against Metabolic and Frailty Syndromes. Advances in Preventive Medicine. 2011;2011.
50.    Cederholm T. Overlaps between Frailty and Sarcopenia Definitions. Nestle Nutrition Institute workshop series. 2015;83:65-69.
51.    Srikanthan P, Karlamangla AS. Relative Muscle Mass Is Inversely Associated with Insulin Resistance and Prediabetes. Findings from The Third National Health and Nutrition Examination Survey. The Journal of Clinical Endocrinology & Metabolism. 2011;96:2898-2903.
52.    Leung PS. The physiology of a local renin–angiotensin system in the pancreas. The Journal of physiology. 2007;580:31-37.
53.    Lang IA, Llewellyn DJ, Alexander K, Melzer D. Obesity, physical function, and mortality in older adults. Journal of the American Geriatrics Society. 2008;56:1474-1478.
54.    McLean RR, Shardell MD, Alley DE, Cawthon PM, Fragala MS, Harris TB, et al. Criteria for clinically relevant weakness and low lean mass and their longitudinal association with incident mobility impairment and mortality: the foundation for the National Institutes of Health (FNIH) sarcopenia project. Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences. 2014;69:576-583.
55.    Barberi L, Scicchitano BM, Musaro A. Molecular and Cellular Mechanisms of Muscle Aging and Sarcopenia and Effects of Electrical Stimulation in Seniors. European Journal of Translational Myology. 2015;25:5227.
56.    Cnop M, Havel PJ, Utzschneider KM, Carr DB, Sinha MK, Boyko EJ, et al. Relationship of adiponectin to body fat distribution, insulin sensitivity and plasma lipoproteins: evidence for independent roles of age and sex. Diabetologia. 2003;46:459-469.
57.    Bischoff-Ferrari HA. Relevance of vitamin D in muscle health. Reviews in Endocrine and Metabolic Disorders. 2012;13:71-77.
58.    Annweiler C, Schott A-M, Berrut G, Fantino B, Beauchet O. Vitamin D-related changes in physical performance: a systematic review. The journal of nutrition, health & aging. 2009;13:893-898.
59.    Dukas L, Staehelin H, Schacht E, Bischoff H. Better functional mobility in community-dwelling elderly is related to D-hormone serum levels and to daily calcium intake. The journal of nutrition, health & aging. 2004;9:347-351.
60.    Wicherts IS, van Schoor NM, Boeke AJP, Visser M, Deeg DJ, Smit J, et al. Vitamin D status predicts physical performance and its decline in older persons. The Journal of Clinical Endocrinology & Metabolism. 2007;92:2058-2065.
61.    Autier P, Boniol M, Pizot C, Mullie P. Vitamin D status and ill health: a systematic review. The Lancet Diabetes & Endocrinology. 2014;2:76-89.
62.    Body JJ, Bergmann P, Boonen S, Devogelaer JP, Gielen E, Goemaere S, et al. Extraskeletal benefits and risks of calcium, vitamin D and anti-osteoporosis medications. Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA. 2012;23:011-1891.
63.    Simpson R, Thomas G, Arnold A. Identification of 1, 25-dihydroxyvitamin D3 receptors and activities in muscle. Journal of Biological Chemistry. 1985;260:8882-8891.
64.    Atlantis E, Martin SA, Haren MT, Taylor AW, Wittert GA, Study ftFAMA. Lifestyle factors associated with age-related differences in body composition: the Florey Adelaide Male Aging Study. The American Journal of Clinical Nutrition. 2008;88:95-104.
65.    Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, et al. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute scientific statement. Circulation. 2005;112:2735-2752.



F. Xu1, S.A. Cohen2, I.E. Lofgren3, G.W. Greene3, M.J. Delmonico1, M.L. Greaney2


1. Department of Kinesiology, University of Rhode Island, Independence Square II, Kingston, Rhode Island, 02881, United States; 2. Health Studies program, University of Rhode Island, Independence Square II, Kingston, Rhode Island, 02881, United States; 3. Department of Nutrition and Food Sciences, University of Rhode Island, Fogarty Hall, Kingston, Rhode Island, 02881, United States.
Corresponding author: Furong Xu, PhD, Department of Kinesiology, 25 West Independence Way, Suite P, The University of Rhode Island, Kingston, RI 02881, Email: fxu2007@uri.edu

J Frailty Aging 2018;in press
Published online October 19, 2018, http://dx.doi.org/10.14283/jfa.2018.34



Background: Physical activity reduces the likelihood of developing metabolic syndrome (MetS). However, the association between different physical activity levels and MetS remains unclear in older adults with obesity. Methods: This cross-sectional study used four waves of data (2007-2008, 2009-2010, 2011-2012, 2013-2014) from two datasets: The National Health and Nutrition Examination Survey and United Sates Department of Agriculture’s Food Patterns Equivalents Database. The sample included adults 60+ years of age (n= 613) with obesity who had physical activity and MetS data.  Physical activity was assessed using the Global Physical Activity Questionnaire and categorized into three physical activity levels (low, medium, and high); and medium or high physical activity levels are aligned with or exceed current physical activity recommendations. Participants were classified as having MetS using a commonly agreed upon definition. Multiple logistic regression models examined the association between the three physical activity levels and MetS risk factors and MetS. All analyses adjusted for potential confounding variables and accounted for complex sampling. Results: Of 613 respondents, 72.1% (n=431) were classified as having MetS, and 44.3% (n = 263) had not met physical activity recommendations. Participants with high levels of physical activity had a lower risk of MetS (OR = 0.31, 95%CI: 0.13, 0.72) and more healthful levels of high-density lipoprotein cholesterol (OR = 0.39, 95%CI: 0.18, 0.84), blood pressure (OR = 0.39, 95%CI: 0.20, 0.77), fasting glucose (OR = 0.34, 95%CI: 0.15, 0.78) than participants categorized as having low physical activity. Conclusions: Physical activity is associated with lower risk of MetS only for participants with the highest level of physical activity, which suggests that physical activity dosage is important to reduce MetS risk in older adults with obesity.

Key words: Older adults, obesity, metabolic syndrome, physical activity.




Currently 35.4% of adults aged 60+ years have obesity [body mass index (BMI) ≥ 30kg/m2] (1).  Increasing age and obesity both contribute to metabolic syndrome (MetS) in older adults (2-4).  MetS, a predictor of cardiovascular mortality (2), is defined by the presence of at least three of the following risk factors: 1) elevated waist circumference, 2) elevated triglycerides, 3) low high-density lipoprotein cholesterol (HDL-C), 4) elevated blood pressure, and 5) elevated fasting glucose (4). The prevalence of MetS and obesity are increasing concurrently (5, 6).  About half (46.7%) of adults 60+ years have MetS (5), and it is estimated that nearly 80% of older adults with obesity have MetS (6). Research is needed to understand how to better address MetS in this population since rates of MetS are greatest among older adults with obesity (5, 6) and older adults are the fast-growing segment of the US population (7).
Physical activity (PA) may prevent or reduce the MetS risk factors as well as the risk of MetS among adults (8-10). Older adults may respond to PA differently than younger adults given that aging-related function declines and various health conditions exist in the older adult population (11). Yet, research examining the relationship between PA and MetS among older adults is lacking (12), and the association between PA dose and MetS in older adults with obesity remains unclear. Furthermore, although physical function limitations may be a barrier to PA participation for all individuals with obesity (13) including older adults (14), the impact of physical function has not been considered in previous studies examining the association between PA and MetS in older adults with obesity. Therefore, the purpose of this study was to examine the association between different levels of PA participation and MetS risk factors and MetS in a representative sample of older adults with obesity independent of physical function limitations.



Study Design

The present study was a cross-sectional analysis of data from the National Health and Nutrition Examination Survey (NHANES) and United Sates Department of Agriculture’s Food Patterns Equivalents Database (USDA-FPED) from 2007 to 2014. Demographic and questionnaire data were collected at respondents’ homes. Examination and laboratory data were collected in the Mobile Examination Center (MEC) (15).  All data used for the present study were de-identified and are publicly available on the Centers for Disease Control and Prevention and USDA-FPED websites (15, 16).

Analytic sample

The sample was limited to respondents 60+ years of age (n=7522) with obesity (23.5%, n=1733) whose PA and MetS risk factors data were available. Respondents were excluded if they had physical function limitations and were physically inactive as defined by a zero-minute PA time (13, 15). Respondents were classified as having physical function limitations if they reported experiencing any of the following: difficulty or challenges in walking, standing, lifting and carrying things up to 10 pounds, stooping, crouching or kneeling, or that they used special equipment for walking (13).  In total, 662 respondents met study inclusion criteria, 49 (7.5%) of whom were excluded due to lack of PA and/or having physical function limitations, yielding a final analytical sample of 613 respondents.

Physical activity

PA during a typical week was assessed in three PA domains (work, travel, recreational) using the 16-item Global Physical Activity Questionnaire (17, 18).  PA time in each domain was expressed in minutes and summed to create minutes of moderate to vigorous PA per week, then converted to metabolic equivalent (MET) minutes which takes PA intensity into account (moderate PA=4 MET, vigorous PA =8 MET) (18). This information was then used to create three PA categories equivalent to the U.S. Department of Health and Human Services PA guidelines: low (< 600 MET-minutes/week), medium (600-1200 MET-minutes/week), and high (> 1200 MET-minutes/week) (19). The medium or high PA levels are aligned with current PA recommendations (≥ 600 MET-minutes/week or 150+ minutes of moderate to vigorous PA) (19).

Metabolic syndrome (MetS) parameters

Five measures were used to classify people as having MetS: triglycerides, HDL-C, fasting glucose, waist circumference and blood pressure. Blood samples were collected at the MEC and then analyzed for triglycerides, HDL-C, plasma glucose at the University of Minnesota following standard laboratory procedures (15). Waist circumference and blood pressure measures were taken at the MEC by trained health technicians using standard methods (15). In addition, participants also reported if they had treatment for triglyceride, or HDL-C, or blood pressure, or blood glucose (15, 20). Participants were classified as having MetS if they had three or more of the following: 1) waist circumference ≥ 102 cm in men or ≥ 88 cm in women, 2) triglyceride ≥ 150 mg/dL or with reported lipid lowering treatment, 3) HDL-C < 40 mg/dL in men or < 50 mg/dL in women or with reported cholesterol treatment, 4) blood pressure, systolic ≥ 130 mm Hg or diastolic ≥ 85 mm Hg or both or with reported blood pressure treatment, 5) fasting glucose ≥ 100 mg/dL or with reported blood glucose treatment (4, 20).


Overall diet quality was determined using National Cancer Institute’s simple Healthy Eating Index-2015 scoring algorithm using information from two 24-hour dietary recalls data from NHANES and USDA-FPED (15, 16, 21). Participants were stratified into three tertiles based on their total dietary quality score (ranging from 0 to 100) with the top tertile (scores >59.0) being classified as a healthier diet and the other two tertiles (scores 0-46.3 or 46.3-59.0) being considered indicative of a less healthful diet.
Comorbid disease status was measured using a single question that asked respondents’ chronic diseases identified by their doctor or other health professional: arthritis, cancer, cardiovascular disease, chronic kidney disease, diabetes, hypertension, pulmonary disease, and stroke (15, 22).  A summary score was created ranging from 0-8 and further divided into four categories: 0, 1, 2, or ≥3.
Alcohol consumption was classified into three categories based on respondents’ reported alcohol consumption per day in the past year: heavy drinkers, average of >2 drinks per day for men and >1 drink per day for women; moderate drinkers, average of ≤2 drinks per day for men and ≤1 drink per day for women; never drinkers, no drinking reported (15, 23).
Demographic information including age, sex, race/ethnicity, smoking status (never, former, current), education (high school degree or less vs. some college or higher). Respondents also reported their family income and family size, which was then used to calculate poverty to income ratio (PIR) (15, 24).  PIR was divided into two categories based on the poverty guidelines: at or above (PIR ≥1) and below federal poverty level (PIR<1) (24).

Statistical analyses

The MEC exam 2-year weights were used as the sample weight for all analyses (25).  Baseline characteristics are presented as weighted means ± standard error for continuous variables and as count and weighted percentages for categorical variables. To assess the associations between PA and MetS risk factors and MetS, the PA was classified into three levels to compare medium and high PA levels with low PA level. Multiple logistic regression models were conducted using PROC SURVEYLOGISTIC, including STRATA, CLUSTER, and WEIGHT statement, to obtain adjusted odds ratios (ORs) with 95% confidence intervals, adjusted for age, sex, race/ethnicity, and current smokers (model 1). As education, poverty levels, alcohol intake, level of comorbidities, BMI, and diet quality may be associated with PA levels, a second logistic regression model was constructed that included these variables as covariates (model 2). Additionally, two sensitivity analyses were conducted in which respondents with physical function limitations were excluded or the drug treatment for elevated triglycerides or low HDL-C as the inclusion criteria to define MetS was excluded, and results did not change substantially. All analyses were performed using SAS 9.4 (SAS Institute Inc., Cary, NC, USA), and p-values <0.05 were considered statistically significant.



Approximately three quarters (72.1%, n = 431) of participants were classified as having MetS, 53.5% of respondents were females, 20.5% were racial/ethnic minorities, 43.1% had a high school degree or less and 9.5% whose household income placed them below the federal poverty line. About half of respondents (44.3%) were at the lowest PA level and had not met PA recommendations. As seen in Table 1, there were no significant differences in demographic characteristics by MetS status. However, compared to individuals without MetS, participants with MetS were more likely to have a low PA level and more likely in groups that have 2+ chronic diseases. In addition, individuals classified as having MetS had larger waist circumference, higher triglycerides, blood pressure, and blood glucose, but lower HDL-C and were more likely to report treatment for lipid profiles and blood glucose.

Table 1 Characteristics of subjects with and without metabolic syndrome (n=613)

Table 1
Characteristics of subjects with and without metabolic syndrome (n=613)

Note: Data are present as weighted mean ± standard error for continuous variables and as count and weighted percentages for categorical variables; PRI = poverty to income ratio, BMI = body mass index, PA= physical activity, MET = metabolic equivalent, HDL-C=high-density lipoprotein cholesterol, Low PA = <600 MET-minutes/week, Medium PA = 600-1,200 MET-minutes/week, High PA > 1,200 MET-minutes/week, # HEI-2015 = Healthy Eating Index-2015, * p < 0.05.


As shown in Table 2, the multiple logistic models adjusted for age, sex, race/ethnicity and current smoking status (model 1) revealed an inverse association between PA level and the risk of MetS (p=0.007). That is, respondents classified as having a high PA level were 63% less likely to have MetS than individuals with low levels of PA (OR = 0.37, 95% CI: 0.20, 0.68). Although not statistically significant, participants with a medium level of PA were 44% less likely to have MetS than individuals with low PA (OR = 0.56, 95% CI: 0.23, 1.34).

Table 2 The association of physical activity with metaboloc syndrome and metabolic syndrome risk factors (n=613)

Table 2
The association of physical activity with metaboloc syndrome and metabolic syndrome risk factors (n=613)

Note: # Model 1: Adjusted for age, sex, race/ethnicity, and current smokers; & Model 2: Additionally adjusted for education levels, poverty levels, alcohol intake, level of comorbidities, BMI, and diet quality; OR=odds ratio, CI=confidence interval, HDL-C = high-density lipoprotein cholesterol, PA= physical activity, Low PA: <600 MET-minutes/week, Medium PA: 600-1,200 MET-minutes/week, Higher PA: > 1,200 MET-minutes/week, * p < 0.05.


Analyses examining the relationships between PA levels and the five MetS risk factors determined that participants with a high level of PA had a lower risk of having hypertriglyceridemia, low HDL-C, and high blood pressure (see Table 2).  The same pattern was observed for HDL-C and blood pressure and participants with a high level of PA had lower risk of elevated fasting glucose (OR = 0.34, 95% CI: 0.15, 0.78) after additionally adjusting for education, poverty levels, alcohol intake, level of comorbidities, BMI, and diet quality (model 2).
Subsequent multiple logistic analyses examined the association of PA levels with MetS risk factors in males and females respectively adjusted for all confounders. Males with a high level of PA had a lower risk of MetS than males with a low level of PA (OR = 0.33, 95%CI: 0.14, 0.80). However, females with a high level of PA did not have a lower risk of MetS. In addition, males with a high level of PA had a reduced risk of low HDL-C (OR = 0.35, 95%CI: 0.13, 0.94) and elevated blood pressure (OR = 0.35, 95%CI: 0.14, 0.87). Female participants with a high level of PA had lower risk of elevated blood glucose (OR = 0.12, 95%CI: 0.03, 0.51) but no other statistically significant differences in MetS risk factors observed (Figure 1).

Figure 1 The association of physical activity and metabolic syndrome by sex

Figure 1
The association of physical activity and metabolic syndrome by sex

PA= physical activity, HDL-C= high-density lipoprotein cholesterol, Medium PA: 600-1,200 MET-minutes/week, High PA: > 1,200 MET-minutes/week, Low PA: < 600 MET-minutes/week and Low PA is reference category, all analyses adjusted for age, race/ethnicity, education level, household income, current smokers, alcohol intake, level of comorbidities, BMI, and diet quality, * p < 0.05.



A large percentage of the sample had MetS (72.1%) and about half of the respondents not met current PA recommendations (44.3%). Study findings suggest that more than 1,200 MET-minutes of PA per week is needed to reduce MetS risk in older adults with obesity who do not have physical function limitations. This information is important given the increasing obesity prevalence among older adults and calls to increase PA among older adults (1, 8-10).
Study findings indicate that just meeting or slightly exceeding PA recommendations did not reduce MetS risk. The risk of MetS did not decline until PA levels exceeded 1,200 MET-minutes per week, which is equivalent to 42+ minutes of moderate or 21+ minutes of vigorous PA per day (18).  Studies with adults of all ages have determined that greater levels of PA are associated with a lower MetS risk (8-10).  Bell et al. (2015) found that among older adults with obesity and those with MetS were less likely to be physically active, although the relationship between PA dosage and MetS risk was not examined and function limitation was not considered (12).  The present study extends existing research by focusing on older adults with obesity without physical function limitations and indicates the importance of a PA dosage in excess of 1,200 MET-minutes of PA a week to reduce MetS risk in this population.
A notable finding from the present study is that participants with a high level of PA had a lower risk of low HDL-C, high blood pressure, and elevated fasting glucose. This finding is consistent with previous studies that have found increased HDL-C, lower risk of high blood pressure and elevated fasting glucose associated with higher PA levels although these studies did not focus on older adults with obesity or the US population (8-10).  Therefore, comparisons should be made with caution due to the differences in the study populations. Nevertheless, the present study controlled for potential confounders that could bias the study findings (e.g., diet, level of comorbidities) which support the benefits of a high PA dosage on MetS risk factors particularly to improve lipid profiles and manage blood pressure and blood glucose.
Additionally, the present study found that males with a high level of PA had a reduced risk of MetS, HDL-C, and blood pressure while a high PA level in females was only associated with reduced blood glucose. Few studies have examined the association of PA dosage with MetS risk factors or with MetS by sex, and available studies are not specific to older adults with obesity (26-29).  Nonetheless, study findings are somewhat consistent with prior research that have observed an inverse relationship between PA and risk of MetS in males (26) and in males and females (27). Previous studies also indicate that a higher PA dosage is associated with lower blood glucose only for males (28) and with higher HDL-C was for females (29).  Findings from the current study suggest that among older adults with obesity, males and females may respond to PA dosage differently. Future research should explore these differences.
The present study has strengths and limitations. This is one of first studies, to our knowledge, to use nationally representative data to examine the association between different PA levels and MetS among older adults with obesity independent of physical function limitations. In addition, the study assessed three domains of PA (work, travel, and recreational) instead of focusing only on leisure time PA (8, 9).  Nonetheless, the study is subject to several important limitations including its cross-sectional design that does not allow causality to be determined. PA was assessed by self-report, although a valid instrument was used (17). Also, the study was limited by using existing datasets thus not all possible confounders associated with aging could be examined (e.g., cognitive function, affective status). Additionally, 7.5% of the eligible sample was excluded due to being physically inactive and/or having physical function limitations. Although, no substantive differences were found with the sensitivity analysis due to this exclusion, this limits our generalizability to older adults with obesity and physical function limitations.
To conclude, the present study indicates that meeting or slightly exceeding PA recommendations may not be sufficient for lowering MetS risk and only participants with a high PA level (>1,200 MET-minutes/week) had a reduced risk of MetS. Thus, PA dosage might be important to address MetS risk in older adults with obesity and improve public health thus people should be encouraged to exceed PA recommendations if capable but PA promotion needs to be sensitive to gender in different contexts.


Acknowledgement: Data used for the current study were collected by the National Center for Health Statistics, Centers for Disease Control and Prevention and combined from two datasets: The National Health and Nutrition Examination Survey and United Sates Department of Agriculture’s Food Patterns Equivalents Database. The content in this paper is the responsibility of the authors and does not necessarily represent the official views of Centers for Disease Control and Prevention or United Sates Department of Agriculture.
Conflict of interest: The authors declare that there are no conflicts of interest regarding this paper.
Ethical standards: The University of Rhode Island Institutional Review Board has determined this study does not meet the definition of human subject research under the purview of federal regulation 45 CFR 46 regarding human subject research since data used for the current study are de-identified and publicly accessible.



