jfa journal

AND option

OR option



A.K. Stuck1, A. Weber2, R. Wittwer1, A. Limacher3, R.W. Kressig4


1. Department of Geriatrics, Inselspital, Bern University Hospital, and University of Bern, Bern, Switzerland; 2. University of Bern, Medical Faculty, Bern, Switzerland; 3. CTU Bern, University of Bern, Bern, Switzerland; 4. University Department of Geriatric Medicine FELIX PLATTER, Basel, and University of Basel, Basel, Switzerland

Corresponding Author: Dr. med. Anna K. Stuck, Department of Geriatrics, Inselspital, Bern University Hospital, and University of Bern, Freiburgstrasse 46, CH-3010 Bern, Switzerland, e-Mail: anna.stuck@insel.ch

J Frailty Aging 2021;in press
Published online October 1, 2021, http://dx.doi.org/10.14283/jfa.2021.40



Objectives: To investigate practicality and repeatability of a handheld compared to a state-of-the-art multisegmental bioelectrical impedance analysis (BIA) device to facilitate screening of sarcopenia in older inpatients.
Design and setting: Cross-sectional study in a geriatric rehabilitation hospital.
Participants: 207 inpatients aged 70+.
Measurements: In a first phase, appendicular skeletal muscle mass index (ASMI) was measured using the handheld Biody xpertZm II BIA device (n=100). In a second phase, ASMI was obtained using the multisegmental Biacorpus RX 4004M device (n=107). Repeatability of BIA devices was compared in subgroups of patients (handheld BIA device: n=36, multisegmental BIA device: n=46) by intra-class correlation (ICC) and Bland-Altman plots.
Results: Overall, measurement failure was seen in 31 patients (31%) tested with the handheld BIA device compared to one patient (0.9%) using the multisegmental BIA device (p<0.001). Main reasons for measurement failure were inability of patients to adopt the position necessary to use the handheld BIA device and device failure. The mean difference of two ASMI measurements in the same patient was 0.32 (sd 0.85) using the handheld BIA device compared to 0.02 kg/m2 (sd 0.07) using the multisegmental device (adjusted mean difference between both groups -0.35, 95% confidence interval (CI) -0.61 to -0.09 kg/m2). Congruently, Bland-Altman plots showed poor agreement with the handheld compared to the multisegmental BIA device.
Conclusion: The handheld BIA device is neither a practical nor reliable device for assessing muscle mass in older rehabilitation inpatients.

Key words: Appendicular skeletal muscle mass, sarcopenia, repeatability, practicality, bioelectrical impedance analysis, geriatric.



Sarcopenia is a common disease primarily affecting older and multimorbid patients (1) and is associated with poor outcomes, such as falls and mortality (2, 3). It is important to identify older patients presenting with sarcopenia in order to begin appropriate targeted interventions (4). According to the most recent guidelines by the European Working group on sarcopenia (EWGSOP2), muscle mass, together with muscle strength, are considered the two criteria necessary to diagnose sarcopenia (5).
Currently, bioelectrical impedance analysis (BIA) using a multisegmental device is considered the portable reference method for measuring muscle mass. Thereby, mutisegmental BIA devices are commonly used in inpatients in the screening work-up of sarcopenia, osteosarcopenia, malnutrition, cachexia, obesity, and frailty (6-11). In older inpatients, validity of multisegmental BIA devices to measure muscle is acknowledged (12, 13).
However, multisegmental BIA devices are typically large, heavy, and require installation on a mobile cart. From a clinical perspective, a handheld pocket device would be more ideal in geriatric inpatient institutions (e.g. geriatric acute and post-acute hospitals, nursing homes), and could eventually facilitate routine screening of sarcopenia in older patients.
A promising handheld pocket BIA device (BIODY XPERTZM II) has recently become available for bedside evaluation of muscle mass (Figure 1). While potentially a useful tool for diagnosing sarcopenia, we were unable to find data demonstrating how this device performs in the assessment of older inpatients.
This study evaluated the practicality and repeatability of this new handheld BIA device compared to a multisegmental BIA device among older patients in a geriatric inpatient facility.

Figure 1a. Schematic figure of the handheld BIA device

Figure 1b. Position for measurement using the handheld BIA device




We conducted a retrospective analysis of cross-sectional anonymized assessment data of all patients admitted to a Geriatric Rehabilitation Hospital in Bern Switzerland, between September and December 2019 (n=207). All patients met the following admission criteria: (1) age>75 years, (2) direct transfer from acute care hospital, (3) living in the community (i.e., not in a nursing home) prior to acute care hospital admission, (4) potential for functional improvement and discharge home following inpatient rehabilitation. No patient data were excluded for this analysis.
Standard geriatric assessment was performed by designated clinically trained assessors upon admission. The assessment validated for older persons included a mini-mental status test examination (MMSE) (14) (cognitive deficit ≤26 points), clock-test (15) (deficit in executive and visuospatial functions ≤5 points), 5-item geriatric depression scale (GDS-5) (16) (depressive symptoms ≥2 points), nutritional risk screening 2002 (17) (nutritional deficit ≥3 points), and vision (18) and hearing impairment testing (19). Gait speed was measured using a standardized protocol (20) (low gait speed ≤0.8m/sec), and frailty was assessed using the clinical frailty scale (21) (frailty ≥5 points). According to recommendation by the “Deutsche Gesellschaft für Geriatrie” we used German versions of the referenced assessments (22). Body weight and height were measured using standard methods and BMI was calculated.
The study was approved by the Ethics committee “Kantonale Ethikkomission Bern”, Switzerland (Req-2020-00125).

Bioelectrical Impedance Analysis

Measurement of muscle mass was also part of the admission geriatric assessment, unless there was an absolute contraindication (i.e, patient had a pacemaker or internal cardioverter defibrillator because of an increased arrhythmic potential on implanted electronic devices). The handheld BIA device was used to measure muscle mass of all patients admitted during time period 1 (September 23 to November 11, 2019) (n=100). The multisegmental BIA device was used for patients admitted during time period 2 (November 11 to December 9, 2019) (n=107). Assessments of muscle mass were performed and monitored according to manufacturer’s guidelines and are described in detail in the following paragraphs.

Handheld BIA device

The BIA-device BIODY XPERTZM II is a unilateral validated handheld pocket device to measure muscle mass with software that uses the appendicular skeletal muscle mass index (ASMI) formula developed by Ursula Kyle (7). The device is designed for optimal testing to be conducted on the right side of the body with the patient in a seated position. The hand and heel were moisturized with a disinfecting standardized towel and patients were asked to remove all jewelry. The patient was instructed to lean over and place the device against their right heel and push the button on the side of the device. Patients were also tested in a supine position if they were unable to safely perform the test in a seated position. Patients with right-side prostheses, an internal metal device, or clinical asymmetry of body sides (e.g., right hemiplegia with muscle atrophy) were tested on the left side. During testing, the assessor insured that the patient did not touch any metal (e.g. chair leg, or bedrail) and that the two sides of the body did not come into contact. If necessary, the assessor provided assistance, and pressed the release button at the thumb electrode wearing isolating rubber gloves.

Multisegmental device

The multisegmental BIA device that we used during the second testing period was the BIACORPUS RX 4004M. The integrated software of the BIA device calculates the ASMI based on the Sergi equation (23, 24). Measurements were performed in a supine position in a standard hospital bed (maximum 30° inclination of the head). The patient rested in a supine position for 5 minutes before measurements were initiated. Both hands and heels were moisturized using a disinfecting standard towel. Two electrodes were placed at each extremity. The upper border of the proximal electrode was placed on the imaginary line between radius and ulna head, and between medial and lateral malleoli, respectively. Distal electrodes were put on within a 5cm distance of the distal border of the proximal electrodes. Whenever possible, measurement was performed bilaterally on both sides. Unilateral measurement (right or left body side) was only performed if the patient had prostheses or internal metal parts in one body side, or if the patient had unilateral atrophy or amputation. The patient was allowed to wear jewelry provided the electrodes were not blocked. As with the handheld device, the assessor insured that the patient did not touch any metal and that the two sides of the body did not come into contact. Mobile telephones had to be placed at least 1m from the measuring device.

Practicality of BIA-devices

To assess the practicality of using each device in a geriatric inpatient setting, the assessor recorded the total number of attempts that were necessary to achieve a valid measurement, any alterations in the testing procedure that were necessary to achieve a valid measurement (e.g., supine vs sitting position, assessor assistance), reasons for test failure and contraindications for use. The major reasons for test failure were categorized post-hoc into the following domains:
– Transmission error of the BIA device after 5 measurement attempts (device failure)
– Musculoskeletal impairment (e.g., limited flexibility of the hip to lean over)
– Cognitive impairment (e.g., patient unable to follow instructions)

Repeatability of BIA-devices

Repeatability was measured in a priori defined subgroup of 36 patients tested with the handheld BIA device and 46 patients tested with the multisegmental BIA device. For each patient, two consecutive measurements were performed by the same assessor using the same BIA device. Repeatability was defined according to the definition by Bartlett et al. (25).

Statistical Analysis

Study sample characteristics from admission data are presented by absolute and relative frequencies or by mean with standard deviation (sd) for categorical and continuous variables, respectively. Categorical variables (practicality) were compared between the handheld and the multisegmental device using chi-squared test and continuous variables (muscle mass) were compared using the Student’s t-test. Measures of repeatability (within and between patient standard deviation and intra-class correlation (ICC)) for the two BIA devices were calculated using one-way analysis-of-variance (ANOVA) models (checked for normality by visual inspection of histograms). Bland-Altman plots were generated displaying the differences between measurements 1 and 2 of ASMI against the mean of the two ASMI measurements. Repeatability coefficients were calculated as 1.96x the standard deviation of the mean difference for each BIA device (25). For analysis of descriptive results of ASMI, all patients who had a BIA measurement yielding an ASMI value were included in the secondary analysis. If a patient had two ASMI results, the lower value of the two was included for descriptive analysis. Linear regression analysis was performed to compare ASMI values between the two groups using the handheld or multisegmental BIA device adjusted for age, sex and frailty status. An a priori decision was made to not perform statistical comparisons among subgroups of patients to avoid type I and II error inflation. Analyses were computed using Stata Version 16.1 (StataCorp LLC, College Station, Texas, USA). A p-value of <0.05 was considered statistically significant.



The mean age of the sample was 84.3 years (standard deviation (sd) 6.4) and 65.7% were female. Descriptive characteristics of patients (n=207) are shown in Table 1. Clinical characteristics of patients measured with the handheld BIA device did not differ from characteristics of patients measured with the multisegmental device.

Table 1. Clinical Characteristics of Patients Stratified for Type of BIA-device

Abbreviations: BIA, bioelectrical impedance analysis; sd, standard deviation; MMS, mini mental status examination; NRS, nutritional risk score; GDS, 5-item geriatric depression scale; BMI, body mass index; a. No data due to inability to perform clock-test in n=18 (handheld BIA), and n=14 (multisegmental); b. No data due to cognitive impairment in n=7 (handheld BIA), and n=14 (multisegmental BIA); c. Missing data in n=1 (handheld BIA); d. P-value indicated for the comparison of patients with handheld BIA vs. patients with multisegmental BIA.



Overall, measurement of muscle mass was not possible in 36 (36%) of the 100 patients with the handheld, and in 9 (8.4%) with the multisegmental device, respectively. This was in part due to a contraindication for BIA measurement (see Methods section for list of contraindications), and in part due to an inability to obtain a measurement value (Table 2). In specific, five patients had a contraindication for use of the handheld BIA device and 8 patients had a contraindication for use of the multisegmental device, but this difference was not statistically significant (5.0% vs. 7.5%, p=0.46) (Table 2). ASMI measurement was unsuccessful in 31 patients (31%) using the handheld BIA and one patient (0.9%) using the multisegmental BIA corresponding to a difference between the groups of 30.1% (95% CI, 20.9 to 39.8%; p<0.001). Reasons for inability to obtain an ASMI measurement with the handheld BIA device, other than a contraindication, included musculoskeletal impairment (n=17), transmission error of the BIA device after five measurement attempts (n=13), and cognitive impairment (n=1). In contrast, the reason for inability to obtain an ASMI result with the multisegmental device in the one patient was due to musculoskeletal impairment.

Table 2. Comparison of practicality and repeatability between the handheld and multisegmental BIA-device

Abbreviations: BIA, bioelectrical impedance analysis; sd, standard deviation; ASMI, appendicular skeletal muscle mass index; ICC, intraclass correlation coefficient; n.a., not applicable; CI, confidence interval; a. Contraindication of BIA include patients with pacemaker or an implantable cardioverter defibrillator (ICD); b. Patients with repeatable ASMI is defined as patients with two ASMI results and thereof a mean relative difference of <2.5%; c. Difference in mean difference of ASMI measurement 1 and 2 between both devices, adjusted for age, sex and frailty status using linear regression analysis



Repeatability of results for each BIA device is shown in Table 2. The ASMI mean absolute difference between two measurements of the same patient was 0.32 (0.85) kg/m2 in the group using the handheld BIA device versus 0.02 (0.07) kg/m2 in the group using the multisegmental BIA device (adjusted ASMI mean difference between the two groups of -0.35, 95% confidence interval (CI) -0.61 to -0.09 kg/m2). The variability of the two handheld BIA measurements within patients was much higher than the two multisegemental BIA measurements (within-patient standard deviation of 0.63 versus 0.05 kg/m2, respectively). Correspondingly, the intra-patient correlation was much lower for the handheld BIA device than for the multisegmental BIA device (ICC of 0.80 versus 0.998, respectively). Bland-Altman plots of the two devices also show a lower agreement for the handheld BIA device compared to the multisegmental device (Figure 2, Panels A and B).

Figure 2a. Bland-Altman Plot of repeatability using the handheld BIA device (n=36)

Figure 2b. Bland-Altman plot of repeatability using the multisegmental BIA device (n=46)

Abbreviations: ASMI, appendicular skeletal muscle mass index; BIA, bioelectrical impedance analysis; The grey horizontal line displays the reference line indicating a difference of 0 kg/m2 between measurement 1 and 2 of ASMI. The dashed green line represents the mean difference between measurement 1 and 2 of ASMI. Shaded area represent the area within the limits of agreement defined as the mean difference ± 1.96 SD of differences.



ASMI means differed significantly between patients measured with the handheld BIA and the multisegmental BIA device (6.9 (1.5) vs. 6.4 (1.3) kg/m2, respectively; mean adjusted difference, 0.51, 95% CI 0.12 to 0.90 kg/m2, p-value = 0.01). Among women, mean ASMI was 6.5 (1.3) kg/m2 using the handheld device and 6.0 (1.1) kg/m2 using the multisegmental device (Figure 3). Among men, mean ASMI was 7.7 (1.2) kg/m2 using the handheld device and 7.3 (1.2) kg/m2 using the multisegmental device.

Figure 3. Appendicular skeletal muscle mass index by gender and type of BIA device

Abbreviations: ASMI, appendicular skeletal muscle mass index; Horizontal lines indicate gender-specific minimum threshold lines of appendicular skeletal muscle mass index (ASMI) according to the European guidelines on diagnosis of sarcopenia (pink line: women=<5.5kg/m2; blue line: men =<7.0kg/m2).



To our knowledge, this is the first study that has evaluated a handheld BIA device in geriatric inpatients. Our findings demonstrated that the handheld BIA device tested was less practical to use and had significantly lower repeatability than the multisegmental BIA device for measuring muscle mass in older inpatients of a geriatric rehabilitation hospital.

Overall, measurement failure of the handheld BIA device occurred in a large proportion of patients compared to a negligible proportion using the multisegmental BIA device. This lack of practicality of the handheld device was observed although two clinical assessors received standard instruction and training prior to using the devices. Moreover, to account for intermittent transmission errors, we allowed five measurement trials using the BIA device. We assume that in clinical practice rates of measurement failure might even be higher since training of assessors and allowance for repetition of measurements may vary.
Proof of successful application is the basic requirement for eventual implementation of a diagnostic tool. While the handheld BIA device is smaller and more portable than the multisegmental BIA device, we observed that its design contributed to measurement failures in nearly one-third of patients. The handheld BIA device requires that the patient have basic flexion of the hip and knee and minimal grip and finger strength to hold the device and to push the activation button. Activation and handling of the BIA device also requires basic cognitive performance to follow instructions. However, both these basic requirements on mobility and cognition are frequently lacking among geriatric inpatients. In contrast, the multisegmental BIA tests patients in a supine position without requiring the patient to manually activate the device, which likely contributed to the low failure rate.

Our results also reveal that the handheld device had much lower repeatability compared to the multisegmental BIA device indicating limited bias of the handheld device. Although prior evidence suggests that the handheld BIA is feasible and repeatable in younger and healthy participants (26) these results cannot be confirmed in our study of older inpatients. The most likely explanation for this finding is that the handheld device is susceptible to small changes of the patient’s position because clear reference points are lacking for the two electrodes. In contrast, standard reference points are provided for placement of the eight electrodes used with the multisegmental BIA device. Therefore, it may be more challenging to identify and maintain the correct position for ASMI measurement using the handheld device for functionally impaired older inpatients.
In our study, we found a mean ASMI of 6.4 kg/m2 using the state-of-the-art multisegmental BIA device which is consistent with previous studies of older inpatients [13, 14]. Prior findings of ASMI in a geriatric rehabilitation hospital reported a mean ASMI of 6.4 kg/m2 [13], while another study in institutionalized patients reported an ASMI of 6.3 kg/m2 (27). Additionally, van Ancum et al. observed in acute inpatients an ASMI of 7.6 kg/m2 in men and 6.5 in women that are close to our results (28).
Overall, our results from subanalyses further indicate that there may be an additional issue of limited validity of the handheld BIA device, although direct comparison between the handheld and multisegmental devices within a patient was not possible due to our study design. However, other studies similarly reported that there are differences between BIA devices suggesting that BIA devices are not necessarily interchangeable. Beaudart et al. found that the prevalence of low muscle mass and sarcopenia was dependent on the diagnostic tool used (29). In another study, Lahav et al. found that the InBody™ BIA device underestimated body fat to a higher degree then the Seca™ BIA device in both genders and in all BMI categories (30). Similarly, Kreissl et al. (31) found that in a pediatric population, the Tanita™ BIA device underestimated fat free mass compared to the Biacorpus™ BIA device used in our study.
There are several limitations to our study. First, we investigated a single handheld BIA device in one sample of older inpatients in a geriatric rehabilitation hospital. Consequently, our results may not be generalizable to other devices or populations. Second, we chose a pragmatic sequential study design, so direct comparison of BIA devices (handheld vs. multisegmental) was not possible in the same patient. We therefore adjusted differences between both groups for potential confounding variables. Nevertheless, our main findings of limited practicality and repeatability of the handheld BIA device are independent of group comparisons and would not alter our findings.
Our findings have implications both for clinical research and practice.
Further research is needed to identify a practical and valid handheld device to measure muscle mass in older inpatients to facilitate routine diagnostic work-up of sarcopenia (32). According to the latest guidelines by the European Working group on Sarcopenia Project 2 (EWGSOP 2) sarcopenia is defined as the combination of low muscle mass and low muscle strength highlighting the importance of measuring muscle mass in this vulnerable population. The key role of identifying low muscle mass is based on longitudinal evidence, that both muscle mass and strength are predictive for significant adverse outcome such as falls (33).
However, the method of bioelectrical impedance analysis itself to measure muscle mass has intrinsic limitations, due to the BIA’s absolute contraindication in patients with an internal pacemaker. Recently, pocket handheld devices using a different technology than BIA are being considered, including ultrasound (34, 35). Additional studies are needed to identify and evaluate methodological approaches with user-friendly benefits for clinical use that could promote rapid screening of muscle mass facilitating diagnosis of sarcopenia.
In conclusion, the handheld BIA device that we evaluated failed practicality and repeatability and cannot be recommended for the use in older inpatients.


Ethical Standards: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The study was approved by the Ethics committee “Kantonale Ethikkomission Bern”, Switzerland (Req-2020-00125).

Acknowledgements: We thank Karen R. Josephson for editing the manuscript.

Funding: This work was supported by the Forschungsfonds Geriatrie, Bern, Switzerland. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Conflict of interest: The authors declare no conflict of interest.

Open Access: This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, duplication, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.



1. Pacifico J, Geerlings MAJ, Reijnierse EM, et al. Prevalence of sarcopenia as a comorbid disease: A systematic review and meta-analysis. Experimental gerontology 2020; 131: 110801. doi:10.1016/j.exger.2019.110801.
2. Beaudart C, Zaaria M, Pasleau F, Reginster JY, Bruyere O. Health outcomes of sarcopenia: A systematic review and meta-analysis. PloS one 2017; 12: e0169548. doi:10.1371/journal.pone.0169548.
3. Zhao Y, Zhang Y, Hao Q, Ge M, Dong B. Sarcopenia and hospital-related outcomes in the old people: a systematic review and meta-analysis. Aging clinical and experimental research 2019; 31: 5-14. doi:10.1007/s40520-018-0931-z.
4. Lee SY, Tung HH, Liu CY, Chen LK. Physical activity and sarcopenia in the geriatric population: A systematic review. Journal of the American Medical Directors Association 2018; 19: 378-383. doi:10.1016/j.jamda.2018.02.003.
5. Cruz-Jentoft AJ, Bahat G, Bauer J, et al. Sarcopenia: revised European consensus on definition and diagnosis. Age and ageing 2019; 48: 601. doi:10.1093/ageing/afz046.
6. Player EL, Morris P, Thomas T, et al. Bioelectrical impedance analysis (BIA)-derived phase angle (PA) is a practical aid to nutritional assessment in hospital in-patients. Clinical nutrition 2019; 38: 1700-1706. doi:10.1016/j.clnu.2018.08.003.
7. Hirose S, Nakajima T, Nozawa N, et al. Phase angle as an indicator of sarcopenia, malnutrition, and cachexia in inpatients with cardiovascular diseases. Journal of clinical medicine 2020; 9, doi:10.3390/jcm9082554.
8. Lunt E, Ong T, Gordon AL, Greenhaff PL, Gladman JRF. The clinical usefulness of muscle mass and strength measures in older people: a systematic review. Age and ageing 2021; 50: 88-95. doi:10.1093/ageing/afaa123.
9. Brunani A, Perna S, Soranna D, et al. Body composition assessment using bioelectrical impedance analysis (BIA) in a wide cohort of patients affected with mild to severe obesity. Clinical nutrition 2021; 40: 3973-3981. doi:10.1016/j.clnu.2021.04.033.
10. Makizako H, Kubozono T, Kiyama R, et al. Associations of social frailty with loss of muscle mass and muscle weakness among community-dwelling older adults. Geriatrics & gerontology international 2019; 19: 76-80. doi:10.1111/ggi.13571.
11. Taniguchi Y, Makizako H, Kiyama R, et al. The association between osteoporosis and grip strength and skeletal muscle mass in community-dwelling older women. International journal of environmental research and public health 2019; 16, doi:10.3390/ijerph16071228.
12. Smoliner C, Sieber CC, Wirth R. Prevalence of sarcopenia in geriatric hospitalized patients. Journal of the American Medical Directors Association 2014; 15: 267-272. doi:10.1016/j.jamda.2013.11.027.
13. Bertschi D, Kiss CM, Beerli N, Kressig RW. Sarcopenia in hospitalized geriatric patients: insights into prevalence and associated parameters using new EWGSOP2 guidelines. European journal of clinical nutrition 2020, doi:10.1038/s41430-020-00780-7.
14. Folstein MF, Folstein SE, McHugh PR. «Mini-mental state». A practical method for grading the cognitive state of patients for the clinician. Journal of psychiatric research 1975; 12: 189-198. doi:10.1016/0022-3956(75)90026-6.
15. Freedman M, Leach L, Kaplan E, et al. Clock drawing: a neuropsychological analysis. Oxford University Press, Inc., 1994. New York
16. Hoyl MT, Alessi CA, Harker JO, et al. Development and testing of a five-item version of the Geriatric Depression Scale. Journal of the American Geriatrics Society 1999; 47: 873-878. doi:10.1111/j.1532-5415.1999.tb03848.x.
17. Kondrup J, Rasmussen HH, Hamberg O, Stanga Z, Ad Hoc EWG. Nutritional risk screening (NRS 2002): a new method based on an analysis of controlled clinical trials. Clinical nutrition 2003; 22: 321-336. doi:10.1016/s0261-5614(02)00214-5.
18. Inouye SK, Bogardus ST, Jr., Charpentier PA, et al. A multicomponent intervention to prevent delirium in hospitalized older patients. The New England journal of medicine 1999; 340: 669-676. doi:10.1056/NEJM199903043400901.
19. Bagai A, Thavendiranathan P, Detsky AS. Does this patient have hearing impairment? Jama 2006; 295: 416-428. doi:10.1001/jama.295.4.416.
20. Stuck AK, Bachmann M, Fullemann P, Josephson KR, Stuck AE. Effect of testing procedures on gait speed measurement: A systematic review. PloS one 2020; 15: e0234200. doi:10.1371/journal.pone.0234200.
21. Rockwood K, Song X, MacKnight C, et al. A global clinical measure of fitness and frailty in elderly people. CMAJ : Canadian Medical Association journal = journal de l’Association medicale canadienne 2005; 173: 489-495. doi:10.1503/cmaj.050051.
22. Deutsche Gesellschaft für Geriatrie. S1-Leitlinie „Geriatrisches Assessment der Stufe 2“. 2019. https://www.awmf.org/leitlinien/detail/ll/084-002.html. Accessed 22.07.2021.
23. Brun JF, Guiraudou M, Mardemootoo C, et al. Validation of the measurement of segmental body composition compared to DEXA: Interest of the measurement of trunk fat mass. Science & Sports 2013; 158-162.
24. Sergi G, De Rui M, Veronese N, et al. Assessing appendicular skeletal muscle mass with bioelectrical impedance analysis in free-living Caucasian older adults. Clinical nutrition 2015; 34: 667-673. doi:10.1016/j.clnu.2014.07.010.
25. Bartlett JW, Frost C. Reliability, repeatability and reproducibility: analysis of measurement errors in continuous variables. Ultrasound in obstetrics & gynecology : the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2008; 31: 466-475. doi:10.1002/uog.5256.
26. Pereira B. Etude de concordance impédancemètrie. Mesure de la composition corporelle à l’aide de l’impédancemètre Biody Xpert: comparaison avec la méthode d’absorptiométrie biphotonique (DXA). 2015. DRCI – CHU Clermont-Ferrand
27. Cebria IIMA, Arnal-Gomez A, Tortosa-Chulia MA, et al. Functional and clinical characteristics for predicting sarcopenia in institutionalised older adults: Identifying tools for clinical screening. International journal of environmental research and public health 2020; 17, doi:10.3390/ijerph17124483.
28. Van Ancum JM, Alcazar J, Meskers CGM, et al. Impact of using the updated EWGSOP2 definition in diagnosing sarcopenia: A clinical perspective. Archives of gerontology and geriatrics 2020; 90: 104125. doi:10.1016/j.archger.2020.104125.
29. Beaudart C, Reginster JY, Slomian J, et al. Estimation of sarcopenia prevalence using various assessment tools. Experimental gerontology 2015; 61: 31-37. doi:10.1016/j.exger.2014.11.014.
30. Lahav Y, Goldstein N, Gepner Y. Comparison of body composition assessment across body mass index categories by two multifrequency bioelectrical impedance analysis devices and dual-energy X-ray absorptiometry in clinical settings. European journal of clinical nutrition 2021, doi:10.1038/s41430-020-00839-5.
31. Kreissl A, Jorda A, Truschner K, Skacel G, Greber-Platzer S. Clinically relevant body composition methods for obese pediatric patients. BMC pediatrics 2019; 19: 84. doi:10.1186/s12887-019-1454-2.
32. Bauer J, Morley JE, Schols A, et al. Sarcopenia: A time for action. An SCWD position paper. Journal of cachexia, sarcopenia and muscle 2019; 10: 956-961. doi:10.1002/jcsm.12483.
33. Van Ancum JM, Pijnappels M, Jonkman NH, et al. Muscle mass and muscle strength are associated with pre- and post-hospitalization falls in older male inpatients: a longitudinal cohort study. BMC geriatrics 2018; 18: 116. doi:10.1186/s12877-018-0812-5.
34. Buckinx F, Landi F, Cesari M, et al. Pitfalls in the measurement of muscle mass: a need for a reference standard. Journal of cachexia, sarcopenia and muscle 2018; 9: 269-278. doi:10.1002/jcsm.12268.
35. Isaka M, Sugimoto K, Yasunobe Y, et al. The usefulness of an alternative diagnostic method for sarcopenia using thickness and echo intensity of lower leg muscles in older males. Journal of the American Medical Directors Association 2019; 20: 1185 e1181-1185 e1188. doi:10.1016/j.jamda.2019.01.152.