1.    Ogden CL, Carroll MD, Kit BK et al. Prevalence of childhood and adult obesity in the United States, 2011-2012. JAMA 2014;311(8):806–814. doi:10.1001/jama.2014.732
2.    Kaur J. A Comprehensive Review on Metabolic Syndrome. Cardiology Research and Practice 2014; 2014:943162. doi:10.1155/2014/943162.
3.    Janssen I. Morbidity and Mortality Risk Associated with an Overweight BMI in Older Men and Women. Obesity 2007; 15: 1827–1840. doi:10.1038/oby.2007.217
4.    Alberti KG, Eckel RH, Grundy SM et al. Harmonizing the metabolic syndrome: a joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation 2009;120(16):1640–5.
5.    Aguilar M, Bhuket T, Torres S et al. Prevalence of the Metabolic Syndrome in the United States, 2003-2012. JAMA 2015;313(19):1973–1974. doi:10.1001/jama.2015.4260
6.    Wildman RP, Muntner P, Reynolds K et al. The obese without cardiometabolic risk factor clustering and the normal weight with cardiometabolic risk factor clustering: prevalence and correlates of 2 phenotypes among the US population (NHANES 1999-2004). Arch Intern Med 2008; 168:1617–24.
7.    Knickman JR, Snell EK. The 2030 Problem: Caring for Aging Baby Boomers. Health Services Research 2002;37(4):849-884. doi:10.1034/j.1600-0560.2002.56.x.
8.    Lee J, Kim Y, Jeon JY. Association between physical activity and the prevalence of metabolic syndrome: from the Korean National Health and Nutrition Examination Survey, 1999–2012. Springer Plus 2016;5(1):1870. doi:10.1186/s40064-016-3514-5.
9.    Rennie KL, McCarthy N, Yazdgerdi S et al. Association of the metabolic syndrome with both vigorous and moderate physical activity. Int J Epidemiol 2003; 32:600–606. doi: 10.1093/ije/dyg179.
10.    Yu Z, Ye X, Wang J et al. Associations of physical activity with inflammatory factors, adipocytokines, and metabolic syndrome in middle-aged and older Chinese people. Circulation 2009; 119: 2969–2977.
11.    Stratton JR, Levy WC, Caldwell JH et al. Effects of aging on cardiovascular responses to parasympathetic withdrawal. J Am Coll Cardiol 2003; 41: 2077–2083.
12.    Bell JA, Hamer M, van Hees VT et al. Healthy obesity and objective physical activity. The American Journal of Clinical Nutrition 2015;102(2):268-275.
13.    Peterson MD, Snih SA, Stoddard J et al. Obesity Misclassification and the Metabolic Syndrome in Adults with Functional Mobility Impairments: NHANES 2003–2006. Preventive medicine 2014; 60:71-76.
14.    Paterson DH, Warburton DE. Physical activity and functional limitations in older adults: a systematic review related to Canada’s Physical Activity Guidelines. The International Journal of Behavioral Nutrition and Physical Activity 2010; 7:38. doi:10.1186/1479-5868-7-38.
15.    Centers for Disease Control and Prevention. National Health and Nutrition Examination Survey. http://wwwn.cdc.gov/nchs/nhanes. Accessed 12 July 2017.
16.    United States Department of Agriculture. Food Patterns Equivalents Database. https://www.ars.usda.gov/northeast-area/beltsville-md/beltsville-human-nutrition-research-center/food-surveys-research-group/docs/fped-databases/. Accessed 12 December 2017.
17.    Armstrong T, Bull F. Development of the world health organization global physical activity questionnaire (GPAQ). J Public Health 2006; 14 (2): 66–70.
18.    World Health Organization. Global physical activity questionnaire (GPAQ) analysis guide. http:// www.who.int/chp/steps/resources/GPAQ_Analysis_Guide.pdf. Accessed 10 August 2017.
19.    The U.S. Department of Health and Human Services. Physical activity guidelines for Americans. http://health.gov/paguidelines/guidelines/chapter5.aspx. Accessed 19 December 2017.
20.    Moore JX, Chaudhary N, Akinyemiju T. Metabolic Syndrome Prevalence by Race/Ethnicity and Sex in the United States, National Health and Nutrition Examination Survey, 1988–2012. Preventing Chronic Disease 2017; 14: E24. doi:10.5888/pcd14.160287.
21.    National Cancer Institute. The Healthy Eating Index – HEI Scoring Algorithm Method. https://epi.grants.cancer.gov/hei/hei-scoring-method.html. Accessed 14 January 2018.
22.    Jindai K, Nielson CM, Vorderstrasse BA, Quiñones AR. Multimorbidity and Functional Limitations Among Adults 65 or Older, NHANES 2005–2012. Preventing Chronic Disease 2016;13: E151. doi:10.5888/pcd13.160174.
23.    Jordan HT, Tabaei BP, Angell SY, Chamany S, Kerker B, Nash D. Metabolic Syndrome Among Adults in New York City, 2004 New York City Health and Nutrition Examination Survey. Preventing Chronic Disease 2012; 9:E04.
24.    U.S. Census Bureau, Population Division, Fertility & Family Statistics Branch. (2004). Current Population Survey: Definitions and explanations. http:// www.census.gov/population/www/cps/cpsdef.html. Accessed 10 August 2017.
25.    CDC National Center for Health Statistics. Specifying weighting parameters. 2013. http://www.cdc.gov/nchs/tutorials/nhanes/surveydesign/weighting/intro.htm. Accessed 11 August 2017.
26.    Brien SE, Katzmarzyk PT. Physical activity and the metabolic syndrome in Canada. Appl Physiol Nutr Metab 2006; 31:40-47.
27.    Zhu S, St-Onge MP, Heshka S et al. Lifestyle behaviors associated with lower risk of having the metabolic syndrome. Metabolism 2004; 53:1503-1511.
28.    Kim J, Choi Y. Physical activity, dietary vitamin C, and metabolic syndrome in the Korean adults: the Korea National Health and Nutrition Examination Survey 2008 to 2012. Public Health 2016 Jun; 135:30-7.
29.    Skoumas J, Pitsavos C, Panagiotakos DB, et al. Physical activity, high density lipoprotein cholesterol and other lipids levels, in men and women from the ATTICA study. Lipids in Health and Disease 2003; 2:3. doi:10.1186/1476-511X-2-3.





1. UMR1027 Inserm, Toulouse, France; 2. University of Toulouse III, Toulouse, France; 3. Gérontopôle Toulouse, Toulouse University Hospital, Toulouse, France; 4. Nutrition, Exercise Physiology, and Sarcopenia Laboratory, Jean Mayer USDA Human Nutrition Research Center on Aging, Tufts University, Boston, MA, USA; 5. Alliance for Aging Research, Aging in Motion (AIM), Washington, DC, USA; 6. Fondation Policlinico A. Gemelli, Roma, Italy; 7. Research Institute, California Pacific Medical Center, San Francisco, CA, USA; 8. Department of Epidemiology and Biostatistics, University of California, San Francisco, CA, USA; 9. Department of Public Health and Caring Sciences, Clinical Nutrition and Metabolism, Uppsala University, and Department of Geriatrics, Uppsala University Hospital, Uppsala, Sweden; 10. Hospital Universitario Ramón y Cajal (IRYCIS). Madrid, Spain; 11. Boiphytis, France; 12. Bluecompanion Ltd, UK; 13. Regneron, Tarrytown, NY, USA ; 14. Division of Geriatric St. Louis, University Medical School, St. Louis, MO, USA; 15. Institute on Aging, University of Florida, Gainesville, FL. USA; 16. University of Liege, Liege, Belgium ; 17. Jefe de Servicio de Geriatría, Hospital Universitario de Getafe, Toledo, Spain; 18. Service de Médecine Interne et Gérontologie, Clinique Gérontopôle, Hôpital La Crave, Casselardit, Toulouse, France; 19. Novartis Institutes for Biomedical Research, Basel, Switzerland; 20. Diabetes and Geriatric Research Unit, University of Luton, Luton, United Kingdom; 21. Gérontopôle, Centre Hospitalier Universitaire de Toulouse, Toulouse, France; 22. Université de Toulouse III Paul Sabatier, Toulouse, France
Corresponding author: Bruno Vellas, MD. Gérontopôle, CHU Toulouse, Service de Médecine Interne et Gérontologie, Clinique, 170 Avenue de Casselardit, 31059 Toulouse, France. Phone: +33 (0) 5 6177-6425; Fax: +33 (0) 6177-6475. Email: vellas.b@chu-toulouse.fr

Task Force members: Manuel Anxo Blanco (Madrid, Spain), Cynthia Bens (Washington, USA), Roberto Bernabei (Roma, Italy), Shalender Bhasin (Boston, USA), Denis Breuillé (Vevey, Switzerland), Ryne Carney (Washington, USA), Peggy Cawthon (San Francisco, USA), Tommy Cederholm (Uppsala, Sweden), Matteo Cesari (Toulouse, France), Alfonso Cruz Jentoft (Madrid, Spain), Susanna Del Signore (Paris, France), Waly Dioh (Paris, France), Stephen Donahue (Tarrytown, USA), Roger Fielding (Boston, USA), Makoto Kashiwa (Tokyo, Japan), Kala Kaspar (Vevey, Switzerland), Tatiana Klompenhouwer (Utrecht, The Netherlands), Valérie Legrand (Nanterre, France), José Maria Lopez (Granada, Spain), Yvette Luiking (Utrecht, The Netherlands), Marie Mc Carthy (Dublin, Ireland), Bradley Morgan (South San Francisco, USA), John Morley (St Louis, USA), Serge Muller (Buc, France), David Neil (King of Prussia, U.S.A.), Marco Pahor (Gainesville, USA), Suzette Pereira (Columbus, USA), Jean-Yves Reginster (Liege, Belgium), Leocadio Rodriguez Manas (Madrid, Spain), Yves Rolland (Toulouse, France), Ronenn Roubenoff  (Basel, Switzerland), Ricardo Rueda (Columbus, USA), Alan Russell (King of Prussia, USA), Peter Schüler (Langen, Germany), Alan Sinclair (Bedfordshire, United Kingdom), Bruno Vellas (Toulouse, France), Kevin Wilson (Marlborough, USA)

J Frailty Aging 2018;7(1):2-9
Published online August 23, 2017, http://dx.doi.org/10.14283/jfa.2017.30



Establishment of an ICD-10-CM code for sarcopenia in 2016 was an important step towards reaching international consensus on the need for a nosological framework of age-related skeletal muscle decline. The International Conference on Frailty and Sarcopenia Research Task Force met in April 2017 to discuss the meaning, significance, and barriers to the implementation of the new code as well as strategies to accelerate development of new therapies. Analyses by the Sarcopenia Definitions and Outcomes Consortium are underway to develop quantitative definitions of sarcopenia. A consensus conference is planned to evaluate this analysis. The Task Force also discussed lessons learned from sarcopenia trials that could be applied to future trials, as well as lessons from the osteoporosis field, a clinical condition with many constructs similar to sarcopenia and for which ad hoc treatments have been developed and approved by regulatory agencies.

Key words: Sarcopenia, frailty, obesity, disability, intervention studies, prevention.



The age-related loss of muscle mass and strength, known as sarcopenia, is a major cause of frailty and disability in older persons worldwide. Nevertheless, progress in developing treatments for sarcopenia has been hindered by a lack of consensus on how the condition is defined and diagnosed (1). A major step forward in correcting this deficiency was achieved on October 1, 2016, when a unique ICD10 code for sarcopenia was established (2).  In April 2017, the International Conference on Frailty and Sarcopenia Research Task Force met in Barcelona, Spain to discuss the meaning and significance of the new ICD-10 code.


Background and History of ICD-10 for Sarcopenia

The International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) is the latest version of a system used by physicians, researchers, and health systems to classify diseases and other health conditions according to recognized diagnoses. Based on the ICD-10 system used by all World Health Organization (WHO) member countries, the ICD-10-CM Code Book is the US version, prepared by the ICD-10 Coordination and Maintenance (C&M) Committee (including representatives from the Centers for Medicare and Medicaid Services [CMS], the Centers for Disease Control and Prevention [CDC], and the National Center for Health Statistics [NCHS]). The ICD-10-CM diagnostic codes are mandated in the US for all health care providers as a means of removing barriers for diagnosis, standardizing recognition of disease conditions, and providing robust data for outcomes research.
The Aging in Motion (AIM) coalition (aginginmotion.org), established by the Alliance for Aging Research in 2011, submitted a proposal to the CDC in 2014 to create an ICD-10 code for sarcopenia. This code was considered crucial for recognizing this age-related condition and characterizing it among the many conditions affecting the older person. The proposal outlined the evolution of sarcopenia as a distinct diagnosis, the efforts to reach an international consensus definition (3, 4), the impact of sarcopenia on function, and the potential for development of drugs to treat the condition. The submission of the proposal was followed by a public meeting with the C&M committee where concerns were raised that sarcopenia could be conflated with muscle and neurological conditions. An extensive literature review allayed these concerns, and a revised version of the proposal addressing these issues was sent to the CDC. Finally, in April 2016, a new code for sarcopenia – M62.84 – was added, and went into effect in October 2016. The code specifies that if underlying conditions such as other muscle diseases are present, they should be coded first, followed by the code for sarcopenia. However, sarcopenia should be coded first if associated with conditions such as generalized weakness or accelerated physical disability. These refinements to the way sarcopenia should be coded are designed to ensure that data are captured accurately.


Implications of ICD-10 Codes for Sarcopenia

Establishing the ICD-10 code allows the recognition of sarcopenia as a separately reportable condition by the US Food and Drug Administration (FDA) and European Medicines Agency (EMA). Indeed, sarcopenia was selected as one of eight conditions to be addressed by patient-focused drug development meetings conducted by the FDA in 2017. Establishment of the code also has the potential to incentivize funders and sponsors to allocate increased resources to address sarcopenia.
Task Force participants noted that establishing ICD10 codes is the first step, allowing for the collection of data demonstrating a change in various metrics of muscle weakness and disability across large population cohorts, which would allow those metrics to be used to support drug development. The FDA has requested qualitative research to validate assessment tools for the measurement of outcomes that are useful to patients. Functional assessments and Patient-reported outcome (PRO) are among the types of endpoints that hold appeal for regulatory agencies, who might accept a quantitative measure of benefit plus a PRO as co-primary endpoints in a confirmatory clinical trial.
Nonetheless, there remain barriers to the use of the ICD-10 among both general practitioners and specialists. Patients may complain of loss of physical function, such as not being able to lift a grandchild, however they do not understand “sarcopenia”, which further hinders translation into clinical practice. Payers may also be slow to embrace the new code given the relatively high prevalence of sarcopenia, in contrast to low-prevalence disorders with clear endpoints that can be modified by the intervention. Task Force members cited the need to communicate with professional societies to raise awareness of the code and ensure clinical recognition and coverage of sarcopenia. On the international front, establishment of an ICD-10-CM code in the US may encourage creation of a unique code in the next version of the WHO code book, ICD-11.
In clinical practice, another challenge with moving ICD-10 forward is creating awareness that sarcopenia interventions can help prevent disability. Sarcopenia has a relevant impact on quality of life over the lifespan, but individuals may not yet be aware of the myriad of ways through which sarcopenia can lead to a loss of independence and increase risk of death. Moreover, individuals and physicians should be made aware that sarcopenia is a problem that can be addressed. With increased awareness, patients and clinicians may begin to see treatment of sarcopenia as a means to avoid disability, similarly to how they were educated to treat hypertension as a means of preventing stroke. To get to this point, however, Task Force members cited the need for more health economics data and the identification of surrogate endpoints (e.g., increased hospitalizations, institutionalizations, healthcare services consumption), which will be facilitated by the introduction of the ICD-10 code.
Improved screening tools for sarcopenia, including self-administered instruments, are also needed to maximize the potential benefits of the ICD-10 code. Given that heightened awareness of sarcopenia in the general public may lead to higher demand for physical therapists and/or dietitians, sarcopenia researchers should devise messages that align with the goals of these allied health practitioners. To increase the efficiency of clinical trials, new models are needed to engage potential participants, which can be particularly challenging in older populations. Clinics that focus on falls or other functional impairments that result from sarcopenia may be one useful approach.


Establishing Evidence-Based Cut-points to Define Sarcopenia

Declining muscle strength is a universal feature of aging that people often dismiss as inevitable. Thus, there is a need to distinguish between normal aging of the skeletal muscle and sarcopenia, a clinical condition can and should be prevented and treated. The first phase of the FNIH Sarcopenia Project established a clinical paradigm for identifying subjects with sarcopenia in which poor physical function should immediately lead to the evaluation of possible weakness. If muscle weakness is excluded, other conditions should be considered, while quantification of muscle mass is recommended in the presence of weakness. Sarcopenia is present when muscle weakness and low muscle mass coexist (4).
To implement this paradigm, assessment tools for weakness and low muscle mass are needed. The Sarcopenia Project used clinical data from over 26,000 individuals in nine studies to define normal and abnormal cut-points for different assessments. The derived cut-points were then used to estimate prevalence and predictive capacity for major clinical outcomes, such as mortality and falls. Establishing a cut-point for a disease that follows a continuum – such as hypertension — always relies on some underlying arbitrary decisions. A cut-point that results in a low prevalence could result in too few individuals identified and treated (determining a high number of false negative results), while a cutpoint that overestimates the prevalence of the condition may result in over-treatment. The choice of a cut-point thus balances sensitivity and specificity (e.g., false negatives and false positives) according to the needs of the evaluation. As a screening tool, sensitivity might be particularly important in order to be more comprehensive in the identification of subjects at risk, whereas specificity may be preferred to filter individuals to be treated with a costly intervention. In establishing cut-points, one also needs to identify what outcome is most important – e.g. mobility (slowness), mortality, falls, or hospitalization. A barrier in sarcopenia research is that no single outcome serves as a gold standard against which potential definitions would be evaluated. Consensus on what outcomes are most important for sarcopenia would help solidify its definition.

Figure 1 Prevalence of Slow Gait in the General United States Population (NHANES)

Figure 1
Prevalence of Slow Gait in the General United States Population (NHANES)

How slow gait is defined substantially affects its prevalence. If it is defined as slower than 1 m/sec, the prevalence of slow gait is substantially greater than if it is defined as less than .6 m/sec.  Adapted from Cummings, et al., JAMA 2014 (5).


Epidemiological data can help answer these questions. For example, using data from the National Health and Nutrition Examination Survey (NHANES), Cummings et al. demonstrated how applying different cut-points to define slow gait speed resulted in different prevalence estimates across age groups, with increasing prevalence of slow gait with increasing age (Figure 1) (5). Batsis et al. also showed that prevalence differs depending on the end-point used, for example, if prevalence is based on low lean mass versus weakness (6-8).
The Sarcopenia Definitions and Outcomes Consortium Project, an ongoing project funded by the US National Institute on Aging and the Foundation for the NIH, included in their analysis data from nine cohort studies, applying different statistical methods to determine the best way to compare strength, muscle mass, and physical performance in a heterogeneous population. Unfortunately, none of the studies included in the FNIH-Sarcopenia Project simultaneously included all three key measures: gait speed, weakness, and lean mass. In addition, other factors that need to be considered include race, ethnicity, and the cost of screening. Body size affects measures of sarcopenia, with obese individuals needing stronger muscles to carry their excess weight. Therefore, the analysis searched for and identified several candidate measures and cut-points that accurately categorized participants as either sarcopenic or non-sarcopenic, regardless of whether they were slim or overweight. These various measures were tested in the large dataset to determine which combination of factors provided the best discriminatory power. The ratio of appendicular lean mass to body mass index was chosen as the most reliable marker for capturing skeletal muscle loss, although other parameters were also possible and further evaluation of these data are underway.
Separate analyses for men and women revealed important differences. For example, while grip strength and slower walking speed appear to correlate with the risk of falls and death in both men and women, slow walking speed increases the risk of mortality more in men than in women, although the prevalence of slow walking speed is higher in women. Women are also much more likely to be disabled. Because of these sex differences the Sarcopenia Definitions and Outcomes Consortium created different cut-points for men and women, although they noted that sex-specific cut-points are not commonly used in other disease areas.
Task Force members raised several caveats about the analyses of the Sarcopenia Definitions and Outcomes Consortium project. Since the studies from which the cohorts were derived mostly required participants to be community dwelling and ambulatory, the analyses completed thus far included relatively few with mobility complaints, cognitive impairment, or other medical conditions that are associated with a high prevalence of sarcopenia. The influence of race and country of origin also needs to be explored, and concerns were expressed about using body mass index in the algorithm because of the high prevalence of obesity in the US (potentially biasing the application of the measure/cut-points in non-US populations). Indeed, sarcopenic obesity may be a different condition (9). The Sarcopenia Definitions and Outcomes Consortium analyses will be presented and discussed at consensus conference later in 2017 to reach agreement on definitions of sarcopenia.
It may be that additional data, including functional data, are needed before a consensus can be reached. Many clinical cohorts are available in Europe that focus on function. These cohorts could enable exploration of the predictive capacity of sarcopenia for disability or hospitalization. Efforts to acquire and combine these datasets could provide important insight into the prevalence of clinically meaningful aspects of sarcopenia. Indeed, refining cut-points for better capturing hard outcomes as well as outcomes valued by older people is essential if the field wants to move forward and effectively address unmet clinical needs.


Learning from Current Trials in Sarcopenia and Osteoporosis

The Task Force also reflected on lessons learned from ongoing trials on sarcopenia and osteoporosis (Table 1). A recent analysis of 123 sarcopenia intervention studies found that most were single-center randomized studies focused on nutrition and exercise. Few used recent consensus definitions of sarcopenia and an extreme variety of endpoints were considered. For example, muscle mass and strength were primary outcome variables in less than 30% of studies and physical performance was included in less than 20% (10). While some studies have demonstrated beneficial effects of resistance exercise training combined with protein supplementation in younger adults, a meta-analysis of 15 studies failed to show a similar effect in older healthy, frail, and sarcopenic adults (11). However, a very recent systematic overview showed that exercise and nutrition improved outcomes in well-defined populations using strict criteria for the diagnosis of sarcopenia and frailty as inclusion criteria (12).
Difficulty recruiting sarcopenic participants was cited as a major challenge in this field of research (e.g., in the trial conducted on the anti-myostatin drug REGN1033). Establishing clinical services specifically designed for sarcopenic patients might provide a solution to this problem. Diverse communication strategies (e.g., mass mailing, raising awareness in primary care) might be used to reach the community as demonstrated by the successful results of the LIFE study (13).

Table 1 Selected Sarcopenia Intervention Studies

Table 1
Selected Sarcopenia Intervention Studies

ADL: Activities of Daily Living, ALM: Appendicular Lean Mass, BMI – Body Mass Index, CaHMB: calcium 3-hydroxy-3-methyl butyrate, DXA – Dual-energy X-ray Absorptiometry, EWGSOP: European Working Group on Sarcopenia in Older People, ICT: Information and Communications Technology. IGF-1: Insulin-like Growth Factor 1, IL-6: Interleukin 6, LMM: leg muscle mass, MMSE: Mini-Mental State Examination, ONS: Oral Nutritional Supplement, QOL: Quality of Life, SPPB: Short Physical Performance Battery, 1-RM: 1 Repetition Maximum, 6MWT: 6 Minute Walk Test, 6CIT: 6-item Cognitive Impairment Test,


Trial design also presented recruitment challenges in the SPRINTT (Sarcopenia & Physical fRailty IN older people: multi-componenT Treatment strategies) trial. This study is designed to compare a multicomponent intervention (consisting of structure physical activity, personalized nutritional counseling/dietary intervention, and an informational/communication technology [ICT] intervention versus a healthy aging lifestyle education program) to prevent mobility disability in 1,500 individuals with physical frailty and sarcopenia (14). In addition to testing the effectiveness of the intervention, SPRINTT was designed to provide a clear operationalization of the theoretical concept of a condition (i.e., “physical frailty and sarcopenia”, PF&S) (15) that can meet the methodological construct required by regulatory agencies; and included the definition of diagnostic and prognostic biomarkers. Thus, when designing the trial, the SPRINTT Consortium conferred with the European Medicines Agency (EMA), which provided its scientific advice and finally endorsed the trial design, statistical approach, and the proposed definition of PF&S. It is noteworthy that the EMA accepted for the first time to consider a condition focused on loss of function (i.e., the skeletal muscle-related loss of mobility) instead of the traditional paradigms of diseases. Specifically, the EMA Committee for Medicinal Products for Human Use (CHMP) agreed on the operational definition of sarcopenia based on the FNIH proposed criteria of low appendicular body mass normalised for body mass index or low appendicular body mass (4) and a low score at the Short Physical Performance Battery (SPPB, formal correspondence on file). The EMA CHMP is then awaiting final refinement of the applied selection/inclusion criteria based on the evaluation of the study final results. Overall, SPRINTT will generate data on the body composition (measured by DXA) and physical function (SPPB and 400 meter walk) from 1,500 frail sarcopenic older persons and their 2-year change. It is paramount to wait for these data becoming available to the scientific community as they could meaningfully impact the ongoing discussions on sarcopenia diagnosis by taking into account a representative European sample of older people at risk of mobility disability. The trial began in January 2016, and sixteen clinical sites are currently recruiting participants in eleven European countries and the recruitment is expected to the completed at the end of September 2017.
A meta-analysis of studies combining exercise and protein supplementation demonstrated additive effects on muscle mass and strength in both younger and older subjects (16), although as mentioned earlier, results in older adults are weaker (11). Some studies have shown benefits of nutritional supplementation. For example, a meta-analysis of high-protein oral nutritional supplements in patients following hospital discharge showed a reduction in complications and re-admissions as well as improvements in weight and grip strength (17). As described in Table 1, nutritional supplementation (alone or in combination with exercise) is a widely-studied treatment strategy (13, 14, 18-21). The variable results obtained from available studies raise many questions about the design of the intervention (e.g., adequacy of the dosing, appropriateness of the specific nutrients), the eligibility criteria (i.e., recruitment of the best candidates to benefit from the supplementation), and the adopted measures for measuring the risk condition and the endpoints.  The aging process is responsible for declines in both bone (osteoporosis) and muscle (sarcopenia) health, contributing to frailty (22) and leading to increased risk of fracture, disability, loss of independence, decreased quality of life, and increased mortality. Yet, the development of treatments for osteoporosis has far outpaced those for sarcopenia. As mentioned earlier, one of the factors that has enabled the development of osteoporosis treatments is the availability of ICD codes and a clear operational definition (based on dual x-ray absorptiometry, or DXA) for the condition. In this way, physicians have been able to diagnose the condition, researchers to collect data about its prevalence and pathophysiological mechanisms, and pharmaceutical companies to design ad hoc interventions. In the field of sarcopenia, things appear more complicated because, whereas osteoporosis (i.e., low bone mineral density, BMD) is naturally related to the fracture endpoint, a similarly strong relationship does not exist between the skeletal muscle and a clinically relevant and organ-specific outcome.
Endpoints that matter to patients, such as falls or hip fracture for osteoporosis, are also less clear for sarcopenia. For osteoporosis trials, the EMA requires demonstration of an effect on both spinal and non-spinal fractures. Possible hard clinical outcomes for sarcopenia clinical trials include mobility disability, activities of daily living (ADL) disability, fractures, recurrent falls, injurious falls, mortality, or hospitalization. A surrogate marker would be ideal. Validating a surrogate endpoint requires demonstrating that it correlates with medically relevant endpoints in the natural course of the disease and in treated subjects. In addition, regulators want to see a demonstration of the magnitude of the relationship between the surrogate and the hard endpoint in treated subjects. Surrogate endpoints that might be acceptable in sarcopenia clinical trials include grip strength, walking speed, or chair stand since they correlate with mortality and other clinical outcomes (23, 24). Regulators are also increasingly requiring as co-primary endpoints patient-reported outcomes (PROs). Various auto-evaluation questionnaires, like the Short Form Health Survey (SF-36), have been extensively tested in similar populations and validated in numerous languages.  Recently, Beaudart and colleagues developed a sarcopenia-specific quality of life questionnaire (SarQoL) that has been shown to be valid, consistent and reliable (25, 26). The SarQoL (www.sarqol.org) can be used for both clinical and research purposes, but still needs to be validated regarding sensitivity to change. It has been translated into 11 languages with another 19 translations in development. SarQol and SF-36 are currently being measured head-to-head in the SARA observational clinical study, a project currently recruiting patients with sarcopenia or sarcopenic obesity (according the FNIH DXA criteria and very low SPPB score) both in Europe and in the US.  The osteoporosis field also benefits from the availability of the International Osteoporosis Foundation’s Fracture Risk Assessment tool, the IOFFRAX®, which has been scientifically validated and translated for global use. This simple questionnaire enables the identification of persons at elevated risk for fracture who may be appropriate subjects for clinical trials. A one-minute osteoporosis risk test is also available as a screening tool. In the sarcopenia and frailty fields, similar screening tools have been developed, including the Gérontopôle Frailty Screening Test (GFST) (27), the SARC-F (28), and FRAIL (29).    We are confident the ongoing initiatives could generate objective data to contribute identifying better methodologies for studying and characterizing age-related sarcopenia in the more concerned population, older persons at increased risk of losing physical function, of hospitalization and other sarcopenia related major outcomes.