M. Serra-Prat1,2, M. Terradellas1, I. Lorenzo1, M. Arús3, E. Burdoy4, A. Salietti4, S. Ramírez5, E. Palomera1, M. Papiol6, E. Pleguezuelos7


1. Research Unit. Consorci Sanitari del Maresme, Mataró, Barcelona, Spain; 2. Centro de Investiación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBEREHD), ISCIII, Madrid, Spain; 3. Dietetics and Nutrition Unit, Hospital of Mataró, Consorci Sanitari del Maresme, Mataró, Barcelona, Spain; 4. ABS Mataró Centre, Consorci Sanitari del Maresme, Mataró, Barcelona, Spain; 5. ABS Cirera-Molins, Consorci Sanitari del Maresme, Mataró, Barcelona, Spain; 6. ABS Argentona, Consorci Sanitari del Maresme, Mataró, Barcelona, Spain; 7. Rehabilitation Department, Hospital of Mataró, Consorci Sanitari del Maresme, Mataró, Barcelona, Spain

Corresponding Author: Mateu Serra-Prat, Research Unit, Hospital de Mataró, Carretera de Cierera s/n, 08304 Mataró, Barcelona, Spain, Tel. + 34 93 741 77 30, mserra@csdm.cat

J Frailty Aging 2021;in press
Published online September 24, 2021, http://dx.doi.org/10.14283/jfa.2021.38



Background: Obesity is a risk factor for frailty and muscle weakness, so weight loss in obese older adults may prevent frailty and functional decline.
Objective: To assess the safety and efficacy of a multimodal weight-loss intervention in improving functional performance and reducing frailty risk in obese older adults.
Design: Randomized controlled trial with 2 parallel arms.
Setting and participants: Community-dwelling obese adults aged 65-75 years with body mass index (BMI) 30-39 kg/m2.
Intervention: 6-month multimodal intervention based on diet and a physical activity program. Control group: Usual care.
Main and secondary outcome measures: Frailty (Fried criteria) rate and functional performance at 6, 12, and 24 months of follow-up, respectively.
Intermediate outcome measures: Weight loss, body composition changes, and metabolic and inflammatory biomarker changes.
Results: N=305. The study intervention increased gait speed at 12 and 24 months of follow-up, but had no significant effect on frailty prevention. It was effective in reducing weight, BMI, fat mass, interleukin 6, and insulin resistance and improving self-reported quality of life.
Conclusions: The study intervention was not demonstrated to be effective in preventing frailty in obese people aged 65-75 years at 24 months of follow-up. However, it allowed weight loss and a reduction in inflammatory and insulin resistance markers, which could have a long-term effect on frailty that requires further research.

Key words: Randomized controlled trial, obese, multimodal intervention, frailty, insulin resistance.



Frailty is a clinical condition in which the individual is in a vulnerable state at increased risk of disease, adverse health outcomes, disability, and death when exposed to a stressor (1). It is considered a geriatric syndrome characterized by a decline in the functioning of different organs and systems. Frailty has a great impact on quality of life (QoL) and increases the risk of falls, functional decline, dependency, and institutionalization (2). Frailty prevalence in the community-dwelling population aged over 65 years has been estimated at 11%, rising to as high as 50% in the population aged over 80 years (3). Frailty causes and pathophysiology are not well understood. Some authors consider frailty to be the result of an accumulation of different unrelated diseases, dysfunctions, and disabilities (4). Other authors consider frailty to be the result of a pathophysiological process involving the breakdown of homeostatic mechanisms mainly expressed as impaired muscle function (5). As risk factors for frailty, several studies have pointed to comorbidities such as diabetes, stroke, dementia, and depression (6, 7), and also to pain and certain physiological changes that appear with age, such as low-intensity inflammatory processes (8), changes in body composition, hormonal imbalances (9), loss of appetite, nutritional disturbances, insulin resistance, and dehydration (10).
Obesity, which is also prevalent in the population of older adults (11), has been related with frailty in both cross-sectional (12-16) and prospective population-based studies (17), with the relationship between obesity and muscle mass and strength described as negative (18). Mechanisms through which obesity can promote frailty and muscle wasting may include: (a) fatty infiltration of the muscle so that it loses quality and strength, (b) pro-inflammatory cytokine release with catabolic and anorectic effects (19), (c) low physical activity levels favouring muscular atrophy, (d) increased cardiovascular risk and increased stroke and heart failure incidence, (e) increased risk of osteoarthritis and pain, limiting physical activity, (f) hormonal changes including reduced oestrogen, testosterone, sex hormone binding globulin, dehydroepiandrosterone, growth hormone, and insulin-like growth factor (IGF-I) levels, and increased prolactin and cortisol levels (20), and (g) insulin resistance and a risk of type 2 diabetes mellitus, associated with a loss of muscle mass and strength (21). Sarcopenic obesity is a clinical condition of reduced muscle mass in the context of excess adiposity, most often reported in older people as both conditions increase with age (22). Sarcopenia has a negative prognostic impact in obese individuals and may lead to frailty, disability, and increased morbimortality (23).
Based on the evidence, it would seem reasonable to think that an intervention based on a healthy diet and physical exercise aimed at reducing body mass index (BMI) below 30 kg/m2 without loss of muscle mass could effectively and safely prevent frailty and disability in obese older adults over the mid-/long-term. While the health benefits of diet and exercise in older people have been widely demonstrated, leaving no doubt as to the validity of the corresponding recommendations (24-27), weight loss to prevent frailty, disability, dependency, and institutionalization in obese older adults has been little studied. The lack of evidence may be due in part to poor adherence to diet and exercise recommendations, the difficulty to change habits and lifestyles (28), and the fear of weight loss accentuating muscle mass loss and muscle wasting in older adults (20). We evaluated the safety and efficacy of a multimodal weight-loss intervention in community-dwelling obese older adults. Primary objectives were to assess improvements in functional performance and reductions in frailty risk, while secondary objectives were to assess changes in anthropometric, body composition, metabolic, and inflammatory parameters.


Materials and Methods

Study design and population

Our open-label randomized controlled trial (RCT) with 2 parallel arms (intervention and control) and 2 years of follow-up recruited adults aged 65-75 years, with a BMI of
30-39 kg/m2 (inclusive), who had at least one of the following obesity-related clinical conditions for which weight loss is advisable: dyslipidaemia, hypertension, diabetes or insulin resistance, obesity-related physical limitations, or sleep apnoea/hypopnoea. Exclusion criteria were dementia, neurodegenerative diseases, severe psychiatric disorders, cancer diagnoses, lower limb amputation, institutionalization, and life expectancy <6 months. Candidate participants were pre-selected from the database corresponding to three Maresme region primary care centres in the province of Barcelona (Catalonia, Spain) according to age, BMI criteria, and inclusion/exclusion criteria. These candidates were invited by telephone to a visit in which selection criteria were verified and individuals received information about the study and granted their written consent. Sample size was calculated usig the GRANMO calculator (https://www.imim.cat/ofertadeserveis/software-public/granmo/) according to the proportion of frail individuals at 24 months of follow-up. For an alpha risk of 0.05 and a beta risk of less than 0.1 in a bilateral comparison, 162 subjects needed to be included in each group to detect a statistically significant difference in frailty prevalence, expected to be 21% for the control group and 7% for the intervention group (6, 17). For an anticipated loss to follow-up of 25%, recruitment of a total of 360 individuals was required. Participants were recruited at the 3 primary care centres from February to June 2017. Individuals were randomized to either the intervention or control group according to identification codes and the sequentially numbered opaque sealed envelope (SNOSE) technique. Randomization was stratified by primary care centre, with each centre receiving 120 envelopes (60 for each study group). Envelopes were opened only once the patient had consented and had been recruited. The study protocol was approved by the local ethics committee (reference number CEIC CSdM 60/16) and the study was registered in the ClinicalTrials.gov database (https://clinicaltrials.gov) under identifier NCT03000907.


The intervention consisted of a 6-month multimodal personalized program combining individual and group sessions. The intervention was as follows:
1. Diet. A dietician assessed nutritional status and nutrition requirements and developed a personalized eating plan for each intervention group individual aimed at achieving BMI<30 or weight loss>10%. The assessment considered how obesity had evolved, possible triggers, current dietary habits, the dietary environment, daily physical activity, and exercise regime, and included a physical examination covering weight, height, waist and hip circumferences, and bioimpedance analysis (BIA). These data were used to estimate the basal metabolic rate (BMR), daily energy expenditure (DEE) based on activity level, and recommended daily intake (RDI) necessary to achieve BMI<30 or weight loss>10% within 6 months. During this period individuals attended monthly individualized follow-up visits to evaluate weight evolution, to assess and reinforce adherence and to make necessary changes to established recommendations. Personalized eating plans were based on a diet as follows: (a) hypocaloric, with a caloric deficit of 300-400 kcal/day with respect to the DEE, (b) balanced in macronutrients, with around 20%, 50% 27% of total energy delivered in the form of proteins (1.2 g/kg/day), carbohydrates, and fat, respectively, and (c) balanced in micronutrients (vitamins and minerals) according to current recommendations, with supplementation of vitamins (D, B6, B12) and minerals (calcium, magnesium, selenium) if deficient.
2. Exercise. A multicomponent physical exercise program included the following: (a) 45 minutes of unsupervised daily aerobic exercise (e.g., walking outdoors) on at least 5 days/week, (b) unsupervised strength, balance, and flexibility exercises for 15-20 minutes/day on 3 days a week (adapted to different ailments and with personalized follow-up to avoid injuries) to be done at home, and (c) health education by a physiotherapist, consisting of 20 theoretical-practical group sessions of 1 hour/week in the primary care centre, aimed at improving adherence and emphasizing the importance of physical exercise and also including Nordic walking in groups twice a month. Unsupervised daily walks and home exercises were not monitored. Participants received a leaflet explaining the exercises they had to do at home with illustrated instructions.

Individuals assigned to the control group received their usual care, which involved no specific weight-loss intervention other than the usual dietary and hygienic recommendations of the primary care team.

Outcome measures and data collection

The main outcome measure was prevalence of frailty at 6, 12 and 24 months according to Fried criteria (5), with participants classified as robust, pre-frail, or frail if they fulfilled none, 1-2, or 3 or more of the following 5 criteria: (a) unintentional weight loss, (b) exhaustion, (c) low physical activity, (d) slow walking speed, and (e) poor grip strength, measured using a handheld Jamar dynamometer (Lafayette Instrument Co). Secondary outcome measures were the following indicators of functional performance: Barthel score, 2-minute walking test (2MWT), timed up-and-go test (TUG), gait speed, unipodal stance test (UST), number of falls in the previous 3 months, and daily hours walking outdoors.
Intermediate outcome measures were as follows: weight loss, change in BMI, change in body composition (BIA-evaluated fat mass, lean mass, and muscle mass), body fat distribution according to waist circumference, hip circumference, and waist/hip ratio (WHR), glycaemic control according to haemoglobin subunit alpha (HbA1) levels, insulin resistance according to the homeostatic model assessment of insulin resistance (HOMA-IR), and serum levels of inflammatory markers – interleukin 6 (IL-6) and C-reactive protein (CRP) – and anabolic hormones – IGF-1 and testosterone – as determined by commercialized kits. Outcome measures were evaluated in follow-up checks at 6, 12, and 24 months. Other data collected were as follows: sociodemographic characteristics, toxic habits (tobacco and alcohol consumption), comorbidities, number of medications, chronic pain as assessed by a visual analogue scale (VAS), and self-reported QoL as assessed by a 0-10 point horizontal VAS using identical question to those for the 5-dimension EuroQoL (EQ5D) VAS. Outcome evaluators, participants, and usual healthcare providers were not blinded to the intervention group.

Data analysis

Data analysis was by intention-to-treat (ITT). The intervention and control groups were compared at baseline and at 6, 12, and 24 months of follow-up using the X2 test or Fisher’s exact test for categorical variables, and the t-test or Mann-Whitney U test for numerical variables. Normality of continuous variables was assessed by the Kolmogorov-Smirnov test. Baseline to follow-up differences in the outcome measures were calculated and compared between the two study groups using the X2 test or Fisher’s exact test for categorical variables, and the t-test or Mann-Whitney U test for numerical ones. Moreover, regarding main numerical outcome measures, a general linear model analysis (GLM; ANOVA of repeated measures) was used to test both; a) if their evolution over time (repeated measures) differ between study groups (interaction between evolution and study group) and b) the group effect (independently of evolution). Statistical significance was established in all cases for a 2-sided p-value of <.05.



A total of 1,014 subjects were pre-selected, 319 of whom fulfilled the selection criteria and granted their informed consent. Of the 305 who attended the baseline visit, 150 were randomly allocated to the intervention group and 155 to the control group. Loss to follow-up was 21.6% at 6 months, 28.2% at 12 months, and 43.6% at 24 months, with no significant differences observed between groups. Figure 1 shows the study flow chart. Two deaths occurred during the study, neither attributed to the intervention (a case of cancer and the other non-specified). No other side effects were reported. The study sample (n = 305) had a mean age of 69.7 years, 65.9% were women, 15% lived alone, 27% had no education, and showed a mean BMI of 34 and a mean Barthel score of 99. Most frequent co-morbidities were arterial hypertension (76.1%), dyslipidaemia (66.9%), arthrosis (66.5%), diabetes (26.2%), gastroesophageal reflux (24.6%), peripheral vascular disease (23.9%) and depression (22.6%). Table 1 summarizes the main baseline characteristics for the sample and the 2 groups. Except for asthma prevalence and gait speed (significantly higher – p=0.025- and lower –p=0.035-, respectively, in the intervention group), no differences were observed in baseline sociodemographic characteristics, toxic habits, comorbidities, medications, functional capacity, physical examination parameters, or inflammatory, metabolic, or anabolic biomarkers between the groups. Baseline frailty prevalence was 2.7% in the intervention group and 1.3% in the control group (p=0.442). Overall, those data indicate that intervention and control groups were homogeneous and comparable.

Figure 1. Study flow Chart


No between-group differences in frailty prevalence were observed at 6, 12, and 24 months. At 6 months, frailty status had worsened (changed from robust to pre-frail or from pre-frail to frail) in 8.3% of intervention group individuals compared to 16.2% of control group individuals (p=0.069). No significant frailty status worsening differences were observed at 12 and 24 months of follow-up. At 12 months, 8 and 9 intervention and control group individuals, respectively, had become frail, for an incidence rate of 7.38 and 7.37 new cases/100 person-years, respectively. Table 2 compares frailty and functional outcome measures for the 2 groups at 6, 12, and 24 months of follow-up. It points to a significant effect of the study intervention in terms of increased daily outdoors walking by 6 months, and increased gait speed by 12 and 24 months of follow-up, with an improvement of 0.07 m/sec maintained at both 12 (p=0.035) and 24 (p=0.036) months.

Table 1. Main baseline characteristics for the sample and intervention (I) and control (C) groups

Data expressed as mean (standard deviation) except where otherwise indicated. 2MWT: 2-minute walking test. BMI: body mass index. HbA1: haemoglobin subunit alpha. HOMA-IR: homeostatic model assessment of insulin resistance. MNA-sf: Short-Form Mini-Nutritional Assessment. QoL: quality of life. TUG: Timed up-and-go test. UST: unipodal stance test. VAS: 0-10 visual analogue scale. WHR: waist-hip ratio; Items in bold are statistically significant at p<.05.


Table 2. Frailty and functional outcome measures for the intervention (I) and control (C) groups at 6, 12, and 24 months of follow-up

Data expressed as mean (standard deviation) except where otherwise indicated. 2MWT: 2-minute walking test. TUG: Timed up-and-go test. UST: unipodal stance test. Items in bold are statistically significant at p<.05.


Table 3 compares intermediate outcome measures between groups at 6, 12, and 24 months of follow-up. It shows a significant effect of the study intervention on weight loss (-4.2 vs -1.1 kg, p<0.001) and BMI reduction (-1.7 vs -0.5 points, p<0.001) at 6 months of follow-up that, however, tended to attenuate over time. It also points to a slight reduction in fat mass, especially in women at 12 months (-1.3 vs +0.6 kg, p=0.013), but no effect on lean mass. The intervention group compared to the control group achieved lower IL-6 levels at 12 months of follow-up (4.1 vs 5.7 mg/dL, p=0.030), lower insulin levels at 12 and 24 months of follow-up (12.7 vs 16.4 µU/mL; p=0.027 and 12.4 vs 15.1 µU/mL; p=0.025, respectively), and lower insulin resistance (HOMA-IR) at 12 and 24 months of follow-up (3.6 vs 4.8; p=0.016 and 3.4 vs 4.3; p=0.021, respectively). Self-reported QoL was better among intervention group individuals, especially in the first 6 months of the intervention (73.3 vs 68.1; p=0.024).

Table 3. Intermediate outcome measures for the intervention (I) and control (C) groups at 6, 12, and 24 months of follow-up

Data expressed as mean (standard deviation) except where otherwise indicated. BMI: body mass index. CRP: C-reactive protein. HbA1: haemoglobin subunit alpha. HOMA-IR: homeostatic model assessment of insulin resistance. IGF-1: insulin-like growth factor 1. IL-6: interleukin-6. QoL: quality of life. TSF: tricipital skinfold. VAS: 0-10 visual analogue scale. WHR: waist-hip ratio; Items in bold are statistically significant at p<.05.


The GLM analysis showed that the evolution over time of 2MWT, weight, BMI, waist circumference, fat mass, lean mass and cholesterol significantly differ between both study groups (p valeues: 0.001, <0.001, <0.001, 0.008, 0.013, 0.006, 0.037, respetively). It also showed a significant group effect on insulin (p=0.011) and HOMA IR (p=0.039).



Our RCT shows that, for obese 65-75 year old individuals, a multimodal weight-loss intervention based on diet and exercise is safe and effective in reducing weight and normalizing BMI. It may also reduce fat mass, improve certain metabolic and inflammatory parameters, and enhance self-reported QoL. However, not only was it not possible to demonstrate an effect in preventing frailty and functional decline, benefits progressively disappeared within a few months of the intervention ending.
There is abundant evidence on the positive effects of physical exercise and healthy diet in improving muscle strength and physical performance in older populations (27). The LIFE study showed that a structured, moderate-intensity physical activity program consisting of aerobic, resistance, and flexibility training activities reduced major disability over 2.6 years among older adults at risk for disability (29). Several RCTs that have assessed the effect of interventions on frailty prevention or reversion as the main outcome measure (30-33) indicate that physical, nutritional, and cognitive interventions are effective in improving frailty scores in pre-frail and frail older people. However, as far as we are aware, no RCTs have assessed the effects of weight loss on mid- and long-term prevention of frailty in obese older people. While not reaching any definitive conclusion, our RCT brings new evidence to bear. During the 6 months of the intervention, the worsening in frailty status in the intervention group (8%) was half that in the control group (16%), for a difference that was very nearly statistically significant (p=0.069). After 24 months of follow-up, however, no clearly significant effect of the study intervention in preventing frailty was observed. On the other hand, at 12 and 24 months, intervention group individuals experienced a higher 0.07 m/s improvement in gait speed in comparison to control group individuals. Although this improvement in functional capacity may seem poor, it is worth remembering the clinical relevance of gait speed as an indicator of functional performance, frailty and health, and noting that gait speed is strongly associated with survival in older adults (34). A pooled analysis of 9 cohort studies shows that survival increases across the full range of gait speeds, with significant increments (hazard ratio 0.88) per 0.1 m/s (34). Our rather modest results regarding the prevention of frailty and functional decline may be explained in several ways. Firstly, the study sample was relatively young (mean age 69 years) and had good baseline functional capacity. Secondly, the intervention was specifically designed to lose weight, not to improve muscle strength; a more intensive physical activity program that included sensorimotor aspects, as well as training in activities of daily living, would likely have improved results. A program to eliminate obesity alone, if not accompanied by a physical strengthening program, would not appear to be enough to enhance strength and functional capacity over the short-to-medium term. Thirdly, follow-up to 24 months was insufficient as this was too short a period for frailty to develop in our relatively young and functional population: obesity’s impact on frailty is a long-term effect, and likewise the effect of reducing obesity on frailty prevention. Fourthly, losses to follow-up were higher than expected and, especially during the second year, occurred mostly among frailer people, consequently limiting the statistical power of the study to identify differences between the intervention and control groups. Finally, intervention effects were evident during it but progressively fell off after its conclusion. The fact that the positive effect on weight, BMI improvement, and waist circumference was lost once the intervention concluded highlights the need to maintain such interventions over more prolonged periods of time and to incorporate the intervention guidelines as lifestyle changes in order to observe long-term effects. A higher duration of the diet and physical exercise intervention could provide a longer effect duration, but maintaining the intervention for prolonged periods in aged population is a recognised difficult challenge (35). While our results do not support the main hypothesis of study, they do not rule it out. Thus, further studies with large enough samples, more prolonged interventions, and longer follow-up periods are required to confirm the hypothesis that weight loss in obese older adults may contribute to the prevention of frailty and functional decline. Furthermore, the additional inclusion of intensive interventions designed to improve muscle strength would contribute to the prevention of frailty and functional decline in this population.
An interesting result of the study intervention is the observed effect in reducing fat mass, especially in women, and correlating with improved insulin resistance (HOMA-IR) and inflammation (IL-6) indicators. Reducing excess weight and fat mass could have a long-term effect in reducing frailty, as both insulin resistance and inflammation have been related with frailty incidence (36). Clinical relevance of weight loss is reflected in the improvement of inflammatory and insulin resistance parameters. In our study, weight loss had a high impact on reducing, and even eliminating, treatment with oral anti-diabetics or insulin in some patients. No improvement was observed in lean mass or in muscle mass, however, which would reaffirm the risk of muscle mass loss deriving from weight loss diets in older people. Such diets should be restricted only to older adults who are obese (BMI≥30 kg/m2) and who have comorbidities, functional limitations, or metabolic complications related to excess weight, should be accompanied by a muscle strengthening program (20), and should be strictly supervised. Moreover, weight loss can carry a special risk in older people with sarcopenic obesity as it can accentuate loss of muscle mass and bone mass and increase the risk of falls and fractures (22). This is why in people with sarcopenic obesity, interventions should prioritize the increase in muscle mass and function over weight loss. Although loss of strength is a more consistent risk for disability and death than is loss of muscle mass (37), both must be targets of interventions to prevent frailty. Finally, it is also worth mentioning that intervention group individuals had better self-reported QoL during the intervention than control group individuals; however, this effect also disappeared within 6 months after the intervention ended. Although it may be difficult for the subjects themselves to perceive the short-term positive effects of physical activity on health (some physiological parameters may improve, but may only be perceptible if analysed), regular exercise has clear psychological benefits for perceptions of health and wellbeing and even for improved socialization and self-esteem (38). These results agree with those reported by the published RCT on the effect of weight loss interventions in older adults with obesity (39). These trails provided evidence of low to moderate quality and suggest that weight reduction, especially fat mass reduction with preserved lean mass, can lead to improvements in physical function and quality of life (39). However, further well-designed RCTs are needed in aged obese population to provide definitive guidance in clinical practice. Adherence of older adults to physical exercise programs tends to be poor. Barriers to exercising include age, being female, fatigue, health problems, pain, and lack of motivation and willpower; in contrast, the desire to spend time outdoors and in nature are protective factors for exercising (28). Poor adherence – defined as attendance at less than 50% of the sessions with the dietician or physiotherapist – was observed in almost half the individuals in our study, a similar percentage to that observed in other studies (32). Compliance with visits with the dietician (individual sessions) was better (67%) than with the physiotherapist (practical group sessions with a less flexible schedule) (40%). Better compliance was associated with a higher effect of the study intervention in terms of weight loss, reduced BMI, fat mass, and insulin levels, and was also related to improved gait speed. Future lines of research in the field of frailty prevention in obese seniors should include; a) the design of new therapeutic strategies with more effective combinations of diet and exercise, b) the establishment of the type of exercise with the greatest impact on functional improvement and its optimal duration, c) the assessment of the interactions with other types of interventions that may potentiate the expected effect, and interactions with other clinical conditions (chronic inflammation, dehydration, etc.), or d) studies on how to improve compliance, adherence, or motivation for a change in healthier habits in aged population.
While the RCT design is a key strength, our study also has some weaknesses. A first limitation is the lack of sufficient statistical power, mainly due to greater losses to follow-up than expected. Follow-up losses were slightly higher in the intervention group and were associated with female gender, sarcopenia, number of medications, no outdoors life, poor physical activity, and depression. While the losses occurred homogeneously in both study groups (reducing possible bias), they mainly reflected people with greater frailty or poorer functional capacity (thereby diluting the possible effects). Another limitation was that it was impossible to keep participants blind to the study intervention and, thus, the difficulty to mask the evaluators. Although not a double-blind study, we think that the main outcome measures followed clear, explicit, and precise criteria, and standardized and validated procedures. However, it must be noted that while frailty is a clinical condition that is well accepted by geriatricians and the scientific community, there is no consensus regarding diagnostic criteria. In this study, we used the very widely used and validated criteria of Linda Fried (5), which, however, have limitations related to their exclusively physical orientation and reliability when applied by non-expert evaluators. Moreover, the study intervention consisted of a combination of exercise and diet at the same time and, therefore, it was impossible distinguishing the effects of these two components. The observed effect must be considered due to the multimodal intervention as a whole. Finally, as mentioned earlier, our relatively young and robust study sample, the non-extension of the intervention beyond 6 months, and an overly short follow-up period (24 months) restricted possibilities of proving the hypothesis regarding the prevention of frailty and functional decline.
In conclusion, this study has not conclusively demonstrated that weight loss and BMI normalization in obese people aged 65-75 years reduces frailty risk at 24 months of follow-up. Even so, weight loss and fat mass reduction were accompanied by an improvement in certain inflammatory and insulin resistance parameters, and this may, in turn, have long-term effects in preventing frailty and functional decline. The working hypothesis remain reasonable and plausible, therefore, and should be tested with larger sample sizes and longer interventions and follow-up periods.


Author Contributions: Conceptualization MS-P; methodology MS-P, EP, MP and EB; formal analysis EP; investigation MT, IL, MA, AS, SR and MP; data curation MS-P, EP; writing—original draft preparation MS-P; funding acquisition, MS-P. All authors have read and agreed to the published version of the manuscript.

Funding: This study was funded by grants from the Spanish Ministry of Health (Instituto de Salud Carlos III, Fondo de Investigación Sanitaria (FIS) program PI16/00750).

Institutional Review Board Statement: The study protocol was approved by the local ethics committee (reference number CEIC CSdM 60/16). The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Ethics Committee of Consorci Sanitari del Maresme (CSdM)(protocol code CEIC CSdM-60/16, 26th October 2016).

Informed Consent Statement: Written informed consent was obtained from all subjects involved in the study.

Conflicts of Interest: The authors have no conflicts of interest to disclose.

Ethical standards: All participants gave their informed consent. The study protocol was approved by the local ethics committee (reference number CEIC CSdM 60/16)



1. Morley J, Vellas B, van Kan GA, Anker SD, Bauer JM, Bernabei R, et al. Frailty consensus: a call to action. JAMDA 2013; 14: 392-7.
2. Brown NA, Zenilman ME. The impact of frailty in the elderly on the outcome of surgery in the aged. Advances in Surgery 2010; 44: 229-49.
3. Collard R M, Boter H, Schoevers RA, Oude Voshaar RC. Prevalence of frailty in community-dwelling older persons: a systematic review. J Am Geriatr Soc 2012; 60: 1487-92.
4. Rockwood K, Mitnitski A. Frailty Defined by Deficit Accumulation and Geriatric Medicine Defined by Frailty. Clin Geriatr Med 2011; 27: 17–26.
5. Fried LP, Tangen CM, Walston J, et al. Frailty in older adults: evidence for a phenotype. Journals of Gerontology Series A: Biological Sciences and Medical Sciences 2001; 56 (3): M146-M156.
6. Serra-Prat M, Papiol M, Vico J, Palomera E, Arús M, cabré M. Incidence and risk factors for frailty in the community-dwelling elderly population. A two-year follow-up cohort study. J Gerontol Geriatr Res 2017; 6:6. DOI: 10.4172/2167-7182.1000452.
7. Tze Pin Ng, Liang Feng, Ma Shwe Zin Nyunt, AnisLarbi, Keng Bee Yap. Frailty in older persons: multisystem risk factors and the frailty risk index (FRI) (2014). JAMDA 15: 635-42.
8. Soysal P, Stubbs B, Lucato P, Luchini C, MarcoSolmiM.D., Roberto Peluso M.D., Ahme TuranIsikM.D., Enzo Manzato M.D., Stefania Maggi M.D., Marcello Maggio M.D.nA. Matthew PrinaPh.D.oTheodore D.CoscoPh.D.pYu-TzuWuPh.D.qNicolaVerones. Inflammation and frailty in the elderly: A systematic review and meta-analysis. Ageing Research Reviews 2016; 31: 1-8.
9. Joshua I. Barzilay, MD; Caroline Blaum, MD, MS; Tisha Moore, BA; et al Qian Li Xue, PhD; Calvin H. Hirsch, MD; Jeremy D. Walston, MD; Linda P. Fried, MD, MPH. Insulin Resistance and Inflammation as Precursors of Frailty. The Cardiovascular Health Study. Arch Intern Med. 2007;167(7):635-641. doi:10.1001/archinte.167.7.635
10. Serra-Prat M, Lorenzo I, Palomera E, Yébenes JC, Campins L, Cabré M. Intracellular water content in lean mass is associated with muscle strength, functional capacity and frailty in community-dwelling elderly individuals. A cross-sectional study. Nutrients 2019, 11, 661; doi:10.390/nu11030661.
11. Gutiérrez-Fisac JL, Guallar-Castillón P, León-Muñoz LM, Graciani A, Banegas JR, Rodríguez-Artalejo F. Prevalence of general and abdominal obesity in the adult population of Spain, 2008-2010: the ENRICA study. Obes Rev 2012;13(4):388-92.
12. 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. J Am Geriatr Soc. 2005;53(6):927-34.
13. Woods NF, LaCroix AZ, Gray SL, Aragaki A, Cochrane BB, Brunner RL, Masaki K, Murray A, Newman AB, Women’s Health Initiative. Frailty: emergence and consequences in women aged 65 and older in the Women’s Health Initiative Observational Study. J Am Geriatr Soc 2005; 53:1321-30.
14. Stenholm S, Strandberg TE, Pitkälä K, Sainio P, Heliövaara 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; 69:73-8.
15. 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 Apr;65(4):377-81.
16. Barzilay JI, Cotsonis GA, Walston J, Schwartz AV, Satterfield S, Miljkovic I, Harris TB, Health ABC Study. Insulin resistance is associated with decreased quadriceps muscle strength in nondiabetic adults aged >or=70 years. Diabetes Care 2009;32(4):736-8. LM
17. García-Esquinas E, García-García FJ, León-Muñoz LM, Carnicero JA, Guallar-Castillón P, Gonzalez-Colaço M, López-García E, Alonso-Bouzón C, Rodríguez-Mañas L, Rodríguez-Artalejo F. Obesity, fat distribution, and risk of frailty in two population-based cohorts of older adults in Spain. Obesity (Silver Spring) 2015 Apr;23(4):847-55.
18. Stenholm S, Tiainen K, Rantanen T, Sainio P, Heliövaara M, Impivaara O, Koskinen S. Long-term determinants of muscle strength decline: prospective evidence from the 22-year mini-Finland follow-up survey. J Am Geriatr Soc. 2012 Jan;60(1):77-85. doi: 10.1111/j.1532-5415.2011.03779.x. Epub 2011 Dec 28. PMID: 22211568.
19. Schrager MA, Metter EJ, Simonsick E, Ble A, Bandinelli S, Lauretani F, Ferrucci L. Sarcopenic obesity and inflammation in the InCHIANTI study. J Appl Physiol (1985). 2007 Mar;102(3):919-25. doi: 10.1152/japplphysiol.00627.2006. Epub 2006 Nov 9. PMID: 17095641; PMCID: PMC2645665.
20. Kennedy RL, Chokkalingham K, Srinivasan R. Obesity in the elderly: who should we be treating, and why, and how? Curr Opin Clin Nutr Metab Care. 2004 Jan;7(1):3-9. doi: 10.1097/00075197-200401000-00002. PMID: 15090896.
21. Kalyani RR, Saudek CD, Brancati FL, Selvin E. Association of diabetes, comorbidities, and A1C with functional disability in older adults: results from the National Health and Nutrition Examination Survey (NHANES), 1999-2006. Diabetes Care. 2010 May;33(5):1055-60. doi: 10.2337/dc09-1597. Epub 2010 Feb 25. PMID: 20185736; PMCID: PMC2858174.
22. Prado CM, Wells JC, Smith SR et al. Sarcopenic obesity: a critical appraisal of the current evidence. Clin Nutr 2012; 31: 583–601.
23. Tian S, Xu Y. Association of sarcopenic obesity with the risk of all-cause mortality: A meta-analysis of prospective cohort studies. Geriatr Gerontol Int 2016; 16: 155–66.
24. Limongi F, Siviero P, Bozanic A, Noale M, Veronese N, Maggi S. The Effect of Adherence to the Mediterranean Diet on Late-Life Cognitive Disorders: A Systematic Review. J Am Med Dir Assoc. 2020 Oct;21(10):1402-1409. doi: 10.1016/j.jamda.2020.08.020. PMID: 32981667.
25. Tsai MC, Lee CC, Liu SC, Tseng PJ, Chien KL. Combined healthy lifestyle factors are more beneficial in reducing cardiovascular disease in younger adults: a meta-analysis of prospective cohort studies. Sci Rep. 2020 Oct 23;10(1):18165. doi: 10.1038/s41598-020-75314-z. PMID: 33097813; PMCID: PMC7584648.
26. Rossi PG, Carnavale BF, Farche ACS, Ansai JH, de Andrade LP, Takahashi ACM. Effects of physical exercise on the cognition of older adults with frailty syndrome: A systematic review and meta-analysis of randomized trials. Arch Gerontol Geriatr. 2020 Dec 10;93:104322. doi: 10.1016/j.archger.2020.104322. Epub ahead of print. PMID: 33360014.
27. Sherrington C, Fairhall N, Kwok W, Wallbank G, Tiedemann A, Michaleff ZA, Ng CACM, Bauman A. Evidence on physical activity and falls prevention for people aged 65+ years: systematic review to inform the WHO guidelines on physical activity and sedentary behaviour. Int J Behav Nutr Phys Act. 2020 Nov 26;17(1):144. doi: 10.1186/s12966-020-01041-3. PMID: 33239019; PMCID: PMC7689963.
28. Sebio R, Serra-Prat M. Opinion of community-dwelling elderly obese about the barriers and facilitators to engage physical exercise. Sport Sci Health 2020; 16: 411–418. https://doi.org/10.1007/s11332-019-00616-3
29. 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):2387-2396.
30. Cesari M, Vellas B, Hsu FC, Newman AB, Doss H, King AC, et al. A physical activity intervention to treat the frailty syndrome in older persons-results from the LIFE-P study. J Gerontol A Biol Sci Med Sci. 2015 Feb;70(2):216-22.
31. Ng TP, Feng L, Nyunt MS, Feng L, Niti M, Tan BY, et al. Nutritional, Physical, Cognitive, and Combination Interventions and Frailty Reversal Among Older Adults: A Randomized Controlled Trial.Am J Med 2015;128(11):1225-1236.
32. Serra-Prat M, Sist X, Domenich R, Jurado L, Saiz A, Roces A, Palomera E, Terradelles M, Papiol M. Effectiveness of an intervention to prevent frailty in community-dwelling older people consulting in primary care. A randomized controlled trial. Age Ageing 2017; 46(3):401-7.
33. Puts MTE, Toubasi S, Andrew MK, Ashe MC, Ploeg J, Atkinson E, Ayala AP, Roy A, Rodríguez Monforte M, Bergman H, McGilton K. Interventions to prevent or reduce the level of frailty in community-dwelling older adults: a scoping review of the literature and international policies. Age Ageing. 2017 May 1;46(3):383-392.
34. Studenski S, Perera S, Patel K, et al. Gait speed and survival in older adults. JAMA. 2011;305(1):50-58.
35. Billot M, Calvani R, Urtamo A, et al. Preserving Mobility in Older Adults with Physical Frailty and Sarcopenia: Opportunities, Challenges, and Recommendations for Physical Activity Interventions. Clin Interv Aging. 2020 Sep 16;15:1675-1690.
36. Barzilay JI, Blaum C, Moore T, Xue QL, Hirsch CH, Walston JD, Fried LP. Insulin resistance and inflammation as precursors of frailty: the Cardiovascular Health Study. Arch Intern Med. 2007 Apr 9;167(7):635-41.
37. Mitchell WK, Williams J, Atherton P, Larvin M, Lund J, Narici M. Sarcopenia, dynapenia, and the impact of advancing age on human skeletal muscle size and strength; a quantitative review. Front Physiol. 2012;3:260.
38. Kazeminia M, Salari N, Vaisi-Raygani A, Jalali R, Abdi A, Mohammadi M, Daneshkhah A, Hosseinian-Far M, Shohaimi S. The effect of exercise on anxiety in the elderly worldwide: a systematic review and meta-analysis. Health Qual Life Outcomes. 2020 Nov 11;18(1):363.
39. Batsis JA, Gill LE, Masutani R, et al. Weight Loss Interventions in Older Adults with Obesity: A Systematic Review of Randomized Controlled Trials Since 2005. J Am Geriatr Soc. 2017; 65(2): 257–268.



A.R. Orkaby1,2, A.B. Dufour3, L. Yang3, H.D. Sesso2, J.M. Gaziano1,2, L. Djousse1,2, J.A. Driver1,2, T.G. Travison3


1. New England GRECC, VA Boston Healthcare System, Boston, MA, USA; 2. Brigham & Women’s Hospital and Harvard Medical School, Boston, MA, USA; 3. Marcus Institute for Aging Research, Hebrew Senior Life, and Harvard Medical School, Boston, MA, USA.

Corresponding Author: Ariela Orkaby, MD MPH, Assistant Professor of Medicine, Harvard Medical School, Geriatric Research, Education and Clinical Center, VA Boston Healthcare System, 150 South Huntington St, Boston, MA 02130, USA, aorkaby@bwh.harvard.edu

J Frailty Aging 2021;in press
Published online September 17, 2021, http://dx.doi.org/10.14283/jfa.2021.36



Background: Mobility limitation is a component of frailty that shares a bidirectional relationship with cardiovascular disease (CVD). Data are limited on the role of established CVD prevention therapies, such as aspirin, for prevention of frailty and mobility limitation.
Objectives: Examine the association between long-term aspirin use and walking speed.
Design, Setting, Participants: Prospective cohort of 14,315 men who participated in the Physicians’ Health Study I, a completed randomized controlled trial of aspirin (1982-1988), with extended post-trial follow-up.
Measurements: Annual questionnaires collected data on aspirin use, lifestyle and other factors. Average annual aspirin use was categorized for each participant: ≤60 days/year and >60 days/year. Mobility was defined according to self-reported walking pace, categorized as: don’t walk regularly (reference), easy/casual <2mph, normal ≥2-2.9mph, or brisk/very brisk ≥3mph. Propensity scoring balanced covariates between aspirin categories. Multinomial logistic regression models estimated odds of being in each self-reported walking category.
Results: Mean age was 70±8 years; mean aspirin use 11 years. There were 2,056 (14.3%) participants who reported aspirin use ≤60 days/year. Aspirin use >60 days/year was associated with drinking alcohol, smoking, hypertension, heart disease and stroke, while ≤60 days/year was associated with anticoagulation use and bleeding history. In all, 13% reported not walking regularly, 12% walked <2 mph, 44% walked ≥2-2.9 mph, and 31% walked ≥3 mph. After propensity score adjustment, regular aspirin use was associated with a faster walking speed. Odds ratios (95% confidence intervals) were 1.16 (0.97 to 1.39), 1.24 (1.08 to 1.43), and 1.40 (1.21 to 1.63) for <2 mph, ≥2-2.9 mph and ≥3 mph, respectively, compared to not walking regularly (p-trend<0.001).
Conclusions: In this cohort of older men, long-term aspirin use is associated with a greater probability of faster walking speed later in life.

Key words: Aspirin, mobility, gait speed, frailty.



One of the most feared consequences of aging is loss of function and resultant loss of independence (1). Mobility is a critical aspect of function (l) independence, and limitations in mobility – specifically as measured by usual walking speed – are associated with an increased risk of morbidity and mortality (2-4). To date, the only proven therapy to prevent mobility limitation is exercise and leisure-time activity (5).
Mobility limitation is often considered a component of frailty, a state of decreased resilience to stressors that is more common in older adults (6) We have previously demonstrated that slow walking speed is associated with cardiovascular disease (CVD) events (7), and it is thought that frailty and dismobility have a bidirectional relationship with CVD, including peripheral vascular disease (PVD) (8-10). This suggests that pharmacologic therapies targeting CVD may prevent frailty and, specifically, mobility limitation, but this has yet to be investigated.
Among such therapies, aspirin is an intriguing option as it has both anti-inflammatory and anti-thrombotic properties which can improve large and small vessel blood flow and muscle function (11-13). Data from the ASPirin in Reducing Events in the Elderly (ASPREE) trial, demonstrated a reduction in persistent ADL disability among older adults randomized to aspirin vs placebo (14). However, in older adults, data on aspirin are limited and conflicting, with increasing risks of major bleeding largely outweighing benefit for those without a history of CVD (15, 16). In prior work we have shown that long term aspirin use in men is associated with a lower risk of frailty (17). We hypothesized that long term aspirin use, especially if started in middle age when risks of aspirin are minimal, would be associated with a lower risk of mobility limitation measured according to walking speed. We used 15-year follow-up data from men enrolled in the extended observational phases of the Physician’s Health Study (PHS) I.




PHS I is a completed double-blind, 2×2 factorial placebo-controlled trial that randomized 22,071 male physicians to either aspirin or placebo, and to beta-carotene or placebo, beginning in 1982 (18-20). At the start of the trial, participants were free of cancer and CVD. The aspirin intervention component of the trial ended early in 1988 at the recommendation of the PHS Data Safety Monitoring Board, due to the highly significant reduction in the rate of total myocardial infarctions among those assigned to aspirin versus placebo (19). Observational follow-up was extended after completion of the trial with annual questionnaires.
Detailed questionnaires were sent to participants annually to collect information on clinical variables, medications, and lifestyle from 1982 through 2012. Questions regarding walking speed were added to the annual questionnaire returned from 2001-2003. Participants responding to this questionnaire were eligible for the cross-sectional analyses described here, which compares participants to one another on the basis of cumulative aspirin exposure irrespective of randomized beta-carotene assignment. Of 14,896 participants who responded to the self-reported walking speed question, 581 were excluded due to missing covariate data. All participants were aged ≥58 years at the time of the walking speed assessment.
All participants executed written informed consent, and the PHS trial and subsequent follow-up were approved by the Institutional Review Board at Brigham and Women’s Hospital in Boston, MA, USA.


For the duration of the aspirin trial, participants were randomized to 325 mg aspirin or placebo every other day. The aspirin arm of the trial was stopped after an average of 60 months of follow-up (19). Once the aspirin arm of the trial ended, treatment crossover to aspirin use among those who had been assigned to placebo was greater than 70% (21). Ongoing use of aspirin was asked on every annual questionnaire, with the following question: “Over the past 12 months, on how many days have you taken aspirin or medication containing aspirin? 0 days, 1-13 days, 14-30 days, 31-60 days, 61-90 days, 91-120 days, 121-180 days, 180+days.” For the analyses reported here, aspirin exposure was computed from responses to the questionnaires sent out 1988-2001, corresponding to the year that walking pace was added. Average annual use of aspirin was summed and categorized for each participant as follows: ≤60 days/year (low use), >61 days/year (regular use) for the follow up years post-trial, as has been done previously in PHS (22). Information on the actual dose of aspirin was not available once the trial had ended.


A previously validated self-assessment of average walking speed was added to the questionnaire in 2001 (23, 24). The question was asked as follows: “What is your usual walking pace?” with the following possible responses: “Don’t walk regularly; Easy, casual (<2 mph); Normal, average (2-2.9mph); Brisk pace (3-3.9mph); Very brisk, striding (4mph or faster)”. After examining the data distribution, we categorized walking pace into 4 categories: don’t walk regularly, easy casual <2mph, normal ≥2-2.9mph, and brisk or very brisk ≥3mph.

Other Baseline Covariates

Information on demographics, comorbidities, and health, including history of bleeding, heart disease, stroke, peripheral artery disease, arthritis, migraine or headache, atrial fibrillation, or anticoagulation use, was drawn from all prior questionnaires to ensure complete capture of data. Smoking status was quantified as “never, past, or current”. Alcohol consumption was categorized as “daily, weekly, monthly, or rarely”. Cumulative non-aspirin non-steroidal anti-inflammatory drug use (NSAID) was quantified as the average number of years of self-reported NSAID use that were >60 days per year.

Statistical Analysis

Because of significant crossover to aspirin after completion of the intervention phase of the trial and dissolution of its randomized groupings, we developed a propensity score to balance the degree of aspirin exposure over a maximum 15 years of the post-intervention follow-up by relevant covariates. Estimation of the propensity score accounted for the initial PHS trial randomization; other variables that might be related to post-intervention aspirin use (e.g. history of bleeding, heart disease) and health-related risk factors and outcomes (e.g. smoking history). Exercise and other potential mediators or moderators of the influence of aspirin on walking speed were excluded from the propensity scoring model. We examined the distribution of propensity scores using kernel density plots to determine whether there was sufficient overlap between exposed vs unexposed (25, 26). We initially considered aspirin exposure in 2 or 3 categories of use (≤60 days vs >60 days or ≤60 days vs 61-180 vs >180 days per years). Two categories had the best overlap and were used for all analyses, as in our prior work (17, 22). Estimation employed inverse probability of treatment weighting (27-29). We compared the standardized differences between baseline covariates before and after propensity score adjustment to ensure <10% difference (25).
Descriptive characteristics were obtained after adjustment for propensity scores. Multinomial logistic regression models estimated the odds of prevalent mobility defined according to self-reported walking speed category among aspirin users relative to non-users (≤60 d/yr) while contrasting each level of walking speed to “Don’t walk regularly” category. Potential effect modification by age, history of heart disease, arthritis, bleeding and exercise frequency were examined by inspection and application of interaction tests, stratification and subgroup analysis.
Model-based estimates of association were accompanied by 95% confidence intervals (95% CI). Prespecified hypothesis testing was performed at an α=0.05 level. All analyses were conducted in SAS version 9.4 (SAS Institute, Cary, NC, USA) and R version 3.3.2 (R Foundation for Statistical Computing, Vienna, Austria).



In total, 14,315 male, predominantly white (93%), physicians were included in this study. At time of walking speed assessment, the mean (standard deviation; SD) age was 70 (8) years with range 58-100 years; mean (SD) duration of aspirin use was 11 (5) years (range 0-18 years). Aspirin use was reported as ≤60 days/year by 2,056 (14.4%) participants. For walking speed, 13.1% reported not walking regularly, 12.2% walked <2 mph, 44.1% walked ≥2-2.9 mph, and 30.7% walked ≥3 mph.
Prior to propensity score adjustment, those who reported greater aspirin use were more likely to report drinking alcohol daily, prior smoking, and a history of atrial fibrillation, diabetes, hypertension, coronary heart disease (CHD), and stroke. Lower frequency of aspirin use was associated with greater anticoagulation use and bleeding history such as gastritis and melena. Following propensity score IPTW adjustment, there were no significant differences between groups (Table 1). With adjustment for propensity score, we found a graded association between regular aspirin use and faster walking speed. The multiplicative increase (95% CI) attributable to high aspirin use in odds of faster walking speed as compared to not regularly walking was 1.16 (0.97 to 1.39), 1.24 (1.08 to 1.43), and 1.40 (1.21 to 1.63) for those reporting walking speeds of <2 mph, ≥2-2.9 mph and ≥3 mph, respectively, (p-trend<0.001) (Table 2). In a sensitivity analysis, after adjusting for the propensity score in the crude model, results were unchanged.

Table 1. Cohort characteristics of 14,315 PHS participants according to average annual aspirin use, before and after propensity score adjustment

Table 2. Association between regular aspirin use and self-reported walking pace in 14,315 PHS participants before and after propensity score weighting

Legend: Sample sizes are provided for each category to allow for calculation of the raw ORs. For example, comparing “normal” walkers with those who “don’t walk” the estimated unadjusted OR is 1.18, as shown in the table. This is interpreted as: for an individual in the high-aspirin group, the relative odds of being in the “normal» walking group than the «don’t walk” group are18% greater with high aspirin exposure vs low aspirin exposure.


Subgroup analyses generally showed little evidence of differential effects across pre-specified categories of risk factors, including age, arthritis, bleeding or exercise frequency. The only subgroup in which there was a statistically significant interaction was among those with history of CHD (p<0.0008). The OR (95% CI) for those with prior CHD and high aspirin use (vs low) was 1.15 (0.80 to 1.64), 1.72 (1.26 to 2.35), and 2.84 (1.93 to 4.16) for those reporting <2 mph, ≥2-2.9 mph and 3+ mph, respectively, compared to 1.18 (0.96 to 1.45), 1.15 (0.98 to 1.34), and 1.24 (1.05 to 1.46) for those without CHD. (Figures 1 and 2).

Figure 1. Forest plot showing the relative odds of increased mobility according to level of aspirin use, stratifying by age and and history of heart disease. The referent mobility group is those who do not engage in regular walking

Figure 2. Forest plot showing the relative odds of increased mobility according to level of aspirin, stratifying on history of arthritis, bleeding, and weekly exercise. The referent mobility group is those who do not engage in regular walking



In this 15-year post-trial follow up study of 14,315 male physicians, long term aspirin use, started in middle age, was associated with a greater probability of faster walking speed in late life. Our results suggest that decreased walking speed might be prevented through regular aspirin use, independent of medical history and health behaviors. Importantly, however, results suggest that after consideration of these factors, the associations due to aspirin exposure are more strongly manifested among older individuals with a history of heart disease. This observation is in line with the hypothesis that those cardiovascular morbidity and frailty have shared underlying pathophysiology and may be targeted by regular aspirin use.
Aspirin lowers the risk of cardiovascular events by both preventing platelet aggregation and reducing inflammation. In patients with PVD, aspirin is an established part of the treatment plan, either alone or in combination with other antiplatelets or anticoagulants (30, 31). However, clinical trial data has not consistently demonstrated a benefit of aspirin of intermittent claudication. Although aspirin has anti-inflammatory and vasodilatory properties which might improve walking speed, prior clinical data in those with PVD does not support this hypothesis. While most older adults do not have clinical PVD, the microvascular changes that occur with aging contribute to decreased blood flow to leg muscles and may explain in part the natural slowing in walking and increased sarcopenia that is seen with physiologic aging (3, 32). We hypothesize that regular use of aspirin over years, in late middle age and early old age, may improve blood flow and circulation to distal muscles particularly in those with CVD, and therefore lead to less mobility limitation in later decades, in those with heart disease in particular. This is likely mediated by the vasodilatory and anti-inflammatory properties of aspirin.Randomized trials of aspirin for mobility are limited. The ASPREE trial examined the effect of aspirin on disability-free survival, CV events, and mortality in adults aged 70 and older, over 5-years (15, 33, 34). In secondary analyses there was a signal for a lower rate of persistent physical disability (HR 0.85, 95% CI 0.70 to 1.03) (33). and possible benefit for mobility. However, there are serious concerns regarding the safety of aspirin that need to be considered, particularly related to fatal bleeding events (15). It is possible that any benefit of aspirin for prevention of mobility limitation would need to begin earlier in life, however we found that the association between long-term use of aspirin and lower risk of slow walking remained even after accounting for age.
In subgroup analyses we explored whether aspirin use is more effective in certain disease states which may be impacted by aspirin use. Although overall results were similar, there was a statistically significant interaction for those with a history of CHD and a non-significant suggestion of greater benefit among those who do not exercise regularly. Cardiovascular disease has multiple causes, including increased inflammation and thrombotic milieu. That those with CHD had evidence of greater benefit from aspirin may suggest that those using aspirin for secondary prevention of CVD might improve walking speed later in life by reducing their burden of CVD over the lifespan by preventing future CV events and resultant disability. On the other hand, those who exercise may have sufficient benefit from improving blood flow to extremities (35) that the addition of aspirin does not further improve walking. Exercise has been shown to be one of the few modalities that can improve function and frailty at all stages (36), and it is possible that the benefits of aspirin are most beneficial in those who are inactive or who have biologically active CVD.
There are several important limitations to consider. First, this cohort is entirely male and largely made up of individuals identifying as white within a US context, and the study should be repeated in a cohort of women. However, women are also more likely to outlive men by 6-8 years and data that is sex specific is needed in aging research. Second, walking speed was assessed using self-report which could either over or under estimate actual walking pace, although the question used has been well validated (23, 24). Although the end-point mixes aspects of physical activity, function, and ability or disability, we interpret this as an important measure of actualized functional capacity in the real-world setting (21). Third, even though we used propensity score methods to address confounding by indication, reverse causality remains possible. Fourth, information on dose of aspirin was not available after completion of the trial and we are unable to examine the role of aspirin dose on the outcome. Future work to understand the relationship between aspirin and mobility could include plasma analysis for inflammatory biomarkers and mediation analysis to understand what role aspirin may play in reducing inflammation and preventing frailty.
In conclusion, in this cohort of older men, we found that long-term aspirin use started in late midlife was associated with lower odds of mobility limitation, as defined by self-reported walking speed. Future work should replicate these findings in women and further explore the protentional role of aspirin as a preventive strategy for frailty and mobility limitation.