Designing Preventive Trials for Sarcopenia

The establishment of ICD-10 codes and related efforts described above to define sarcopenia and establish evidence-based cut-points to be used diagnostically should enable more productive and efficient clinical trials of sarcopenia interventions. However, there are also efforts underway to prevent sarcopenia, both by targeting people at-risk for the disease (because of a sedentary lifestyle, inadequate energy intake, and other intrinsic factors) as well as individuals with specific conditions characterized by accelerated and/or accentuated aging (30-32). A key issue in geriatric medicine is whether to focus on treatment or prevention. In sarcopenia, public health interventions should follow a life-course approach in order to positively affect the earlier phases of the skeletal muscle decline (roughly starting after the age of 40 years). If lifestyle and behavioral interventions (e.g., nutrition, physical activity) might be foreseen on the large scale given their likely cost-effectiveness and public health interest, the development of drugs for sarcopenia might represent short-term and intense interventions reserved for individuals affected by specific sarcopenia conditions, or target a higher risk sub-population not responding to the life-style intervention and deserving long-term pharmacological treatment.
Designing a prevention trial requires targeting of a risk factor. For example, the University of Florida Institute on Aging is conducting a prevention trial called ENRGISE (ENabling Reduction of lowGrade Inflammation in Seniors) that targets age-related inflammation as a risk factor for mobility loss, frailty, and sarcopenia (33). After conducting a systematic review of anti-inflammatory interventions, they selected an approach that combines an angiotensin receptor blocker (i.e., losartan) with omega-3 fatty acids, two widely available and low-cost interventions. If proven efficacious, this combined intervention could be relatively easy to deliver to older adults at high risk of mobility disability. At the time of the Task Force meeting, the trial was nearing its recruitment goal for a pilot study that will include the assessment of novel inflammatory biomarkers and could provide preliminary data to design a definitive clinical trial.



Sarcopenia is a highly prevalent condition of older age, and a major contributor to frailty and disability. It thus presents a considerable social and economic burden. Establishing an ICD-10 code for sarcopenia is an important first step towards developing effective treatments, but there are significant gaps in knowledge and tools related to risk assessment, and regulatory guidelines are needed. Large clinical trials (SPRINTT, ENRGIZE, etc.) are currently ongoing in age-related sarcopenia and age-related inflammation that will generate meaningful data to better characterize this therapeutic area and feed regulatory appraisal. It will be important to integrate PROs in next coming initiatives, in order to link objective measurement of physical function to what is meaningful for the older person.
Moreover, ICFSR Task Force participants suggested building a risk model similar to FRAX for osteoporosis and/or investigating the value of indexing threshold values for sarcopenia measures and outcomes using a risk-based analysis for one of the strong clinical endpoints. They also mentioned the need to reach consensus on a core outcome set to bring standardization and comparability to research and improve the evidence base (34).
Finally, the Task Force discussed specific characteristics that every trial on sarcopenia should include in its design. Factors that may contribute to the failure of studies to demonstrate benefit include insufficient exposure due to short duration trials and heterogeneity among participants. In addition, there is a clear need to identify endpoints that are clinically meaningful and that are associated with improved clinical outcomes such as reduced disability and mortality. Some workshop participants advocated an increased focus on conducting sarcopenia trials in primary care centers. Certainly, this will require increased attention to issues such as 1) training and providing the necessary tools for general practitioners to conduct grip strength and other evaluations, 2) facilities improvements to handle large numbers of older people coming to the clinics, 3) relief for the increased administrative burden, and 4) strategies to address the transportation needs of trial participants. Other strategies suggested to improve intervention trials for sarcopenia, included conducting trial in well-defined populations with sarcopenia and identifying subpopulations where medical need is addressed, identifying confounding factors, combining treatment modalities in trials, establishing and implementing clear requirements for study sites, and optimizing/standardizing regulations for IRB/ethics approvals.


Conflicts of interest: Dr. Fielding reports grants, personal fees and other from Axcella Health, personal fees from Cytokinetics, grants and personal fees from Biophytis, personal fees from Amazentis, grants and personal fees from Nestle’, grants and personal fees from Astellas, grants from Lonza, personal fees from Glaxo Smithkline, outside the submitted work. Dr. Fielding is partially supported by the U.S. Department of Agriculture, under agreement No. 58-19500-014. Any opinions, findings, conclusion, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Dept of Agriculture. Dr. Roubenoff is a full-time employee of Novartis.  Drs. Cederholm have nothing to disclose. Dr. Del Signore is employee of Biophytis and founder of Bluecompanion

Acknowledgements: The authors thank Lisa Bain for assistance in preparing this manuscript.



1.     Cesari M, Fielding R, Benichou O, et al. Pharmacological Interventions in Frailty and Sarcopenia: Report by the International Conference on Frailty and Sarcopenia Research Task Force. J Frailty Aging. 2015;4(3):114-20.
2.    Anker SD, Morley JE, von Haehling S. Welcome to the ICD-10 code for sarcopenia. J Cachexia Sarcopenia Muscle. 2016;7(5):512-4.
3.    Cruz-Jentoft AJ, Baeyens JP, Bauer JM, et al. Sarcopenia: European consensus on definition and diagnosis: Report of the European Working Group on Sarcopenia in Older People. Age Ageing. 2010;39(4):412-23.
4.    Studenski SA, Peters KW, Alley DE, Cawthon PM, et al. The FNIH sarcopenia project: rationale, study description, conference recommendations, and final estimates. J Gerontol A Biol Sci Med Sci. 2014;69(5):547-58.
5.    Cummings SR, Studenski S, Ferrucci L. A diagnosis of dismobility–giving mobility clinical visibility: a Mobility Working Group recommendation. JAMA. 2014;311(20):2061-2.
6.    Masanes F, Rojano I. Luque X, et al. Cut-off points for muscle mass — not grip strength or gait speed — determine variations in sarcopenia prevalence. J Nutr Health Aging. 2016.
7.    Batsis JA, Mackenzie TA, Jones JD, Lopez-Jimenez F, Bartels SJ. Sarcopenia, sarcopenic obesity and inflammation: Results from the 1999-2004 National Health and Nutrition Examination Survey. Clin Nutr. 2016.
8.    Batsis JA, Mackenzie TA, Lopez-Jimenez F, Bartels SJ. Sarcopenia, sarcopenic obesity, and functional impairments in older adults: National Health and Nutrition Examination Surveys 19992004. Nutr Res. 2015;35(12):1031-9.
9.    Wannamethee SG, Atkins JL. Muscle loss and obesity: the health implications of sarcopenia and sarcopenic obesity. Proc Nutr Soc. 2015;74(4):405-12.
10.    Pena Ordonez GG, Bustamante Montes LP, Ramirez Duran N, Sanchez Castellano C, Cruz-Jentoft AJ. Populations and outcome measures used in ongoing research in sarcopenia. Aging Clin Exp Res. 2016.
11.    Thomas DK, Quinn MA, Saunders DH, Greig CA. Protein Supplementation Does Not Significantly Augment the Effects of Resistance Exercise Training in Older Adults: A Systematic Review. J Am Med Dir Assoc. 2016;17(10):959 e1-9.
12.    Lozano-Montoya I, Correa-Perez A, Abraha I, et al. Nonpharmacological interventions to treat physical frailty and sarcopenia in older patients: a systematic overview – the SENATOR Project ONTOP Series. Clin Interv Aging. 2017;12:721-40.
13.    Pahor M, Guralnik JM, Ambrosius WT, et al. Effect of structured physical activity on prevention of major mobility disability in older adults: the LIFE study randomized clinical trial. JAMA. 2014;311(23):2387-96.
14.    Landi F, Cesari M, Calvani R, et al. The «Sarcopenia and Physical fRailty IN older people: multi-componenT Treatment strategies» (SPRINTT) randomized controlled trial: design and methods. Aging Clin Exp Res. 2017;29(1):89-100.
15.    Cesari M, Landi F, Calvani R, et al. Rationale for a preliminary operational definition of physical frailty and sarcopenia in the SPRINTT trial. Aging Clin Exp Res. 2017;29(1):81-8.
16.    Cermak NM, Res PT, de Groot LC, Saris WH, van Loon LJ. Protein supplementation augments the adaptive response of skeletal muscle to resistance-type exercise training: a meta-analysis. Am J Clin Nutr. 2012;96(6):1454-64.
17.    Cawood AL, Elia M, Stratton RJ. Systematic review and meta-analysis of the effects of high protein oral nutritional supplements. Ageing Res Rev. 2012;11(2):278-96.
18.    Bauer JM, Verlaan S, Bautmans I, et al. Effects of a vitamin D and leucine-enriched whey protein nutritional supplement on measures of sarcopenia in older adults, the PROVIDE study: a randomized, double-blind, placebo-controlled trial. J Am Med Dir Assoc. 2015;16(9):740-7.
19.    Cramer JT, Cruz-Jentoft AJ, Landi F, et al. Impacts of HighProtein Oral Nutritional Supplements Among Malnourished Men and Women with Sarcopenia: A Multicenter, Randomized, Double-Blinded, Controlled Trial. J Am Med Dir Assoc. 2016;17(11):1044-55.
20.    Tieland M, Dirks ML, van der Zwaluw N, et al. Protein supplementation increases muscle mass gain during prolonged resistance-type exercise training in frail elderly people: a randomized, double-blind, placebo-controlled trial. J Am Med Dir Assoc. 2012;13(8):713-9.
21.    Tieland M, van de Rest O, Dirks ML, et al. Protein supplementation improves physical performance in frail elderly people: a randomized, doubleblind, placebo-controlled trial. J Am Med Dir Assoc. 2012;13(8):720-6.
22.    Fried LP, Tangen CM, Walston J, et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56(3):M146-56.
23.    Cooper R, Kuh D, Cooper C, et al. Objective measures of physical capability and subsequent health: a systematic review. Age Ageing. 2011;40(1):14-23.
24.    Simmonds SJ, Syddall HE, Westbury LD, Dodds RM, Cooper C, Aihie Sayer A. Grip strength among community-dwelling older people predicts hospital admission during the following decade. Age Ageing. 2015;44(6):954-9.
25.    Beaudart C, Biver E, Reginster JY, et al. Validation of the SarQoL(R), a specific health-related quality of life questionnaire for Sarcopenia. J Cachexia Sarcopenia Muscle. 2017;8(2):238-44.
26.    Beaudart C, Biver E, Reginster JY, et al. Development of a selfadministrated quality of life questionnaire for sarcopenia in elderly subjects: the SarQoL. Age Ageing. 2015;44(6):960-6.
27.    Vellas B, Balardy L, Gillette-Guyonnet S, et al. Looking for frailty in community-dwelling older persons: the Gerontopole Frailty Screening Tool (GFST). J Nutr Health Aging. 2013;17(7):629-31.
28.    Malmstrom TK, Miller DK, Simonsick EM, Ferrucci L, Morley JE. SARC-F: a symptom score to predict persons with sarcopenia at risk for poor functional outcomes. J Cachexia Sarcopenia Muscle. 2016;7(1):28-36.
29.    Morley JE, Malmstrom TK, Miller DK. A simple frailty questionnaire (FRAIL) predicts outcomes in middle aged African Americans. J Nutr Health Aging. 2012;16(7):601-8.
30.    Janssen I, Heymsfield SB, Wang ZM, Ross R. Skeletal muscle mass and distribution in 468 men and women aged 18-88 yr. J Appl Physiol (1985). 2000;89(1):81-8.
31.    Saini A, Faulkner S, Al-Shanti N, Stewart C. Powerful signals for weak muscles. Ageing Res Rev. 2009;8(4):251-67.
32.    Vellas B, Fielding R, Bhasin S, et al. Sarcopenia Trials in Specific Diseases: Report by the International Conference on Frailty and Sarcopenia Research Task Force. J Frailty Aging. 2016;5(4):194-200.
33.    Fougere B, Boulanger E, Nourhashemi F, Guyonnet S, Cesari M. Chronic Inflammation: Accelerator of Biological Aging. J Gerontol A Biol Sci Med Sci. 2016.
34.    Reginster JY, Cooper C, Rizzoli R, et al. Recommendations for the conduct of clinical trials for drugs to treat or prevent sarcopenia. Aging Clin Exp Res. 2016;28(1):47-58.





1. Gérontopôle, Centre Hospitalier Universitaire de Toulouse, Toulouse, France; 2. INSERM UMR1027, Université de Toulouse III Paul Sabatier, Toulouse, France; 3. Jean Mayer USDA, Human Nutrition Research Center, Boston, MA, USA; 4. Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA; 5. European Medicines Agency (EMA), London, United Kingdom; 6. Division of Endocrinology and Metabolism, University of Pittsburgh, Pittsburgh, PA, USA; 7. Department of Epidemiology and Public Health, University of Maryland School of Medicine, Baltimore, MD, USA; 8. The Sticht Center on Aging, Wake Forest School of Medicine, Winston-Salem, NC, USA; 9. VP Project Management, CNS pain & Aging disorders, Icon Plc, Nanterre, France; 10. Dir, Medical Safety Services Firecrest, Icon Plc, Limerick, Ireland; 11. Dept of Epidemiology and public health, School of medicine, University of Maryland, Baltimore, USA; 12. Division of Geriatric St Louis, University Medical School, St Louis, MO, USA; 13. Servicio de Geriatria, Hospital Universitario de Getafe, Getafe, Spain; 14. Global Translational Medicine, Novartis Institutes for Biomedical Research, Basel, Switzerland; 15. National Institute on Aging, Baltimore, MD, USA; 16. Michael E. DeBakey VA medical center, Baylor college of medicine, Houston, TX, USA.
Corresponding author: Bruno Vellas, MD. Gérontopôle, CHU Toulouse; Service de Médecine Interne et Gérontologie Clinique. 170 Avenue de Casselardit, 31059 Toulouse, France. Phone: +33 (0)5 6177-6425; Fax: +33 (0)5 6177-6475. E-mail: vellas.b@chu-toulouse.fr

Task Force members: Sandrine Andrieu (Toulouse, France), Patricia Belissa-Mathiot (Suresnes, France), Aouatef Bellamine (Allendale, USA), Roberto Bernabei (Roma, Italy), Shalender Bhasin (Boston, USA), Jesse Cedarbaum (Cambridge, USA), Francesca Cerreta (London, United Kingdom), Matteo Cesari (Toulouse, France), Rosaly Correa-de-Araujo (Bethesda, USA), Susanna Del Signore (London, United Kingdom), Waly Dioh (Romainville, France), Stephen Donahue (Tarrytown, USA), Roger Fielding (Boston, USA), Caroline Forkin (Limerick, Ireland), Bret Goodpaster (Pittsburgh, USA), Jack Guralnick (Baltimore, USA), Joyce Harp (Tarrytown, USA), Michaela Hoehne (Lausanne, Switzerland), Makoto Kashiwa (Tokyo, Japan), Tatiana Klompenhouwer (Utrecht, The Netherlands), Stephen Kritchevsky (Winston-Salem, USA), Valérie Legrand (Nanterre, France), Jay Magaziner (Baltimore, USA), Claudio Mejia (Madison, USA), Badley Morgan (San Francisco, USA), John Morley (St Louis, USA), Vikkie Mustad (Columbus, USA), David Neil (King of Prussia, USA), Dimitris Papanicolaou (Tenafly, USA), Suzette Pereira (Columbus, USA), Robert Pordy (Tarrytown, USA), Chris Rinsch (Lausanne, Switzerland), Leocadio Rodriguez-Manas (Madrid, Spain), Yves Rolland (Toulouse, France), Ronenn Roubenoff (Basel, Switzerland), Alan Russell (King of Prussia, USA), Pierre-Philippe Sagnier (Lausanne, Switzerland), Peter Schüler (Langen, Germany), Anurag Singh (Lauranne, Switzerland), Stephanie Studenski (Baltimore, USA), Min Tian (Columbus, USA), Jacques Touchon (Montpellier, France), Thomas Travison (Boston, USA), Bruno Vellas (Toulouse, France), Sjors Verlaan (Utrecht, The Netherlands), Dennis Villareal (Houston, USA), Thomas Webb (Madison, USA), Sander Wijers (Utrecht, The Netherlands)

J Frailty Aging 2016;in press

Published online October 5, 2016, http://dx.doi.org/10.14283/jfa.2016.110



Abstract: Muscle atrophy occurs as a consequence of a number of conditions, including cancer, chronic obstructive pulmonary disease (COPD), diabetes mellitus, heart failure, and other chronic diseases, where it is generally a predictor of poor survival. It also occurs as a consequence of disuse and an age-related loss of muscle mass and strength (sarcopenia). The aims of the 2016, International Conference on Frailty and Sarcopenia Research (ICFSR) Task Force were to examine how these specific chronic conditions have been employed in treatment trials thus far and how future trials using these patient groups might be designed for efficient identification of effective sarcopenia interventions. Functional limitations assessed as gait speed, distance walked over a set time period, or other attributes of physical performance have been suggested as outcome measures in sarcopenia trials. Indeed, such measures have already been used successfully in a number of trials aimed at preventing disability in older adults.


Key words: Sarcopenia, chronic obstructive pulmonary disease, type 2 diabetes, hip fracture, obesity, frailty.



In recent years, there has been increased awareness about the role of muscle disorders in age-related disability. Skeletal muscles are essential for maintaining whole-body health and longevity, yet muscle wasting occurs as a normal part of aging, although some people lose muscle at a faster rate and cross a threshold defined as sarcopenia (1). Sarcopenia, the loss of muscle mass and strength, frequently observed during aging, occurs in up to 29% of community-dwelling older adults and 33% of long-term care populations (2), although it may go unrecognized.
Sarcopenia also occurs as a consequence of a number of chronic conditions, and may be a predictor of increased mortality in these conditions, including cancer (3), chronic obstructive pulmonary disease (COPD) (4), diabetes mellitus (5), heart failure (6), and other chronic diseases. It also occurs as a consequence of disuse due to prolonged bed rest or immobility, such as that due to hip fracture. The overlap of these conditions with age-related muscle loss may provide clues about the multifactorial mechanisms underlying sarcopenia, e.g., increased cytokine activity, decreased anabolic hormones (1) . It may also point to candidate populations for clinical trials where early intervention would be possible and an efficacy signal could be readily discerned.
In 2015, the International Conference on Frailty and Sarcopenia Research (ICFSR) Task Force proposed targeting a number of specific conditions for sarcopenia trials, including individuals over age 75 with frailty, recent fallers, and inactive people; as well as patients with hip fracture, COPD, diabetes, heart failure, and stroke (7).  Meeting again in 2016 in Philadelphia, Pennsylvania, USA, the Task Force examined how these conditions have been employed in treatment trials thus far and how future trials might be designed for efficient identification of effective sarcopenia interventions.


Specific diseases: COPD, Diabetes, Hip Fracture, and Obesity

The development of bimagrumab illustrates both how treatments for sarcopenia may be applicable to a range of conditions, and how the lessons learned from treating one condition may help elucidate the mechanisms underlying muscle growth and atrophy.  Bimagrumab is a monoclonal antibody that stimulates muscle growth by blocking the binding of myostatin. Developed by Novartis, bimagrumab received breakthrough therapy approval from the U.S. Food and Drug Administration (FDA) in 2013 for the treatment of the rare muscle disease inclusion body myositis (IBM) (8), and has also been shown to promote recovery from disuse atrophy and increase muscle mass in healthy young men, sedentary middle-aged adults, and in elderly people. Approximately one-third of patients with COPD have sarcopenia with a marked decline in Type I muscle fibers (4).
In elderly patients, bimagrumab resulted in a functional benefit: increased distance walked in the 6-minute walking distance (6MWD) test. While this benefit was seen in elderly patients with sarcopenia and baseline slow gait speed, no improvement in distance walked was seen in patients with normal baseline gait speed. In contrast, in a study of patients with COPD-induced cachexia, bimagrumab induced a comparable increase in thigh muscle volume, but no effect on 6MWD among both slow and fast walkers. Additional studies are underway to explore the mechanisms underlying these differences. For example, it may have something to do with the preferential loss of type I vs. type II muscle fibers in COPD.  A combination of exercise and nutritional therapy has been demonstrated to improve functional outcomes (9).