Acknowledgments: The authors gratefully acknowledge the patients and research staff who participated in the Physicians’ Health Study. We are grateful for Natalya Gomelskaya for her statistical support. All authors contributed to study conception, writing, and editing the manuscript. L.Y., A.B.D., and T.G.T. conducted the statistical analysis. The sponsor had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript.

Funding: This work was supported by a Career Development award to A.R.O. from the Boston Claude D. Pepper Older Americans Independence Center, National Institute on Aging grant P30-AG013679. A.R.O. is also supported by Veterans Affairs CSR&D CDA-2 IK2-CX001800 and National Institute on Aging grant R03-AG060169. The Physicians’ Health Study is funded by grants CA-34944, CA-40360, and CA-097193 from the National Cancer Institute and grants HL-26490 and HL-34595 from the National Heart, Lung, and Blood Institute, Bethesda, MD.

Conflicts of Interest: J.M.G. reports serving as a consultant and receiving honoraria for speaking for Bayer.

Ethical standards: All participants provided written informed consent. The trial was approved by the Institutional Review Board at Brigham and Women’s Hospital.



1. Parry SW, Finch T and Deary V. How should we manage fear of falling in older adults living in the community? BMJ (Clinical research ed). 2013;346:f2933. doi: 10.1136/bmj.f2933
2. Studenski S, Perera S, Patel K, Rosano C, Faulkner K, Inzitari M, Brach J, Chandler J, Cawthon P, Connor EB, Nevitt M, Visser M, Kritchevsky S, Badinelli S, Harris T, Newman AB, Cauley J, Ferrucci L and Guralnik J. Gait speed and survival in older adults. Jama. 2011;305:50-8. doi: 10.1001/jama.2010.1923
3. Fritz S and Lusardi M. White paper: «walking speed: the sixth vital sign». Journal of geriatric physical therapy (2001). 2009;32:46-9.
4. Middleton A, Fritz SL and Lusardi M. Walking speed: the functional vital sign. Journal of aging and physical activity. 2015;23:314-22. doi: 10.1123/japa.2013-0236
5. Brown CJ and Flood KL. Mobility limitation in the older patient: a clinical review. Jama. 2013;310:1168-77. doi: 10.1001/jama.2013.276566
6. Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, Seeman T, Tracy R, Kop WJ, Burke G and McBurnie MA. Frailty in older adults: evidence for a phenotype. The journals of gerontology Series A, Biological sciences and medical sciences. 2001;56:M146-56. doi: 10.1093/gerona/56.3.M146
7. Imran TF, Orkaby A, Chen J, Selvaraj S, Driver JA, Gaziano JM and Djoussé L. Walking pace is inversely associated with risk of death and cardiovascular disease: The Physicians’ Health Study. Atherosclerosis. 2019;289:51-56. doi: 10.1016/j.atherosclerosis.2019.08.001
8. Afilalo J, Alexander KP, Mack MJ, Maurer MS, Green P, Allen LA, Popma JJ, Ferrucci L and Forman DE. Frailty assessment in the cardiovascular care of older adults. Journal of the American College of Cardiology. 2014;63:747-62. doi: 10.1016/j.jacc.2013.09.070
9. Bouillon K, Batty GD, Hamer M, Sabia S, Shipley MJ, Britton A, Singh-Manoux A and Kivimaki M. Cardiovascular disease risk scores in identifying future frailty: the Whitehall II prospective cohort study. Heart (British Cardiac Society). 2013;99:737-42. doi: 10.1136/heartjnl-2012-302922
10. Newman AB, Gottdiener JS, McBurnie MA, Hirsch CH, Kop WJ, Tracy R, Walston JD and Fried LP. Associations of subclinical cardiovascular disease with frailty. The journals of gerontology Series A, Biological sciences and medical sciences. 2001;56:M158-66. doi: 10.1093/gerona/56.3.M158
11. Vane JR and Botting RM. The mechanism of action of aspirin. Thrombosis research. 2003;110:255-8. doi: 10.1016/S0049-3848(03)00379-7
12. Oh J, Sinha I, Tan KY, Rosner B, Dreyfuss JM, Gjata O, Tran P, Shoelson SE and Wagers AJ. Age-associated NF-kappaB signaling in myofibers alters the satellite cell niche and re-strains muscle stem cell function. Aging. 2016;8:2871-2896. doi: 10.18632/aging.101098
13. Morris T, Stables M, Hobbs A, de Souza P, Colville-Nash P, Warner T, Newson J, Bellingan G and Gilroy DW. Effects of low-dose aspirin on acute inflammatory responses in humans. Journal of immunology (Baltimore, Md : 1950). 2009;183:2089-96. doi: 10.4049/jimmunol.0900477
14. Woods RL, Espinoza S, Thao LTP, Ernst ME, Ryan J, Wolfe R, Shah RC, Ward SA, Storey E, Nelson MR, Reid CM, Lockery JE, Orchard SG, Trevaks RE, Fitzgerald SM, Stocks NP, Williamson JD, McNeil JJ, Murray AM and Newman AB. Effect of Aspirin on Activities of Daily Living Disability in Community-Dwelling Older Adults. The journals of gerontology Series A, Biological sciences and medical sciences. 2020. doi: 10.1093/gerona/glaa316
15. McNeil JJ, Wolfe R, Woods RL, Tonkin AM, Donnan GA, Nelson MR, Reid CM, Lockery JE, Kirpach B, Storey E, Shah RC, Williamson JD, Margolis KL, Ernst ME, Abhayaratna WP, Stocks N, Fitzgerald SM, Orchard SG, Trevaks RE, Beilin LJ, Johnston CI, Ryan J, Radziszewska B, Jelinek M, Malik M, Eaton CB, Brauer D, Cloud G, Wood EM, Mahady SE, Satterfield S, Grimm R and Murray AM. Effect of Aspirin on Cardiovascular Events and Bleeding in the Healthy Elderly. The New England journal of medicine. 2018;379:1509-1518. doi: 10.1056/NEJMoa1805819
16. Bibbins-Domingo K. Aspirin Use for the Primary Prevention of Cardiovascular Disease and Colorectal Cancer: U.S. Preventive Services Task Force Recommendation Statement. Annals of internal medicine. 2016;164:836-45. doi: 10.7326/M16-0577
17. Orkaby AR, Yang L, Dufour AB, Travison TG, Sesso HD, Driver JA, Djousse L and Gaziano JM. Association Between Long-Term Aspirin Use and Frailty in Men: The Physicians’ Health Study. The journals of gerontology Series A, Biological sciences and medical sciences. 2020. doi: 10.1093/gerona/glaa233
18. Findings from the aspirin component of the ongoing Physicians’ Health Study. The New England journal of medicine. 1988;318:262-4. doi: 10.1056/NEJM198801283180431
19. Final report on the aspirin component of the ongoing Physicians’ Health Study. Steering Committee of the Physicians’ Health Study Research Group. The New England journal of medicine. 1989;321:129-35. doi: 10.1056/NEJM198907203210301
20. Hennekens CH, Buring JE, Manson JE, Stampfer M, Rosner B, Cook NR, Belanger C, LaMotte F, Gaziano JM, Ridker PM, Willett W and Peto R. Lack of effect of long-term supplementation with beta carotene on the incidence of malignant neoplasms and cardiovascular disease. The New England journal of medicine. 1996;334:1145-9. doi: 10.1056/NEJM199605023341801
21. Sturmer T, Glynn RJ, Lee IM, Manson JE, Buring JE and Hennekens CH. Aspirin use and colorectal cancer: post-trial follow-up data from the Physicians’ Health Study. Annals of internal medicine. 1998;128:713-20. doi:10.7326/0003-4819-128-9-199805010-00003
22. Driver JA, Logroscino G, Lu L, Gaziano JM and Kurth T. Use of non-steroidal anti-inflammatory drugs and risk of Parkinson’s disease: nested case-control study. BMJ (Clinical research ed). 2011;342:d198. doi: 10.1136/bmj.d198
23. Wolf AM, Hunter DJ, Colditz GA, Manson JE, Stampfer MJ, Corsano KA, Rosner B, Kriska A and Willett WC. Reproducibility and validity of a self-administered physical activity questionnaire. International journal of epidemiology. 1994;23:991-9. doi: 10.1093/ije/23.5.991
24. Hu FB, Sigal RJ, Rich-Edwards JW, Colditz GA, Solomon CG, Willett WC, Speizer FE and Manson JE. Walking compared with vigorous physical activity and risk of type 2 diabetes in women: a prospective study. Jama. 1999;282:1433-9. doi: 10.1001/jama.282.15.1433
25. Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Statistics in medicine. 2009;28:3083-107. doi: 10.1002/sim.3697
26. Crump RK, Hotz VJ, Imbens GW and Mitnik OA. Dealing with limited overlap in estimation of average treatment effects. 2009;96:187-199. doi: 10.1093/biomet/asn055
27. Elze MC, Gregson J, Baber U, Williamson E, Sartori S, Mehran R, Nichols M, Stone GW and Pocock SJ. Comparison of Propensity Score Methods and Covariate Adjustment: Evaluation in 4 Cardiovascular Studies. Journal of the American College of Cardiology. 2017;69:345-357. doi: 10.1016/j.jacc.2016.10.060
28. Curtis LH, Hammill BG, Eisenstein EL, Kramer JM and Anstrom KJ. Using inverse probability-weighted estimators in comparative effectiveness analyses with observational databases. Medical care. 2007;45:S103-7. doi: 10.1097/MLR.0b013e31806518ac
29. Kim DH, Pieper CF, Ahmed A and Colon-Emeric CS. Use and Interpretation of Propensity Scores in Aging Research: A Guide for Clinical Researchers. Journal of the American Geriatrics Society. 2016;64:2065-2073. doi: 10.1111/jgs.14253
30. Gerhard-Herman MD, Gornik HL, Barrett C, Barshes NR, Corriere MA, Drachman DE, Fleisher LA, Fowkes FG, Hamburg NM, Kinlay S, Lookstein R, Misra S, Mureebe L, Olin JW, Patel RA, Regensteiner JG, Schanzer A, Shishehbor MH, Stewart KJ, Treat-Jacobson D and Walsh ME. 2016 AHA/ACC Guideline on the Management of Patients With Lower Extremity Peripheral Artery Disease: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2017;135:e686-e725. doi: 10.1161/CIR.0000000000000470
31. Hess CN and Hiatt WR. Antithrombotic Therapy for Peripheral Artery Disease in 2018. Jama. 2018;319:2329-2330. doi: 10.1001/jama.2018.5422
32. Proctor DN, Shen PH, Dietz NM, Eickhoff TJ, Lawler LA, Ebersold EJ, Loeffler DL and Joyner MJ. Reduced leg blood flow during dynamic exercise in older endurance-trained men. J Appl Physiol (1985). 1998;85:68-75. doi: 10.1152/jappl.1998.85.1.68
33. McNeil JJ, Woods RL, Nelson MR, Reid CM, Kirpach B, Wolfe R, Storey E, Shah RC, Lockery JE, Tonkin AM, Newman AB, Williamson JD, Margolis KL, Ernst ME, Abhayaratna WP, Stocks N, Fitzgerald SM, Orchard SG, Trevaks RE, Beilin LJ, Donnan GA, Gibbs P, Johnston CI, Ryan J, Radziszewska B, Grimm R and Murray AM. Effect of Aspirin on Disability-free Survival in the Healthy Elderly. The New England journal of medicine. 2018;379:1499-1508. doi: 10.1056/NEJMoa1800722
34. McNeil JJ, Nelson MR, Woods RL, Lockery JE, Wolfe R, Reid CM, Kirpach B, Shah RC, Ives DG, Storey E, Ryan J, Tonkin AM, Newman AB, Williamson JD, Margolis KL, Ernst ME, Abhayaratna WP, Stocks N, Fitzgerald SM, Orchard SG, Trevaks RE, Beilin LJ, Donnan GA, Gibbs P, Johnston CI, Radziszewska B, Grimm R and Murray AM. Effect of Aspirin on All-Cause Mortality in the Healthy Elderly. The New England journal of medicine. 2018;379:1519-1528. doi: 10.1056/NEJMoa1803955
35. Hildebrandt W, Schwarzbach H, Pardun A, Hannemann L, Bogs B, König AM, Mahnken AH, Hildebrandt O, Koehler U and Kinscherf R. Age-related differences in skeletal muscle microvascular response to exercise as detected by contrast-enhanced ultrasound (CEUS). PLoS One. 2017;12:e0172771. doi: 10.1371/journal.pone.0172771
36. Clegg A, Young J, Iliffe S, Rikkert MO and Rockwood K. Frailty in elderly people. Lancet. 2013;381:752-62. doi: 10.1016/S0140-6736(12)62167-9



S. Tejiram1,2, C. Galet3, J. Cartwright4, V.H. Hatcher3, D.A. Skeete3, C. Cocanour2, K.S. Romanowski1


1. Department of Surgery, Division of Burn Surgery, University of California, Davis, Sacramento, CA. 2315 Stockton Blvd, Sacramento, CA, USA; 2. Department of Surgery, Division of Trauma Surgery, University of California, Davis, 2315 Stockton Blvd, Sacramento, CA, USA; 3. Department of Surgery, Division of Acute Care Surgery, University of Iowa, 200 Hawkins Dr., Iowa City, IA, USA; 4. University of Michigan School of Medicine. 1500 E Medical Center Dr, Ann Arbor, MI, USA.

Corresponding Author: Kathleen S. Romanowski, MD, University of California, Davis, and Shriners Hospitals for Children Northern CaliforniaN Stockton Blvd. Suite 718, Sacramento, CA 95762, Phone: 916-453-5022, Email: ksromanowski@ucdavis.edu

J Frailty Aging 2021;in press
Published online June 17, 2021, http://dx.doi.org/10.14283/jfa.2021.25



Purpose: The older population is particularly vulnerable to traumatic injury. Frailty scores, used to estimate the physiologic status of an individual, are key to identifying those most at risk for injury. Global health measures such as the Veterans RAND 12 Item Health Survey (VR-12) are quality of life measures that assess older adults’ overall perception of their health and may serve as a useful adjunct when predicting frailty. Herein, we evaluated whether components of the VR-12 correlated with worse frailty scores over time.
Methods: Older adults (≥65) admitted to burn, trauma, or emergency general surgery services were prospectively enrolled. Demographics, frailty determined using the Trauma Specific Frailty Index (TSFI), and VR-12 surveys were collected at enrollment and 3, 6, 9, and 12-month follow-ups. A physical component score (PCS) and mental component score (MCS) was produced by VR-12 surveys for comparison purposes.
Results: Fifty-eight patients were enrolled, of which 8 died. No significant changes in PCS (p = 0.25) and MCS (p = 0.56) were observed over time. PCS (p = 0.97) and MCS (p = 0.78) at enrollment did not predict mortality. PCS (OR = 0.894 [0.84-0.95], p = 0.0004) and age (OR = 1.113 [1.012-1.223], p = 0.03) independently predicted enrollment frailty.
Conclusion: These global measures of health could be utilized in lieu or in addition to frailty scores when assessing patients in the setting of acute injury. Studies are warranted to confirm this association.

Key words: Frailty, older adults, trauma, VR-12.



Health related quality of life measures obtained from individuals can provide information on a wide range of outcomes such as disease burden and treatment effectiveness (1, 2). These self-reported outcomes can be collected and tracked to better predict resource utilization and mortality (3, 4). Global health measures are typically generic and concise to allow application to a wide variety of patient populations (2). Though these measures are not explicitly used in economic evaluations, there is increasing interest in incorporating global health perceptions into cost analysis to better inform health care decision making (5, 6). The Centers for Medicare & Medicaid Services have adopted global health perception assessments in their cost-effectiveness analysis and the Health Outcomes Survey (HOS) became the first Medicare managed outcomes measure in the Healthcare Effectiveness Data and Information Set (HEDIS) (7). In this context, surveys such as the Veteran RAND 12-Item Health Survey (VR-12) have become commonly used in population-based studies like the Medicare HOS to more efficiently utilize health care resources (8, 9).
Global health measures like the VR-12 are quality of life measures that assess an individual’s overall perception of their health and were originally developed from the Short Form-36 (SF-36) Health Survey. The SF-36 is a 36-item questionnaire that assesses patient health across eight dimensions such as health perceptions, physical and social functioning, and mental health to produce summary scores for physical and mental health. Individuals provide self-assessment and perceptions of their health by answering questions about their perceived limitations. For example, one may answer a question about their ability to ambulate as limited by a lot, a little, or not at all. These answers correspond to a score of 1, 2, or 3, respectively, to calculate a score range from 10 to 30 that is later transformed to a 100-point scale for comparison purposes. In this survey, each answer is given equal weight for score determination. Previous literature has established the significance of these scores to detect meaningful differences following intervention among clinic patient groups with various disease processes (2, 5, 10-12). The Veterans RAND 36-Item Health Survey (VR-36) was developed from the SF-36 and both were reduced and adapted from a 36-item survey to 12 items with minimal information loss to create the VR-12 survey (13-17).
The VR-12 is freely available to assess health related quality of life and disease burden through a 12 item questionnaire focused again on eight dimensions: general health, physical functioning, changes and limitations in physical and emotional health, pain, energy, mental health, and social functioning (2, 15, 18). Answers to these questions produce a physical component score (PCS) and a mental component score (MCS). The PCS focuses more on questions about general health, physical functioning and pain whereas the MCS places more emphasis on emotional, mental health, and social functioning questions (18).
Quality of life measures that assess an individual’s overall perception of their health may serve as a useful adjunct when predicting frailty. Frailty utilizes the physiologic, environmental, and behavioral characteristics of an individual to determine their vulnerability to an adverse event. Frailty scores determined by these characteristics can be used to estimate the physiologic status of an individual and are key in identifying those most at risk. Frailty has been identified as a major predictor of poor outcomes and has been incorporated as a prognostication tool for an adverse event in the acute injury setting (19-23). The older population is particularly vulnerable to traumatic injury. Fall-related injuries represent a significant portion of traumatic injury experienced by older adults (24-26). Fall related injury and frailty are inter-related. Frailty itself may help better identify those more at risk for fall injury, but fall injuries can also be considered an adverse event secondary to frailty (27). For those individuals who sustain a fall injury, their frailty may worsen and lead to increasing recidivism with repeat traumatic injuries related to falls or other adverse events. The fear of falling itself can also lead to decreased activity, thereby increasing frailty. With the older population expected to double in size by the year 2060, the incidence of trauma in older adults is expected to increase considerably (28). There is a myriad of reasons for this, related to multiple dynamics that include the increasing amount of care needed for this rising population, increased use of facility assistance like nursing homes and assisted living facilities, increasing need to recognize therapy interactions like benzodiazepine use, limited staff, and geriatrician availability, and limited resources that increase the risk of traumatic injury in this population. Regardless, the rising population density of older individuals has resulted in an increased number of geriatric trauma overall. As a result, there is increasing interest in identifying models and scoring systems that better predict and identify older individuals most at risk for falls.
While frailty scores are typically assessed by providers, patient self-assessment by global health surveys may provide additional information on their risk burden at the time of presentation in the acute injury setting. Much like frailty assessments, self-reported measures of global health usually occur at the time of injury or admission and does not consider the long-term dynamics of patient management or rehabilitation as their course and disposition evolve over time. As a result, there is a paucity of literature examining global health measures in those with acute injury. The aim of this study was to evaluate measures of global health in acutely injured adults by VR-12 assessment, to assess whether components of the VR-12 changed over time, and whether VR-12 assessment correlated with frailty scores.



This prospective observational study was approved by our Institutional Review Board. After determining eligibility by assessing whether patients met the study’s inclusion and exclusion criteria, they were consented and observed for one year following hospital discharge. A written informed consent was obtained from all participants.

Inclusion and Exclusion Criteria

Patients included for study were those aged 65 years and older who presented with an acute injury and admitted to our institution’s burn, trauma, or emergency general surgery (EGS) services from December 2016 to June 2017. Patients that were non-English speakers, unable to participate in, discuss, or assess their own health and social situation, or those who did not have participating family members that could answer questions about a patient’s functional status or medical history were excluded from study.

Study Design

A questionnaire was provided to a patient or appropriate surrogate at enrollment and at 3, 6, 9, and 12 months intervals following discharge to collect information on demographics, general health and comorbidities, and objective information necessary for global health assessment. Patients completed questionnaires by mail, email, phone, or at scheduled follow up appointments. Data collected from the questionnaires was utilized to determine frailty using the Trauma Specific Frailty Index (TSFI) and VR-12 scoring. The TSFI is a validated fifteen variable assessment that considers patient attributes such as comorbidities, ability to perform daily activities, and general life attitudes to ultimately determine frailty (29). Each variable can score up to one point depending on the severity of disease and the total score is divided by 15 to determine a final score. The score can range from 0 to 1 with a score of 0.27 or higher designating frailty. Patients identified as frail by TSFI score have been shown to be more likely to have an unfavorable discharge disposition (29, 30). For VR-12 determination, PCS and MCS was produced for comparison purposes.

Demographic Data Collection

Medical records and surveys were reviewed for demographic information that included age, gender, race, co-morbidities, substance abuse history, occupation, height, and weight. Additional information pertaining to hospital course and management included length of stay, follow up and repeat admissions, any associated complications related to their hospital stay, nutritional status and assessments, participation with therapy, and any trauma-related data were collected.

Statistical Analysis

SAS statistical software, version 9.4 (SAS Institute, Cary, NC, USA) was used to analyze the data. Student’s t-test was used to differentiate differences between groups for continuous data. Simple linear regression was performed for PCS and MCS at each time point measured with respect to TSFI categorical frailty. Pearson correlation coefficients were utilized to examine the relationship between the continuous variables (age, PCS, MCS, and TSFI) at each time point. Multivariate logistic regression was used to evaluate for an association of the VR-12 components, PCS and MCS, and frailty as determined by TSFI at enrollment. Age was included as a covariate to control for this variable in this analysis. To account for correlation within each subject, generalized estimating equations were used to estimate parameters and the covariance matrix estimated with a robust sandwich estimator. An independence structure was assumed for the working correlation matrix. Age was centered at 65. Weeks post-discharge was modeled as a continuous variable as were all frailty metrics. Models were fit for each frailty metric using Proc Genmod. Frail patients were defined as patients with a TSFI score of 0.27 or greater. P < 0.05 was considered significant. A linear marginal model with Gaussian errors was used to relate each VR-12 metric to time expressed in months as a categorical variable. To account for the correlation of observations from the same subject, the correlation structure was modeled. Several correlation structures (unstructured, compound symmetry, autoregressive, and Toeplitz) were evaluated for each frailty metric and model fit compared using Akaike Information Criterion (AIC) (31, 32). The correlation structure yielding the lowest AIC value was used to fit the final model. Models were fit using Proc Mixed in SAS version 9.4. Subjects lost to follow up were not included in this analysis.




Overall, fifty-eight patients were enrolled for study, of which 31 were trauma patients, 19 were EGS patients, and eight were burn surgery patients. Of the 31 trauma patients, 25 (80.7%) suffered a fall, four (12.9%) a motor vehicle accident, one suffered an ATV accident, and one was struck by someone falling from height. Patients were prospectively followed for the duration of their hospital course and for one year after hospital discharge. One patient was dropped from analysis after it was found that they did not have enrollment frailty metrics. A second patient was excluded from the analysis as they did not have any PCS or MCS measurements. The median TSFI score was 0.18 (median range 0.1–0.67). Eighteen patients were identified as frail by TSFI score. The most common past medical history findings were cardiovascular disease (67.2% of patients), cancer (24.1%), and diabetes (24.1%). Six patients in our cohort had previously sustained a stroke while two patients had a history of Parkinson’s Disease and one patient had a history of dementia noted at enrollment. Demographics data is summarized in Table 1.

Table 1. Demographic characteristics of the study participants (n = 58)

Abbreviations: BMI – Body Mass Index; GCS – Glasgow Coma Scale; ISS – Injury Severity Score; LOS – Length of Stay; Median was used for qualitative variables


When grouping patients based on mortality, 8 patients died and 44 patients lived, no significant differences in frailty scoring, age, and PCS (p = 0.97) or MCS (p = 0.78) at enrollment were observed between patients who ultimately lived and those who died.

Analysis of PCS and MCS Scoring

PCS and MCS scores were evaluated to assess differences at various time points and changes over time. Fifty-six patients were included in the analysis of PCS and MCS scoring as two patients did not have any scores recorded. An overall downtrend in overall mean PCS from 47.59 ± 13.71 to 42.67 ± 13.79 was noted from the time of enrollment up to 12 months following hospital discharge. Significant differences in PCS were noted between frail patients and non-frail patients at enrollment (31.91 ± 10.88 vs. 50.44 ± 13.10, p = 0.03) and at 12 months following hospital discharge (26.51 ± 11.48 vs. 47.43 ± 12.54, p = 0.03; Table 2). In comparison, the overall mean MCS remained relatively stable over the course of the study period. Significant differences were noted in MCS between frail and non-frail patients at enrollment only (38.91 ± 6.16 vs. 42.55 ± 5.27, p = 0.04; Table 2).

Table 2. Comparison of PCS and MCS for All Patients and By Frailty

Abbreviations: MCS – Mental Component Score; PCS – Physical Component Score; TSFI – Trauma Specific Frailty Index; Frailty was defined by TSFI with score greater than 0.27; p <0.05 was considered significant; Frailty was defined by TSFI score greater than 0.27; p < 0.05 was considered significant)


A pairwise comparison of PCS and MCS was then conducted between time points. In direct comparison of time points for PCS, no significant differences were noted between enrollment and the 3 month (p = 0.78), 6 month (p = 0.42), and 9 month (p = 0.34) time points. However, PCS was significantly different at 12 months compared to enrollment (p = 0.03). No other significant differences were noted in further comparisons of time points within the study period. Pairwise comparison of MCS showed significant differences between enrollment and the 6 month (p < 0.01), 9 month (p < 0.01), and 12 month (p < 0.01) time points. Similarly, no significant differences were noted with any other time point comparisons. Pairwise comparisons of PCS and MCS are summarized in Table 3 and Table 4, respectively.

Table 3. Pairwise Comparisons of PCS Between Time Points

Abbreviations: MCS – Mental Component Score; PCS – Physical Component Score; TSFI – Trauma Specific Frailty Index; Frailty was defined by TSFI with score greater than 0.27; p <0.05 was considered significant; Frailty was defined by TSFI score greater than 0.27; p < 0.05 was considered significant)


Table 4. Pairwise Comparisons of MCS Between Time Points

Abbreviations: MCS – Mental Component Score; p < 0.05 was considered significant.

Correlation of PCS and MCS with Frailty

The data was then assessed for associations between PCS and MCS with frailty using Pearson correlation. At enrollment, TSFI significantly correlated with PCS (r = -0.72, p < 0.01), MCS (r = -0.35, p < 0.01), and age (r = 0.34, p = 0.02). The correlation with frailty continued only for PCS and age at 3 months (PCS: r = -0.64, p < 0.01; age: r = 0.40, p = 0.02), 6 months (PCS: r = -0.79, p < 0.01; age: r = 0.34, p = 0.03), 9 months (PCS: r = -0.75, p < 0.01; age: r = 0.35, p = 0.03), and 12 months (PCS: r = -0.79, p < 0.01; age: r = 0.19, p = 0.23) following enrollment. MCS only correlated with age at the 12 month time point (r = 0.32, p = 0.04).
A linear regression model was performed using PCS and MCS at enrollment with age as a covariable and enrollment TSFI was the dependent variable. Age (p < 0.01), PCS (p < 0.01), and MCS (p < 0.01) were all independent predictors of enrollment frailty (Table 5). Univariate and multivariate analysis showed an association between PCS and age with frailty. MCS was not associated with frailty on multivariate analysis. PCS (OR = 0.894 [0.84-0.95], p = 0.0004) and age (OR = 1.113 [1.012-1.223], p = 0.03) independently predicted enrollment frailty. The area under the curve for this model was 0.89.

Table 5. Analysis of Maximum Likelihood Parameter Estimates

Abbreviations: PCS – Physical Component Score; MCS – Mental Component Score; p <0.05 was considered significant.



Our data shows that quality of life measures that assess a patient’s overall perception of their health significantly correlated with frailty in this cohort. In addition to clinician frailty assessment, the information provided by these measures may further inform healthcare providers on the degree of risk burden their patients have for repeat falls or other injury following their initial injury. Others have examined the use and accuracy of patient reported global health measures as a link to clinical disorders. Lapin et al. assessed the accuracy of general health cross-walk tables in a clinical sample of patients with spine disorders using VR-12 for general health evaluation. They noted that VR-12 could be accurately linked within a sample of patients with spine disorders but noted high bias and low precision (2). Further combinations with other global health scores have strengthened that linkage. Schalet et al. found that the combined set of VR-12 and PROMIS global health measures were linkable. They noted that linking worked better between physical and mental health scores using VR-12 item responses than with linkage based on algorithmic scores (18).
Much like frailty assessments, measures of global health typically occur at the time of injury or admission but does not consider the long-term dynamics of patient management or rehabilitation as their course and disposition change over time. As a result, most literature typically utilizes an initial evaluation of either frailty or global health measures to determine outcomes. This study presents a unique approach to patient reported global health assessment as well as its association with frailty. In our cohort, 18 patients were identified as frail as defined by the TSFI. An overall downtrend was observed in PCS scores following admission and up to twelve months after. Additionally, significant differences were observed in PCS scores between patients identified as frail and those who were not. This significant difference was present twelve months after as well. MCS scores, in comparison, remained steady throughout the survey period. Our data additionally noted significant correlations between frailty, PCS, and MCS. This significance continued following enrollment for PCS. These data suggest a relationship between PCS and frailty and that repeated assessment of an individual’s physical component score over time may be crucial to determining risk burden or those more at risk for additional injury following acute injury.
Identifying an individual’s risk for additional injury easily and observing significant changes over time through the use of frequent reassessments is crucial to developing prevention and clinical intervention strategies. Prospective interventions in frail older populations consisting of delirium prevention, nutritional and geriatrician consultation, physical therapy, early ambulation, and pain control have demonstrated significant decreases in 30-day readmission and delirium compared to no intervention in frail patients (33). Increasing the frequency of patient reported global health measures may help identify changes in both frailty and clinical progress earlier and may lead to more rapid intervention.
This study presents several limitations. It is a prospective initial analysis of patient reported global health measures and its association with frailty in acutely injured older patients. It should be noted that this is a single center study performed on a mainly rural, Caucasian population in the Midwest and, as such, these findings may not be generalizable to other populations. While this study provides significant observations to consider in the course of frailty assessment and the study and management of frail patients, the study itself remains underpowered. A higher-powered study over an even longer period of time could tease out associations and outcomes heretofore unidentified. Furthermore, while this study encompasses older patients with acute injury, the diversity of injury types sustained required trauma surgery, emergency general surgery, or burn surgery management. Older patients who sustain a fall will require different management and rehabilitation compared to older patients who sustain large total body surface area burns. Each injury type may affect patient reported global health measures differently over time and, as such, must be considered on an individual basis in future studies. Additionally, the subjectiveness of the questionnaire must be considered when assessing patient frailty and risk burden. While TSFI considers degree of dementia in its frailty assessment, global health measures like the VR-12 are dependent on patient self-assessment and may be unreliable depending on a patient’s medical history and clinical course.
Despite these limitations, this study provides a novel look at global health measures as a metric to consider when assessing frailty in acutely injured older patients. It is further bolstered by its methodology whereupon repeated assessments are performed over a period of time. This methodology may prove useful in more quickly and accurately determining physiologic changes that require intervention.


Conclusions and Future Directions

PCS scores were strongly associated with frailty. Our results suggest that these global measures of health could be utilized in addition to or in lieu of frailty scores when assessing older individuals. Further studies are warranted to confirm this association.

Plain summary

This study provides a novel look at global health measures as a metric to consider when assessing frailty in acutely injured older adults. Global health measures such as the Veterans RAND 12 Item Health Survey (VR-12) are quality of life measures that assess an individual’s overall perception of their health and may serve as a useful adjunct when predicting frailty. Our results show that the physical component score (PCS) produced by VR-12 surveys was strongly associated with frailty. These global measures of health could be utilized in addition to or in lieu of frailty scores when assessing patients in the clinical setting.


Acknowledgements: We thank Ella Born for assisting with follow up phone calls to participants.

Conflict of Interest: The authors have no conflicts.

Sources of funding: This research was supported by the University of Iowa Injury Prevention Research Center and funded in part by grant # R49 CE002108-05 of the National Center for Injury Prevention and Control/CDC. This project was further supported by the National Center for Advancing Translational Sciences, National Institutes of Health, through grant number UL1 TR001860 and UL1 TR002537. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Sponsor’s Role: The sponsor had no role in the design, methods, subject recruitment, data collection, and preparation of the manuscript.

Ethics Approval: This prospective observational study was approved by the University of Iowa Institutional Review Board (IRB # 201611731).

Consent to participate: A written informed consent was obtained from all participants.

Consent for publication: not applicable.

Availability of data and material: Data are available upon request to the corresponding author.

Author Contributions: All authors have made equal contributions to the conception, design, analysis, interpretation, drafting, and revision of the manuscript for this study.



1. Ahmed S, Berzon RA, Revicki DA, Lenderking WR, Moinpour CM, Basch E, et al. The Use of Patient-reported Outcomes (PRO) Within Comparative Effectiveness Research: Implications for Clinical Practice and Health Care Policy. Medical Care, 2012. 50(12): 1060-70.
2. Lapin BR, Kinzy TG, Thompson NR, Krishnaney A, and Katzan IL. Accuracy of Linking VR-12 and PROMIS Global Health Scores in Clinical Practice. Value Health, 2018. 21(10): 1226-33.
3. Bjorner J, Fayers P, and Idler E, Self-rated health. Assessing Quality of Life in Clinical Trials: Methods and Practice 2nd edn (ed Fayers,P. M.;Hays,R. D.). 2005, Oxford: Oxford University.
4. Han PK, Lee M, Reeve BB, Mariotto AB, Wang Z, Hays RD, et al. Development of a prognostic model for six-month mortality in older adults with declining health. J Pain Symptom Manage, 2012. 43(3): 527-39.
5. Brazier J, Roberts J, and Deverill M. The estimation of a preference-based measure of health from the SF-36. J Health Econ, 2002. 21(2): 271-92.
6. Eichler H, Kong S, Gerth W, Mavros P, and Jonsson B. Use of Cost-Effectiveness Analysis in Health-Care Resource Allocation Decision-Making: How Are Cost-Effectiveness Thresholds Expected to Emerge? Value in Health Value in Health, 2004. 7(5): 518-28.
7. Kazis LE, Bierman A, Diehr P, Hays R, Holzman J, Iqbal S, et al. A White Paper report commissioned by CMS with Recommendations for the Medicare Health Outcomes Survey. 2007.
8. Selim AJ, Rogers W, Qian SX, Brazier J, and Kazis LE. A preference-based measure of health: the VR-6D derived from the veterans RAND 12-Item Health Survey. Qual Life Res, 2011. 20(8): 1337-47.
9. Neumann PJ, Rosen AB, and Weinstein MC. Medicare and cost-effectiveness analysis. N Engl J Med, 2005. 353(14): 1516-22.
10. Ware J, Kosinski M, and Gandek B, SF-36 Health Survey: Manual and Interpretation Guide. 2005, Lincoln, RI: Quality Metric Inc.
11. Garratt AM, Ruta DA, Abdalla MI, and Buckingham JK. The SF 36 Health Survey Questionnaire: An Outcome Measure Suitable for Routine Use Within the NHS? British Medical Journal, 1993. 306(6890): 1440.
12. Revicki DA and Kaplan RM. Relationship between psychometric and utility-based approaches to the measurement of health-related quality of life. Quality of life research: An international journal of quality of life aspects of treatment, care and rehabilitation, 1993. 2(6): 477-87.
13. Ware J, Kosinski M, and Keller SD. A 12-Item Short-Form Health Survey: Construction of Scales and Preliminary Tests of Reliability and Validity. Medical Care Philadelphia, 1996. 34(3): 220-33.
14. Kazis LE, Miller DR, Skinner KM, Lee A, Ren XS, Clark JA, et al. Applications of Methodologies of the Veterans Health Study in the VA Healthcare System: Conclusions and Summary. Journal of Ambulatory Care Management, 2006. 29(2): 182-8.
15. Selim AJ, Rogers W, Fleishman JA, Qian SX, Fincke BG, Rothendler JA, et al. Updated U.S. population standard for the Veterans RAND 12-item Health Survey (VR-12). Quality of Life Research, 2009. 18(1): 43-52.
16. Kazis LE, Lee A, Spiro A, Rogers W, Ren XS, Miller DR, et al. Measurement Comparisons of the Medical Outcomes Study and Veterans SF-36 Health Survey. Health Care Financing Review, 2004. 25: 43-58.
17. Kazis LE, Miller DR, Clark JA, Skinner KM, Lee A, Ren XS, et al. Improving the Response Choices on the Veterans SF-36 Health Survey Role Functioning Scales Results From the Veterans Health Study. Journal of Ambulatory Care Management Journal of Ambulatory Care Management, 2004. 27(3): 263-80.
18. Benjamin D, Nan E, Ron D, Lewis E, Karon F, Joshua P, et al. Linking Physical and Mental Health Summary Scores from the Veterans RAND 12-Item Health Survey (VR-12) to the PROMISE Global Health Scale. Journal of General Internal Medicine, 2015. 30(10): 1524-30.
19. Joseph B, Hassan A, and Pandit V. Frailty in trauma: A systematic review of the surgical literature for clinical assessment tools. J Trauma Acute Care Surg, 2016. 81(4): 805.
20. Maxwell CA, Mion LC, Mukherjee K, Dietrich MS, Minnick A, May A, et al. Preinjury physical frailty and cognitive impairment among geriatric trauma patients determine postinjury functional recovery and survival. J Trauma Acute Care Surg, 2016. 80(2): 195-203.
21. Clegg A and Young J. The frailty syndrome. Clin Med Clinical medicine (London, England), 2011. 11(1): 72-5.
22. Lang P, Michel J, and Zekry D. Frailty Syndrome: A Transitional State in a Dynamic Process. Gerontology, 2009. 55(5): 539-49.
23. Morley JE, Vellas B, Abellan van Kan G, Anker SD, Bauer JM, Bernabei R, et al. Frailty Consensus: A Call to Action. Journal of the American Medical Directors Association, 2013. 14(6): 392-7.
24. Curtis E, Romanowski K, Sen S, Hill A, and Cocanour C. Frailty score on admission predicts mortality and discharge disposition in elderly trauma patients over the age of 65 y. J Surg Res, 2018. 230: 13-9.
25. Moschny A, Platen P, Klaaflen-Mielke R, Trampisch U, and Hinrichs T. Barriers to physical activity in older adults in Germany: A cross-sectional study. International Journal of Behavioral Nutrition and Physical Activity, 2011. 8: 121.
26. Rasinaho M, Hirvensalo M, Leinanen R, Lintunen T, and Rantanen T. Motives for and Barriers to Physical Activity Among Older Adults With Mobility Limitations. Journal of Aging and Physical Activity, 2007. 15(1): 90.
27. Hatcher VH, Galet C, Lilienthal M, Skeete DA, and Romanowski KS. Association of Clinical Frailty Scores With Hospital Readmission for Falls After Index Admission for Trauma-Related Injury. JAMA Netw Open, 2019. 2(10): e1912409.
28. Galet C, Zhou Y, Eyck PT, and Romanowski KS. Fall injuries, associated deaths, and 30-day readmission for subsequent falls are increasing in the elderly US population: A query of the WHO mortality database and National Readmission Database from 2010 to 2014. Clin Epidemiol, 2018. 10: 1627-37.
29. Joseph B, Pandit V, Zangbar B, Kulvatunyou N, Tang A, O’Keeffe T, et al. Validating trauma-specific frailty index for geriatric trauma patients: a prospective analysis. J Am Coll Surg, 2014. 219(1): 10-7 e1.
30. Maxwell CA, Patel MB, Suarez-Rodriguez LC, and Miller RS. Frailty and Prognostication in Geriatric Surgery and Trauma. Clin Geriatr Med, 2019. 35(1): 13-26.
31. Akaike H. New Look at Statistical-Model Identification. IEEE Transactions on Automatic Control, 1974. AC19(6): 716-23.
32. Akaike H. Information theory and an extension of the maximum likelihood principle. 2nd International Symposium on Information Theory (Akademia Kiado, Vudapest), 1973: 267-81.
33. Bryant EA, Tulebaev S, Castillo-Angeles M, Moberg E, Senglaub SS, O’Mara L, et al. Frailty Identification and Care Pathway: An Interdisciplinary Approach to Care for Older Trauma Patients. J Am Coll Surg, 2019. 228(6): 852-9 e1.




M. Cesari1,2, R. Calvani3,4, M. Canevelli5,6, I. Aprahamian7,8, P. de Souto Barreto9,10, D. Azzolino1,2, R.A. Fielding11, N. Vanacore5, M. Inzitari12,13, E. Marzetti3,14


1. Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy; 2. Healthy Aging Lab, IRCCS Istituti Clinici Scientifici Maugeri, Milan, Italy;
3. Fondazione Policlinico Universitario «Agostino Gemelli» IRCCS, Rome, Italy; 4. Aging Research Center, Department of Neurobiology, Care Sciences and Society, Karolinska Institutet and Stockholm University, Stockholm, Sweden; 5. National Centre for Disease Prevention and Health Promotion, Italian National Institute of Health, Rome, Italy; 6. Department of Human Neuroscience, Sapienza University of Rome, Rome, Italy; 7. Geriatrics Division, Internal Medicine Department, Jundiaí Medical School, Jundiaí, Brazil; 8. University of Groningen, University Medical Center Groningen, Groningen, The Netherlands; 9. Gérontopôle de Toulouse, Institut du Vieillissement, Centre Hospitalo-Universitaire de Toulouse, Toulouse, France; 10. CERPOP INSERM 1295, University of Toulouse III, INSERM, UPS, Toulouse, France; 11. Nutrition, Exercise Physiology, and Sarcopenia Laboratory, Jean Mayer USDA Human Nutrition Research Center on Aging, Tufts University, Boston, MA, USA; 12. RE-FIT Barcelona research group, Parc Sanitari Pere Virgili and Vall d’Hebron Institute of Research (VHIR), Barcelona, Spain; 13. Department of Health Sciences, Universitat Oberta de Catalunya (UOC), Barcelona, Spain; 14. Università Cattolica del Sacro Cuore, Institute of Internal Medicine and Geriatrics, Rome, Italy.

Corresponding Author: Matteo Cesari, MD, PhD. IRCCS Istituti Clinici Scientifici Maugeri; Via Camaldoli 64, 20138 Milan, Italy. Email: macesari@gmail.com; Twitter: @macesari

J Frailty Aging 2021;10(4)310-312
Published online May 24, 2021, http://dx.doi.org/10.14283/jfa.2021.23


Key words: Research, methodology, publishing, medicine, science.


From the beginning of 2020, the world has been fighting the SARS-Cov-2 outbreak. The life of each one of us has profoundly changed. Unavoidably, our clinical routine has drastically modified in its priorities and methodologies (1). The COVID-19 pandemic has also raised significant issues in the field of research. The investigators’ responsibility has increased with the need to thoughtfully weigh the risk-benefit ratio for each protocol in an emergency and evolving scenario (2).
Many projects that had started in the pre-COVID-19 era were halted in the past months, mainly because of 1) the restrictions that governments put in place to limit the spread of the Coronavirus, 2) the inadequacy/inappropriateness of the research activities as planned before the pandemic, and/or 3) the participants’ reluctancy at attending the scheduled visits in facilities usually located within the hospital premises. Some investigators managed to redesign their study methods and make them compatible with the new situation. Someone also proposed possible solutions to tackle the most common issues of the “new” research routine (3). Others faced challenging (if not insurmountable) stalls to solve and eventually decided to stop their studies. Indeed, it was not rare the closure of research activities in some countries, especially during the early period of the pandemic.
How should we judge the findings generated by studies that suffered the effects of the COVID-19 pandemic? In principle, the modification of a protocol “on the way” or the premature ending of a study are deviations that, in “normal” times, would be flagged as major methodological flaws. Purists of research methodology may indeed find these deviations hard to accept. But, can we judge today’s research activities without considering the new context of COVID-19 and the global consequences we all have been experiencing? What degree of flexibility should we apply when reviewing an article containing issues clearly due to the pandemic? In this context, it is also essential to consider the increasing interest around pragmatic research for achieving a wise compromise between the potentially decontextualized findings from rigid methodologies and the heterogenous “real life” scenarios where the study results will be applied (4). Perhaps, the pandemic may lead us to introduce sufficient doses of pragmatism and creativity as core elements of future research. For sure, scientific projects will have to include backup plans considering the possibility of postponing/modifying recruitment strategies, maintaining protocol adherence despite environmental changes, and designing alternative strategies for intervention delivery.
In this issue of The Journal of Frailty & Aging, Tavoian and colleagues (5) present the results of a pilot study initiated before the COVID-19 outbreak. The trial was focused on exploring whether stationary bicycle high-intensity interval training was a more efficient standalone exercise strategy to improve cardiovascular and muscular function compared with resistance or aerobic training in older adults. Unfortunately, the restrictions imposed by the emergent pandemic situation led to a premature stop of the program, precluding the achievement of the expected sample size and leaving the researchers with partial results.
The data could appear insufficient to convey a complete and clear message (at least coherent with the original study aims). For example, the unbalanced representation of women in the study sample (11 vs. only 3 men) might already render the findings of difficult interpretation. On the other hand, we might say that something is always better than nothing! Those partial results could still feed the discussion in the scientific community and serve to support further steps in the field (6). The Chinese philosopher Lao Tzu once said: “The journey of a thousand miles begins with one step”.
The study by Tavoian and colleagues was well-presented and conducted rigorously, methodologically speaking. It also had its record on clinicaltrials.gov, having been registered before the initiation of recruitment. The topic was of interest and novel. Last but not least, the trial was framed as a pilot study with the primary aim of collecting preliminary data for future larger-scale activities. Under these premises, it could have been unfair to reject it simply because an external event (i.e., the COVID-19 pandemic) had disrupted the researchers’ original plans. Are we sure that what is (partially) presented by Tavoian and colleagues may not be sufficient to address questions raised by other researchers? After all, when a study is well-conducted, it is always important to publish it, independently of issues that might have affected its conduct. Of course, the expected impact of an unforeseen event disrupting a pilot study is different from that occurring in a phase 3 trial. This implies that a case-by-case approach is crucial but needs to be structured on specific criteria. Otherwise, the risk could be to fall into a dangerous level of subjectivity or to reject everything in search of a utopic perfection.
No study is perfect, and every research has an own value. It is a matter of clearly and fairly presenting the experience, constructively discussing the strengths and weaknesses. At the same time, the readers should be transparently informed of the deviations imposed by the COVID-19 outbreak and provided with all the information necessary to judge the value of findings and discern potentially relevant results from “background noise”. The need for clarity and details is important for discriminating vague justifications (e.g., “due to the pandemic”) from founded decisions based on political, sanitary, and/or scientific choices (2).
It is also noteworthy that the numbers we report in our tables are persons who have trusted in our research and devoted part of their time to it. The restitution of the study results has to be considered an ethical duty of the researcher in front of the participant. Consistently, many investigations benefit from funds allocated by private or public institutions for specific research purposes and potentially taken from other priorities. It is, thus, also a matter of financial clarity and accountability behavior.
With the aim of providing the correct legitimacy to those valuable research activities critically affected by the COVID-19 pandemic, it has been felt that a set of standardizing criteria could be helpful. Table 1 presents aspects that might support clinicians, researchers and policymakers in better judging the quality of the information provided by partial scientific results.
The post-COVID-19 world we live in is putting us in the position of reconsidering many aspects of our (professional) lives. Different priorities are today modifying our decisions. How we conduct research today is not the same as before COVID-19. The population we are studying is not the same either, and its values and needs might be different (7). Some adaptations of research to the new scenario (e.g., more extensive use of technologies, higher involvement of participants in collecting their data) are necessary to continue advancing in research and avoiding that what was initiated before COVID-19 could go lost.

Table 1. Specific criteria to additionally consider when judging if an article is still worth being published despite COVID-19-related deviations from original plans


Conflicts of interest: MC has received honoraria from Nestlè Health Sciences for presentations at scientific meetings and serving as a member of Expert Advisory Boards. MI has received honoraria from Nestlé Health Sciences for serving as an expert advisor. EM has received honoraria from Abbott, Nestlè, Nutricia, and Thermofisher for presentations at scientific meetings. RAF received grants and personal fees from Nestlé. RAF also reports grants from National Institutes of Health (National Institute on Aging), during the conduct of the study; grants, personal fees and other from Axcella Health, other from Inside Tracker, grants and personal fees from Biophytis, grants and personal fees from Astellas, personal fees from Cytokinetics, personal fees from Amazentis, personal fees from Glaxo Smith Kline, personal fees from Juvicell, outside the submitted work. No conflict of interest reported by the other authors.

Acknowledgments: RAF is partially supported by the US Department of Agriculture (USDA), under agreement No. 58-8050-9-004 and by NIH Boston Claude D Pepper Center (OAIC; 1P30AG031679). Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the USDA.




1. Astrone P, Cesari M. Integrated Care and Geriatrics: A Call to Renovation from the COVID-19 Pandemic. J Frailty Aging. 2021;10(2):182-183. doi: 10.14283/jfa.2020.59
2. Fleming TR, Labriola D, Wittes J. Conducting Clinical Research During the COVID-19 Pandemic: Protecting Scientific Integrity. JAMA. 2020;324(1):33-34. doi:10.1001/jama.2020.9286
3. Kroenke K, Bair MJ, Sachs GA. Continuing Research During a Crisis. J Gen Intern Med. 2021;36(4):1086-1088. doi:10.1007/s11606-021-06636-5
4. Ford I, Norrie J. Pragmatic Trials. N Engl J Med. 2016;375(5):454-463. doi:10.1056/NEJMra1510059
5. Tavoian D, Russ DW, Law TD, et al. Effects of three different exercise strategies for optimizing aerobic capacity and skeletal muscle performance in older adults: a pilot study. J Frailty Aging. doi:10.14283/jfa.2021.21
6. Karmakar S, Dhar R, Jee B. Covid-19: research methods must be flexible in a crisis. BMJ. 2020:m2668. doi:10.1136/bmj.m2668
7. Palmer K, Monaco A, Kivipelto M, et al. The potential long-term impact of the COVID-19 outbreak on patients with non-communicable diseases in Europe: consequences for healthy ageing. Aging Clin Exp Res. 2020;32(7):1189-1194. doi:10.1007/s40520-020-01601-4



M. Gagesch1,2, P.O. Chocano-Bedoya2,3, L.A. Abderhalden1,2, G. Freystaetter1,2, A. Sadlon1,2, J.A. Kanis4, R.W. Kressig5, S. Guyonnet6, J.A. P. DaSilva7, D. Felsenberg8, R. Rizzoli9, M. Blauth10, E.J. Orav11, A. Egli2, H.A. Bischoff-Ferrari1,2,12


1. Department of Geriatrics, University Hospital Zurich, Zurich, Switzerland; 2. Centre on Aging and Mobility, University of Zurich, Zurich, Switzerland; 3. Institute of Primary Health Care (BIHAM), University of Bern, Bern, Switzerland; 4. Mary McKillop Institute for Health Research, Australian Catholic University, Melbourne, Australia and Centre for Metabolic Bone Diseases, University of Sheffield Medical School, Sheffield, UK; 5. University Department of Geriatric Medicine FELIX PLATTER, Basel, Switzerland; 6. Gérontopôle, Department of Geriatrics, CHU Toulouse, Toulouse, France; 7. Institute for Clinical and Biomedical Research (i.CBR), Faculty of Medicine, University of Coimbra, Department of Rheumatology, Centro Hospitalar e Universitário de Coimbra, Coimbra, Portugal; 8. Centre for Muscle and Bone Research, Department of Radiology, Charité Universitätsmedizin Berlin, Berlin, Germany (Prof. Felsenberg is deceased); 9. Service of Bone Diseases, Geneva University Hospitals and Faculty of Medicine, Geneva, Switzerland; 10. University of Innsbruck, Innsbruck, Austria; 11. Department of Medicine, Harvard Medical School, Boston, MA, USA; 12. University Clinic for Acute Geriatric Care, City Hospital Waid and Triemli, Zurich, Switzerland

Corresponding Author: Michael Gagesch, MD, Department of Geriatrics, University Hospital Zurich, Ramistrasse 100, 8091 Zurich, Switzerland, Email: michael.gagesch@usz.ch, https://orcid.org/0000-0003-3089-5768

J Frailty Aging 2021;in press
Published online May 3, 2021, http://dx.doi.org/10.14283/jfa.2021.18



BACKGROUND: Frailty is a geriatric syndrome associated with multiple negative health outcomes. However, its prevalence varies by population and instrument used. We investigated frailty and pre-frailty prevalence by 5 instruments in community-dwelling older adults enrolled to a randomized-controlled trial in 5 European countries.
METHODS: Cross-sectional baseline analysis in 2,144 DO-HEALTH participants recruited from Switzerland, Austria, France, Germany, and Portugal with complete data for frailty. Frailty status was assessed by the Physical Frailty Phenotype [PFP], SOF-Frailty Index [SOF-FI], FRAIL-Scale, SHARE-Frailty Instrument [SHARE-FI], and a modified SHARE-FI, and compared by country, age, and gender. Logistic regression was used to determine relevant factors associated with frailty and pre-frailty.
RESULTS: Mean age was 74.9 (±4.4) years, 61.6% were women. Based on the PFP, overall frailty and pre-frailty prevalence was 3.0% and 43.0%. By country, frailty prevalence was highest in Portugal (13.7%) and lowest in Austria (0%), and pre-frailty prevalence was highest in Portugal (57.3%) and lowest in Germany (37.1%). By instrument and overall, frailty and pre-frailty prevalence was highest based on SHARE-FI (7.0% / 43.7%) and lowest based on SOF-FI (1.0% / 25.9%). Frailty associated factors were residing in Coimbra (Portugal) [OR 12.0, CI 5.30-27.21], age above 75 years [OR 2.0, CI 1.17-3.45], and female gender [OR 2.8, CI 1.48-5.44]. The same three factors predicted pre-frailty.
CONCLUSIONS: Among relatively healthy adults age 70 and older enroled to DO-HEALTH, prevalence of frailty and pre-frailty differed significantly by instrument, country, gender, and age. Among instruments, the highest prevalence of frailty and pre-frailty was documented by the SHARE-FI and the lowest by the SOF-FI.