Older people not only have a high prevalence of sarcopenia but of diabetes as well, yet the intersection or synergy between these two conditions has not been fully elucidated. Disuse atrophy may play a role, since individuals with type 2 diabetes (T2D) are more likely to be hospitalized and have longer durations of hospitalization than those without diabetes.  Adults with T2D also have a higher prevalence of mobility-related disability (10) and a lifestyle-intervention study aimed at improving weight loss and fitness showed an improvement in mobility among overweight adults with T2D (11). Other factors that may explain a connection between diabetes and sarcopenia include impaired glucose tolerance and insulin resistance (12). Obesity is also strongly related to T2D, particularly in middle aged and younger adults, as well as adolescents (13), obesity is less prevalent in the elderly (14). A study by Goodpaster and colleagues suggested that regional fat distribution, rather than obesity per se, contributes to the increased risk of T2D (15)
The Health, Aging, and Body Composition (Health ABC) study demonstrated that older people with T2D have an accelerated loss of leg muscle mass and strength (16). Leenders et al (17) showed that persons with type 2 diabetes had a greater decline in muscle mass, muscle strength and functional capacity with aging.  People with diabetes are also have increased intracellular lipids (18) and decreased mitochondrial function in muscles (19), suggesting a possible link between energetics and atrophy. Both increase lipid accumulation and impaired mitochondrial function have also been linked to insulin resistance in skeletal muscle. Further, persons with Type II diabetes mellitus have lower testosterone, greater oxidative damage to muscle and decreased blood flow (5).  These studies thus suggest a number of potential treatment targets for sarcopenia. They also indicate that individuals with T2D may represent a useful target population for clinical trials of sarcopenia treatments. However, investigators should keep in mind that the pathophysiologic mechanisms in diabetes may differ between younger versus older individuals; thus treatment targets may also differ.

Hip fracture

In the United States, over 290,000 older adults are hospitalized each year for hip fracture, with three-quarters of these fractures among women (20). Worldwide, the incidence of hip fracture varies substantially (21), totaling about 3.9 million. This number is expected to increase markedly over the next 30 years as the population ages (22). Moreover, despite advances in surgical procedures and post-operative care, hip fractures are substantial causes of death, disability, functional decline, and pain and suffering. The morbidity associated with hip fracture results in large part from changes in body composition: fat mass, lean mass, and bone mineral density (BMD), with marked changes in lean mass and BMD being observed during the first two months after hip fracture (23).
The prevalence of sarcopenia in patients with hip fracture varies depending on sex and the definition of sarcopenia used. Men appear to be at higher risk of sarcopenia after hip fracture, and the prevalence of sarcopenia stabilizes over time; while in women the prevalence of sarcopenia decreases over 12 months. The effects on mobility and other measures of physical function also varies depending on the definition of sarcopenia used. Other functional consequences of hip fracture include effects on cognition, mood, and socialization. Moreover, functional recovery varies across different activities (24). In addition to functional losses, hip fracture also leads to changes in the inflammatory response (25). Frailty and osteoporosis further complicate recovery from hip fracture (26).
Over the past 30 years, the Baltimore Hip Studies have enrolled and followed more than 4,000 hip fracture patients admitted to Baltimore-area hospitals through observational and interventional studies. These studies have captured data on many outcomes, including mortality and functional recovery; changes in BMD, muscle mass and composition, and bone and muscle strength; and caregiver burden, health care use, and cost.
Obesity. Sarcopenia in combination with obesity increases the risk of disability in the elderly (27), yet as with other prevalence statistics, depends in large part on the definition used. Using the Foundation for the National Institutes of Health (FNIH) definitions for both sarcopenia and obesity, an analysis of data from the 1999-2004 National Health and Nutrition Examination Surveys (NHANES) of non-institutionalized older adults found a prevalence of sarcopenic obesity of 22.7% when sarcopenia was defined as appendicular lean mass (ALM) adjusted for body mass index (BMI) and 24.4% when sarcopenia was defined by ALM alone (28). To be included in the analysis, subjects had to function at a high level at baseline. In both cases, obesity was defined as body fat greater than 25% percent in men and 35% in women. In another study of healthy elderly women, high body fat and high BMI were associated with poorer physical function assessed using a modified version of the Activities of Daily Living Scale, and leg strength measurement (29). Body composition measures, particularly thigh intermuscular fat and total thigh muscle volume, are also associated with gait-speed decline in both men and women (30).
Sarcopenic obesity also may contribute to exercise intolerance in older adults with the most common form of heart failure, i.e., heart failure with preserved ejection fraction (HFPEF) (31).  Indeed, these patients show abnormalities in skeletal muscle composition, specifically increased fat infiltration into the skeletal muscle, which appears to result in reduced peak oxygen uptake (Peak VO2) and thus, exercise intolerance (32). Investigators at Wake Forest School of Medicine thus targeted these patients in a clinical trial that explored the effects of caloric restriction and aerobic exercise on exercise capacity and quality of life in obese older patients with HFPEF (NCT00959660). Using a 2X2 factorial design, they randomized 100 patients to one of four 20-week interventions: supervised exercise alone, caloric restricted diet alone, exercise plus diet, or an attention control group that received telephone calls every two weeks. The primary outcome measure was peak VO2. Both exercise alone and diet alone significantly improved peak VO2, and the combination of exercise and diet was additive. In addition, the change in peak VO2 was accompanied by improvements in total body mass, fat mass, percent lean mass, and skeletal muscle to intermuscular fat ratio (33).

Table 1 Evaluating sarcopenia treatments across specialties – Target populations

Table 1
Evaluating sarcopenia treatments across specialties – Target populations


Lifestyle interventions across specialties

While bimagrumab and other pharmacologic therapies being developed represent biological intervention aimed at addressing muscle dysfunction, many treatment trials for sarcopenia across all specialties have assessed the impact of comprehensive lifestyle interventions, including diet and exercise, as well as other components such as counseling, stress management, and smoking cessation. The rationale for this approach is two-fold: Unhealthy lifestyles are thought to contribute to the development of sarcopenia and lifestyle factors may be more controllable than systemic changes such as inflammation or oxidative stress. In 2005, the American Society for Nutrition and the Obesity Society recommended diet, behavioral therapy, and regular exercise for older obese persons (34). Subsequently, a randomized controlled trial showed that a combination of diet and exercise resulted in greater improvement in physical function than either intervention alone (35). Similarly, clinical trials of comprehensive lifestyle interventions have demonstrated benefits for persons with obesity and T2D (36); and physical activity has been shown to benefit people with sarcopenia (37). Singh et al. (38) also found that exercise twice a week markedly reduces the negative effects of hip fracture.  In all of these trials, sarcopenia is just one of several potential causes of poor function being addressed.


Roadblocks to clinical trials for sarcopenia

Task Force members agreed that the lack of a common nomenclature across the field has stymied the development of effective therapies for sarcopenia. One advance in this regard came in April 2016, when the Centers for Disease Control and Prevention established an ICD-10-CM code for sarcopenia. This new code, M62.84, will be available beginning in October, 2016, according to the Aging in Motion Coalition, which led efforts to secure an ICD-10 code for sarcopenia. However, while the code should promote increased recognition of the validity of sarcopenia as a condition requiring diagnosis and treatment, more work is needed to reach consensus on a universal definition and the clinically relevant aspects of the disease.
An international group of sarcopenia experts recently proposed a framework for classifying muscle wasting diseases  (39). This framework incorporates the concepts of muscle wasting, sarcopenia, frailty, and cachexia, as well as the various etiologies such as those discussed above. Yet this framework, as well as the criteria proposed by the European Working Group on Sarcopenia in Older People (EWGSOP) (40) and a consensus conference convened by the Society of Sarcopenia, Cachexia, and Wasting Disorders (41), fail to address the role of adiposity in muscle wasting diseases.  Original studies of sarcopenic obesity suggested that muscle mass was the determining factor in function (42, 43). However, studies done in establishing cut-points using the FNIH criteria show that adiposity alters the relationships between muscle mass and weakness, especially in women (44, 45). In COPD, sarcopenic obesity is related to worse physical performance and a higher inflammatory burden (46). Adiposity is related to insulin resistance and markers of inflammation (28), aerobic capacity and physical function in patients with HFPEF (47), and prevalence of sarcopenia in patients with COPD (48). Muscle wasting and adiposity also affect outcomes in colon cancer (49).
These data suggest that two sets of sarcopenia definitions are needed: a wasting form and an adiposity form. Such a framework would have to acknowledge that some conditions involve both wasting and adiposity, and could also incorporate poor muscle quality to distinguish older people with sarcopenic obesity from those with obesity but no sarcopenia.
Alternatively, sarcopenia could be viewed as a geriatric syndrome rather than a disease. This would enable the recognition of multiple risk factors and perhaps disentangle the links among sarcopenia, frailty, disability, and mortality (40). It would also allow the use of less selective inclusion and exclusion criteria, which could make recruitment easier and result in more representative, “real-life” populations. Currently, the inclusion and exclusion criteria developed for many clinical trials make it nearly impossible to recruit older people (50), in part because these populations have multiple conditions and diseases (51). As a result, there are currently few clinical trials that enroll older people (52).
The multi-modal intervention in diabetes in frailty (MID-Frail) study took a different approach, designing inclusion and exclusion criteria that are as simple and few as possible (53) in order to optimize recruitment, follow-up, and compliance, while ensuring the exclusion of those with unacceptable risks, co-morbidities that would interfere with the intervention or measurement of the outcome, as well as those unlikely to receive benefit from the treatment or unlikely to adhere to the intervention. It is suggested that the SARC-F may be a rapid screening tool for clinical trials (54, 55).
The tradeoffs that accompany a widening of inclusion criteria include the potential for a high number of adverse events and increased variability, which can increase sample size and decrease power. However, it may be that without these changes in how trials are done, it will be impossible to test drugs in older populations for whom these treatments are intended.


Primary endpoints for sarcopenia trials across specialties

In addition to reaching consensus on a definition of sarcopenia, developing efficacious interventions requires broad agreement on the most appropriate endpoints to be used in clinical trials of new treatments (56). Functional limitations assessed as gait speed, distance walked over a set time period, or other attributes of physical performance have been suggested as outcome measures in sarcopenia trials. Indeed, such measures have already been used successfully in a number of trials aimed at preventing disability in older adults, as well as in observational studies. For example, the Health ABC study showed that an increase of one-minute in the time it takes to walk 400 meters is significantly associated with mortality, cardiovascular disease, mobility limitations, and disability (57). Specialty areas have also begun to look at functional limitations as screening tools and outcome measures for clinical trials. For example, in a study of adults receiving chemotherapy, the Short Physical Performance Battery (SPPB) was shown to be the strongest predictor of survival (58).
The FDA approves clinical outcome assessments either as part of a drug application review or under the Drug Development Tool Qualification Process (59). The Aging in Motion Coalition proposed, in a letter of intent to the FDA, qualification of SPPB and gait speed as outcome measures acceptable in clinical trials related to sarcopenia. However, the FDA responded by recommending targeting of particular instruments to particular diseases or conditions, with a clear definition of the disease or condition for which qualification is sought. Thus, it appears that qualification will have to be sought one condition at a time, in contrast to the usual scenario in older adults where several conditions are usually present in combination, participating jointly in producing loss of function. A previous Task Force meeting outlined the critical next steps that will be needed to achieve acceptance of outcome measures in hip fracture trials, and these steps remain true for other areas as well (60): 1) development of an evidence base of key measures and their behavior in diverse target populations over time; 2) correlations of physical performance measures to self-report information; 3) identification of minimum clinically relevant thresholds and biomarkers of skeletal muscle mass modifications; and 4) development of better patient-reported outcome scales.  It will also be necessary to describe clearly what each endpoint is thought to mean from both clinical and pathophysiologic perspectives; for example, the clinical meaningfulness and pathophysiologic meaning of an improvement in gait speed or another marker of performance.


Biomarkers and imaging

Biomarkers have greatly advanced the drug development process in a number of fields, but in regards to drug development for sarcopenia, validation of a biomarker has proven elusive.  Theoretically, a biomarker could be a symptom or a clinical, laboratory, or imaging measure of muscle mass, muscle performance, or physical function that has been shown to be predictive of health outcomes. Indeed, muscle mass and strength  meet many of the criteria for an ideal biomarker, as defined by Baker and Sprott in 1988 (61), to identify people at risk of disability or a bad health outcome. That is, muscle mass and strength can be measured accurately and precisely in a clinical setting and are related to the biological pathway of the disease. However, much work remains to establish and validate muscle mass and strength cutpoints. Moreover, their relation to change in functional performance and clinical endpoints has not been demonstrated clearly and it is unclear how they may be affected by comorbidities and other physiologic impairments.
The selection of appropriate biomarkers for sarcopenia could be guided by the putative mechanistic pathway by which promyogenic agents might improve outcomes (Figure 1). Imaging modalities such as dual energy X-ray absorptiometry (DXA), computed tomography (CT), magnetic resonance imaging (MRI), magnetic resonance spectroscopy (MRS) may be useful to assess muscle mass and composition, body composition, skeletal muscle fiber tracking, sarcomere size, and muscle biochemistry and energetics.  Indeed, DXA and CT/MRI have proven useful in phase 1 and 2 studies of promyogenic molecules to demonstrate increases in muscle mass and guide early go-no-go decisions for further development of these as therapeutic agents. However, these methods require additional data to demonstrate their relation to function. Tracer methods for tracking changes in muscle mass and protein turnover are promising techniques, but currently in early stages of development and used primarily for research.

Figure 1 Putative Pathways by Which Promyogenic Drugs Might Improve Health Outcomes

Figure 1
Putative Pathways by Which Promyogenic Drugs Might Improve Health Outcomes

HRQOL: Health-related Quality of Life


For regulators, the context of use and the clinical meaningfulness of biomarkers are of paramount importance. Regarding context of use, for example, biomarkers might be used for prognosis; demonstration of disease activity, progression, or severity; or as surrogate markers or predictors of a treatment response.


Regulatory considerations for sarcopenia as a potential indication

A major roadblock to gaining regulatory approval of a treatment for sarcopenia has been the fact that sarcopenia is not recognized as a condition (62). The acceptance of an ICD-10 code, discussed earlier, may help address this roadblock. However, there remain many other issues that will need to be clarified: How to define the populations to be treated, what characteristics (such as severity) justify pharmacologic intervention and how those will be measured, how a clinically meaningful effect will be measured, and how confounding factors will be dealt with, including those associated with comorbidities discussed in this paper, such as hip fracture, COPD, diabetes, and obesity. In terms of a clinically meaningful effect, more data are needed to demonstrate whether, for example, improved strength leads to functional improvement and what is the minimal level of improvement that provides benefit.
Other questions that will need to be answered include whether a minimum data set should be required for all sarcopenia trials; whether interventions should be tested on a background exercise and dietary regimen; to what extent comorbidities should be used as inclusion and/or exclusion criteria; and whether co-primary outcomes of a performance-based measure plus a patient-reported outcome should be required.



A large body of data from prospective observational studies and randomized controlled trials is beginning to shed light on the optimal ways to conduct trials of treatments for sarcopenia. While Task Force members agreed that it is important to choose the correct definition of sarcopenia, there may be different forms of sarcopenia, for example an adiposity-predominant form and a wasting-predominant form.
Sarcopenia is also closely related to frailty, and recently there have been efforts to merge the two into a new clinical entity, the physical frailty and sarcopenia syndrome (PF&S) (63). These individuals are at especially high risk for mobility disability. The Innovative Medicines Initiatives has funded a research project to test multi-component treatment strategies (the SPRINT-T project) in this population (64). It should be recognized that not all frail persons are sarcopenic and their causes of frailty, e.g., sleep apnea, depression, will not necessarily respond to treatments for sarcopenia (46, 65).
The preponderance of co-morbidities in sarcopenia both complicates clarification of the condition and provides potential opportunities for testing interventions. However, designing clinical trials in populations with particular co-morbidities such as COPD, T2D, obesity, and hip fracture also introduces additional complexity in trial design and interpretation of results.


Disclaimer: The views expressed in this article are the personal views of the authors and may not be understood or quoted as being made on behalf of or reflecting the position of the EMA or one of its committees or working parties.
Conflicts of interest: Dr. Bhasin reports personal fees from Novartis, grants and personal fees from Regeneron, grants and personal fees from AbbVie, grants from Transition Therapeutics, grants from Lilly, outside the submitted work; he also has equity interest in FPT, LLC, which has a patent on selective anabolic therapy. Dr. Cesari is Workpackage leader of an Innovative Medicines Initiatives-funded research project (SPRINTT), which has Sanofi, Servier, Novartis, Lilly, and GSK among the partners. Dr. Fielding reports grants and personal fees from Nestle Inc., grants and personal fees from Pronutria Inc., grants and personal fees from Astellas Inc., personal fees from Cytokinetics Inc., personal fees from Biophytis Inc., outside the submitted work. Dr. Magaziner reports personal fees from Ammonett, personal fees from Novartis, personal fees from Sanofi, personal fees from Viking, personal fees from Scholar Rock, personal fees from Pluristem, outside the submitted work. Dr. Roubenoff reports personal fees from Novartis, during the conduct of the study; other from Novartis, outside the submitted work;  In addition, Dr. Roubenoff has a patent Bimagrumab for treatment of muscle wasting issued. Drs. Cerreta, Forkin, Goodpaster, Guralnik, Kritchevsky, Legrand, Morley, Studenski, Vellas, and Villareal have nothing to disclose.



1.    Morley JE, Baumgartner RN, Roubenoff R, Mayer J, Nair KS. Sarcopenia. J Lab Clin Med. 2001;137(4):231-43.
2.    Cruz-Jentoft AJ, Landi F, Schneider SM, et al. Prevalence of and interventions for sarcopenia in ageing adults: a systematic review. Report of the International Sarcopenia Initiative (EWGSOP and IWGS). Age Ageing. 2014;43(6):748-59.
3.    Tan BH, Birdsell LA, Martin L, Baracos VE, Fearon KC. Sarcopenia in an overweight or obese patient is an adverse prognostic factor in pancreatic cancer. Clin Cancer Res. 2009;15(22):6973-9.
4.    van de Bool C, Gosker HR, van den Borst B, Op den Kamp CM, Slot IG, Schols AM. Muscle Quality is More Impaired in Sarcopenic Patients With Chronic Obstructive Pulmonary Disease. J Am Med Dir Assoc. 2016;17(5):415-20.
5.    Morley JE, Malmstrom TK, Rodriguez-Manas L, Sinclair AJ. Frailty, sarcopenia and diabetes. J Am Med Dir Assoc. 2014;15(12):853-9.
6.    von Haehling S. The wasting continuum in heart failure: from sarcopenia to cachexia. Proc Nutr Soc. 2015;74(4):367-77.
7.    Cesari M, Fielding R, Benichou O, et al. Pharmacological Interventions in Frailty and Sarcopenia: Report by the International Conference on Frailty and Sarcopenia Research Task Force. J Frailty Aging. 2015;4(3):114-20.
8.    Amato AA, Sivakumar K, Goyal N, et al. Treatment of sporadic inclusion body myositis with bimagrumab. Neurology. 2014;83(24):2239-46.
9.    van Wetering CR, Hoogendoorn M, Broekhuizen R, et al. Efficacy and costs of nutritional rehabilitation in muscle-wasted patients with chronic obstructive pulmonary disease in a community-based setting: a prespecified subgroup analysis of the INTERCOM trial. J Am Med Dir Assoc. 2010;11(3):179-87.
10.    Gregg EW, Gerzoff RB, Caspersen CJ, Williamson DF, Narayan KM. Relationship of walking to mortality among US adults with diabetes. Arch Intern Med. 2003;163(12):1440-7.
11.    Rejeski WJ, Ip EH, Bertoni AG, et al. Lifestyle change and mobility in obese adults with type 2 diabetes. N Engl J Med. 2012;366(13):1209-17.
12.    Cleasby ME, Jamieson P, Atherton PJ. Insulin resistance and sarcopenia: mechanistic links between common co-morbidities. J Endocrinol. 2016.
13.    Kopelman PG. Obesity as a medical problem. Nature. 2000;404(6778):635-43.
14.    Mokdad AH, Serdula MK, Dietz WH, Bowman BA, Marks JS, Koplan JP. The spread of the obesity epidemic in the United States, 1991-1998. JAMA. 1999;282(16):1519-22.
15.    Goodpaster BH, Krishnaswami S, Resnick H, et al. Association between regional adipose tissue distribution and both type 2 diabetes and impaired glucose tolerance in elderly men and women. Diabetes Care. 2003;26(2):372-9.
16.    Park SW, Goodpaster BH, Strotmeyer ES, et al. Accelerated loss of skeletal muscle strength in older adults with type 2 diabetes: the health, aging, and body composition study. Diabetes Care. 2007;30(6):1507-12.
17.    Leenders M, Verdijk LB, van der Hoeven L, et al. Patients with type 2 diabetes show a greater decline in muscle mass, muscle strength, and functional capacity with aging. J Am Med Dir Assoc. 2013;14(8):585-92.
18.    Goodpaster BH, Theriault R, Watkins SC, Kelley DE. Intramuscular lipid content is increased in obesity and decreased by weight loss. Metabolism. 2000;49(4):467-72.
19.    Kelley DE, He J, Menshikova EV, Ritov VB. Dysfunction of mitochondria in human skeletal muscle in type 2 diabetes. Diabetes. 2002;51(10):2944-50.
20.    Centers for Disease Control. Health Data Interactive, Health Care Use and Expenditures. National Hospital Discharge Survey (NHDS). In: Statistics NCfH, editor.
21.    Kanis JA, Oden A, McCloskey EV, et al. A systematic review of hip fracture incidence and probability of fracture worldwide. Osteoporos Int. 2012;23(9):2239-56.
22.    Gullberg B, Johnell O, Kanis JA. World-wide projections for hip fracture. Osteoporos Int. 1997;7(5):407-13.
23.    D’Adamo CR, Hawkes WG, Miller RR, et al. Short-term changes in body composition after surgical repair of hip fracture. Age Ageing. 2014;43(2):275-80.
24.    Magaziner J, Hawkes W, Hebel JR, et al. Recovery from hip fracture in eight areas of function. J Gerontol A Biol Sci Med Sci. 2000;55(9):M498-507.
25.    Miller RR, Cappola AR, Shardell MD, et al. Persistent changes in interleukin-6 and lower extremity function following hip fracture. J Gerontol A Biol Sci Med Sci. 2006;61(10):1053-8.
26.    Rolland Y, Abellan van Kan G, Benetos A, et al. Frailty, osteoporosis and hip fracture: causes, consequences and therapeutic perspectives. J Nutr Health Aging. 2008;12(5):335-46.
27.    Baumgartner RN. Body composition in healthy aging. Ann N Y Acad Sci. 2000;904:437-48.
28.    Batsis JA, Mackenzie TA, Jones JD, Lopez-Jimenez F, Bartels SJ. Sarcopenia, sarcopenic obesity and inflammation: Results from the 1999-2004 National Health and Nutrition Examination Survey. Clin Nutr. 2016.
29.    Zoico E, Di Francesco V, Guralnik JM, et al. Physical disability and muscular strength in relation to obesity and different body composition indexes in a sample of healthy elderly women. Int J Obes Relat Metab Disord. 2004;28(2):234-41.
30.    Beavers KM, Beavers DP, Houston DK, et al. Associations between body composition and gait-speed decline: results from the Health, Aging, and Body Composition study. Am J Clin Nutr. 2013;97(3):552-60.
31.    Upadhya B, Haykowsky MJ, Eggebeen J, Kitzman DW. Sarcopenic obesity and the pathogenesis of exercise intolerance in heart failure with preserved ejection fraction. Curr Heart Fail Rep. 2015;12(3):205-14.
32.    Haykowsky MJ, Kouba EJ, Brubaker PH, Nicklas BJ, Eggebeen J, Kitzman DW. Skeletal muscle composition and its relation to exercise intolerance in older patients with heart failure and preserved ejection fraction. Am J Cardiol. 2014;113(7):1211-6.
33.    Kitzman DW, Brubaker P, Morgan T, et al. Effect of Caloric Restriction or Aerobic Exercise Training on Peak Oxygen Consumption and Quality of Life in Obese Older Patients With Heart Failure With Preserved Ejection Fraction: A Randomized Clinical Trial. JAMA. 2016;315(1):36-46.
34.    Villareal DT, Apovian CM, Kushner RF, Klein S. Obesity in older adults: technical review and position statement of the American Society for Nutrition and NAASO, The Obesity Society. Am J Clin Nutr. 2005;82(5):923-34.
35.    Villareal DT, Chode S, Parimi N, et al. Weight loss, exercise, or both and physical function in obese older adults. N Engl J Med. 2011;364(13):1218-29.
36.    Houston DK, Leng X, Bray GA, et al. A long-term intensive lifestyle intervention and physical function: the look AHEAD Movement and Memory Study. Obesity (Silver Spring). 2015;23(1):77-84.
37.    Liu CK, Leng X, Hsu FC, et al. The impact of sarcopenia on a physical activity intervention: the Lifestyle Interventions and Independence for Elders Pilot Study (LIFE-P). J Nutr Health Aging. 2014;18(1):59-64.
38.    Singh NA, Quine S, Clemson LM, et al. Effects of high-intensity progressive resistance training and targeted multidisciplinary treatment of frailty on mortality and nursing home admissions after hip fracture: a randomized controlled trial. J Am Med Dir Assoc. 2012;13(1):24-30.
39.    Anker SD, Coats AJ, Morley JE, et al. Muscle wasting disease: a proposal for a new disease classification. J Cachexia Sarcopenia Muscle. 2014;5(1):1-3.
40.    Cruz-Jentoft AJ, Baeyens JP, Bauer JM, et al. Sarcopenia: European consensus on definition and diagnosis: Report of the European Working Group on Sarcopenia in Older People. Age Ageing. 2010;39(4):412-23.
41.    Morley JE, Abbatecola AM, Argiles JM, et al. Sarcopenia with limited mobility: an international consensus. J Am Med Dir Assoc. 2011;12(6):403-9.
42.    Baumgartner RN, Wayne SJ, Waters DL, Janssen I, Gallagher D, Morley JE. Sarcopenic obesity predicts instrumental activities of daily living disability in the elderly. Obes Res. 2004;12(12):1995-2004.
43.    Rolland Y, Lauwers-Cances V, Cristini C, et al. Difficulties with physical function associated with obesity, sarcopenia, and sarcopenic-obesity in community-dwelling elderly women: the EPIDOS (EPIDemiologie de l’OSteoporose) Study. Am J Clin Nutr. 2009;89(6):1895-900.
44.    Alley DE, Shardell MD, Peters KW, et al. Grip strength cutpoints for the identification of clinically relevant weakness. J Gerontol A Biol Sci Med Sci. 2014;69(5):559-66.
45.    Cawthon PM, Peters KW, Shardell MD, et al. Cutpoints for low appendicular lean mass that identify older adults with clinically significant weakness. J Gerontol A Biol Sci Med Sci. 2014;69(5):567-75.
46.    Joppa P, Tkacova R, Franssen FM, et al. Sarcopenic Obesity, Functional Outcomes, and Systemic Inflammation in Patients With Chronic Obstructive Pulmonary Disease. J Am Med Dir Assoc. 2016.
47.    Haykowsky MJ, Brubaker PH, Morgan TM, Kritchevsky S, Eggebeen J, Kitzman DW. Impaired aerobic capacity and physical functional performance in older heart failure patients with preserved ejection fraction: role of lean body mass. J Gerontol A Biol Sci Med Sci. 2013;68(8):968-75.
48.    Jones SE, Maddocks M, Kon SS, et al. Sarcopenia in COPD: prevalence, clinical correlates and response to pulmonary rehabilitation. Thorax. 2015;70(3):213-8.
49.    Malietzis G, Aziz O, Bagnall NM, Johns N, Fearon KC, Jenkins JT. The role of body composition evaluation by computerized tomography in determining colorectal cancer treatment outcomes: a systematic review. Eur J Surg Oncol. 2015;41(2):186-96.
50.    Nyman SR, Victor CR. Older people’s recruitment, sustained participation, and adherence to falls prevention interventions in institutional settings: a supplement to the Cochrane systematic review. Age Ageing. 2011;40(4):430-6.
51.    Barnett K, Mercer SW, Norbury M, Watt G, Wyke S, Guthrie B. Epidemiology of multimorbidity and implications for health care, research, and medical education: a cross-sectional study. Lancet. 2012;380(9836):37-43.
52.    Broekhuizen K, Pothof A, de Craen AJ, Mooijaart SP. Characteristics of randomized controlled trials designed for elderly: a systematic review. PLoS One. 2015;10(5):e0126709.
53.    Rodriguez-Manas L, Bayer AJ, Kelly M, et al. An evaluation of the effectiveness of a multi-modal intervention in frail and pre-frail older people with type 2 diabetes–the MID-Frail study: study protocol for a randomised controlled trial. Trials. 2014;15:34.
54.    Malmstrom TK, Miller DK, Simonsick EM, Ferrucci L, Morley JE. SARC-F: a symptom score to predict persons with sarcopenia at risk for poor functional outcomes. J Cachexia Sarcopenia Muscle. 2016;7(1):28-36.
55.    Woo J, Leung J, Morley JE. Validating the SARC-F: a suitable community screening tool for sarcopenia? J Am Med Dir Assoc. 2014;15(9):630-4.
56.    Cooper C, Fielding R, Visser M, et al. Tools in the assessment of sarcopenia. Calcif Tissue Int. 2013;93(3):201-10.
57.    Newman AB, Simonsick EM, Naydeck BL, Boudreau RM, Kritchevsky SB, Nevitt MC, et al. Association of long-distance corridor walk performance with mortality, cardiovascular disease, mobility limitation, and disability. JAMA. 2006;295(17):2018-26.
58.    Klepin HD, Geiger AM, Tooze JA, et al. Geriatric assessment predicts survival for older adults receiving induction chemotherapy for acute myelogenous leukemia. Blood. 2013;121(21):4287-94.
59.    Food and Drug Administration. Guidance for Industry: Qualification Process for Drug Development Tools, accessed 12/18/12 http://www.fda.gov/downloads/Drugs/GuidanceComplicanceRegulatoryInformationi/Guidances/UCM230597.pdf. 2010.
60.    Vellas B, Pahor M, Manini T, et al. Designing pharmaceutical trials for sarcopenia in frail older adults: EU/US Task Force recommendations. J Nutr Health Aging. 2013;17(7):612-8.
61.    Baker GT, 3rd, Sprott RL. Biomarkers of aging. Exp Gerontol. 1988;23(4-5):223-39.
62.    Roman D, Mahoney K, Mohamadi A. Sarcopenia: What’s in a Name? J Am Med Dir Assoc. 2012.
63.    Marzetti E, Calvani R, Cesari M, et al. Operationalization of the physical frailty & sarcopenia syndrome: rationale and clinical implementation. Transl Med UniSa. 2015;13:29-32.
64.    Marzetti E, Calvani R, Landi F, et al. Innovative Medicines Initiative: The SPRINTT Project. J Frailty Aging. 2015;4(4):207-8.
65.    Mijnarends DM, Schols JM, Meijers JM, et al. Instruments to assess sarcopenia and physical frailty in older people living in a community (care) setting: similarities and discrepancies. J Am Med Dir Assoc. 2015;16(4):301-8.





1. Faculty of Physical Activity Sciences, University of Sherbrooke, Sherbrooke, Quebec, Canada; 2. Research Centre on Aging, Institute of Geriatrics of Sherbrooke, Sherbrooke University, Sherbrooke, Quebec, Canada; 3. The Manitoba Institute of Child Health, University of Manitoba, Manitoba, Canada; 4. Faculty of Medicine, University of Manitoba, Winnipeg, Manitoba, Canada; 5. Institut de Recherche de l’Hopital Monfort, Ottawa, Ontario, Canada; 6. School of Human Kinetics, Faculty of Health Sciences, University of Ottawa, Ottawa, Ontario, Canada; 7. Department of Nutrition, Université de Montréal, Montreal, Quebec, Canada; 8. Institut de Recherches Cliniques de Montréal (IRCM), Montreal,
Quebec, Canada.

Corresponding author: Martin Brochu, Ph.D., Research Centre on Aging,1036 Belvédère Sud, Sherbrooke, Québec, Canada, J1H 4C4, Tel. (819) 780-2220 # 45326, Fax. (819) 829-7141, E-mail. martin.brochu@usherbrooke.ca

J Frailty Aging 2015;4(3):155-162
Published online November 10, 2015, http://dx.doi.org/10.14283/jfa.2015.54



Objective: The dynapenic (DYN)-obese phenotype is associated with an impaired metabolic profile. However, there is a lack of evidences regarding the effect of lifestyle interventions on the metabolic profile of individual with dynapenic phenotype. The objective was to investigate the impact of caloric restriction (CR) with or without resistance training (RT) on body composition, metabolic profile and muscle strength in DYN and non-dynapenic (NDYN) overweight and obese menopausal women. Design: 109 obese menopausal women (age 57.9 ± 9.0 yrs; BMI 32.1 ± 4.6 kg/m2) were randomized to a 6-month CR intervention with or without a RT program. Participants were categorized as DYN or NDYN based on the lowest tertile of relative muscle strength in our cohort (< 4.86 kg/BMI). Measurements: Body composition was measured by DXA, body fat distribution by CT scan, glucose homeostasis at fasting state and during an euglycemic-hyperinsulinemic clamp, fasting lipids, resting blood pressure, fasting inflammation markers and maximal muscle strength. Results: No difference was observed between groups at baseline for body composition and the metabolic profile. Overall, a treatment effect was observed for all variables of body composition and some variables of the metabolic profile (fasting insulin, glucose disposal, triglyceride levels, triglycerides/HDL-Chol ratio and resting diastolic blood pressure) (P between 0.05 and 0.001). No Group X Treatment interaction was observed for variables of body composition and the metabolic profile. However, an interaction was observed for muscle strength; which significantly improved more in the CR+RT NDYN group (all P ≤ 0.05). Conclusions: In the present study, dynapenia was not associated with a worse metabolic profile at baseline in overweight and obese menopausal women. DYN and NDYN menopausal women showed similar cardiometabolic benefit from CR or CR+RT interventions. However, our results showed that the addition of RT to CR was more effective in improving maximal strength in DYN and NDYN obese menopausal women.


Key words: Dynapenia, muscle strength, obesity, caloric restriction, resistance training.



The rapid increase in obesity prevalence has been reported in individuals aged between 55 and 75 years; and particularly in menopausal women (1). The transition to menopause is often accompanied by increased central adiposity, a more atherogenic lipid profile, alterations in glucose metabolism and a sharp rise in type 2 diabetes and cardiovascular disease risk (2, 3). In addition, the consequences of a deteriorated cardiometabolic profile seem to be amplified in older individuals. It is possibly explained by physiologic changes associated with the aging process and/or the concurrence of chronic conditions more frequently observed during this period of life (4). Hence, fat mass increases and muscle mass and strength decrease with the aging process (5). Dynapenia, which refers to the normal age-related loss of muscle strength and power that usually occurs as part of aging, is now recognized as a debilitating and life threatening condition in older persons (6). Dynapenia has been negatively related to functional capacity, metabolic profile (7,8) and mortality (9). The combination of high fat mass accumulations and low muscle strength, phenotype called ‘dynapenic-obesity’ (10), is increasingly prevalent among older individuals (11). Interestingly, this phenotype has been associated with worse metabolic abnormalities compared to both conditions alone (11).

The few studies that have investigated the effects of DYN-obese phenotype in older individuals mostly looked at functional capacity (10) disability and mortality (11). A recent cross-sectional study has quantify the independent and additive effects of dynapenia and abdominal obesity on the metabolic syndrome in older men and women (12). Results showed that DYN-obese individuals have more metabolic alterations than those displaying dynapenia alone or those with neither abdominal obesity nor dynapenia. Furthermore, to our knowledge, only one study has examined the effect of a lifestyle intervention in DYN-obese individuals on the metabolic profile (13). Results of this study suggest that caloric restriction (CR) with or without resistance training (RT) is effective in improving the metabolic profile [total cholesterol (chol), low-density lipoprotein cholesterol (LDL-chol), systolic and diastolic blood pressure] in DYN-obese menopausal women. However, the study has some limitations such as a limited number of participants in each group, and no measures of visceral fat (VF) and insulin sensitivity. Also, because of the study design, they did not compare subjects displaying DYN- and non-dynapenic (NDYN-) obesity phenotypes. Consequently, the comparison of the effect of a weight loss intervention on DYN and NDYN obese individuals needs to be further investigated.

Thus, the current study’s objective was to compare the effect of CR alone or in combination with RT on body composition, the metabolic profile and muscle strength in DYN and NDYN overweight and obese menopausal women. We hypothesized that interventions aimed to increase muscle strength and decrease obesity levels would have greater effects on the metabolic profile and muscle strength in DYN overweight and obese women.



This article presents secondary analyses from a randomized controlled trial, originally designed to determine the effects of a 6-month RT in combination with CR on body composition, body fat distribution, and the metabolic profile in a large cohort of overweight and obese menopausal women (14). The present study population consisted of 109 non-diabetic overweight and obese menopausal women aged between 46 and 70 years at baseline. We excluded subjects who did not complete the 6-month intervention. Consequently, data of 87 women were used for these secondary analyses.

Eligible participants met the following criteria: BMI between 27 and 40 kg/m2, no menstruation for 1 > year, having a follicle stimulating hormone level > 30 U/l, <2h per week of structured exercise, non-smoker, low to moderate alcohol consumers (≤ 2 drinks/day), and no use of hormone replacement therapy. All participants were apparently healthy and had no history or evidence on physical examination or laboratory testing of (i) cardiovascular disease, peripheral vascular disease, or stroke; (ii) diabetes (2h standard 75-g oral glucose tolerance test (OGTT)); (iii) severe hypertension (resting blood pressure >170/100mm Hg); (iv) orthopedic limitations; (v) body weight fluctuation >5 kg in the previous 6 months; (vi) uncontrolled thyroid or pituitary disease; and (vii) medication that could affect cardiovascular function or metabolism. The study was approved by the Université de Montréal ethics committee. After reading and signing the consent form, each participant was submitted to a series of tests.

Weight stabilization period

Before and after the 6-month weight loss protocol, subjects were submitted to a weight stabilization period (±2 kg of body weight) before testing. The goal of this approach was to stabilize the various metabolic variables of interest that could be altered by body weight fluctuations.

Caloric restriction intervention

Study participants entered a 6-month weight loss program aimed at reducing body weight by 10%. To determine the level of CR, 500 to 800 kcal were subtracted from baseline resting metabolic rate (determined by indirect calorimetry) multiplied by a physical activity factor of 1.4, corresponding to a sedentary state (15).

Macronutrient composition of the diets was standardized: 55%, 30% and 15% of energy intake from carbohydrates, total fat and proteins accordingly to the American Heart Association (16). Each participant met with the study dietitian to receive the diet prescription and recommendations. In addition, participants were invited to meet bi-monthly with the study dietitian for nutrition classes of 1-1.5 hours. All participants in the CR group were instructed not to change their usual daily physical activity habits during the weight loss protocol.

Exercise intervention

The 6-month RT program consisted of four progressive phases and was performed weekly on 3 nonconsecutive days under the supervision of an exercise physiologist, as previously described (14). The workload was adjusted by the exercise physiologist (when necessary) to maintain the intensity prescribed. Lower body and upper body strength was assessed using leg press and bench press weight training equipment from Atlantis Precision Series (Atlantis Inc., Laval, Que.).

Each training session included a warm-up of low intensity walking on a treadmill for 10 min. The RT program consisted of the following exercises: 1) leg press; 2) chest press; 3) lateral pull downs; 4) shoulder press; 6) arm curls; and 7) triceps extensions. These exercises provide a total body RT program for all of the major muscle groups of the body.  

Anthropometric and body composition measures

Body weight was measured to the nearest 0.1 kg on a calibrated balance (Balance Industrielle Montréal, Montréal, Québec, Canada) and participant’s height was obtained with a standard stadiometer (Perspective Enterprises, Portage, Michigan, USA). Waist circumference (WC) was measured using a measuring tape to the nearest 0.1 cm at the highest point of the iliac crest at minimal expiration. Total fat mass (FM), percentage of FM (%FM) and total lean body mass (LBM) were measured using dual energy X-ray absorptiometry (DXA) (General Electric Lunar Prodigy, Madison, Wisconsin; software version 6.10.019), as previously described (17). During the procedure, participants were asked to wear only a standard hospital gown while in the supine position. Calibration was executed daily with a standard phantom. In our laboratory, the intra-class coefficient correlation for test–retest for FM and LBM was 0.99 (n = 18).

A CT scanner (GE LightSpeed 16, General Electric Medical Systems, Milwaukee, WI) was used to measure the VF and the abdominal subcutaneous fat (AScF) area. Participants were examined in the supine position with both arms stretched above their head. The position of the scan was established at the L4-L5 vertebral disc using a scout image of the body (17). We quantified VF by delineating the intra-abdominal cavity at the internal most aspect of the abdominal and oblique muscle walls surrounding the cavity and the posterior aspect of the vertebral body. The AScF area was quantified by highlighting fat located between the skin and the external most aspect of the abdominal muscle wall. Deep AScF (DAScF) and superficial ScF (SAScF) areas were measured by delineating the subcutaneous fascia within the AScF and by computing areas of the layers of fat on each side of the fascia. The cross-sectional areas of fat were highlighted and computed using an attenuation range of -190 to -30 Hounsfield Units (HU).

Muscle attenuation (MA) was calculated by delineating the regions of interest and then computing the surface areas using attenuation range of -190 to -30 HU for fat, and 0 to 100 HU for skeletal muscle. In our laboratory, test-retest measures of the different body fat distribution indices on 10 CT scans yielded a mean absolute difference of + 1%.

Characterization of subjects

Participants were first randomly assigned to one of the two groups [CR (n= 52) or CR+RT (n= 35)]. For the present secondary analyses, participants were characterized as DYN or NDYN.  Since a correction for anthropometric variability has been recommended to define dynapenia (5, 18), we computed a muscular relative strength index with the ratio of strength (lower body strength + upper body strength) on BMI (19, 20). We included both measures of lower and upper extremity muscle strength since previous studies suggested that the rate of age-associated decline in muscle strength is quite different in these two anatomic regions (21).  Women in the lowest tertile of muscular relative strength index were considered DYN, while those in the second and third tertiles were considered as NDYN. In our sample, the cutoff point was 4.86 kg/BMI. Four groups of subjects were then created [group 1: CR/DYN (n= 13); group 2: CR/NDYN (n= 48); group 3: CR + RT/DYN (n= 18); group 4: CR + RT/NDYN (n= 30)].  

Fasting blood samples

Venous blood samples were collected to measure fasting total chol, HDL-chol, LDL-chol, triglycerides (TG), glucose and insulin levels after a 12-h overnight fast. Plasma was analyzed on the day of collection. Analyses were done on the COBAS INTEGRA 400 (Roche Diagnostic, Montreal, Canada) analyzer for total cholesterol, HDL-chol, triglycerides and glucose. Total chol, HDL-chol and TG levels were used in the Friedewald formula (22) to calculate LDL-chol concentrations. Insulin levels were determined by automated radioimmunoassay (Linco Research Inc. (St-Charles, MO, USA). Serum high-sensitivity CRP (hs-CRP) concentrations were assessed by immunonephelometry on IMMAGE analyzer (Beckman Coulter, Villepinte, France).

Glucose disposal

The test began at 07h30 after a 12-h overnight fast, as previously described by DeFronzo et al. (23). An antecubital vein was cannulated for the infusion of 20 % dextrose and insulin (Actrapid®, Novo-Nordisk, Toronto, Canada). The other arm was cannulated for sampling of blood. Plasma glucose was measured every 10 min with a glucose analyzer (Beckman Instruments, Fullerton, CA) and maintained at fasting level with a variable infusion rate of 20 % dextrose. Insulin infusion was initiated at the rate of 75 mU/m2/min for 180 min. Glucose disposal was calculated as the mean rate of glucose infusion measured during the last 30 min of the clamp (steady state).

Resting blood pressure 

Sitting blood pressure was measured in the left arm after participants rested quietly for 10 min using a Dinamap automatic machine (Welch Allyn, San Diego, CA, USA). An appropriate cuff size was selected for each subject based on arm circumference. Procedure was carefully standardized (24). 

Statistical analyses

Data in tables are presented as means ± standard deviation (SD). Univariate Analyses of Variance (ANOVA) were performed to compare means for each variable of interest at baseline and after the intervention. Repeated ANOVA analyses were performed to quantify the effect of treatment. Then, the Games-Howell test was used for posteriori group comparisons when a main model effect was noted. P-value of ≤0.05 was considered statistically significant. Statistical analyses were performed using the SPSS Statistical Package (version 15.0, SPSS, Chicago, Il, USA).


Anthropometric measures and body composition 

No differences were observed at baseline among groups for anthropometric and body composition variables, with the exception of muscle attenuation (P= 0.03) (Table 1). Body weight, BMI, WC, FM, FMI, LBM, LBMI, AScF and VF significantly and similarly decreased in the four groups after the intervention (all P≤ 0.05).

Table 1 Groups’ comparisons for body composition and body fat distribution

Values are means ± standard deviation (SD); Δ = delta; BMI= body mass index; LBM= lean body mass; LBMI= LBM index, FM= fat mass; FMI= FM index; CT= computed tomography; AScF= abdominal subcutaneous fat; VF= visceral fat; MA= muscle attenuation; *= Do not include bone mass; NS= non-significant; Dynapenic ≤ 4.86 kg/BMI; Non-Dynapenic > 4.86 kg/BMI.

Metabolic profile 

No differences were observed at baseline among groups for the metabolic profile. TG, TG/HDL-chol ratio, fasting insulin, glucose disposal, relative glucose disposal and diastolic blood pressure significantly decreased in the four groups after the intervention (all P≤ 0.05); with no difference among groups. No improvement was observed in any groups for hs-CRP, IL-6, total cholesterol, HDL-chol and LDL-chol, Chol/HDL ratio, fasting glucose and systolic blood pressure after the interventions.

Table 2 Groups’ comparisons for metabolic profile

Values are means ± standard deviation (SD); Δ = delta; HR= heart rate; IL-6= interleukin-6; Chol= cholesterol; HDL= high-density lipoprotein; LDL= low-density lipoprotein; NS=non-significant; Dynapenic ≤ 4.86 kg/BMI; Non-Dynapenic > 4.86 kg/BMI.

Muscle strength 

By design, strength variables were significantly different between DYN and NDYN groups at baseline (all P≤ 0.001); with significantly higher values in NDYN groups (Table 3). All measures of strength were significantly improved in RT groups (DYN and NDYN subjects) and the CR-DYN group after the intervention (all P≤ 0.001); with greater improvements in the CR+RT/NDYN group compared to the others (all P≤ 0.001). Moreover, a significant decrease for all the measures of strength was observed in the CR/NDYN group compared to the CR+RT/NDYN group following the intervention.

Table 3 Groups’ comparisons for measures of strength

Values are means ± standard deviation (SD); Δ = delta; a = significantly different from Diet/Dynapenic; b = significantly different from Diet/Non-Dynapenic; c = significantly different from Diet + Resistance training/Dynapenic; d = significantly different from Diet + Resistance training/Non-Dynapenic; Dynapenic ≤ 4.86 kg/BMI; Non-Dynapenic > 4.86 kg/BMI.



The aim of this secondary analysis was to compare the effect of a 6-month lifestyle intervention on body composition, the metabolic profile, and strength variables between DYN and NDYN overweight and obese menopausal women. We hypothesized that interventions aimed to increase muscle strength and decrease obesity levels would have greater effects on the metabolic profile and muscle strength in DYN overweight and obese women. Results from the present study showed that both DYN and NDYN obese women improved similarly body composition and the metabolic profile following CR alone or in combination with RT. However, the addition of RT to CR has superior beneficial effect on muscle strength.