Key words: Frailty, community-dwelling, prevalence, clinical trials, epidemiology.



Frailty is an age-related medical syndrome of reduced functional capacity and one of the most burdensome conditions for the growing segment of the older population, and society as a whole (1). Frailty is often preceded by sarcopenia (2), the age-associated loss of muscle mass and function and both concepts overlap, e.g. in regard to reduced muscle strength, physical function (e.g. gait speed), and unintentional weight loss (3). In addition, frailty and sarcopenia are often promoted by factors such as malnutrition (4, 5) and multimorbidity (6, 7).
Over the last two decades, frailty has been identified as a stronger predictor of acute-care complications, length of stay, and hospital readmission than multimorbidity or chronological age alone (8-11). E.g. in a 2016 systematic review and meta-analysis of 31 prospective studies, Vermeiren et al. found that frailty increases the likelihood of mortality by 1.8-2.3-fold, hospital admissions by 1.2-1.8-fold, incident functional impairment related to basic activities of daily living by 1.6-2.0, and falls and fractures by 1.2-2.8-fold (12).
The development of frailty in older adults should be recognized as a dynamic process including transitions from robust, to pre-frail and to frail and vice versa over time (13). Therefore, early detection of the at-risk group of pre-frail individuals appears as a window of opportunity for the prevention of overt frailty, and perhaps prevention of sarcopenia as well (3, 14, 15).
While the identification of frailty is becoming increasingly recognized as important, the choice of frailty instrument might be guided by medical practice within a specialty, the population, setting, and outcome under investigation (16). Frailty is actively screened for in various clinical settings, including oncology and surgical specialties. However, the lack of consensus for defining frailty still hinders the comparison of frailty prevalence between different populations (1, 17-19).
Over time, the concept of the Physical Frailty Phenotype (PFP), introduced by Fried and colleagues in 2001 (20) and the deficit accumuation approach (i.e. Frailty Index) introduced by Mitnitski and Rockwood (21, 22) have emerged as the most frequently cited ways to operationalize frailty (23). While the Frailty Index approach scores accumulating deficits in age-associated variables covering multipe body systems (24), the five components of the PFP (fatigue, weight loss, slowness, weakness, and low activity level) aggregate to a clinical syndrome (20). Subsequently, many researchers have modified the PFP in order to accommodate variables available from research studies’ data sets (25). Among those, the Study of Osteoporotic Fractures (SOF)-Frailty Index (26), a condensed variation of only 3 items, and the FRAIL-Scale (27), a self-report based 5-item screening tool substituting the low activity criterion with multimorbidity (defined as the presence of five or more chronic diseases) have been frequently cited.
To our knowledge, the prevalence of frailty and pre-frailty in older Europeans according to the PFP has not been studied in a single multi-national dataset with the exception of the Survey of Health Aging and Retirement in Europe (SHARE) database in 2009 and 2019 (28, 29). The reported frailty prevalence by the latter study in SHARE ranged from 3.0% (Switzerland) to 15.6% (Portugal), thus indicating an uneven distribution in the older European population, plus indicating a need for further studies on frailty and its assessment concepts in older Europeans.
DO-HEALTH is a clinical trial in relatively healthy community-dwelling participants age 70 years and older without major health events (cancer or myocardial infarction) in the 5 years prior to enrollment, sufficient mobility to come to the study centers without help, and intact cognitive function, recruited from 5 European countries (30). Examining the prevalence of frailty and pre-frailty in DO-HEALTH at baseline aims to contribute important insights to the participants’ clinical presentation upon enrolment, and also with regard to several primary endpoints of the study, including cardiovascular health and physical function (30).
In view of the variety of instruments derived from the PFP, and to estimate the variability of frailty prevalence based on the instrument used, we investigated the PFP as our primary measure and four alternative derived instruments in DO-HEALTH. However, one has to acknowledge that this is not a probabilisitc sample of the countries investigated: in addition to the DO-HEALTH eligibility criteria listed above, each countries sample consisted of older adults volunteering to participate in a clinical trial, and recruited according to the methods of each study site.
In addition to examine the variability in frailty according to instrument, we examine whether the prevalence of frailty and pre-frailty differs by country, age, and gender.


Materials and methods

This is a cross-sectional study taking advantage of the baseline examination of 2,144 of 2,157 participants of the DO-HEALTH clinical trial recruited from Switzerland (Zurich, Basel, Geneva), Austria (Innsbruck), France (Toulouse), Germany (Berlin), and Portugal (Coimbra), with complete data for frailty. As a multicenter randomized controlled trial (RCT), DO-HEALTH (clinical trials: NCT01745263) was designed to investigate the effect of Vitamin D3 and Omega 3-fatty acids supplementation and a simple home exercise program in a 2x2x2 factorial design over a three year follow-up addressing multiple endpoints associated with aging, including frailty (30).
Inclusion criteria for DO-HEALTH were age 70 years and older, a Mini-Mental Status Examination Score of at least 24 points (31), living in the community and being sufficiently mobile to visit the study center, i.e. being capable to walk 10 meters and getting in and out of a chair without help. In order to capture the broad scope of the trial’s endpoints, the participants were meticulously phenotyped with standardized assessments of multiple organ, cognitive, and physical function at each of the four whole-day clinical study visits (baseline and follow-up at 12, 24, and 36 months). Assessments followed a strict protocol with regular site visits for inter-center quality control, overseen by the primary investigator. Further details of the design, aims and scope, exclusion criteria, and primary endpoints of DO-HEALTH are described elsewhere (30).

Assessments of Physical Function and Comorbidities

For the assessment of physical function, we used the Short Physical Performance Battery (SPPB), a validated assessment tool for lower extremity function in older adults (32). The SPPB includes a gait speed test, a chair-rise test, and a balance test. Its three components each score 0–4, with 4 points indicating the highest level of performance, and add up to an overall score of maximum 12 points. The number of comorbidities was recorded by a self-administered comorbidity questionnaire (33).

Fried Physical Frailty Phenotype

The physical frailty phenotype (PFP) has been validated in community dwelling older adults from numerous countries and in many in- and outpatient settings by various operationalizations (17, 25, 34). We built our PFP model with variables available from the baseline DO-HEALTH dataset as our primary frailty assessment tool. Unintentional weight loss was defined as a self-reported loss of more than 5% of body weight over the prior 12 months. Fatigue was operationalized as a positive answer to the self-reported question: “In the last month, have you had too little energy to do things you wanted to?” from the SHARE study original questionnaire. For weakness, grip-strength was taken in kilopascal (kPa) from the best of three consecutive trials at the dominant hand using a Martin Vigorimeter (KLS Martin Group, Tuttlingen, Germany) (35). Across all countries and ages, we used cut points by the lowest quintile approach by gender as did Fried and colleagues in their landmark study, but for simplification, irrespective of BMI (20). Slowness was defined as a gait speed below 0.67 m/s and 0.7 m/s respectively, according to gender and height as in the original study by Fried et al (20). Low activity level was conceptualized as done earlier by Santos-Eggimann and colleagues analyzing the SHARE dataset (28). An answer of “less than once a week” to 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 doing a walk?” fulfilled the criterion. Pre-frailty was fulfilled with 1-2 positive criteria and frailty with ≥3 positive criteria.

SOF-Frailty Index

The Study of Osteoporotic Fractures-Frailty Index (SOF-FI), introduced by Ensrud et al. is a condensed three-item variant of the Fried PFP including only the variables for weakness, involuntary weight loss, and fatigue (26, 36). It has been validated in community dwelling participants in the MOBILIZE Boston Study (37). In DO-HEALTH, we used data from the chair-rise test of the Short Physical Performance Battery for the SOF-FI weakness criterion.[38] Weight loss was defined by a self-reported loss of more than 5% of body weight over the prior 12 months. Fatigue was defined by a positive answer to the question: “Do you feel full of energy?” from the Geriatric Depression Scale as in the original study (26). Pre-frailty was defined by the presence of one, and frailty by ≥2 positive criteria.


The FRAIL-Scale (Fatigue, Resistance, Ambulation, Illnesses, and Loss of Weight-Scale) consists completely of self-reported items (39). Since its introduction, it has been validated for the prediction of most relevant frailty associated negative outcomes in various settings (40-42). For building the FRAIL-Scale in DO-HEALTH, we used the same question for the fatigue criterion as described above for the PFP. For resistance (weakness) we used the question for difficulties climbing stairs from SHARE-FI (see below). For the ambulation criterion, we used the question: “Because of a health problem, do you have difficulty (expected to last more than 3 months) walking 100 meters”, also from SHARE-FI. Illnesses (i.e. multimorbidity) was defined by the presence of >5 diagnoses). Weight loss was defined by a self-reported loss of more than 5% of body weight over the prior 12 months. On the FRAIL-Scale, pre-frailty was defined by 1-2 positive criteria, and frailty by ≥3 positive criteria.

SHARE-Frailty Instrument (SHARE-FI1)

To match the available variables from DO-HEALTH, we used a marginally modified version of the original SHARE-Frailty Instrument (SHARE-FI), validated in community dwelling older adults and in primary care.[43-45] Weight loss was defined by a self-reported >5% loss of body weight over the prior 12 months. Fatigue was operationalized in the same way as for the PFP, by a positive answer to the question: “In the last month, have you had too little energy to do things you wanted to do?” For indicating low grip strength on Martin Vigorimeter readings, we used the lowest quintile approach as described for the PFP above. Slowness was defined as originally described by Santos-Eggimann et al. by a positive answer to either i) “Because of a health problem, do you have any difficulty walking 100 meters? (Exclude any difficulties that you expect to last less than three months.)”, or ii) “Because of a health problem, do you have any difficulty climbing one flight of stairs without resting? (Exclude any difficulties that you expect to last less than three months.)” (28). Low activity was fulfilled by a response of “less than once a week” (i.e. “one to three times a month” or “hardly ever or never”) to the same question as for the PFP: “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?” Pre-frailty was defined as 1-2 positive criteria and frailty with ≥3 positive criteria. The SHARE-FI1 differs from our implementation of the PFP only in the way that slowness was assessed by self-report only.


For a fully self-report based variant of the SHARE-FI in DO-HEALTH, we substituted grip strength measurement by a positive answer to the question: “Because of a health problem, do you have any difficulty climbing one flight of stairs without resting (exclude any difficulties that you expect to last less than three months)?” This question was usually meant only to depict slowness in the original SHARE-FI operationalization. However, measuring resistance (or weakness) by a single question of difficulty walking one flight of stairs has been validated as a surrogate marker for weakness (46), as it is also used as the weakness criterion in the FRAIL-Scale introduced by Abellan van Kan et al. (see above) (39). For SHARE-FI2, pre-frailty was fulfilled with 1-2 positive criteria and frailty with ≥3 positive criteria.
Supplementary Table 1 provides an additional overview of the single components comprising the described instruments.

Statistical Analysis

Data is reported as means and standard deviations (SD) for continuous variables and frequencies and percentages for categorical variables. We estimated the prevalence of fraily and pre-frailty as the proportion of our study population who could be classified as either frail or pre-frail according to the PFP and the derived instruments and reported as frequencies and percentages. We then compared the prevalences by country, age sub-groups (<75y, ≥75y), and gender using a Chi-square (χ2) test or Fisher’s exact test when a frequency was less than five. We report on the general prevalence by screening tool including participants with a maximum of two missing values, as instruments indicate frailty as having ≥3 out of 5 criteria present (with the exception of the SOF where ≥2 positive items indicate frailty). We constructed a logistic regression model to compare frail and non-frail patients according to the PFP by study site, age group and gender using Zurich as our reference site. A separate model was run with pre-frailty as the outcome. All analyses were performed with SAS v9.4 (SAS Inc., Carey, North Carolina, USA).



Overall, mean age was 74.9 ±4.4 years, and 61.6% were women. Mean Mini-Mental Status Examination score was 28.5 points (SD 1.5). Mean BMI was 26.3kg/m2 (SD 4.3), and 28.5% (n=569) of participants had three or more comorbidities at baseline. Baseline characteristics of all study participants by gender are presented in Table 1. Of the total 2,157 participants in DO-HEALTH, 2,144 (99.4%) had complete data at baseline on at least three of the variables for all frailty measures (Supplementary Table 2). In all, 1,005 participants were included from Switzerland (1 participant was excluded from Zurich), 350 from Germany, 300 from Portugal (1 participant excluded), 289 from France (11 participants excluded), and 200 from Austria.

Table 1. Baseline Characteristics of the DO-HEALTH participants

Values are means and standard deviations unless otherwise noted. BMI (Body Mass Index), MMSE (Mini-Mental Status Examination, used to measure cognitive impairment, range of 0-30 points, where higher scores are better, and scores >24 suggest normal cognitive function), kPa (Kilopascal), SPPB (Short Physical Performance Battery, used to measure lower extremity function, scores range from 0-12 points, where higher scores indicate better functioning), aaccording to a Self-Administered Comorbidity Questionnaire (23).


Prevalence of frailty and pre-frailty in all DO-HEALTH participants by instrument

According to our primary measure PFP, 3.0 % (n=64) of DO-HEALTH participants were frail, and 43.0% (n=922) were pre-frail (i.e. at-risk for frailty). See Figure 1 for the overall prevalence of frailty and pre-frailty based on the PFP and the other four derived instruments. Prevalence of frailty in the total population was highest based on SHARE-FI1 (7.0%), and lowest based on SOF-FI (1.0%). Also, prevalence of pre-frailty in the total population was highest based on SHARE-FI1 (43.7%), and lowest based on SOF-FI (25.9%).

Figure 1. Prevalence of frailty and pre-frailty by frailty instrument among DO-HEALTH participants (n=2,144 with available data out of n=2,157 total participants)


Prevalence of frailty and pre-frailty by country, age, and gender

With regard to country, the prevalence of frailty was highest using the SHARE-FI1 (range 1.5-24.3%), and lowest with the SOF-FI (range 0-3.0%). With regard to country, the prevalence of pre-frailty was highest with the SHARE-FI1 tool for all countries (range 40.3-51.3%), except for Portugal (51.3% vs. 57.3% on the PFP). With regard to country, the prevalence of pre-frailty was lowest with the SOF-FI for all countries (range 23.2-32.0%), except for Austria (27.5% vs. 24.0% on the FRAIL-Scale).
Figure 2a shows all prevalence rates for frailty by instrument and country. Figure 2b shows all prevalence rates for pre-frailty by instrument and country.

Figure 2a. Prevalence of frailty by country and investigated frailty instrument

Figure 2b. Prevalence of pre-frailty by country and investigated frailty instrument


With regard to age group (<75 years vs. ≥75 years), persons 75 years and older were more often frail compared to younger persons, regardless of the frailty instrument used (P <0.05 for all instruments). In both age groups, the prevalence of frailty was highest on the SHARE-FI1 (4.4%; 10.4%, P <0.0001) and lowest on the SOF-FI (0.5%, 1.7%, P = 0.0069). With regard to age group (<75 years vs. ≥75 years), there were significant differences in pre-frailty with a higher proportion of pre-frail participants among those age 75 years and older (P <0.05 for all instruments). In the younger age group, the prevalence of pre-frailty was highest using the SHARE-FI1 (42.0%), and lowest using the SOF-FI (22.9%), while in the older age group, the prevalence of pre-frailty was highest using the PFP (46.2%), and lowest using the SOF-FI (30.1%).
With regard to gender, there were significant differences for frailty by all investigated instruments with women having an overall higher prevalence of frailty compared to men (P <0.01), with the exception of the SOF-FI (P = 0.3505). With regard to pre-frailty, there were significant differences by gender for all investigated instruments with women having an overall higher prevalence of pre-frailty compared to men (P <0.001).
Results for frailty by country (study site) and each subgroup are shown in detail in Table 2. Results for pre-frailty by country (study site) and each subgroup are shown in detail in Supplementary Table 3.

Table 2. Prevalence of Frailty by Instrument in DO-HEALTH at baseline according to country (study site), age, and gender

All values are n (%); of the total 2,157 participants of DO-HEALTH, 2,144 (99.4%) had complete data at baseline on at least three of the variables for all frailty measures (Fried PFP: 42 missing 1 variable, 3 missing 2 variables, 10 missing 3 variables; SOF-FI: 37 missing 1 variable, 10 missing 2 variables; FRAIL-Scale: 34 missing 1 variable, 1 missing 2 variables, 12 missing ≥3 variables; SHARE-FI1: 5 missing 1 variable, 1 missing 2 variables, 12 missing ≥3 variables; SHARE-FI2: 13 missing ≥3 variables) and were included in this analysis. In all, 1,005 participants were from Switzerland (1 participant was excluded from Zurich), 350 from Germany, 300 from Portugal (1 participant excluded), 289 from France (11 participants excluded), and 200 from Austria). Fisher’s exact test was used to compare subgroups by country. Chi-square tests were used to compare frailty between subgroups by age and gender.


In our logistic regression model, we compared frail and non-frail participants according to the PFP by study center (using Zurich as our reference), age and gender. In regard to frailty, we found a 12-fold increased odds for participants from Coimbra (Portugal) [OR 12.00; CI 5.30-27.21] compared to Zurich, a 2-fold increased odds for older participants (≥75 years of age) [OR 2.00; CI 1.17-3.45], and a 2.8-fold increased odds for female participants [OR 2.84; CI 1.48-5.44]. In regard to pre-frailty, we found an almost 2-fold increased odds for participants from Coimbra (Portugal) [OR 1.90; CI 1.43-2.53] compared to Zurich, a 1.2-fold increased odds for older participants (≥75 years of age) [OR 1.20; CI 1.00-1.42], and a 1.45-fold increased odds for female participants [OR 1.45; CI 1.21-1.73] as shown in Supplementary Table 4.



In this large study of 2,144 relatively healthy, community dwelling European adults age 70 and older recruited from 5 European countries, and based on our primary measure PFP instrument, 3% of participants were frail and close to half of all participants were pre-frail. With regard to subgroups, prevalence of frailty and pre-frailty was highest in Portuguese participants, women, and adults age 75 and older. Notably, prevalence of frailty and pre-frailty varied substantially by the instrument used and was lowest by the SOF-FI, and highest by the SHARE-FI1.
Earlier studies in community dwelling older adults from the US and Germany reported frailty prevalence rates ranging from 2.5 to 17% using the PFP definition (26, 47-49). Of note, our reported frailty prevalence in DO-HEALTH with a mean age of 74.9 years was substantially lower than in similar age groups of population-based studies such as SHARE (1.3 vs. 4.4% for Switzerland, 13.8 vs. 28.6% for Portugal, 2.9 vs. 15.6% for France, 0.6 vs. 9.6% for Germany, and 0 vs. 11.6% for Austria) (29). However, the prevalence appeared more comparable to the age group between 65-74 years in the same study hereby reflecting the relatively good health of the DO-HEALH participants. At the same time our results of a higher odds for being frail and pre-frail by higher age and female gender are in line with the prior literature (1).
While our findings do not arise from a representative sample of the population, our results of a higher frailty prevalence in our Portugal center compared to the more central European countries in DO-HEALTH are consistent with the SHARE cohort findings (waves 1 and 6) and supported by a higher rate of physical inactivity in southern European countries, including Portugal (28, 29, 50). This is further supported by a recent study investigating limitations in instrumental activities of daily living (iADL) from SHARE data, highlighting the uneven distribution of burdens in self-care with higher rates in southern European countries, including Portugal (51). Additionally, the observed health disparity with regard to frailty prevalence between Portugal and Austria or Switzerland in DO-HEALTH may be explained by a potential difference in educational level (28)and linked to national economic indicators (52, 53), or specific patient sampling. In fact, DO-HEALTH participants from Portugal had less years of education compared to the four other countries. Further, the 2015 OECD (Organization for Economic Co-operation and Development) report on the attainment of overall goals in health care supports our findings by ranking Switzerland (No. 2), France (No. 6), Austria (No. 10), and Germany (No. 14) relatively high, compared with Portugal (No. 32) (54). Moreover, the observed pattern of a higher frailty prevalence in Portugal is supported by the 2015 OECD data where older adults from southern European countries had worse health states than central European older adults (54).
To the best of our knowledge, we lack population-based data on the prevalence of frailty and pre-frailty for the DO-HEALTH countries comparing different PFP operationalizations with test-based measurements of grip-strength and gait speed (i.e. PFP, SOF-FI) to mainly self-report instruments (i.e. FRAIL-Scale, SHARE-FI). In our analysis, the mainly self-report based instruments indicated a higher frailty prevalence compared with our primary measure PFP that included the test-based measurements of grip-strength and gait speed. This is in contrast to prior research indicating lower levels of frailty by self-report, e.g. in a longitudinal study from Ireland, and thus warrants further investigation (55).
With regard to instruments, the SOF-FI differed most from the PFP in terms of prevalence overall, but also in terms of differences by country, gender, and age. A possible explanation may be that SOF-FI only covers three of the original five PFP components (56, 57). A higher prevalence of frailty and pre-frailty among Portuguese, female, and participants age 75+ was observed, regardless of the frailty instrument used. Therefore, our findings caution its use as a frailty prevalence tool among relatively healthy older adults.
Our study has several strengths. First, we were able to utilize a large RCT baseline data set of 2,144 meticulously phenotyped older adults with standardized assessments of all participants from five European countries. Second, given the rich DO-HEALTH data set, we were able to match five frailty operationalizations with self-report and test-based assessments derived from Fried’s original physical frailty phenotype. Third, despite that fact that DO-HEALTH does not reflect population-based data and selected relatively healthy older adults, our data provides a conservative comparison of frailty and pre-frailty prevalence by five different frailty operationalizations and between five European countries.
Our study also has its limitations. First, we have to acknowledge that this is not a probabilistic sample and participants were recruited as relatively-healthy volunteers from the regions where the study was performed. Thus, suggesting that the actual prevalence of frailty may be subject to self selection and healthy participant bias (e.g. more frail participants may refrain from participating). Further, adaptions were necessary to apply the original Fried PFP to DO-HEALTH, due to the availability of variables from our data set. However, this has been necessary in most other studies following the initial publication by Fried et al (25, 58). A conceptual limitation of our study is the expected low overall prevalence of frailty in the relatively healthy community dwelling older adults enrolled in DO-HEALTH. In addition, DO-HEALTH data does not include information on differences in health care systems or access to primary care between participating centers. However, we still document significant variability between countries, and subgroups.
In summary, our study contributes unique data on the prevalence of pre-frailty and frailty in adults age 70 and older from five European countries enrolled in a large-scale RCT. With regard to the high prevalence of pre-frailty, we add important data indicating that interventions at this early stage possibly modulate the progression to frailty in a substantial share of the older adult population. Thus, our findings may help trigger efforts in the prevention of frailty, not only in Portugal, but also in the central European countries involved in DO-HEALTH. Our results support the fact that fraily prevalence is significantly influenced by the choice of instrument among relatively healthy community dwelling older adults. Further, all instruments but SOF-FI applied in the DO-HEALTH data set, appear to have construct validity with regard to country, gender, and age differences in prevalence of pre-frailty and frailty.
In conclusion, our findings suggest that the choice of frailty instrument among relatively healthy older adults is relevant with regard to frailty prevention and care planning efforts. Given the suggested large health care disparities with regard to frailty among participants from Portugal versus the central European countries in DO-HEALTH, further studies are needed to explore these differences with regard to other aspects of health and at the population-based level.


Funding: This work was supported by the European Commission’s 7th framework programme (FP7, HEALTH.2011.2.2.2-1) for investigator-driven clinical trials for therapeutic interventions in elderly populations (Project No: 278588-2).

Acknowledgements: MG drafted the manuscript and interpeted the data. POCB and LAA analysed and interpreted the data and contributed to the drafting of the manuscript. GF and AS provided critical revision of the manuscript. JAK, RWK, BV, JAPD, DF, RR, MB, EJO, AE and HABF designed the study concept, aquised the data and critically revised the manuscript. We would like to thank all DO-HEALTH investigators, researchers, and study personnel at the seven partner sites throughout Europe for their contribution.

Conflict of Interest: The authors declare no conflict of interest in regard to this work.

Ethical standards: The DO-HEALTH study was approved by ethics and regulatory agencies of all 5 countries and the study protocol has been previously published (30). A data and safety monitoring board oversaw the study.

Open Access: This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, duplication, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.