As previously mentioned, there is only one study that directly compared the effect of CR and RT on the combined condition of obesity and dynapenia on body composition, metabolic profile and strength in older women (13). Their results showed that both CR and CR+RT groups improved several variables of the metabolic profile to the same extent (total chol, TG and systolic blood pressure) after the intervention. They showed that a moderate weight loss (≈5 kg) was associated with significant improvements in the metabolic profile. Our results support their findings. Of all metabolic components measured in our study, we observed similar improvements for TG, TG/HDL ratio, glucose homeostasis (fasting insulin, glucose disposal (mg/min/kg) and relative glucose disposal) and diastolic blood pressure to the same extent in DYN and NDYN women following both interventions. Moreover, the addition of RT to the diet had no supplementary value on the metabolic profile in the present study. Overall, our results are in line with previous studies in overweight menopausal women (25) and DYN obese older women (13).

Although weight-loss therapy is recommended to improve obesity-related metabolic complications, a prevailing concern in the clinical community is that the use of CR alone could have negative effects on LBM (26). In fact, it is recognized that CR induced weight loss results in a decreased LBM, which correspond in general to approximately 25% of total body weight loss (27). Furthermore, it has been reported that the addition of RT can attenuate this loss by half (28). RT has also been proposed to be an interesting approach to counteract the decrease in LBM observed in older adults, and particularly during caloric restriction-induced weight loss (29). Our data showed however that CR was associated with a reduced LBM to a similar extent to CR+RT in our DYN and NDYN obese women. This result is in disagreement with those of Frimel et al. (30), which is the only other study comparing CR to CR+RT in older (70 ± 5 yr) obese (BMI 37 ± 5 kg/m2) adults. They showed significantly greater decreases in LBM in the CR only group compare to the CR+RT (3.5 ± 2.1 vs. 1.8 ± 1.5 kg; P= 0.02). Regarding the absence of difference in LBM loss between the CR vs. CR+RT in our study, it is important to note that even though RT is known to be a potent stimulus for acute increases in circulating anabolic hormones such as testosterone, growth hormone and insulin-like growth factor-1 in men (31), the increase in anabolic hormones is minor or absent in older women following 6-month RT (32). A low level of testosterone in older women may be a limiting factor for muscle hypertrophy in response to RT (32), and hence could partly explain why we did not observe an increase or a lower reduction in LBM following RT. Furthermore, our results may also suggest an age-related resistance to alterations in metabolic function in response to RT. This assumption is supported by Dionne et al. and Hakkinen et al. (33, 34) who previously observed that older women experience significantly lower increases in LBM compared to younger women following a 6-month RT program.

Significant changes were also observed in our study for body weight, BMI, WC, FM, VF and AScF; which significantly and similarly decreased in DYN and NDYN obese women after both interventions. These results are in agreement with those of Senechal et al. (13) who showed similar decreased in body weight and trunk FM between CR or CR+RT in DYN obese women. Finally, Frimel et al. (30) also observed similar losses in body weight and FM following CR and CR+RT in frail obese older adults.

The progressive decline in muscle strength is perhaps the most ineluctable anatomical change occurring with aging. While muscle strength and muscle mass are related, it has previously been suggested that muscle quality (e.g., strength) is more important than muscle quantity (e.g., mass) in aging humans (35) as a determinant of functional limitation and poor health in older age (5, 21). Although the loss of LBM was not prevented with RT in our study, RT was still associated with increases in muscle strength in DYN women. These findings demonstrate that despite a decrease in LBM induced by CR, it did not translate into a decrease in muscle strength in DYN women. Very few studies have reported the effects of weight-loss intervention on muscle strength or muscle quality in obese women (36). Our results are in line with those of Wang et al. (36) showing that CR induced weight loss combined with moderate strength training maintained absolute knee extensor muscle strength in older persons with osteoarthritis. Like our results, these effects occurred despite a significant loss of LBM during the intervention. Taken together, our findings suggest that other factors than muscle mass contribute to gains in muscle strength in older obese women following a RT program. Finally, muscular strength also appears to be an important protecting factor for cardiometabolic health (37). Actually, some cross sectional and prospective data have showed an inverse association between strength and cardiometabolic risk factors (7, 12). However, our data do not showed a superior beneficial effect of RT on the metabolic profile.

Our study presents limitations. First, generalization of our data applies only to non-diabetic, sedentary, overweight and obese menopausal women. Second, the small number of participants in each group limits the power of the study. Third, comparisons with other studies are difficult because of the use of different cut off points for defining dynapenia (5). For example, the Foundation for National Institutes of Health Sarcopenia Project (FNIH) recommends the use of a single measure, grip strength, to characterize and individual’s overall strength status (38). However, it has been observed that the use of a single action (such as grip strength) may not always be a valid representative of and individual’s overall strength (39). Hence, the combination of upper and lower body strength might have superior potential value as an indicator of overall strength than handgrip strength alone (39). Fourth, the relatively normal metabolic profile at baseline in our subjects is likely to explain the absence of effect on total cholesterol, HDL-chol, LDL-chol, fasting glucose and resting systolic blood pressure observed. Five, measures of muscle fiber types and areas, capillaries, biochemical properties of skeletal muscle and hormonal profile would have been interesting. Finally, although another study from the MONET cohort examined the relationship between muscle strength and metabolic disturbances (insulin sensitivity) (40), its design was cross-sectional and participants were not categorized as DYN or NDYN. Hence, we believe our findings add to the body of literature regarding the effects of lifestyle intervention in DYN overweight and obese women.

Despite limitations, strengths of our findings include the use of a randomized design as well as the use of gold standard techniques available for the measurement of body composition, body fat distribution and cardiometabolic profile. We also used a 1-month weight stabilization period before testing to minimize the impact of body weight fluctuations on the metabolic profile. Finally, the use of the lowest tertile of strength is a relatively simple measure, which enhance the application in clinical practice (41). Overall, we consider that the methodology used strengthens the validity of our results.

In conclusion, results from the present study show that weight loss following CR and RT improves similarly the metabolic profile and body composition in DYN and NDYN overweight and obese menopausal women. The addition of RT had no additional effect on body composition and the metabolic profile, but was associated with improved strength in DYN and NDYN women despite losses of LBM. Further studies are needed to evaluate the effect of lifestyle interventions on the metabolic profile of individual with DYN phenotype.

Acknowledgements: RRL is a senior FRQS (Fonds de Recherches en Santé du Québec). MS is supported by the Manitoba Health Research Council, the Canadian Institute of Health Research (CIHR), and by the CIHR-Integrated and Mentored Pulmonary and Cardiovascular Training (CIHR-strategic Training Program). The Montreal Ottawa New Emerging Team in Obesity Group thanks Lyne Messier (study coordinator, Registered dietitian); Maxime St-Onge, Benoit Tousignant, and Philippe Carrier (training supervision); Isabelle Vignault and Jennifer Levasseur, R.N.; and the patients for their exceptional involvement in this study. This work was supported by grants from the Canadian Institute of Health Research (CIHR) New and Emerging Teams in Obesity (Université de Montréal and University of Ottawa; Montreal Ottawa New Emerging Team in Obesity project). OHN—63279 The MONET group thanks patients for their exceptional involvement in this study.

Conflict of Interest: The authors declared no conflict of interest.


1. Flegal, K.M., M.D. Carroll, B.K. Kit, and C.L. Ogden, Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999-2010. JAMA, 2012. 307(5): p. 491-7.

2. Burke, G.L., A.G. Bertoni, S. Shea, et al., The impact of obesity on cardiovascular disease risk factors and subclinical vascular disease: the Multi-Ethnic Study of Atherosclerosis. Arch Intern Med, 2008. 168(9): p. 928-35.

3. Pan, W.H. and W.T. Yeh, How to define obesity? Evidence-based multiple action points for public awareness, screening, and treatment: an extension of Asian-Pacific recommendations. Asia Pac J Clin Nutr, 2008. 17(3): p. 370-4.

4. Butler, J., N. Rodondi, Y. Zhu, et al., Metabolic syndrome and the risk of cardiovascular disease in older adults. J Am Coll Cardiol, 2006. 47(8): p. 1595-602.

5. Manini, T.M. and B.C. Clark, Dynapenia and aging: an update. J Gerontol A Biol Sci Med Sci, 2012. 67(1): p. 28-40.

6. Clark, B.C. and T.M. Manini, What is dynapenia? Nutrition, 2012. 28(5): p. 495-503.

7. Jurca, R., M.J. Lamonte, C.E. Barlow, et al., Association of muscular strength with incidence of metabolic syndrome in men. Med Sci Sports Exerc, 2005. 37(11): p. 1849-55.

8. Senechal, M., J.M. McGavock, T.S. Church, et al., Cut points of muscle strength associated with metabolic syndrome in men. Med Sci Sports Exerc, 2014. 46(8): p. 1475-81.

9. Xue, Q.L., B.A. Beamer, P.H. Chaves, J.M. Guralnik, and L.P. Fried, Heterogeneity in rate of decline in grip, hip, and knee strength and the risk of all-cause mortality: the Women’s Health and Aging Study II. J Am Geriatr Soc, 2010. 58(11): p. 2076-84.

10. Bouchard, D.R. and I. Janssen, Dynapenic-obesity and physical function in older adults. J Gerontol A Biol Sci Med Sci, 2010. 65(1): p. 71-7.

11. Stenholm, S., D. Alley, S. Bandinelli, et al., The effect of obesity combined with low muscle strength on decline in mobility in older persons: results from the InCHIANTI study. Int J Obes (Lond), 2009. 33(6): p. 635-44.

12. Senechal, M., I.J. Dionne, and M. Brochu, Dynapenic abdominal obesity and metabolic risk factors in adults 50 years of age and older. J Aging Health, 2012. 24(5): p. 812-26.

13. Senechal, M., D.R. Bouchard, I.J. Dionne, and M. Brochu, The effects of lifestyle interventions in dynapenic-obese postmenopausal women. Menopause, 2012. 19(9): p. 1015-21.

14. Brochu, M., M.F. Malita, V. Messier, et al., Resistance training does not contribute to improving the metabolic profile after a 6-month weight loss program in overweight and obese postmenopausal women. J Clin Endocrinol Metab, 2009. 94(9): p. 3226-33.

15. Tremblay, A., C. Pelletier, E. Doucet, and P. Imbeault, Thermogenesis and weight loss in obese individuals: a primary association with organochlorine pollution. Int J Obes Relat Metab Disord, 2004. 28(7): p. 936-9.

16. Krauss, R.M., R.J. Deckelbaum, N. Ernst, et al., Dietary guidelines for healthy American adults. A statement for health professionals from the Nutrition Committee, American Heart Association. Circulation, 1996. 94(7): p. 1795-800.

17. Brochu, M., M.E. Mathieu, A.D. Karelis, et al., Contribution of the lean body mass to insulin resistance in postmenopausal women with visceral obesity: a Monet study. Obesity (Silver Spring), 2008. 16(5): p. 1085-93.

18. Manini, T.M., M. Visser, S. Won-Park, et al., Knee extension strength cutpoints for maintaining mobility. J Am Geriatr Soc, 2007. 55(3): p. 451-7.

19. Alley, D.E., M.D. Shardell, K.W. Peters, et al., Grip strength cutpoints for the identification of clinically relevant weakness. J Gerontol A Biol Sci Med Sci, 2014. 69(5): p. 559-66.

20. McLean, R.R., M.D. Shardell, D.E. Alley, et al., Criteria for clinically relevant weakness and low lean mass and their longitudinal association with incident mobility impairment and mortality: the foundation for the National Institutes of Health (FNIH) sarcopenia project. J Gerontol A Biol Sci Med Sci, 2014. 69(5): p. 576-83.

21. Lauretani, F., C.R. Russo, S. Bandinelli, et al., Age-associated changes in skeletal muscles and their effect on mobility: an operational diagnosis of sarcopenia. J Appl Physiol (1985), 2003. 95(5): p. 1851-60.

22. Friedewald, W.T., R.I. Levy, and D.S. Fredrickson, Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem, 1972. 18(6): p. 499-502.

23. DeFronzo, R.A., J.D. Tobin, and R. Andres, Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol, 1979. 237(3): p. E214-23.

24. Beevers, G., G.Y. Lip, and E. O’Brien, ABC of hypertension. Blood pressure measurement. Part I-sphygmomanometry: factors common to all techniques. BMJ, 2001. 322(7292): p. 981-5.

25. Fox, A.A., J.L. Thompson, G.E. Butterfield, et al., Effects of diet and exercise on common cardiovascular disease risk factors in moderately obese older women. Am J Clin Nutr, 1996. 63(2): p. 225-33.

26. Villareal, D.T., G.I. Smith, K. Shah, and B. Mittendorfer, Effect of weight loss on the rate of muscle protein synthesis during fasted and fed conditions in obese older adults. Obesity (Silver Spring), 2012. 20(9): p. 1780-6.

27. Villareal, D.T., C.M. Apovian, R.F. Kushner, and S. Klein, Obesity in older adults: technical review and position statement of the American Society for Nutrition and NAASO, The Obesity Society. Am J Clin Nutr, 2005. 82(5): p. 923-34.

28. Weinheimer, E.M., L.P. Sands, and W.W. Campbell, A systematic review of the separate and combined effects of energy restriction and exercise on fat-free mass in middle-aged and older adults: implications for sarcopenic obesity. Nutr Rev, 2010. 68(7): p. 375-88.

29. Yarasheski, K.E., Exercise, aging, and muscle protein metabolism. J Gerontol A Biol Sci Med Sci, 2003. 58(10): p. M918-22.

30. Frimel, T.N., D.R. Sinacore, and D.T. Villareal, Exercise attenuates the weight-loss-induced reduction in muscle mass in frail obese older adults. Med Sci Sports Exerc, 2008. 40(7): p. 1213-9.

31. Migiano, M.J., J.L. Vingren, J.S. Volek, et al., Endocrine response patterns to acute unilateral and bilateral resistance exercise in men. J Strength Cond Res, 2010. 24(1): p. 128-34.

32. Hakkinen, K., A. Pakarinen, W.J. Kraemer, R.U. Newton, and M. Alen, Basal concentrations and acute responses of serum hormones and strength development during heavy resistance training in middle-aged and elderly men and women. J Gerontol A Biol Sci Med Sci, 2000. 55(2): p. B95-105.

33. Dionne, I.J., M.O. Melancon, M. Brochu, P.A. Ades, and E.T. Poelhman, Age-related differences in metabolic adaptations following resistance training in women. Exp Gerontol, 2004. 39(1): p. 133-8.

34. Hakkinen, K. and A. Pakarinen, Serum hormones and strength development during strength training in middle-aged and elderly males and females. Acta Physiol Scand, 1994. 150(2): p. 211-9.

35. Stephen, W.C. and I. Janssen, Sarcopenic-obesity and cardiovascular disease risk in the elderly. J Nutr Health Aging, 2009. 13(5): p. 460-6.

36. Wang, X., G.D. Miller, S.P. Messier, and B.J. Nicklas, Knee strength maintained despite loss of lean body mass during weight loss in older obese adults with knee osteoarthritis. J Gerontol A Biol Sci Med Sci, 2007. 62(8): p. 866-71.

37. Artero, E.G., D.C. Lee, C.J. Lavie, et al., Effects of muscular strength on cardiovascular risk factors and prognosis. J Cardiopulm Rehabil Prev, 2012. 32(6): p. 351-8.

38. Studenski, S.A., K.W. Peters, D.E. Alley, et al., The FNIH sarcopenia project: rationale, study description, conference recommendations, and final estimates. J Gerontol A Biol Sci Med Sci, 2014. 69(5): p. 547-58.

39. Bohannon, R.W., Is it legitimate to characterize muscle strength using a limited number of measures? J Strength Cond Res, 2008. 22(1): p. 166-73.

40. Karelis, A.D., B. Tousignant, J. Nantel, et al., Association of insulin sensitivity and muscle strength in overweight and obese sedentary postmenopausal women. Appl Physiol Nutr Metab, 2007. 32(2): p. 297-301.

41. Visser, M., Obesity, sarcopenia and their functional consequences in old age. Proc Nutr Soc, 2011. 70(1): p. 114-8.




Colective Health Department, University of Campinas – School of Medical Sciences, São Paulo, Brazil.

Corresponding author: Bruna Fernanda do Nascimento Jacinto de Souza. Colective Health Department, University of Campinas – School of Medical Sciences, Rua Tessália Vieira de Camargo, 126, Campinas, São Paulo, CEP 13083-887, Brazil. Phone: 05519 35219569, E-mail: brunajs@fcm.unicamp.br

J Frailty Aging 2015;4(4):198-206
Published online June 10, 2015, http://dx.doi.org/10.14283/jfa.2015.55


 Background: The epidemiological and nutritional transition processes in the last decades underlie the rising trend of obesity in the elderly and is related to increased risk of chronic non-communicable diseases and decreased functional status. Objective: To analyze the association of demographic, socioeconomic, lifestyle and health-related factors with overweight and obesity in elderly. Design: Cross-sectional study. Setting: Carried out in Campinas-São Paulo, Brazil, in 2011. Participants: 452 non-institutionalized elderly (aged ≥60 years), half were users of a government-run soup kitchen and the other half were neighbors of the same sex. Results: Overweight frequency (BMI ≥25 and <30 kg/m2) was 44.5% and obesity (BMI ≥30 kg/m2) was 21.7%. In the multiple multinomial logistic regression model adjusted for sex, age group and economic class, there was greater chance of overweight among those that reported dyslipidemia; those that reported arthritis/ arthrosis/rheumatism and that once or more per week replaced supper by a snack were more likely to be obese. Elderly who did not leave home daily and reported diabetes had higher chance of overweight and obesity. Conclusions: Overweight and obesity are associated with worse living and health-related conditions, such as physical inactivity, changes in eating behaviors, and chronic diseases. Public health policies should encourage regular physical activity and healthy eating behaviors, focusing on traditional diet, through nutritional education, in order to reduce the prevalence of overweight and obesity and chronic diseases.

Key words: Overweight, obesity, aging, food intake, chronic disease.


Two major trends are taking place in global health in the last decades: the aging of the population and the increasing obesity (1-3). Data from the 2010 Census show that the elderly aged 60 and over represent 10.8% of the Brazilian population (4) and 12.4% of the population in Campinas – São Paulo (5).

Changes in the nutritional status of the elderly, such as malnutrition or obesity may be associated with health risks. Several studies have pointed out the association of malnutrition and mortality among the elderly (6, 7). Malnutrition and sarcopenia among elderly people are considered risk factors for death (8, 9).

Meanwhile, obesity has become a global epidemic over the past 20 years, affecting genders, all ages and socioeconomic groups. Obesity is characterized by excessive accumulation of body fat due to the imbalance between intake and energy expenditure (10). Obesity is considered a chronic, complex and multifactorial condition that leads to metabolic disease (10)   and increases the risk of chronic non-communicable diseases (NCDs). Among them cardiovascular disease, type 2 diabetes, endocrine and metabolic disorders, sleep apnea, osteoarthritis, certain cancers, various psychological problems and functional disability are responsible for decreased quality of life, high social and economic cost and  increased mortality (10-13).

Studies with elderly have estimated the prevalence of overweight and obesity. In Korea, the prevalence of overweight/ obesity was 31.7% in men and 42.2% in women (12). A multicenter study observed prevalence of overweight/ obesity of 43.8% in Havana-Cuba, 56.3% in Bridgetown-Barbados, 65.1% in São Paulo-Brazil, 70% in Montevideo-Uruguay, 72.4% in México City-Mexico and 75.6% in Santiago-Chile (13).

Thus, considering the growing trend of aging and obesity, it is important to evaluate the nutritional status of the elderly, so the aim of this study was to analyze the associations between demographic, socioeconomic, lifestyle and health-related variables with overweight and obesity in elderly people.


A cross-sectional study was carried out in Campinas-SP with non-institutionalized elderly, with 60 years or more, in 2011. All study participants were individually informed about the study and interviewed only after they signed a free and informed consent form. The project was approved by the Research Ethics Committee of the School of Medical Sciences of the Universidade Estadual de Campinas in April 2010 (CEP nº 169/2009).

Initially, the research was designed as a case-control study with the objective of comparing elderly soup kitchen users with elderly neighbors, to evaluate the soup kitchen restaurant as a tool to fight hunger and food insecurity. The sample size was calculated using Epi-Info, version 6.0.4 giving 210 elderly cases and controls. Including a 10% loss, the sample was 462 elderly. A total of 457 elderly living in permanent dwellings were interviewed and five of them, classified as «underweight» were excluded, thus the final sample was 452.

Half of the sample was selected among the users of the soup kitchen restaurant Bom Prato, which offers healthy and low cost meals (14, 15). The invitation to participate in the research was done for all the seniors who were in the queue; approximately 15% refused to participate. The anthropometric measurements were taken before the meal and the interviews after it.  The other half of the sample was composed by neighbors of the same sex of those selected in the restaurant. The selection process started to the right of the reference household. In the absence of a subject in the block, the search continued in the next block to the right of the first one. In households with more than one subject of the same sex, just one was selected, using as selection criteria the age, it was included in the study the elderly with the nearest age to the elderly soup kitchen user. Among neighbors there were approximately 4% of refusals to participate. For all participants the following exclusion criteria were applied: cognitive disorders, bed ridden, wheelchair users, and problems for standing up in a right position.

The interviews were conducted by trained interviewers between January and August 2011, after a pilot study. The questionnaire was developed by the researchers, with closed or semi-open questions about demographic and socioeconomic characteristics, food intake, use of the soup kitchen restaurant, health-related conditions and anthropometric measurement.

The demographic variables were sex (male; female), age (60-69 years; 70-79 years, and 80 years or more), race/color (Caucasian; not Caucasian), marital status (married; single/separated/ divorced/widowed) and being the head of the household (yes; no).

The socioeconomic variables were total family income (≤2 minimum salaries; >2 minimum salaries), job status (yes; no), and economic class according to the Economic Classification Criterion Brazil 2008 (16): A+B; C; D+E, being A+B the richest and D+E the poorest.

The health-related variables were self-perceived health (very good/good; regular; bad/ very bad); smoking status (current smoker, former smoker, never smoked), self-reported diseases were obtained with the following question «Have you been told by a doctor or other health professional that you have any of the following diseases?» (high blood pressure; arthrosis /arthritis /rheumatism; dyslipidemia; diabetes; heart disease); daily use of medications (yes; no); and hospitalization in the past year (yes; no). Physical activity was estimated by the following question: «Do you leave your home daily?» (yes; no); and disability was estimated with the question «For how long can you walk without getting tired?» (15 minutes or less; above 15 minutes).

Food security was determined by Escala Brasileira de Insegurança Alimentar (EBIA, Brazilian Household Food Insecurity Measurement Scale) (17) (food security, mild food insecurity, moderate/severe food insecurity). This scale was adapted from the USDA Food Insecurity Scale, and validated in Brazil after tested for content and face validity in expert and focus groups made up of community members. Chronbach’s α was 0.91 and the scale item response curves were parallel across the 4 household income strata. After it was shortened using the one-parameter logistic (Rasch) model, EBIA psychometric behavior was analyzed with respect to acceptable adjustment values ranging from 0.7 to 1.3, and to severity scores of the items with theoretically expected gradients.

Food intake was determined by a food frequency questionnaire with seven foods or food groups (fruits, leafy vegetables, other vegetables, deep-fried food, sausages, soda, artificial juice) and their respective intake frequencies (never; 1-2 times a month, 1-2 times week, 3-6 times/week, and daily). For the analysis, for each item, recommended frequencies in a healthy diet were set, then, for fruit, leafy vegetables and other vegetables consumption was dichotomized into daily and less than 7 times a week. Additionally, the study investigated how often supper was replaced by a snack (never; at least once a week); and whether the participants frequented the soup kitchen Restaurante Popular -Bom Prato (yes; no).

Weight and height were measured according to procedures recommended by Lohman (18) and body mass index (BMI) was calculated (kg/m2). Nutritional status was classified using BMI cutoffs of the World Health Organization: Underweight (BMI <18.5kg/m2), Normal weight (BMI 18.5kg/m2 to 24.9kg/m2), Overweight (BMI ≥25kg/m2 to 30 kg/m2), and Obesity (BMI ≥30 kg/m2) (19).

Waist circumference was measured with a tape measure on an imaginary horizontal line passing through the umbilicus, according to Heyward procedures (20) and the parameters of the IV Brazilian Guideline for Dyslipidemia and Atherosclerosis Prevention (21) cutoffs were used, values greater than 94 centimeters in men and greater than 80 in women, are indicative of increased cardiovascular risk.

Statistical analysis

The data analysis was performed using Stata version 9.1 software. In order to rule out high correlation, a matrix of correlation was elaborated with all variables using Spearman correlation coefficient. The dependent variable was the nutritional status, categorized as normal weight, overweight and obesity, being «eutrophic» considered the reference category. The nutritional status was described for each independent variable included in the study. Univariate multinomial regression was used to estimate Odds Ratio (OR) and 95% Confidence Interval (95%CI) to assess the existence of an association between nutritional status and each variable. The variables with p<0.20 in the univariate analysis were selected for a multiple multinomial regression model. Only the statistically significant variables (p<0.05) adjusted for sex, age and economic class remained in the final model.


The study included 452 subjects of whom, 60.6% were men, 44.5% were overweight and 21.7% were obese. Women and people aged 60-69 years had higher chance of obesity (Table 1).

Table 1 Frequency of overweight and obesity according to demographic and socioeconomic variables and Odds Ratio (OR) of overweight and obesity using normal weight as the reference category*

* OR calculated by univariate multinomial logistic regression; † 39 Did not inform. 

Nutritional status was also associated with health-related conditions, with a greater chance of obesity in those who reported poor or very poor health, did not leave home daily, could not walk 15 minutes without getting tired, had hypertension, arthrosis/arthritis or rheumatism, hospitalization in the last year and referred daily use of medication. Elderly with dyslipidemia, diabetes, and heart disease were more likely to be overweight or obese. It is worth mentioning that only eight elderly with overweight showed no increase in waist circumference (Table 2).

Table 2 Frequency of overweight and obesity according to lifestyle and health-related conditions and Odds Ratio (OR) of overweight and obesity using normal weight as the reference category*

* OR calculated by univariate multinomial logistic regression; † Not applied 

The intake pattern was very similar between normal weight, overweight and obese. There was a higher intake of deep-fried foods, sausages, soda and artificial juice among overweight subjects, but the difference was not statistically significant at the confidence interval of the odds ratio. Elderly nonusers of the soup kitchen restaurant and those who reported the replacement of supper by a snack once or twice per week had higher odds of obesity (Table 3). There was a higher frequency of economic class D + E among users of the soup kitchen restaurant (33.5% vs 17.7%), that is the soup kitchens users were poorer than elderly neighbors (data not shown). High correlation coefficients were observed between the following variables: class and income (0.432), hypertension and medicine use (0.498), obesity and waist circumference (0.548), consumption of legumes and vegetables (0.688).

Table 3 Frequency of overweight and obesity according to feeding behavior and food intake and Odds Ratio (OR) of overweight and obesity using normal weight as the reference category*

* OR calculated by univariate multinomial logistic regression; † 30 Did not inform

In the final model adjusted for sex, age and economic class, there was a greater chance of overweight among elderly who reported dyslipidemia; those who reported arthrosis/arthritis/rheumatism and that replaced supper by snacks in one or more times per week were more likely to be obese. Elderly who did not leave home daily and who had diabetes were more likely to be overweight and obese (Table 4).

Table 4 Final multinomial regression model for overweight and obesity, Odds Ratio (OR) adjusted for sex, age and economic class*

* OR calculated by multiple multinomial logistic regression; † Reference category: normal weight

Using additional cross-tabulation, it was observed how significant was the participation of soup kitchen users in the variables that remained in the final model presented in table 4. For variable ‘not leaving home daily’ there was significant difference just for the overweight category, among those in this condition 37.8% were soup kitchen users and 62.2% were neighbors (p=0.04). For variable ‘replace supper by a snack’ also there was significant difference for the overweight category, among those in this condition 46.7% were soup kitchen users and 53.3% were neighbors (p=0.01). For variable ‘arthrosis/arthritis/rheumatism’ there was significant difference for the obese category, among those in this condition 20% were soup kitchen users and 80% were neighbors (p<0.01). For variable ‘diabetes’ also there was significant difference for the obese category, among those in this condition 36.8% were soup kitchen users and 63.2% were neighbors (p=0.03). No significant difference was observed for dyslipidemia among soup kitchen users and neighbors in any category of nutritional status.


Similarly to other authors, there was a higher chance of obesity in women, and a decrease in both sexes with increasing age (22-25). Although there is loss of bone,  muscle mass, and appetite inherent to advancing age (26), we postulate the existence of the possibility of a survival bias, due to healthier diet and a more active life style, among elderly that maintain a normal weight and predominate in the older group.

Moreover, another hypothesis is that the negative impact of obesity and its associated pathologies may contribute to higher mortality in younger obese elderly, although the overall mortality is higher among malnourished elderly and in older age groups (23).

The economic status was inversely associated with overweight and obesity. Monteiro et al. (11), in a review article, claim that the lack of food and the common high energy expenditure among the poorest may explain this protection against obesity in groups of lower socioeconomic status. Despite this, it has already been pointed out changes in the epidemiology of obesity, it has been reported an increase in groups of lower socioeconomic status (27). Although the association between overweight and food insecurity was not significant in our study, several studies have shown this association, especially in females (28-30). The overweight among members of food insecure households is related to the monotony of the diet, rich in simple carbohydrates and sugars, high in fats, and usually low cost foods, that are widely used as a strategy to resist to food insecurity (31). In our study, the fruits and meats intake was lower among food insecure elderly than among food secure ones (32).

Compared with users of the soup kitchen Bom Prato, the neighbors were 20% more likely to be obese. It must be considered that the economic status of the neighbors was better than the soup kitchen users, and this condition may be associated with obesity (11). In addition, eating at the soup kitchen can provide better nutrition due to its balanced and healthy menus. It also must be stated that to get to the soup kitchen located in the city center, walking is necessary and thus these elderly have energy expenditure.

Several categories of fat and soft drinks intake were tested, but no significant differences according to nutritional status was observed. Due to the high frequency of obesity in patients with chronic diseases, it is possible that these people have received guidance regarding obesity control, and that their responses incorporate healthy diet, because in fact they are following it or at least trying to. It is worth noting that in general, the quality of the diet of elderly independent of nutritional status, was inadequate, both for the high frequency of fats, sausages and soda consumption, and the low proportion of daily consumption of vegetables and fruits.

A greater chance of obesity among those that replaced supper by a snack, points out the importance of this behavior on nutritional status, thus it may be considered a synthetic indicator of diet quality, because it proved more significant than the consumption of fruits, vegetables and beans. Many elderly had instant noodles for supper, referring to ease of preparation and low price. This behavior reflects the action of transnational companies that control the production and distribution of ultra-processed foods throughout the world, foods which are high in carbohydrates, refined-grains, fats and additives (33). National surveys report that the traditional Brazilian habit of consuming rice and beans has decreased (34-36), as a result of increased consumption of processed foods (34, 37). Elderly traditionally ate rice and beans, and some legume or vegetable, this food combination is considered nutritionally adequate, gives satiety and is relatively cheap. The replacement of supper by snacks, when made, must have nutritionally adequate foods that address the nutritional needs of elderly, but this is more expensive than traditional diet and thus among poor people the affordable snacks options are obesogenic and cheap. These behavioral changes can cause major impact on public health by increasing the incidence of obesity and chronic diseases, and affect food culture, national identity and local economy (33).

Elderly people with poor or very poor health and hospitalization in the last year, similar to the study of Barreto et al. (24) had increased chance of obesity. Also as in other studies, there was a higher chance of obesity among elderly people with physical disability, a condition that in this study was indirectly estimated by fatigue before 15 minute walk. Worse performance on walks was described between Chinese obese (BMI ≥ 30 kg/m²), age 65 years or more (22). In the Netherlands and in Brazil there was an association between physical inactivity and obesity (24, 38).

Walls et al. (39), suggest that the isolated use of BMI may underestimate the effects of obesity on health, and they advise the complementary use of waist circumference. Likewise, Mota et al. (40) state that the WC presents itself as the best risk marker for metabolic abnormalities associated with obesity and Testa et al. (41) observed a direct association between WC and mortality, with a higher mortality from coronary heart disease. In the present study, all elderly with obesity had increased waist circumference, this being indicative of high risk for mortality, especially in those with heart disease. We must point out that over half of the normal weight also had raised waist circumference, confirming that the use of the BMI isolated could underestimate the association of overweight and some independent variables.

The lack of physical activity, indicated in our study by the category «not leaving home daily», beyond the higher chance of overweight and obesity observed, can promote the loss of muscle mass and increase the risk of sarcopenia (42). This lack of physical activity would be an early indicator of frailty. Therefore, in the elderly it is fundamental to encourage physical activity, however, urban security must be considered to practice this activity, whether for purposes of leisure, exercising or active transport (43-45).

Osteoarthritis, dyslipidemia and diabetes are associated with overweight and obesity, similar to other studies (24, 46), existing common metabolic mechanisms that explain the association (47). In an article review on the effects of obesity, Walls et al. (48) suggest that its impact is greater in life expectancy free of disability than in life expectancy itself, representing future challenges for the health system.

In the present study, because it is a cross-sectional study we cannot point out the direction of causality, but in the clinical setting, obesity is considered a risk for hypertension, diabetes and dyslipidemia and weight loss helps to lower blood pressure levels, blood glucose and lipids. Likewise, considering that this is not a population-based study, a limitation is the inability to estimate the representative prevalence for the elderly population of the city. Moreover, the diseases here, although self-reported required the diagnosis of a health professional, so it depended on the access to health services for diagnosis, thus the likelihood of diagnosis of diseases decreased among residents in areas with smaller supply of health services of adequate quality. Given that the sample size of normal weight with normal waist circumference, was relatively small (n=73), and also that no biochemical analysis were available, it could not be justify the inclusion of normal weight with raised waist circumference in the overweight group.  Thus, it is possible that there has been underestimation of the association of overweight with some independent variables.

The participation of soup kitchen users might have made evident some variables due to special characteristics of this group, even though this variable did not remain in the final model. Although the soup kitchen users from the social point of view are vulnerable because most of them live alone, and also have worse economic condition, it would be expected that they would have poor eating habits, and so a worse nutritional status, but from the physical point of view they are supposed to have a better functional status. Comparing soup kitchen users with neighbors, in each of the three nutritional status categories, no significant difference in the percent that got tiered walking 15 minutes or less was observed. The only difference between these groups was among those that do not leave home daily and get tired of walking 15 minutes or less, the frequency of this condition was significantly higher among neighbors (40%) than among soup kitchen users (23.7%). In relation to the variables that remained in the final model, the frequency of neighbors was significantly higher than soup kitchen users for overweight or obesity but not both for all variables except dyslipidemia. Thus, although there really exist differences among soup kitchen users and neighbors, they are not totally consistent and of a very big magnitude.

The strength of this study is that it points that in Brazil, among a population that is mainly not fragile because of the selection criteria (no bedridden, no wheelchair users, no problem for standing up in a right position and no undernutrition), the changes in traditional dietary habits, evaluated through the replacement of supper by snacks increased the chance of obesity. It also corroborates findings already described by other authors in association to overweight and obesity, such as dyslipidemia, chronic osteoarticular diseases, diabetes and the lack of activity indirectly estimated here with the variable ‘not leaving home daily’.

The results of this study suggest the association of overweight and obesity with lifestyle patterns and health-related conditions as sedentary lifestyle, chronic diseases, including arthrosis/arthritis/rheumatism, diabetes and dyslipidemia, as well as changes in eating behaviors. A large population of obese elderly with chronic diseases represents a major burden for the health system. Thus, public health policy should encourage regular physical activity and promote healthy eating behaviors, focusing on traditional diet through nutritional education starting in childhood, to decrease the prevalence of overweight, obesity and associated chronic diseases in the old age.

Contributors: BFNJ SOUZA wrote this article and performed the statistical analysis. L MARÍN-LEÓN guided and reviewed all sections and versions of the article.

Funding: We thank Fundação de Amparo à Pesquisa do Estado de São Paulo for sponsoring the study (Process nº 2010/51185-2). We thank all the participants that kindly shared their experiences with us.

Declaration of conflicting interests: The authors declared no potential conflicts of interest with respect to the research and publication of this article.

Ethical standards: All study participants were individually informed about the study and interviewed only after they signed a free and informed consent form. The project was approved by the Research Ethics Committee of the School of Medical Sciences of the University of Campinas in April 2010 (CEP nº 169/2009).


1. Popkin BM. The nutrition transition and obesity in the developing world. J Nutr 2001; 131(3):871S-3S.

2. Bloom DE, Boersch-Supan A, McGee P, Seike A. Population aging: facts, challenges, and responses. Program on the Global Demography og Aging. Working Paper Series. PGDA. Working Paper N. 71. 2011.

3. Houston DK, Nicklas BJ, Zizza CA. Weighty concerns: the growing prevalence of obesity among older adults. J Am Diet Assoc 2009; 109(11):1886-95. doi: 10.1016/j.jada.2009.08.014

4. Ministério da Saúde. Informações de Saúde. População residente – Brasil. [Accessed  May 6 2014] . Available in: http://tabnet.datasus.gov.br/cgi/tabcgi.exe?ibge/cnv/popuf.def 

5. Ministério da Saúde. DATASUS. Informações de Saúde. Indicadores de Saúde. Indicadores de Dados Básicos – IDB 2010. Indicadores demográficos. População residente por faixa etária segundo regional de saúde/município. Município: Campinas.[Accessed May 6 2014]. Available in: http://tabnet.datasus.gov.br/cgi/tabcgi.exe?ibge/cnv/popsp.def

6. Flegal KM, Graubard BI, Williamson DF, Gail MH. Cause-specific excess deaths associated with underweight, overweight, and obesity. JAMA. 2007 Nov 7; 298(17):2028-37. doi:10.1001/jama.298.17.2028

7. Locher JL, Roth DL, Ritchie CS, Cox K, Sawyer P, Bodner EV, et al. Body Mass Index, Weight Loss, and Mortality in Community-Dwelling Older Adults. J Gerontol A Biol Sci Med Sci. 2007; 62(12): 1389–1392. 

8. Ferreira LS, Amaral TF, Marucci MF, Nascimento LF, Lebrão ML, Duarte YA. Undernutrition as a major risk factor for death among older Brazilian adults in the community-dwelling setting: SABE survey. Nutrition 2011; 27(10): 1017-22. doi: 10.1016/j.nut.2010.11.008

9. Alexandre TS, Duarte YAO, Santos JLF, Wong R, Lebrão ML. Sarcopenia according to the European working group on sarcopenia in older people (EWGSOP) versus dynapenia as a risk factor for mortality in the elderly. J Nutr Health Aging 2014;18(5):547-53. doi: 10.1007/s12603-013-0424-x.

10. Chambers JA, Swanson V. A health assessment tool for multiple risk factors for obesity: age and sex differences in the prediction of body mass index. Br J Nutr 2010; 104(02):298-307. doi: 10.1017/S0007114510000607

11. Monteiro CA, Moura EC, Conde WL, Popkin BM. Socioeconomic status and obesity in adult populations of developing countries: a review. Bulletin of the World Health Organization. 2004(a); 82:940-6.

12. Kim ILH, Chun H, Kwon J-W. Gender differences in the effect of obesity on chronic diseases among the elderly Koreans. J Korean Med Sci 2011; 26(2):250-7. doi: 10.3346/jkms.2011.26.2.250

13. Al Snih S, Graham JE, Kuo Y-F, Goodwin JS, Markides KS, Ottenbacher KJ. Obesity and disability: relation among older adults living in Latin America and the Caribbean. Am J Epidemiol  2010; 171(12):1282-8. doi: 10.1093/aje/kwq087

14. Brasil. Presidência da República Federativa do Brasil. Fome Zero. Programas e ações. Acesso à Alimentação. [Accessed Oct 14 2011]. Available in: http://www.fomezero.gov.br/estrutura/estrutura/acesso-a-alimentacao

15. Brasil. Ministério do Desenvolvimento Social e Combate à Fome. Restaurantes populares. [Accessed Oct 14 2011]. Available in: http://www.mds.gov.br/segurancaalimentar/equipamentos/restaurantespopulares

16. ABEP. Associação Brasileira de Empresas de Pesquisa. Critério de Classificação Econômica Brasil. 2011. [Accessed Aug 5 2011]. Available in: http://www.abep.org/novo/Utils/FileGenerate.ashx?id=46

17. Marín-León L, Segal-Corrêa AM, Panigassi G, Maranha LK, Sampaio MdFA, Pérez-Escamilla R. Food insecurity perception in families with elderly in Campinas, São Paulo, Brazil. Cad Saúde Pública 2005; 21:1433-40. doi: http://dx.doi.org/10.1590/S0102-311X2005000500016  

18. Lohman T, Roche A, Martorell R. Anthopometric Standardization Reference Manual. Champaign, IL: Human Kinetics; 1988.

19. World Health Organization. The use and interpretation of anthropometry: report of a WHO expert committee. Geneva: WHO1995.

20. Heyward V, Wagner D. Applied body composition assessment. 2nd ed. Champaign, IL: Human Kinetics; 2004.

21. SBC. Sociedade Brasileira de Cardiologia. IV Diretriz Brasileira de Dislipidemias e Prevenção da Aterosclerose. Departamento de Aterosclerose da Sociedade Brasileira de Cardiologia. Arq Bras Cardiol 2007; 88(Supl.1).

22. Woo J, Leung J, Kwok T. BMI, body composition, and physical functioning in older adults. Obesity 2007; 15(7):1886-94.

23. Silveira EA, Kac G, Barbosa LS. Obesity prevalence and associated factors in the elderly in Pelotas, Rio Grande do Sul State, Brazil: obesity classification according to two cutoff points for body mass índex. Cad Saúde Pública 2009; 25:1569-77. doi: http://dx.doi.org/10.1590/S0102-311X2009000700015 

24. Barreto SM, Passos VMA, Lima-Costa MFF. Obesity and underweight among Brazilian elderly: the Bambuí Health and Aging Study. Cad Saúde Pública 2003; 19:605-12. doi: http://dx.doi.org/10.1590/S0102-311X2003000200027 

25.  Andrade FBd, Caldas Junior AF, Kitoko PM, Batista JEM, Andrade TB. Prevalence of overweight and obesity in elderly people from Vitória-ES, Brazil. Cien Saude Colet 2012; 17:749-56. doi: http://dx.doi.org/10.1590/S1413-81232012000300022.

26. Brownie S. Why the elderly individuals at risk of nutritional deficiency? Int J Nurs Pract 2006; 12:110-118.

27. Monteiro CA, Conde WL, Popkin BM. The burden of disease from undernutrition and overnutrition in countries undergoing rapid nutrition transition: a view from Brazil. Am J Public Health 2004(b); 94(3):433-4.

28. Adams EJ, Grummer-Strawn L, Chavez G. Food insecurity is associated with increased risk of obesity in California women. J Nutr 2003; 133(4):1070-4.

29. Wilde PE, Peterman JN. Individual weight change is associated with household food security status. J Nutr 2006; 136(5):1395-400.

30. Mohammadi F, Omidvar N, Harrison GG, Ghazi-Tabatabaei M, Abdollahi M, Houshiar-Rad A et al. Is Household Food Insecurity Associated with Overweight/Obesity in Women? Iranian J Publ Health 2013, 42 (4): 380-390.

31. Panigassi G, Segall-Corrêa AM, Marin-León L, Pérez-Escamilla R, Maranha LK, Sampaio MdFA. Intra-family food insecurity and profile of food consumption. Rev Nutr 2008; 21:135s-44s. doi: http://dx.doi.org/10.1590/S1415-52732008000700012  

32. Souza BFNJ, Marin-Leon L. Food insecurity among the elderly: cross-sectional study with soup kitchen users. Rev Nutr. 2013; 26 (6): 679-691. doi: http://dx.doi.org/10.1590/S1415-52732013000600007  

33.  Monteiro CA, Cannon G. The impact of transnational “Big Food” Companies on the South: a view from Brazil. PLoS Med 2012; 9(7):e1001252. doi:10.1371/journal.pmed.1001252

34. Schmidt MI, Duncan BB, e Silva GA, Menezes AM, Monteiro CA, Barreto SM, et al. Chronic non-communicable diseases in Brazil: burden and current challenges. Lancet 2011; 377(9781):1949-61. doi: http://dx.doi.org/10. 1016/S0140-6736(11)60135-9

35. IBGE. Instituto Brasileiro de Geografia e Estatística. Pesquisa de Orçamentos Familiares 2008-2009. Aquisição alimentar domiciliar per capita. Brasil e grandes regiões. Rio de Janeiro; 2010.

36. IBGE. Instituto Brasileiro de Geografia e Estatística. Pesquisa de Orçamentos Familiares 2002-2003. Primeiros resultados. Brasil e grandes regiões. Rio de Janeiro; 2004.

37. IBGE. Instituto Brasileiro de Geografia e Estatística. Brasileiro come menos arroz com feijão e mais comida industrializada em casa. Comunicação Social. 16 de dezembro de 2010.  Rio de Janeiro; 2010. [Accessed Jul 26 2012]. Available in: http://www.ibge.gov.br/home/presidencia/noticias/noticia_visualiza.php?id_noticia=1788&id_pagina=1

38. Walter S, Kunst A, Mackenbach J, Hofman A, Tiemeier H. Mortality and disability: the effect of overweight and obesity. Int J Obes 2009; 33(12):1410-8. doi: 10.1038/ijo.2009.176.

39. Walls HL, Stevenson CE, Mannan HR, Abdullah A, Reid CM, McNeil JJ, et al. Comparing trends in BMI and Waist Circumference. Obesity 2011; 19(1):216-9. doi: 10.1038/oby.2010.149

40. Mota JF, Rinaldi AEM, Pereira AF, Orsatti FL, Burini RC. Anthropometric indicators as risk markers for metabolic abnormalities. Cien Saude Colet 2011; 16:3901-8. doi: http://dx.doi.org/10.1590/S1413-8123201100100002

41. Testa G, Cacciatore F, Galizia G, Della-Morte D, Mazzella F, Langellotto A, et al. Waist Circumference but not Body Mass Index predicts long-term mortality in elderly subjects with chronic heart failure. J Am Geriatr Soc 2010; 58(8):1433-40. doi: 10.1111/j.1532-5415.2010.02979.x

42. Hirani V. Generalised and abdominal adiposity are important risk factors for chronic disease in older people: results from a nationally representative survey. J Nutr Health Aging. 2011;15(6):469-78.

43. Corseuil MW, Hallal PC, Corseuil HX, Schneider IJC, d’Orsi E. Safety from crime and physical activity among older adults: a population-based study in Brazil. J Environ Public Health 2012; 12:1-7. doi: http://dx.doi.org/10.1155/2012/641010

44. Belon AP, Nieuwendyk LM, Vallianatos H, Nykiforuk CIJ. How community environment shapes physical activity: Perceptions revealed through the PhotoVoice method. Soc Sci Med 2014; 116:10-21. doi: 10.1016/j.socscimed.2014.06.027

45. Gallagher NA, Clarke PJ, Gretebeck KA. Gender Differences in Neighborhood Walking in Older Adults J Aging Health 2014 26: 1280-1300. doi:10.1177/0898264314532686

46. Machado GPM, Barreto SM, Passos VMA, Lima-Costa MFF. Health and aging study: prevalence of chronic joint symptoms among the elderly in Bambui. Rev Assoc Med Bras 2004; 50:367-72. doi: http://dx.doi.org/10.1590/S0104-42302004000400024  

47. Billington CJ, Epstein LH, Goodwin NJ, Hill JO, Pi-Sunyer FX, Rolls BJ et al. National Task Force on the Prevention and Treatment of Obesity. Overweight, obesity, and health risk. Arch Intern Med. 2000; 160 (7): 898-904.