1. Hoogendijk, E.O., et al., Frailty: implications for clinical practice and public health. The Lancet, 2019. 394(10206): p. 1365-1375.
2. Landi, F., et al., Sarcopenia as the Biological Substrate of Physical Frailty. Clin Geriatr Med, 2015. 31(3): p. 367-74.
3. Cruz-Jentoft, A.J., et al., Sarcopenia: revised European consensus on definition and diagnosis. Age and Ageing, 2018: p. afy169-afy169.
4. Verlaan, S., et al., High Prevalence of Physical Frailty Among Community-Dwelling Malnourished Older Adults-A Systematic Review and Meta-Analysis. J Am Med Dir Assoc, 2017.
5. Vanpottelbergh G, Vaes B, Adriaensen W, Mathei C, Legrand D, Wallemacq P, Degryse JM: The glomerular filtration rate estimated by new and old equations as a predictor of important oucomes in elderly patients. BMC Medicine 2014, 12(27):doi:10.11186/11741.
6. Rockwood, K. and A. Mitnitski, Frailty Defined by Deficit Accumulation and Geriatric Medicine Defined by Frailty. Clinics in Geriatric Medicine, 2011. 27(1): p. 17-26.
7. Pacifico, J., et al., Prevalence of sarcopenia as a comorbid disease: A systematic review and meta-analysis. Experimental Gerontology, 2020. 131: p. 110801.
8. Fimognari, F.L., et al., The Severity of Acute Illness and Functional Trajectories in Hospitalized Older Medical Patients. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 2017. 72(1): p. 102-108.
9. Lin, H.S., et al., Frailty and post-operative outcomes in older surgical patients: a systematic review. BMC Geriatr, 2016. 16(1): p. 157.
10. Wou, F., et al., The predictive properties of frailty-rating scales in the acute medical unit. Age Ageing, 2013. 42(6): p. 776-81.
11. Apóstolo, J., et al., Predicting risk and outcomes for frail older adults: an umbrella review of frailty screening tools. JBI Database System Rev Implement Rep, 2017. 15(4): p. 1154-1208.
12. Vermeiren, S., et al., Frailty and the Prediction of Negative Health Outcomes: A Meta-Analysis. J Am Med Dir Assoc, 2016. 17(12): p. 1163.e1-1163.e17.
13. Gill, T.M., et al., Transitions between frailty states among community-living older persons. Arch Intern Med, 2006. 166(4): p. 418-23.
14. Subra, J., et al., The integration of frailty into clinical practice: preliminary results from the Gerontopole. J Nutr Health Aging, 2012. 16(8): p. 714-20.
15. Oliveira, J.S., et al., Evidence on Physical Activity and the Prevention of Frailty and Sarcopenia Among Older People: A Systematic Review to Inform the World Health Organization Physical Activity Guidelines. J Phys Act Health, 2020: p. 1-12.
16. Faller, J.W., et al., Instruments for the detection of frailty syndrome in older adults: A systematic review. PLoS One, 2019. 14(4): p. e0216166.
17. Theou, O., et al., What do we know about frailty in the acute care setting? A scoping review. BMC Geriatr, 2018. 18(1): p. 139.
18. Morley, J.E., et al., Frailty consensus: a call to action. J Am Med Dir Assoc, 2013. 14(6): p. 392-7.
19. Rodriguez-Manas, L., et al., Searching for an operational definition of frailty: a Delphi method based consensus statement: the frailty operative definition-consensus conference project. J Gerontol A Biol Sci Med Sci, 2013. 68(1): p. 62-7.
20. Fried, L.P., et al., Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci, 2001. 56(3): p. M146-56.
21. Mitnitski, A.B., A.J. Mogilner, and K. Rockwood, Accumulation of deficits as a proxy measure of aging. ScientificWorldJournal, 2001. 1: p. 323-36.
22. Rockwood, K., A. Mogilner, and A. Mitnitski, Changes with age in the distribution of a frailty index. Mech Ageing Dev, 2004. 125(7): p. 517-9.
23. Buta, B.J., et al., Frailty assessment instruments: Systematic characterization of the uses and contexts of highly-cited instruments. Ageing Res Rev, 2016. 26: p. 53-61.
24. Searle, S.D., et al., A standard procedure for creating a frailty index. BMC Geriatr, 2008. 8: p. 24.
25. Theou, O., et al., Modifications to the frailty phenotype criteria: Systematic review of the current literature and investigation of 262 frailty phenotypes in the Survey of Health, Ageing, and Retirement in Europe. Ageing Res Rev, 2015. 21: p. 78-94.
26. Ensrud, K.E., et al., Comparison of 2 frailty indexes for prediction of falls, disability, fractures, and death in older women. Arch Intern Med, 2008. 168(4): p. 382-9.
27. Abellan van Kan, G., et al., The I.A.N.A Task Force on frailty assessment of older people in clinical practice. J Nutr Health Aging, 2008. 12(1): p. 29-37.
28. Santos-Eggimann, B., et al., Prevalence of frailty in middle-aged and older community-dwelling Europeans living in 10 countries. J Gerontol A Biol Sci Med Sci, 2009. 64(6): p. 675-81.
29. Manfredi, G., et al., Prevalence of frailty status among the European elderly population: Findings from the Survey of Health, Aging and Retirement in Europe. Geriatr Gerontol Int, 2019. 19(8): p. 723-729.
30. Bischoff-Ferrari, H.A., et al., DO-HEALTH: Vitamin D3 – Omega3 – Home exercise – Healthy aging and longevity trial – Design of a multinational clinical trial on healthy aging among European seniors. Contemporary Clinical Trials, 2020: p. 106124.
31. Folstein, M.F., S.E. Folstein, and P.R. McHugh, “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psych Res, 1975. 12.
32. Guralnik, J.M., et al., Lower extremity function and subsequent disability: consistency across studies, predictive models, and value of gait speed alone compared with the short physical performance battery. J Gerontol A Biol Sci Med Sci, 2000. 55(4): p. M221-31.
33. Sangha, O., et al., The Self-Administered Comorbidity Questionnaire: a new method to assess comorbidity for clinical and health services research. Arthritis Rheum, 2003. 49(2): p. 156-63.
34. Bouillon, K., et al., Measures of frailty in population-based studies: an overview. BMC Geriatr, 2013. 13: p. 64.
35. Sipers, W.M., et al., The Martin Vigorimeter Represents a Reliable and More Practical Tool Than the Jamar Dynamometer to Assess Handgrip Strength in the Geriatric Patient. J Am Med Dir Assoc, 2016. 17(5): p. 466.e1-7.
36. Ensrud, K.E., et al., A comparison of frailty indexes for the prediction of falls, disability, fractures, and mortality in older men. J Am Geriatr Soc, 2009. 57(3): p. 492-8.
37. Kiely, D.K., L.A. Cupples, and L.A. Lipsitz, Validation and comparison of two frailty indexes: The MOBILIZE Boston Study. J Am Geriatr Soc, 2009. 57(9): p. 1532-9.
38. Guralnik, J.M., et al., A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol, 1994. 49(2): p. M85-94.
39. Abellan van Kan, G., et al., Frailty: toward a clinical definition. J Am Med Dir Assoc, 2008. 9(2): p. 71-2.
40. Morley, J.E., T.K. Malmstrom, and D.K. Miller, A simple frailty questionnaire (FRAIL) predicts outcomes in middle aged African Americans. J Nutr Health Aging, 2012. 16(7): p. 601-8.
41. Vellas, B., et al., Designing pharmaceutical trials for sarcopenia in frail older adults: EU/US Task Force recommendations. J Nutr Health Aging, 2013. 17(7): p. 612-8.
42. Woo, J., et al., Frailty Screening in the Community Using the FRAIL Scale. Journal of the American Medical Directors Association, 2015. 16(5): p. 412-419.
43. Cuenoud, P., Prevalence of the Frailty Phenotype in community-dwelling population aged 50 and over in the Survey of Health, Ageing and Retirement in Europe, in IEMS Lausanne – Faculté des Hautes Etudes Commerciales et Faculté de Biologie et de Médecine. 2007, Université de Lausanne.
44. Romero-Ortuno, R., The SHARE operationalized frailty phenotype: a comparison of two approaches. Eur Geriatr Med, 2013. 4(4).
45. Pialoux, T., J. Goyard, and B. Lesourd, Screening tools for frailty in primary health care: a systematic review. Geriatr Gerontol Int, 2012. 12(2): p. 189-97.
46. Verghese, J., et al., Self-reported difficulty in climbing up or down stairs in nondisabled elderly. Arch Phys Med Rehabil, 2008. 89(1): p. 100-4.
47. Woods, N.F., et al., Frailty: emergence and consequences in women aged 65 and older in the Women’s Health Initiative Observational Study. J Am Geriatr Soc, 2005. 53(8): p. 1321-30.
48. Cawthon, P.M., et al., Frailty in older men: prevalence, progression, and relationship with mortality. J Am Geriatr Soc, 2007. 55(8): p. 1216-23.
49. Fuchs, J., M. Busch, and C. Scheidt-Nave, [Frailty: Prevalence in the Resident Population of Germany 70 – 79 Years of Age – a Population-Based Approach]. Anasthesiol Intensivmed Notfallmed Schmerzther, 2017. 52(11-12): p. 758-763.
50. Gomes, M., et al., Physical inactivity among older adults across Europe based on the SHARE database. Age Ageing, 2017. 46(1): p. 71-77.
51. Portela, D., et al., Instrumental Activities of Daily Living (iADL) Limitations in Europe: An Assessment of SHARE Data. Int J Environ Res Public Health, 2020. 17(20).
52. Theou, O., et al., Exploring the relationship between national economic indicators and relative fitness and frailty in middle-aged and older Europeans. Age Ageing, 2013. 42(5): p. 614-9.
53. World Health Organization. World report on ageing and health. 2015 30.04.2017]; Available from: http://apps.who.int/iris/bitstream/10665/186463/1/9789240694811_eng.pdf?ua=1.
54. OECD, Health at a Glance 2015. OECD Publishing.
55. Theou, O., et al., Measuring frailty using self-report and test-based health measures. Age Ageing, 2015. 44(3): p. 471-7.
56. Fried, L.P., et al., Nonlinear multisystem physiological dysregulation associated with frailty in older women: implications for etiology and treatment. J Gerontol A Biol Sci Med Sci, 2009. 64(10): p. 1049-57.
57. Bandeen-Roche, K., et al., Phenotype of frailty: characterization in the women’s health and aging studies. J Gerontol A Biol Sci Med Sci, 2006. 61(3): p. 262-6.
58. Saum, K.U., et al., Development and evaluation of a modification of the Fried frailty criteria using population-independent cutpoints. J Am Geriatr Soc, 2012. 60(11): p. 2110-5.



R. Jiwani1,2, J. Wang1,3, C. Li1,3, B. Dennis3,4, D. Patel1,5,6, J. Gelfond7, Q. Liu7, N. Siddiqui3, C. Bess1, S. Monk1, M. Serra2,8,9, S. Espinoza2,8,9


1. School of Nursing, University of Texas Health Science Center at San Antonio, Texas, USA; 2. Geriatric Research, Education & Clinical Center (GRECC), South Texas Veterans Health Care System, San Antonio, Texas, USA; 3. Center on Smart and Connected Health Technologies, School of Nursing, University of Texas Health Science Center at San Antonio, Texas, USA; 4. Graduate School of Biomedical Sciences, University of Texas Health Science Center at San Antonio, Texas, USA; 5. Biobehavioral Research Laboratory, School of Nursing, University of Texas Health Science Center at San Antonio, Texas, USA; 6. Mays Cancer Center, University of Texas Health Science Center at San Antonio, Texas, USA; 7. Department of Epidemiology & Biostatistics, University of Texas Health Science Center at San Antonio, Texas, USA; 8. Sam and Ann Barshop Institute for Longevity and Aging Studies, University of Texas Health Science Center at San Antonio, Texas, USA; 9. Department of Medicine, Division of Geriatrics, Gerontology & Palliative Medicine, University of Texas Health Science Center at San Antonio, Texas, USA

Corresponding Author: Rozmin Jiwani, PhD, RN, 7703 Floyd Curl Drive, San Antonio, Texas, 78229, Phone: 210-450-8498, Fax: 210-567-5822, Jiwani@uthscsa.edu

J Frailty Aging 2021;in press
Published online April 24, 2021, http://dx.doi.org/10.14283/jfa.2021.17



Background: Older adults with Type 2 diabetes (T2D) are more likely to be frail, which increases the risk for disability and mortality.
Objectives: To determine the feasibility of a behavioral lifestyle intervention, enhanced with mobile health technology for self-monitoring of diet and activity, to improve frailty in overweight/obese older adults (≥65 years) diagnosed with T2D.
Design, Setting, and Participants: Single arm, 6-month study of a behavioral lifestyle intervention in 20 overweight/obese (BMI>25) older adults (≥ 65 years) with self-reported T2D diagnosis who owned a smartphone. A Fitbit tracker was provided to all participants for self-monitoring of diet and physical activity. Our primary outcome of feasibility was measured by session attendance, adherence to Fitbit usage to self-monitor diet and physical activity, and study retention. Secondary outcomes included the preliminary efficacy of the intervention on frailty, physical function, quality of life, and T2D-related outcomes.
Results: Eighteen participants completed the study. The mean age was 71.5 (SD ± 5.3) years, 56% were female, and half were Hispanic. At baseline, 13 (72%) were pre-frail, 4 (22%) were frail, and 1 (6%) were non-frail. At follow-up, frailty scores improved significantly from 1.61 ± 1.15 to 0.94 ± 0.94 (p=0.01) and bodyweight improved from 205.66 ± 45.52 lbs. to 198.33 ± 43.6 lbs. (p=<0.001).
Conclusion: This study provides evidence for the feasibility of a behavioral lifestyle intervention in overweight/obese older adults with T2D and preliminary results support its potential efficacy in improving frailty score.

Keywords: Type 2 diabetes, lifestyle intervention, self-monitoring, frailty, personal fitness technology.



Frailty is characterized as a geriatric syndrome of vulnerability and progressive physical decline, which significantly increases risks for falls, disability, and death (1). Frailty prevalence is approximately 10% population over 60 years old and reaches over 25% in those ages 80 years and older (2). Type 2 Diabetes (T2D) and overweight/obesity are highly prevalent in older adults and are significant predictors of both onset and worsening of frailty in older adults (3-5). Frailty prevalence is much higher in older adults diagnosed with T2D (32% to 48%) than in the general older population (5% to 10%), and frail individuals with T2D have higher mortality rates (6). Indeed, frailty is emerging as another category of T2D complications, in addition to the traditional micro-and macrovascular sequelae (7). To date, there are no widely accepted evidence-based interventions to improve frailty (8). Such interventions would be especially useful in older adults with T2D as frailty is highly associated with diabetes.
Lifestyle modification is the first-line treatment for T2D and there is evidence to suggest that weight loss and exercise improve components of frailty, such as gait speed, muscle strength, and physical function (8). The Look AHEAD (Action for Health in Diabetes) randomized clinical trial study demonstrated the efficacy of behavioral intensive lifestyle intervention (ILI) of caloric restriction and physical activity, targeting a 7% weight loss over the control group to reduce adiposity, improve glycemic control (9), reduce risk factors for cardiovascular disease (10), microvascular complications (11), and physical function (12, 13) over the follow-up years (range 8-11 years). In addition, ILI participants had less need for diabetes medications (14) and lowered health care costs (15). The ILI intervention was also associated with lower frailty index (FI) scores throughout follow-up (16). The FI operationalizes frailty as deficit accumulation and calculates the proportion of health deficits present in a person. This differs from the Fried frailty phenotype criteria, which scores frailty based on the presence of five criteria, including unintentional weight loss, exhaustion, low physical activity, poor muscle strength, and slow gait speed. The Look AHEAD trial did not assess frailty by Fried phenotype criteria at study baseline or during the intervention but did have this measure at later follow-up examinations. Therefore, the effect of the ILI intervention on frailty characteristics as measured by the Fried criteria is not known.
Self-monitoring is a cornerstone of behavioral lifestyle interventions. The rise in popularity and accessibility of personal fitness technology, such as wearable mobile health technology, apps, and mobile devices have become an informative and useful option to positively reinforce health, exercise, and nutritional behavior (17). Technology use has been steadily increasing in older adults, as smartphone ownership in this population has increased from 18% in 2013 to 43% in 2017, and the number of older adults who access the internet has risen to 67% (18). The integration of personal fitness technologies into behavioral interventions has been shown to increase physical activity, decrease sedentary behaviors, and improve objective measures (i.e., glycemic control) in older adults (19-23).
This study aimed to determine the feasibility of Look AHEAD behavioral lifestyle intervention, enhanced with mobile health technology for self-monitoring of diet and physical activity, to improve frailty as assessed by Fried criteria. Our primary outcome of feasibility was measured by session attendance, adherence to Fitbit usage to self-monitor diet and physical activity, and study retention. Secondary outcomes included the preliminary efficacy of the intervention on frailty, physical function, quality of life, and T2D-related outcomes. The primary hypothesis was the study is feasible and behavioral lifestyle intervention will improve frailty in community-dwelling overweight/obese older adults (≥ 65 years) diagnosed with T2D.



Study Design

We conducted a single-arm, 6-month study of a behavioral lifestyle intervention program enhanced with mobile health technology. A Fitbit wristband activity tracker was provided to all participants for self-monitoring of diet and physical activity and the companion app was downloaded on their smartphone by the research team during the baseline screening visit. The study was approved by the Institutional Review Board at the University of Texas Health Science Center at San Antonio (UTHSCSA).

Participant recruitment

We recruited participants using a combination of methods supported by the San Antonio Claude D. Pepper Older Americans Independence Center, including a call center, community advisory board, and research volunteer registry. We also utilized our University’s Find-a-Study recruitment website, advertisement in our primary care clinics, local senior centers, local newspapers, local health fairs/community events, outreach to community leaders, and word of mouth. A study flow diagram using the Consolidated Standards of Reporting Trials (CONSORT) model is shown in Figure 1.

Figure 1. CONSORT flow-diagram for a single arm non-randomized study


Sample and setting

We recruited 20, community-dwelling, overweight/obese (BMI ≥ 25 kg/m2), older adults (≥ 65 years) diagnosed with T2D (self-report of provider-diagnosis). We planned that all participants would receive a total of 10 face-to-face group sessions for the behavioral lifestyle intervention over 6 months. However, due to local and institutional COVID-19 restrictions, 3 of our 4 groups (groups 2, 3, and 4) attended 2-3 virtual (WebEx) group educational sessions to complete the study. They were then seen in-person for their end-of-study assessments. See Table 1 for detailed inclusion and exclusion criteria. We conducted all the study-related visits at the Biobehavioral Research Laboratory at UTHSCSA School of Nursing.

Table 1. Inclusion and exclusion criteria


Sample size

For this feasibility study, a sample size of n=20 was used, based on Moore et al. (24) which provide recommendations for planning pilot studies in clinical and translational research. We assumed a 40% attrition rate and enrolled 20 participants.

Study procedures

Potential participants were informed about the study over the phone and, if interested, proceeded with phone screening for eligibility based on inclusion and exclusion criteria (Table 1). Eligible participants were scheduled for an in-person baseline screening visit. T2D was ascertained by participant self-report of provider-diagnosed T2D in response to the question, “Has a medical doctor or other provider told you that you have Type 2 diabetes?” We also performed medication review and baseline lab work (fasting blood glucose and Hemoglobin A1c [HbA1c]) to verify the reliability of the self-reported T2D diagnoses. After informed consent, we collected baseline data. These same assessments were repeated at 6-month at the end of the study visit.

Data collection

Health history and physical assessment: A standardized history and physical assessment were performed and participants’ self-reported medical, surgical, and medication history were recorded. We measured participants’ vital signs (heart rate, blood pressure, temperature) and anthropometric measurements (waist circumference [inches], height [inches], weight [pounds]. Body mass index was calculated as weight (kg)/height (m2).
Frailty assessment: We classified frailty status using the Fried phenotype criteria (1): 1) Self-reported unintentional weight loss of ≥10 pounds in the past year; 2) Self-reported exhaustion; 3) Low energy expenditure using the Minnesota Leisure Time Physical Activity Questionnaire (MLTQ) to assess physical activity (duration and frequency) (25); 4) Weakness measured via grip strength using a handheld dynamometer in the dominant hand; 5) Timed gait at usual speed over a 10-foot walk as previously described (26). Any intentional weight loss that may have occurred due to the lifestyle intervention would not meet the criteria for frailty-related weight loss, as frailty-related weight loss is defined as unintentional. A frailty score was calculated as the number (0 – 5) of frailty characteristics present. Those with ≥ 3 of these 5 characteristics are categorized as frail; those with 1 or 2 are categorized as pre-frail, and those with none are categorized as non-frail (1).
Physical function: We administered the Short Physical Performance Battery (SPPB), a reliable and valid tool to assess lower extremity physical functions. The SPPB consists of 1) standing balance (ability to stand with the feet together in side-by-side, semi-and full-tandem positions for 10 seconds each); 2) a 4-meter walk (gait speed test) to assess the time it takes to complete the walk at a usual pace; 3) time to complete five repeated chair stand without using hands. Each of the 3 performance measures was assigned a score ranging from 0 to 4 and summed to create a score ranging from 0 (worst) to 12 (best). The SPPB is sensitive to change over time (27, 28).
Self-reported measures of health and quality of life: Participants completed the Patient-Reported Outcome Measurements Information System (PROMIS-57 and PROMIS Global Health) questionnaires to assess self-reported health and quality of life (29). The PROMIS-57 assesses seven health domains (anxiety, depression, fatigue, pain interference, physical function, satisfaction with participation in social roles, and sleep disturbance) using eight items for each domain. PROMIS Global Health measures are generic, rather than disease-specific, and are intended to globally reflect individuals’ assessment of health. The questionnaires are ranked on a 5-point Likert scale and include an additional pain intensity scale (0-10 numeric rating). The feasibility of using PROMIS tools has been supported among ethnic minority, predominantly African American, overweight/obese adults with T2D who had greater symptom burden and poorer physical functioning than the general US population (22).
Clinical laboratory measures: Fasting blood glucose and HbA1c were measured at baseline and at 6-months to monitor safety and the impact of adherence to the interventions. Blood was collected from participants via venipuncture in the antecubital space of the preferred arm. Samples were allowed to sit at room temperature for 30 minutes after which they were centrifuged for 15 mins at 1000xg. Analyses were performed by Quest Diagnostics (Dallas, TX).

Study Intervention

Behavioral Lifestyle Intervention: We used the publicly available behavioral lifestyle intervention available from the Look AHEAD and Diabetes Prevention Program Group Lifestyle Balance websites in this study. The PI (RN, PhD) delivered all the sessions. All participants received 10 behavioral lifestyle intervention sessions over 6 months as shown in Table 2. The frequency and design of these sessions (weekly in month 1, biweekly in months 2 and 3, and monthly in months 4 to 6), allowed participants to master new skills gradually, then eventually adopt the behaviors as part of their daily lives (Table 2). Each session lasted 60 to 90 minutes. Missed sessions were replaced with either an individual make-up session or phone consultation depending on the participant’s availability. All 10 group sessions focused on adherence to the behavioral strategies, such as self-monitoring, goal setting, feedback, mindful eating, talkback negative thoughts, social support, problem-solving, relapse prevention, and handling holidays, among others (Table 2). Participants were given a tailored weight loss goal of 5-7% based on calorie and fat intake based on their current weight, and physical activity goals to gradually increase to 175 minutes (about 3 hours) per week by the end of the study following Look AHEAD guidelines (9). In every group session, we reinforced the message to adhere to self-monitoring of diet, physical activity, and follow the behavioral education provided during the session. We recorded participants’ weight in every in-person group session, but we did not require them to check their weight at home.

Table 2. Topics and schedules for group sessions


Mobile health technology

A Fitbit wristband activity tracker was provided to all participants at the baseline screening visit, for self-monitoring of diet and activity, and the companion Fitbit app was downloaded on each participants’ smartphone. Participants received training on how to record their food (portion size, calories, and fat) and physical activity (duration and type of activity) using the Fitbit wristband activity tracker application on their smartphones. Participants were reminded during group sessions to self-monitor their diet and activity. We entered participants’ tailored bodyweight loss and activity goals in their device and encouraged them to record their food and activity daily. We also gave a written/pictorial step-by-step “cheat sheet” of instructions on how to record their food and activity and we reinforced these instructions throughout the study. Participants were asked to wear the device at all times. The research team sent two text messages per month, either for positive feedback or for a gentle reminder to those who were struggling or not recording their diet or physical activity data based on the data gathered from the connected health platform. Adherence to Fitbit usage for physical activity was inferred from step count data. If participants generated step count data for a given day, we determined that the participant had worn the device.

Statistical Analyses

Feasibility was measured via session attendance, adherence to Fitbit usage to self-monitor diet and physical activity (step count data), and retention at 6 months. We examined the preliminary efficacy of the intervention on frailty phenotype (Fried criteria), including physical function using the Short Physical Performance Battery (SPPB), T2D-related outcomes (body weight, waist circumference, fasting blood glucose, and HbA1c), including self-reported health and quality of life using the Patient-Reported Outcome Measurements Information System (PROMIS) questionnaires. We used a paired t-test to perform the efficacy analysis on the Fried frailty score (0 – 5), which was selected based upon Type I and II error performance with similar ordinal variables (30). Other exploratory endpoints of frailty, SBBP, and PROMIS component scores were assessed with paired t-tests. The five frailty components (e.g., grip strength, gait speed, physical activity, exhaustion, unintentional weight loss) were analyzed using McNemar’s test. Factors (age, gender, and ethnicity) may have moderated intervention effects on outcomes (frailty, body weight, and SPPB) using interaction terms (factor x time) within linear mixed-effects models. All testing was two-sided with a significance threshold of <0.05 for the p-values. All analyses were conducted in R (v3.5+, Vienna, Austria) within an accountable data analysis process.



Participant characteristics

Participant characteristics are detailed in Table 3. A total of 20 participants were enrolled with a mean age of 71.5 (± 5.3) years. Mean BMI was in the obese range (mean=33.7 kg/m2 ± 6.3), mean frailty score was 1.61 ± 1.15 (range: 0-4), and mean SBBP score was 9.22 ± 2.13 (range: 7-12) at study baseline. The majority of participants were pre-frail (72%) at the start of the study.

Table 3. Participant characteristics (N=18)


Feasibility outcome

Eighteen (90%) participants completed the study. Two participants dropped out after the first educational session and were no longer interested in participating. Data from the 18 participants who completed the study have been included in all analyses. Of the 18 participants who completed the study, 11 (61.1%) attended all 10 sessions, 6 missed 1 session, and 1 missed 3 sessions. The mean number of sessions attended was 9.5. The median [interquartile range] adherence rate for Fitbit activity tracker device was 95.6% [79.5%, 100%]. The maximum and minimum adherence rates were 100% and 36.4%, respectively. The median [interquartile range] of diet logging rate (days with at least 1 diet item logged/total days in study) was 81.5% [40.7%, 84.1%]. For the physical activity logging rate, these numbers were 8.4% [2.7%, 48.7%].

Effect of the study intervention

Table 4 shows the effect of the study intervention on frailty, physical function, quality of life, and T2D-related outcomes (tailored weight loss, waist circumference, fasting blood glucose, and HbA1c). Frailty total score improved significantly by 0.67 points (95% CI [1.15, 0.18], p=0.01). Frailty sub-scores for the 10-foot timed gait improved significantly at 0.30 points (95% CI [0.61, -0.01], p = 0.05), while physical activity and grip strength did not change significantly. Although the SBBP total score did not significantly change, there was a trend toward improvement by 0.83 (95% CI [0.07, -1.74] p=0.069). SBBP sub-scores for the balance tests improved significantly 0.50 (95% CI [-0.15, -0.85], p=0.008), while gait speed and repeated chair stand scores did not change significantly.