48. Walls HL, Backholer K, Proietto J, McNeil JJ. Obesity and trends in life expectancy. J Obesity 2012; 2012:1-4. doi: http://dx.doi.org/10.1155/2012/107989



1. Department of Biology. Universidad Autónoma de Madrid, Madrid, Spain.

Corresponding author: Santiago Rodríguez López. Calle Darwin 2. Departamento de Biología, Universidad Autónoma de Madrid. Campus Universitario de Cantoblanco. 28049. Madrid, Spain. Phone: +34 914972561 – Fax: +34 914978638. e-mail: santiago.rodriguez@uam.es

J Frailty Aging 2014;3(2):120-125
Published online December 8, 2014, http://dx.doi.org/10.14283/jfa.2014.12


Background: Evaluate how obesity is associated with the development of frailty among older adults is important. However, few studies have examined the relation between obesity and frailty within different educational backgrounds. Objectives: This study aims to investigate the association between educational level and frailty and to evaluate whether obesity explains any possible associations among Spanish adults. Design, participants and settings: This is a cross-sectional study including 2,319 50-years-old and older community- dwelling Spanish adults, who participated in the first wave (2004/05) of the Survey of Health, Ageing and Retirement in Europe (SHARE). Measurements: Educational differences in frailty phenotypes –defined by the SHARE’s operationalized criterion– and their association with obesity –estimated through self-reports of weight and height– were evaluated using multinomial logistic regression analyses. Results: Women experienced frailty in a larger proportion than men (22.3% vs. 13.3%). After adjusting for all confounders, we found a marked educational gradient in frailty, where individuals with non-formal education showed increased odds of a frailty phenotype than individuals with higher education. Moreover, obesity was significantly related to frailty and the effect of obesity is similar at all levels of education after testing for interaction effects. Although there is a mediation effect of obesity, the educational gradient in frailty is robust to controls for obesity. Conclusions: Our findings suggest a somehow independent effect of both educational background and obesity on frailty among Spanish individuals. This adds to the evidence of the frailty-obesity association among different educational backgrounds, and has implications for future interventions leading to reduce health disparities in elders.

Key words: Frailty, education, obesity, SHARE, Spain.


Frailty is a common condition in older persons and has been described as a geriatric syndrome resulting from age-related cumulative declines across multiple physiological systems (1) and a good predictor of disability among older adults (2). It has been recently suggested that probably the most common phenotype of frailty in the near future will be characterized by the concurrent and interacting presence of obesity (3). Thus, the evaluation of the role of obesity in the development of frailty among older adults is one of the main gaps and future directions in frailty research (4).

Evaluating the importance of health behaviors such as obesity in the magnitude of health inequalities in older adults is important, as risk factors are unequally distributed among the social classes and serve as potential pathways through which socioeconomic status (SES) may influence adult health (5). Whereas some studies conclude that health behaviors make a relatively minor contribution to the social gradient in health (6), others have shown that differences in lifestyles can explain a relevant part of health inequalities (7-8). Previous studies have examined associations between SES and frailty (9, 10-12), while others investigating the relationship between obesity and frailty have been comparatively minor (13-14). However, few studies have examined how obesity is related to frailty within different educational backgrounds.

Growing evidence suggests that obesity can exacerbate the age-related decline in physical function, which potentially causes frailty (15). Body mass index (BMI) may change in parallel with the development of frailty, and assessing how it relates –mainly through obesity– to frailty states among individuals within different SE backgrounds i relevant to frailty prevention/delay. This is particularly important in a country like Spain, where one of the largest life expectancies in Europe coexist with high levels of frailty, disability and obesity. This health paradox exists together with a great proportion of poorly educated older adults.

Within this context, the present study aims to i) identify potential educational inequalities in the prevalence of frailty phenotypes in Spanish adults and ii) evaluate whether obesity explains any possible association. Our study is appropriate by examining potential educational inequalities in frailty among the Spanish population, which exhibits a large proportion of older adults with poor educational background and one of the highest prevalences of disability and obesity within European countries.




This is a cross-sectional study, focusing on 50-years-old and older Spanish individuals who participated in the first wave (W1) of the Survey of Health, Ageing and Retirement in Europe (SHARE), carried out in 2004/05. SHARE is a multidisciplinary cross-country longitudinal survey providing micro data on health, SES, and retirement of older Europeans. The details of the data collection and sampling procedures have been previously described (16). The response rate reached up to 53.3% for Spain (17). We included a relatively young population in order to assess pre-frailty in addition to frailty phenotypes. Originally, 30,816 individuals participated in W1 of SHARE. The Spanish subsample in W1 includes 2,396 participants. After excluding individuals with missing data for demographic characteristics and frailty states, our sample comprises 2,319 individuals (mean age 66.8 years old): 979 men (42.2%) and 1,340 women (57.8%).


Dependent variable

Frailty phenotype was used as the dependent variable in this study. We used the SHARE’s definition of frailty described by Santos-Eggimann et al. (18), based on the following five conditions:

  1. Exhaustion: identified as a positive response to the question: ‘In the last month, have you had too little energy to do the things you wanted to do?’. A positive answer was recoded as 1, and a negative as 0.
  2. Weight loss: this criterion was fulfilled by reporting a ‘diminution in desire for food’ in response to the question: ‘What has your appetite been like?’ or, in the case of a non- specific or uncodeable response to this question, by responding ‘less’ to the question: ‘So, have you been eating more or less than usual?’. The presence of this criterion was coded as 1 and its absence as 0.
  3. Weakness: assessed by handgrip strength (kg) using a dynamometer. Weakness was considered according to the cut- off points for grip strength criterion for frailty, stratified by gender and BMI (19).
  4. Slowness: defined as a positive answer to either of the following two questions: ‘Because of a health problem, do you experience difficulty (expected to last more than 3 months) walking 100 metres?’ or ‘… climbing one flight of stairs without resting?’ One or two positive answers received the score of 1, and two negative answers received the score of 0.
  5. Low activity: assessed by the question: ‘How often do you engage in activities that require a low or moderate level of energy such as gardening, cleaning the car, or going for a walk?’. Four possible answers were considered: 1=‘More than once a week’; 2=‘Once a week’; 3=‘One to three times a month’ and 4=‘Hardly ever or never’. Low activity criterion was considered when a response of 3 or 4 was obtained.

The presence of frailty was defined following Fried’s definition (20). An individual is considered frail when having at least three out of the five components of the frailty index. An intermediate or pre-frail phenotype corresponds to the presence of one or two of the conditions mentioned above. Previous

studies (21) have validated the adapted criteria proposed by Santos-Eggimann (18). Among all the items available in SHARE, the above choice of variables is the closest possible selection to the original variables in Fried’s phenotype. However, we acknowledge two significant departures from Fried’s theoretical framework, such as ‘weight loss’ (replaced by appetite) and ‘slowness’ (measured by questions on functional limitation) (22).


Sociodemographic characteristics are summarized by gender, age, and level of education (23). Education has been recoded into four broad groups: none (no formal education, without distinguishing illiterate from non-illiterate individuals), low (primary school), medium (high school) and high (university studies).

Health variables and health behaviors

General health variables are represented by i) self-reported health (SRH), coded as a five-point continuous variable with 1=excellent and 5=poor; ii) the sum of self-reports of chronic conditions diagnosed by a physician (range 0-9); iii) the number of limitations in basic (ADL; range 0-6) and iv) instrumental (IADL; range 0-7) activities of daily living.

Health-related behaviors are represented by smoking, alcohol consumption and BMI. Smoking was categorized as 1=current/former vs. 0=never. Alcohol consumption 1=drinking more than 2 glasses of any alcoholic beverage 5/6 days a week, vs. 0=other. BMI was estimated from self-reports of weight and height and recoded into four groups: underweight (BMI<18.5 kg/m2), normal weight (18.5-24.9 kg/m2), overweight (25.0-29.9 kg/m2) and obese (BMI>30.0 kg/m2).


Statistical analyses

Descriptive statistics were performed stratified by gender. Associations were tested between frailty phenotypes and sociodemographic characteristics, health variables and health- related behaviors. Multinomial logistic regression was used to analyze odds ratios (ORs) of pre-frail and frail phenotypes (outcomes) as a function of covariates. We estimated four different models: Model 1: unadjusted, including level of education and BMI; Model 2 adds gender and age to Model 1; Model 3 adds health variables and health-related behaviors to Model 1, but excluding BMI. Finally, Model 4 adds BMI to Model 3. The comparison of Models 4 and 3 demonstrates the confounding influence of BMI in the association between education and frailty, after adjusting for health variables and health-related behaviors. In addition to main effects, we also tested for interaction effects between level of education and categories of BMI, in order to show whether the effect of obesity is similar or varies at different levels of education.

We included ADL and IADL in order to control for disability. This allowed us to obtain disabled-free frailty outcomes, after controlling for the effect of physically impaired people. Even though from a conceptual point of view this is probably not the best way to control for disability –as frailty is supposed to increase its risk– we preferred this rather than excluding those disabled individuals from our sample.



Compared to men, a larger proportion of women show frailty (22.3% vs. 13.3%) and pre-frailty phenotypes (50.4% vs. 46.5%, respectively) (χ2=55.6; p<0.001; data not shown). Table 1 shows the distribution of the study characteristics by frailty phenotypes and stratified by gender. Compared to those who are non-frail, pre-frail and frail individuals are older, have fewer years of education –a larger proportion of them have a poorer educational level– more chronic conditions, ADL and IADL, and poorer SRH. On the other hand, they seem to have relatively healthier behaviors related to smoking and alcohol consumption, more apparent among women.

Figure 1 shows the distribution of frailty scores by categories of BMI and stratified by educational background. Both underweight and obesity are associated with higher frailty scores, whereas there is an educational gradient in frailty scores.

Table 2 describes the ORs of both pre-frail and frailty phenotypes by educational level and BMI. The crude analyses show a marked educational gradient in pre-frail and frailty phenotypes (Model 1). After adjusting for gender and age (Model 2) the associations are markedly attenuated, but remain statistically significant. Adjusting for possible confounders (Model 3) and BMI (Model 4) attenuate the associations for the pre-frail phenotype to a non-significant statistical level, while a poorer education stays significantly associated to frailty phenotype in the fully adjusted model (Model 4).

The analysis for interaction effects between level of education and categories of BMI was not statistically significant, indicating a homogenous effect of obesity at all levels of education (data not shown). We found that obesity remains associated to the frailty phenotype in the fully adjusted model (Model 4), while such association was attenuated for the pre-frailty phenotype, after adjusting for all confounders. On the other hand, compared to individuals with normal weight, overweighed people seemed to have a lower risk associated to pre-frailty and frailty phenotypes, although not significantly.

Table 1: Study characteristics by frailty states and stratified by gender

a χ2 and ANOVA/t-test or non-parametric analyses for each particular case; ADL: limitations in basic activities of daily living; IADL: limitations in instrumental activities of daily living; SRH: self-reported health; BMI: Body Mass Index; Alcohol consumption: drinking more than 2 glasses of any alcoholic beverage 5/6 days a week.


Table 2: ORs (95% CI) for pre-frailty and frailty states by educational level and BMI

Non-frail participants (not shown) are considered as the referent category. Model 1: crude odds ratios for the association between level of education and BMI, respectively, with frailty phenotype; Model 2: Adds gender and age to Model 1; Model 3: Adjusted for gender, age, level of education, number of chronic conditions, ADL, IADL, SRH, smoke and alcohol consumption; Model 4: Adjusted for gender, age, level of education, number of chronic conditions, ADL, IADL, SRH, smoke and alcohol consumption and BMI. 


Figure 1: Frailty score (range 0 – 5), body mass index and educational background in the study sample. Low, medium and high educations correspond to primary, high school and university studies, respectively




Our purpose was to assess potential educational differences in frailty phenotype in 50-year-old and older community-dwelling Spanish adults, and to evaluate to what extent BMI –mainly through obesity– could explain such an association.

We found marked educational differences in pre-frail and frailty phenotypes: those with a lower educational background showed increasing likelihood of being frail, even after adjusting for all confounders, including obesity. Furthermore, the effect of obesity is similar at all levels of education after testing for interaction effects. Although there is a mediation effect of obesity, the educational gradient in frailty is robust to controls for obesity, which may suggest a somehow independent effect of both educational background and obesity on frailty individuals.

In line with previous studies (9, 18) we found that, compared to men, a larger proportion of women showed pre-frail and frailty. It has been hypothesized that the higher incidence of frailty among women may be partly due to the marked gender roles still present in older people, for which most women had a restrained social life and little economic independence (24). Moreover, we found educational differences in frailty in our sample. Adjusting for confounders attenuated the associations for the pre-frailty phenotype. In the case of Spain, the large educational differences seem to make education a good indicator of SES that may reflect living conditions, life chances, etc., associated to differences in health (25). Thus, it was expected that these inequalities were differently related to frailty phenotypes among men and women. This finding is in line with general evidence that less educated persons are at increased risk of experiencing frailty (9-10). Although previous research suggests that education contributes to differences in frailty and pre-frailty in most European countries (18), it is important to establish whether such associations are consistently found or are specific to certain countries.

The association between BMI and frailty has been previously described as a U-shaped curve (14): frailty scores are lowest in those with normal weight and overweight individuals, and higher in those with underweight and in those with obesity. We found similar results in the case of Spain (Figure 1), although the low proportion of low BMI (underweight) in the sample conditions the measurement of frailty (0% for most groups, and only 4% in the frail group). However, the low proportion of underweight may indicate that weight loss is not a big issue in this population.

Obesity was associated with pre-frail and frailty phenotypes, and these associations remained also for frailty after controlling for sociodemographic characteristics, health-related variables and health behaviors. Similar findings were previously reported (13). In spite of the fact that there may be differences in findings between our study and others due to both the use of different methods for measuring frailty and the adults’ age considered, obesity also seems to be associated with the frailty syndrome in other cross-sectional studies (14). Similar findings were observed in another study where obesity was related to frailty in women but not in men, suggesting that differential exposure and vulnerability over the life course may partially explain differences between men and women (10). Our results suggest that in Spain obesity is similarly associated to different levels of education, which made the contribution of the former to the odds of pre-frailty and frailty somehow independent of the educational background.

Several limitations need to be highlighted. Probably the most important limitation of this study is the use of self-reports of weight and height to estimate BMI. Despite the large amount of evidence regarding the appropriateness of using self-reported data to estimate nutritional status (26), its use might result in an underestimation of BMI, since Spanish adults seem to over- report higher heights than other Europeans (27). Assuming this as true, our findings related to the association between obesity and frailty could be hampered, but in any case they may be underestimated. Similarly, SHARE’s definitions of frailty (18) rely considerably on self-report data, which forms the basis for four out of five items (i.e. all except handgrip strength). The differences between the definition of frailty used in this study and Fried’s criteria have already been described (18) and there is no guarantee that ours actually measures the same as Fried’s definition of frailty. Consequently, our findings may not be totally comparable to others using Fried’s criteria, although recent evidence support the use of SHARE’s operationalized frailty approach (21, 28).

Moreover, the cross-sectional design does not allow us to treat the factors considered as causal, as they may simply occur together with frailty phenotypes. More prospective studies are needed to elucidate the specific role of obesity in order to prevent the onset and development of frailty, particularly among poorly educated women. This could have a positive impact on their quality of life, by reducing the incidence of disabilities and the period of dependence near the end of life. Encouragement of older men and women to maintain or increase their regular exercise can help to lose weight (1) and reduce obesity, which in turn may help prevent the onset of frailty. Despite the limitation of our data, our study provides evidence on how obesity and educational differences are associated in frailty, leading towards curbing existing trends in health inequalities in the adult Spanish population. Further studies in this area, especially in countries where SE inequalities in health have been examined less –including Spain– are therefore justified.

In conclusion, we have evaluated the association between educational attainment and frailty in Spanish adults and whether obesity could explain such association. We found a graded educational inequality associated to pre-frail and frail phenotypes: those with lower educational background showed an increasing likelihood of being frail, even after adjusting for all confounders, including obesity. The fact that the influence of obesity was similar at all levels of education suggests an independent effect of both education and obesity on frailty. These findings add to the evidence of the frailty-obesity association among different educational backgrounds and have implications for future interventions leading to reduce health disparities among older adults. The association of obesity with higher odds of frailty supports the notion that for preventing or retarding frailty, benefits can be obtained by promoting regular exercise among older individuals.


Acknowledgements: This paper uses data from SHARE wave 4 release 1, as of November 30th 2012 or SHARE wave 1 and 2 release 2.5.0, as of May 24th 2011 or SHARELIFE release 1, as of November 24th 2010. The SHARE data collection has been primarily funded by the European Commission through the 5th Framework Programme (project QLK6-CT-2001-00360 in the thematic programme Quality of Life), through the 6th Framework Programme (projects SHARE-I3, RII-CT-2006-062193, COMPARE, CIT5-CT-2005-028857, and SHARELIFE, CIT4-CT-2006-028812) and through the 7th Framework Programme (SHARE-PREP, N° 211909, SHARE-LEAP, N° 227822 and SHARE M4, N° 261982). Additional funding from the U.S. National Institute on Aging (U01 AG09740-13S2, P01 AG005842, P01 AG08291, P30 AG12815, R21 AG025169, Y1-AG-4553-01, IAG BSR06-11 and OGHA 04-064) and the German Ministry of Education and Research as well as from various national sources is gratefully acknowledged (see www.share-project.org for a full list of funding institutions). The sponsors had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; in the preparation of the manuscript; or in the review or approval of the manuscript.

Conflict of interest statement: Authors declare no conflict of interest on this paper.



1. Landi F, Abbatecola AM, Provinciali M, et al. Moving against frailty: Does physical activity matter? Biogerontology 2010;11:537-545.

2. Vermeulen J, Neyens, JCL, van Rossum, E, Spreeuwenberg, MD, de Witte, LP. Predicting ADL disability in community-dwelling elderly people using physical frailty indicators: A systematic review. BMC Geriatr 2011;11:33.

3. Shah K, Villarreal DT. Preventing frailty in obese older adults. J Frailty & Aging 2012;2:47-49.

4. Mohandas A, Reifsnyder J, Jacobs M, Fox T. Current and future directions in frailty research. Popul Health Manag 2011;14:277-283.

5. Dalstra J, Kunst A, Borrell C, et al. Socioeconomic differences in the prevalence of common chronic diseases: An overview of eight European countries. Int J Epidemiol 2005;34:316-326.

6. Lantz P, Golberstein E, House JS, Morenoff J. Socioeconomic and behavioral risk factors for mortality in a national 19-year prospective study of US adults. Soc Sci Med 2010;70:1558-1566.

7. Laaksonen M, Talala K, Martelin T, et al. Health behaviours as explanations for education level differences in cardiovascular and all- cause mortality: a follow-up of 60,000 men and women over 23 years. Eur J Public Health 2008;18:38-43.

8. Stringhini S, Sabia S, Shipley MJ, et al. Association of socioeconomic position with health behaviors and mortality. JAMA 2010;303:1159-1166.

9. Etman A, Burdorf A, Van der Cammen TJM, Mackenbach JP, Van Lenthe FJ. Socio-demographic determinants of worsening in frailty among community-dwelling older people in 11 European countries. J Epidemiol Community Health 2012;66:1116-1121.

10. Alvarado BE, Zunzunegui M, Beland F, Bamvita J. Life course social and health conditions linked to frailty in Latin American older men and women. J Gerontol A Biol Sci Med Sci 2008;63:1399-1406.

11. Jagger C, Weston C, Cambois E, et al. Inequalities in health expectancies at older ages in the European Union: Findings from the Survey of Health and Retirement in Europe (SHARE). J Epidemiol Community Health 2011;6:1030-1035.

12. Szanton SL, Seplaki CL, Thorpe RJ, Allen JK, Fried L. Socioeconomic status is associated with frailty: The women’s health and aging studies. J Epidemiol Community Health 2010;64:63-67.

13. Blaum C, Xue Q, Michelon E, Semba R, Fried L. The association between obesity and the frailty syndrome in older women: The women’s health and aging studies. JAMA 2005;53:927-934.

14. Hubbard RE, Lang IA, Llewellyn DJ, Rockwood K. Frailty, body mass index, and abdominal obesity in older people. J Gerontol A Biol Sci Med Sci 2010;65:377-381.

15. Villareal D, Apovian C, Kushner R, et al. Obesity in older adults: Technical review and position statement of the American Society for Nutrition and NAASO, The Obesity Society. Obes Res 2005;82:923-934.

16. Börsch-Supan A, Hank K, Jürges H. A new comprehensive and international view on aging: Introducing the Survey of Health, Ageing and Retirement in Europe. Eur J Ageing 2005;2:245-253.

17. De Luca G, Peracchi F. Survey participation in the first wave of SHARE. In Börsch-Supan A, Jürges H (eds) The Survey of Health, Aging, and Retirement in Europe – Methodology. 2005. Manheim Research Institute for the Economics of Ageing, Manheim, Germany, pp 88-104.

18. Santos-Eggimann B, Cuenoud P, Spagnoli J, Junod J. Prevalence of frailty in middle-aged and older community-dwelling Europeans living in 10 countries. J Gerontol A Biol Sci Med Sci 2009;64:675-681.

19. Cuénoud P. Prevalence of the frailty phenotype in community-dwelling population aged 50 and over in the Survey of Health, Ageing and Retirement in Europe. 2007. Dissertation, University of Lausanne.

20. Fried L, Tangen CM, Walston J, et al. Frailty in older adults: Evidence for a phenotype. J Gerontol A Biol Sci Med Sci 2001;56:M146-M56.

21. Macklai NS, Spagnoli J, Junod J, Santos-Eggimann B. Prospective association of the SHARE-operationalized frailty phenotype with adverse health outcomes: evidence from 60+ community-dwelling Europeans living in 11 countries. BMC Geriatr 2013;13:3.

22. Romero-Ortuno R, Walsh CD, Lawlor BA, Kenny RA. A frailty instrument for primary care: Findings from the Survey of Health, Ageing and Retirement in Europe (SHARE). BMC Geriatr 2010;10:57.

23. International Standard Classification of Education: ISCED 1997. 2006. UNESCO publication, Montreal, Canada.

24. Fernández-Bolaños M, Otero A, Zunzunegui MV, et al. Sex differences in the prevalence of frailty in a population aged 75 and older in Spain. J Am Geriatr Soc 2008;56:2370-2371.

25. Rodríguez López S, Nilsson C, Lund R, Montero P, Fernández-Ballesteros R, Avlund K. Social inequality in dynamic balance performance in an early old age Spanish population: The role of health and lifestyle associated factors. Arch Gerontol Geriatr 2012;54:e139-e145.

26. McAdams MA, Van Dam RM, Hu FB. Comparison of self-reported and measured BMI as correlates of disease markers in US adults. Obesity 2007;15:188-196.

27. Gorber SC, Tremblay M, Moher D, Gorber B. A comparison of direct vs. self-report measures for assessing height, weight and body mass index: A systematic review. Obes Rev 2007;8:307-326.

28. Romero-Ortuno R. The SHARE operationalized frailty phenotype: A comparison of two approaches. Eur Geriatr Med 2013;4:255-259.