Table 4. Effect of the study intervention on frailty, physical function, quality of life, and T2D-related outcomes


About T2D-related outcomes, on average, participants lost 7.3 lbs. (95% CI [10.73, 3.92], p<0.001) and reduced their BMI by 1.15 kg/m2 (95% CI [1.69, 0.61], p<0.001). Age and ethnicity did not moderate treatment effects (p>0.05). In exploratory analyses, we found that males were more responsive to the intervention, with a decrease in weight (-7.9 lbs., p<0.05) and SBBP scores (2.8, p<0.001) relative to females (p<0.05). The intervention led to a 3.0% reduction in HbA1c, a 2.3% reduction in fasting blood glucose, a 1.0% reduction in waist circumference, but these changes were not statistically significant. PROMIS scores for self-reported health and quality of life were not affected by the study intervention (p-values > 0.05), however, we noticed some improvements in PROMIS Global Physical and Mental health scores (Table 4). Individual participant’s weight for every session is shown in Figure 2, which depicts weight/weight loss normalized by baseline weight. Most of the participants (except for two participants from group one) participated in virtual monthly educational sessions (sessions 8 to 10) due to the local COVID-19 quarantine. Seven out of 18 (38%) participants regained an average of 4.5 lbs. of weight during the virtual sessions, even though their weight had appropriately declined before this point (Figure 2).

Figure 2. Individual participants’ weight/weight loss normalized by baseline weight. The y-axis is weight percentage of baseline and x-axis is week number at the bottom and session number on the top for behavioral lifestyle intervention



To our knowledge, this is the first study to determine the feasibility of a behavioral lifestyle intervention, enhanced with mobile health technology for self-monitoring of diet and activity, for the goal of improving frailty in overweight/obese older adults with T2D. We demonstrated a high retention rate with 90% of participants completing the study, even though we encountered an unexpected obstacle due to the COVID-19 pandemic at the latter part of the study. Additionally, our participants were highly adherent to the use of the mobile health technology device (Fitbit) for self-monitoring of diet and activity. Although median adherence was high, we did observe a wide range with the lowest adherence rate of 36.4%, which indicates difficulty by some participants with self-monitoring and/or use of the technology. The procedures for logging diet and physical activity were similar in that they were both entered through the Fitbit app. Step counts were automatically logged when participants wore the Fitbit. We had a high adherence rate for wearing the Fitbit (95.6%). Although both logging of diet and physical activity was emphasized during group sessions, it is possible that participants could have assumed that since the Fitbit automatically calculates daily step count, it would also automatically detect all other types of physical activity. This could potentially have resulted in a lower logging rate for physical activity. As logging foods is tedious and requires more sophisticated user operations (selecting food items, portions, etc.), the high diet logging rate is encouraging and reaffirms that mobile technology is acceptable in this population. On the other hand, the low logging rate of physical activity suggests that the intervention may need further improvements to increase physical activity, or that there is a barrier to logging any activities that were performed. Similar findings related to a decrease in self-monitoring adherence in older adults have been reported and authors have suggested that more tailored technology instructions and further engineering may optimize user interface features (31). Future studies might consider providing direct exercise training and/or additional counseling, such as motivational interviewing, on the use of the activity tracker device. In a review of the literature of self-monitoring (diet, exercise, self-weighing) in therapeutic weight loss studies, Burke et al. reported that there was a correlation between adherence to self-monitoring and weight loss. However, there was a gradual decline in self-monitoring adherence throughout the study, which worsened when the treatment sessions decreased in frequency (32). This may, in part, explain our participants’ gradual weight increase starting with the monthly group sessions (sessions 8 to 10). However, the weight increase occurred simultaneously with the start of the local COVID-19 pandemic, which also likely impacted progress on weight loss.
The COVID-19 pandemic began during the implementation of this study, and local regulations required shelter-in-place protocols to be implemented throughout the region. Stay-at-home orders directly impacted the design of this study and led to our decision to use a virtual format to continue group sessions. This shift affected data collection for some outcomes (such as body weight measurement for every group session) and reduced in-person contact during the latter part of the study, however, we were able to continue using the group format virtually and participants recorded their self-reported weight (recording weight at home was not required for the study) for that session. The COVID-19 pandemic (started from monthly sessions 8-10) also likely had an impact on the behaviors of self-monitoring that we sought to modify with behavioral lifestyle intervention, even though bodyweight had appropriately declined prior to this point (Figure 2). This may also explain non-significant scores for self-reported health and quality of life PROMIS (e.g., anxiety, depression, participation in social activities) questionnaires score. We explored the impact of the COVID-19 experience and acceptability of the behavioral lifestyle intervention in our recently published focus group study wherein participants noted additional challenges to engaging in healthy behaviors as compared to before COVID-19 (33).
At baseline, most of our participants were characterized as pre-frail (72%) and frail (22%), supporting the high prevalence of frailty in older adults with T2D, as noted in other studies (34). Our pilot findings suggest that the Look AHEAD lifestyle intervention improves frailty characteristics (Table 3), and total frailty score (Table 4) in older adults diagnosed with T2D. In addition, our study participants started with a lower SPPB total score of 9.2 ± 2.1. A lower SPPB (score ≤ 9) is independently associated with falls (35) and frailty (36). We noticed a significant improvement in the SPPB balance score (p=0.008). Although there was a positive trend in SPPB total score it was not statistically significant (p=0.069). Similar improvements in SPPB total score have been reported by a 12-month multimodal (diet and exercise) intervention in older adults in the intervention group (37).
The small sample size and the lack of a control group were limitations of this study that should be mentioned. A small sample size limits generalizability to a broader population, and a lack of a control group limits a direct comparative set of data in individuals receiving the intervention and those who did not, which does limit the inferences that can be made. However, in spite of these limitations, this data provides information regarding the feasibility of delivering a lifestyle intervention in older adults with T2D and provides preliminary efficacy data for the effect of this intervention on frailty in this population. Secondly, the use of Fitbit to collect self-reported diet and physical activity may have led to imprecise reporting and may have allowed patient biases to affect the study’s outcome; however, this limitation is inherent to all studies using self-reported measures. We attempted to minimize imprecision and bias by using validated objective assessments during the follow-up visit. Lastly, the COVID-19 pandemic did interfere with in-person group intervention as previously discussed.


Conclusion and Implications

Our study provides evidence for the feasibility of using a behavioral lifestyle intervention, enhanced with mobile health technology, for the goal of reducing frailty and improving T2D outcomes in community-dwelling overweight/obese older adults. Our results are particularly encouraging as they were observed despite our study being impacted by the COVID-19 pandemic. We demonstrated a high retention rate and observed that personal fitness technology (Fitbit activity tracker) is acceptable and maybe a valuable intervention tool for self-monitoring in this population. Pre-frailty and frailty were common in our sample of older adults with T2D, in line with prior work showing that T2D and overweight/obesity are highly correlated with frailty. Detecting and managing frailty at its initial stages is ideal to improve healthspan and prevent disability and death. Further, screening for frailty and related measures, such as physical function and self-reported measures of health, may be beneficial for clinicians to identify high-risk patients who may benefit from the interventions to prevent or delay frailty. In the future, larger randomized controlled trials will be needed for definitive evidence of this intervention’s efficacy to reduce frailty-in this high-risk population. Future, more robust studies may include a longer intervention phase and follow-up period, and more precise measurements of diet and physical activity, such as actimetry and 24-hour diet recall. In addition, cognitive assessments may also be important to examine the impact of such an intervention on cognitive function, in addition to frailty and physical function.


Acknowledgments: We would like to thank the University of Texas Health Science Center at San Antonio’s School of Nursing for the use of the Center for Simulation Innovation, Office for Nursing Research Biobehavioral Lab, Center on Smart and Connected Health Technologies, and Geriatric Research, Education, and Clinical Center at the South Texas Veterans Health Care System, San Antonio, Texas for their support for this work.

Funding: This work was supported by the RL5 Mentored Research Career Development Award (Trainee: Rozmin Jiwani, Mentors: Sara Espinoza, Jing Wang) through the San Antonio Claude D. Pepper Older Americans Independence Center (P30AG044271) and the United States Department of Education through Title V Grant P031S150048 (Trainees: Edward Monk, Chandler Bess, Mentors: Darpan Patel, Rozmin Jiwani). This study was also supported by research infrastructure from the Geriatric Research, Education, and Clinical Center at the South Texas Veterans Health Care System.

Conflicts of Interest: There were no conflicts of interest declared by the authors for this study.

Ethics declaration: The University of Texas Health Science Center’s Institutional Review Board for the protection of human participants approved all procedures.



1. Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, et al. Frailty in older adults: evidence for a phenotype. J Gerontology A Biol Sci Med Sci. 2001;56(3):M146-56.
2. Collard RM, Boter H, Schoevers RA, Oude Voshaar RC. Prevalence of frailty in community-dwelling older persons: a systematic review. J Am Geriatr Soc. 2012;60(8):1487-92.
3. Assar ME, Laosa O, Rodríguez Mañas L. Diabetes and frailty. Curr Opin Clin Nutr Metab Care. 2019;22(1):52-7.
4. Espinoza SE, Jung I, Hazuda H. Frailty transitions in the San Antonio Longitudinal Study of Aging. J Am Geriatr Soc. 2012;60(4):652-60.
5. Li G, Prior JC, Leslie WD, Thabane L, Papaioannou A, Josse RG, et al. Frailty and Risk of Fractures in Patients With Type 2 Diabetes. Diabetes Care. 2019;42(4):507-13.
6. Morley JE, Malmstrom TK, Rodriguez-Manas L, Sinclair AJ. Frailty, sarcopenia and diabetes. JAMDA. 2014;15(12):853-9.
7. Ulley J, Abdelhafiz AH. Frailty predicts adverse outcomes in older people with diabetes. Practitioner. 2017;261(1800):17-20.
8. Espinoza SE, Jiwani R, Wang J, Wang CP. Review of Interventions for the Frailty Syndrome and the Role of Metformin as a Potential Pharmacologic Agent for Frailty Prevention. Clinical Therapeutics. 2019;41(3):376-86.
9. Pi-Sunyer X, Blackburn G, Brancati FL, Bray GA, Bright R, Clark JM, et al. Reduction in weight and cardiovascular disease risk factors in individuals with type 2 diabetes: one-year results of the look AHEAD trial. Diabetes Care. 2007;30(6):1374-83.
10. Wing RR. Long-term effects of a lifestyle intervention on weight and cardiovascular risk factors in individuals with type 2 diabetes mellitus: four-year results of the Look AHEAD trial. Arch Intern Med. 2010;170(17):1566-75.
11. Look Ahead Research Group. Effect of a long-term behavioural weight loss intervention on nephropathy in overweight or obese adults with type 2 diabetes: a secondary analysis of the Look AHEAD randomised clinical trial. Lancet Diabetes Endocrinol. 2014;2(10):801-9.
12. Rejeski WJ, Bray GA, Chen SH, Clark JM, Evans M, Hill JO, et al. Aging and physical function in type 2 diabetes: 8 years of an intensive lifestyle intervention. J Gerontology A Biol Sci Med Sci. 2015;70(3):345-53.
13. Rejeski WJ, Ip EH, Bertoni AG, Bray GA, Evans G, Gregg EW, et al. Lifestyle change and mobility in obese adults with type 2 diabetes. N Engl J Med. 2012;366(13):1209-17.
14. Redmon JB, Bertoni AG, Connelly S, Feeney PA, Glasser SP, Glick H, et al. Effect of the look AHEAD study intervention on medication use and related cost to treat cardiovascular disease risk factors in individuals with type 2 diabetes. Diabetes Care. 2010;33(6):1153-8.
15. Espeland MA, Glick HA, Bertoni A, Brancati FL, Bray GA, Clark JM, et al. Impact of an intensive lifestyle intervention on use and cost of medical services among overweight and obese adults with type 2 diabetes: the action for health in diabetes. Diabetes Care. 2014;37(9):2548-56.
16. Simpson FR, Pajewski NM, Nicklas B, Kritchevsky S, Bertoni A, Ingram F, et al. Impact of Multidomain Lifestyle Intervention on Frailty Through the Lens of Deficit Accumulation in Adults with Type 2 Diabetes Mellitus. J Gerontology A Biol Sci Med Sci. 2019.
17. Agarwal S, LeFevre AE, Lee J, L’Engle K, Mehl G, Sinha C, et al. Guidelines for reporting of health interventions using mobile phones: mobile health (mHealth) evidence reporting and assessment (mERA) checklist. BMJ (Clinical research ed). 2016;352:i1174.
18. Anderson M, Perrin A. Technology Adoption Climbs Among Older Adults. Washington, DC: Pew Research Center. 2017. https://www.pewresearch.org/internet/2017/05/17/technology-use-among-seniors/
19. Elavsky S, Knapova L, Klocek A, Smahel D. Mobile Health Interventions for Physical Activity, Sedentary Behavior, and Sleep in Adults Aged 50 Years and Older: A Systematic Literature Review. J Aging Phys Act. 2019;27(4):565-93.
20. Espinoza SE, Orsak B, Wang CP, MacCarthy D, Kellogg D, Powers B, et al. An Individualized Low-Intensity Walking Clinic Leads to Improvement in Frailty Characteristics in Older Veterans. J Frailty Aging. 2019;8(4):205-9.
21. Jiwani R, Wang J, Berndt A, Ramaswamy P, Mathew Joseph N, Du Y, et al. Changes in Patient-Reported Outcome Measures With a Technology-Supported Behavioral Lifestyle Intervention Among Patients With Type 2 Diabetes: Pilot Randomized Controlled Clinical Trial. JMIR Diabetes. 2020;5(3):e19268.
22. Wang J, Cai C, Padhye N, Orlander P, Zare M. A Behavioral Lifestyle Intervention Enhanced With Multiple-Behavior Self-Monitoring Using Mobile and Connected Tools for Underserved Individuals With Type 2 Diabetes and Comorbid Overweight or Obesity: Pilot Comparative Effectiveness Trial. JMIR mHealth and uHealth. 2018;6(4):e92.
23. Jiwani R, Wang C-P, Orsak B, MacCarthy D, Kellogg D, Powers B, et al. A geriatrics walking clinic improves hemoglobin A1c and time gait in veterans with type 2 diabetes. Geriatric Nursing. 2020.
24. Moore CG, Carter RE, Nietert PJ, Stewart PW. Recommendations for planning pilot studies in clinical and translational research. Clin Transl Sci. 2011;4(5):332-7.
25. Folsom AR, Jacobs DR, Jr., Caspersen CJ, Gomez-Marin O, Knudsen J. Test-retest reliability of the Minnesota Leisure Time Physical Activity Questionnaire. J Chronic Dis. 1986;39(7):505-11.
26. Espinoza SE, Hazuda HP. Frailty in older Mexican-American and European-American adults: is there an ethnic disparity? J Am Geriatr Soc. 2008;56(9):1744-9.
27. Guralnik JM, Ferrucci L, Simonsick EM, Salive ME, Wallace RB. Lower-extremity function in persons over the age of 70 years as a predictor of subsequent disability. N Engl J Med. 1995;332(9):556-61.
28. Guralnik JM, Simonsick EM, Ferrucci L, Glynn RJ, Berkman LF, Blazer DG, et al. A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontology A Biol Sci Med Sci. 1994;49(2):M85-94.
29. Cella D, Riley W, Stone A, Rothrock N, Reeve B, Yount S, et al. The Patient-Reported Outcomes Measurement Information System (PROMIS) developed and tested its first wave of adult self-reported health outcome item banks: 2005-2008. J Clin Epidemiol. 2010;63(11):1179-94.
30. Meek G, Ozgur C, Dunning K. Comparison of the t vs. Wilcoxon Signed-Rank Test for Likert Scale Data and Small Samples. J. Mod. Appl. Stat. Methods. 2007;6(1).
31. Rye Hanton C, Kwon Y-J, Aung T, Whittington J, High RR, Goulding EH, et al. Mobile Phone-Based Measures of Activity, Step Count, and Gait Speed: Results From a Study of Older Ambulatory Adults in a Naturalistic Setting. JMIR mHealth and uHealth. 2017;5(10):e104.
32. Burke LE, Wang J, Sevick MA. Self-monitoring in weight loss: a systematic review of the literature. J Am Diet Assoc. 2011;111(1):92-102.
33. Jiwani R, Dennis B, Bess C, Monk S, Meyer K, Wang J, et al. Assessing acceptability and patient experience of a behavioral lifestyle intervention using fitbit technology in older adults to manage type 2 diabetes amid COVID-19 pandemic: A focus group study. Geriatric Nursing. 2021;42(1):57-64.
34. Pritchard JM, Kennedy CC, Karampatos S, Ioannidis G, Misiaszek B, Marr S, et al. Measuring frailty in clinical practice: a comparison of physical frailty assessment methods in a geriatric out-patient clinic. BMC Geriatr. 2017;17(1):264.
35. Lauretani F, Ticinesi A, Gionti L, Prati B, Nouvenne A, Tana C, et al. Short-Physical Performance Battery (SPPB) score is associated with falls in older outpatients. Aging Clin Exp Res. 2019;31(10):1435-42.
36. Bandinelli S, Lauretani F, Boscherini V, Gandi F, Pozzi M, Corsi AM, et al. A randomized, controlled trial of disability prevention in frail older patients screened in primary care: the FRASI study. Design and baseline evaluation. Aging Clin Exp Res. 2006;18(5):359-66.
37. Rodriguez-Mañas L, Laosa O, Vellas B, Paolisso G, Topinkova E, Oliva-Moreno J, et al. Effectiveness of a multimodal intervention in functionally impaired older people with type 2 diabetes mellitus. J Cachexia Sarcopenia Muscle. 2019;10(4):721-33.




I. Ader1, L. Pénicaud1, S. Andrieu2, J.R. Beard3, N. Davezac4, C. Dray5, N. Fazilleau6, P. Gourdy5, S. Guyonnet7, R. Liblau6, A. Parini4, P. Payoux8, C. Rampon5, I. Raymond-Letron1, Y.Rolland2,7, P. de Souto Barreto2,7, P. Valet5, N. Vergnolle9, F. Sierra10, B. Vellas2,7, L. Casteilla1

1. RESTORE, UMR 1301-Inserm 5070 Etablissement Français du Sang-Occitanie (EFS), Inserm 1031, University of Toulouse III, National Veterinary School of Toulouse (ENVT), CNRS, Toulouse, France; 2. Inserm UMR 1027, Toulouse, France; University of Toulouse III, Toulouse, France; Department of Epidemiology and Public Health, CHU Toulouse, Toulouse, France; 3. Centre of Excellence in Population Ageing Research, University of New South Wales, Sydney, Australia; 4. Centre de Recherches sur la Cognition Animale (CRCA), Centre de Biologie Intégrative (CBI), CNRS, University of Toulouse III, Toulouse, France ; 5. Institut des Maladies Métaboliques et Cardiovasculaires, Inserm/Université Paul Sabatier UMR 1048 – I2MC 1 avenue Jean Poulhès BP 84225 31432 Toulouse Cedex 4 – France; 6. Infinity – Institut Toulousain des Maladies Infectieuses et Inflammatoires, CNRS U5051, INSERM U1291, University of Toulouse III, Toulouse, France; 7. Gérontopôle, Department of Geriatrics, CHU Toulouse, Toulouse, France; 8. ToNIC, Toulouse NeuroImaging Center, Université de Toulouse III, Inserm, UPS, France; 9. IRSD, Université de Toulouse, INSERM, INRA, ENVT, Université de Toulouse III, U1220, CHU Purpan, CS60039, 31024, Toulouse, France; 10. Institut du Vieillissement, Université de Toulouse, CHU Toulouse France

Corresponding Author: Louis Casteilla, RESTORE, UMR 1301-Inserm 5070 Etablissement Français du Sang-Occitanie (EFS), Inserm 1031, University of Toulouse III, National Veterinary School of Toulouse (ENVT), CNRS, Toulouse, France; louis.casteilla@inserm.fr
J Frailty Aging 2021;10(4)313-319
Published online April 24, 2021, http://dx.doi.org/10.14283/jfa.2021.15



The find solutions for optimizing healthy aging and increase health span is one of the main challenges for our society. A novel healthcare model based on integration and a shift on research and care towards the maintenance of optimal functional levels are now seen as priorities by the WHO. To address this issue, an integrative global strategy mixing longitudinal and experimental cohorts with an innovative transverse understanding of physiological functioning is missing. While the current approach to the biology of aging is mainly focused on parenchymal cells, we propose that age-related loss of function is largely determined by three elements which constitute the general ground supporting the different specific parenchyma: i.e. the stroma, the immune system and metabolism. Such strategy that is implemented in INSPIRE projects can strongly help to find a composite biomarker capable of predicting changes in capacity across the life course with thresholds signalling frailty and care dependence.

Key words: Biomarkers, healthy aging, frailty, dependence.




The result of the balance between biological damage accumulation and compensatory mechanisms (1). As time passes, the compensatory mechanisms become less and less effective, which leads to more and more damage, a gradual decrease in physiological reserves and, phenotypically, this manifests as aging.

Healthy aging

Healthy aging is defined by The World Health Organization (WHO) as the process of developing and maintaining the functional ability that enables well-being in older age, including not just absence of disease, but also happiness, satisfaction and fulfillment (2). This requires to consider health span as final outcome instead of life span that corresponds to the duration of life whatever the quality of this life.


When an individual displays a maximal intrinsic capacity and resilience, she/he is considered as robust.


Frailty is a progressive age-related clinical condition characterized by a deterioration of physiological capacity leading to an increased vulnerability of the individual (3) and a higher risk of having poor health as well as a faster entry into care dependence (4). Frailty status occurs when the mobilization of the physiological reserve is not able to overcome the challenge (see adaptive capacity and resilience).

Intrinsic capacity (IC)

IC is the composite of 5 major physiological and mental capacities of an individual that can be assessed in a day-to-day environment (5, 6). Taking into account data from large cohort studies (7, 8), the WHO has suggested 5 key domains of capacity to maintain autonomy: sensorial (vision, audition), locomotor, cognitive, psychological and underlying physiological/cellular processes (9). These domains influence each other and are in turn influenced by environmental determinants.

Functional ability

Comprises those health-related attributes that enable people to be able to do what they want to do. This ability comprises not just the intrinsic capacity of the individual, but also the interactions between the individual and their environment.
Adaptive capacity, reserve and resilience: For a function, the ability to handle stress depends on the ability to mobilize adaptive capacity. For an individual, reserves are her/his maximal adaptive capacity. When adaptive capacity is more related to a physiological perspective, the term resilience is more generic and is coming from psychological field. At any time when the reserve exceeds the required adaptation capacity necessary to face a challenge, the individual is resilient. Frailty takes place as soon as these reserves become limited and resilience is diminished, resulting in a high sensitivity to any challenge that could lead to dependency (10).


Geroscience aims to identify the underlying molecular causes of chronic diseases and aging-related conditions. It focuses on hallmarks of the aging process, irrespective of the tissue, since most of these hallmarks are molecular and not specific to a given cell type. For example, a large literature is emerging that indicates the importance of cellular senescence in defining the aging phenotype (11).


Irreversible form of long-term cell-cycle arrest, caused by excessive intracellular or extracellular stress or damage. Senescent cells secrete a panel of molecules deleterious for surrounding cells.


The challenges of healthy aging and the search for biomarkers

While life expectancy has increased in recent decades, healthy life expectancy has not increased to the same extent, which means that people are living more years with functional losses generating crucial socioeconomical issues (12). As proposed by the WHO, a novel healthcare model must be privileged focused on heathy aging (glossary). This new paradigm requires an integrative and multi-disciplinary view based on the functional ability of individual to be able to do what they have reason to value. This corresponds to an interplay between the intrinsic capacity (IC, glossary) of each individual with its own environment. Intrinsic capacity can be described as the interactions between 5 domains (cognition, psychosocial, locomotion, sensory, vitality) that permits each individual to be autonomous. Consistent with this view, recent analysis suggests that quantifying an individual IC is a more useful predictor of future care dependence that the number of clinical morbidities an individual may be experiencing (9). Altogether this requires a shift of current research and healthcare models generally reactive disease-and organ centered towards the maintenance of optimal functional levels defining preventive and function-centered view. In this perspective, identifying frail persons (glossary) and especially when their IC declines and their resilience is limited (pre-frail state, glossary), as well as monitoring their individual evolution in order to propose solutions to maintain or recover IC and to prevent dependency is crucial. Pre-frailty was identified as a potentially reversible condition. Having this function-centered approach in mind, aging can be understood as the result of a subtle and progressive dysregulation of different balances that change overtime even at infra-clinic level rather than the result of a sudden appearance of one molecular or cellular defect. There is great individual diversity in the rate of this deterioration. This makes healthy aging research particularly tricky as well as in need of new multidisciplinary and global approaches to cover this unmet clinical need (13).
While chronological age (civil age, date of birth) is an easy-to-determine number and only reflects the time a person has spent on Earth, no precise definition is yet available for biological age. The basic idea behind biological aging is that aging occurs as damage to various cells and thus tissues in the body, which accumulate gradually with time and decrease the capacity to handle stress. It is the result of the interactions between intrinsic (eg. genetic, epigenetic, physiological…) and extrinsic (eg. lifestyle, infections…) factors that vary individually. In consequence, biological age differs from individual to individual for the same chronological age and thus the kinetics and slope of the biological aging process are unique to each of us (14). Very recently, the Snyder’s team has started to decipher certain elements that could explain biological age (15). They defined different types of biological aging patterns in different individuals, termed “ageotypes”. They discovered that people tend to fall into one of four biomolecular pathways associated with kidney, liver, metabolic and immune dysregulations. Some people fall squarely into one category, but others may meet the criteria of the four, depending on how their biological systems resists against aging. Unfortunately this work lacks long-term follow-up to assess whether these molecular phenotypes were ultimately associated with physiological failures.
Altogether, these reports imply that prediction, prevention and care must be personalized and adjusted to the individual’s biological, not just chronological age. This would be in line with the vision of 4P medicine (personalized, preventive, predictive and participatory) as proposed by Hood and colleagues (16-18). Even though biomarkers related to aging are at present widely used in medicine and a first list of 258 putative candidate biomarkers of aging was proposed by the TAME workgroup (19), identifying relevant biomarkers parameters in the context of biological aging and more specifically to the prefrail and frail state represents a new challenge, simply because those biomarkers do not fit within the disease-centred approaches of the current medicine.


Biomarkers for healthy aging: Contribution of the Occitanie Toulouse INSPIRE initiative

It is now widely recognized that the decline of health and the appearance of aging phenotypes would be the result of simultaneous deregulations of multiple physiological systems and their interactions (20). Thus, it is likely that only the use of multiple or composite biomarkers is relevant to assess these global body disturbances. In the field of healthy aging, ideal biomarkers would be to monitor the evolution towards frailty in order to intervene as soon as possible. It would be a single composite measurement used throughout the lifetime of individuals that encompasses and combines many physiological parameters. Such biomarkers should be capable of predicting changes in capacity across the life course with thresholds signalling the evolution from one state to another. Through deconvolution into their individual components, it should suggest the most appropriate pathway of care (prevention or/and treatment) for a given individual. Furthermore, the use of these biomarkers must be able to not just assess overt expressions of capacity but the resilience of an individual’s capacities in the face of additional stressors particularly for pre-frail biomarkers. The assessment of resilience is an important and challenging issue because this requires identifying the best challenge suitable to reveal the maximal adaptive capacity of the organism as a whole. Furthermore, in the best-case scenario, ideal biomarkers should be able to inform the clinician about the best-adapted care pathway for the individual (Figure 1). As usual, ideal biomarkers should also be robust, specific, as non-invasive as possible and cost-effective. Obviously, such biomarkers are much closer to fantasy than to reality, but such a definition can help the community to direct progress towards this tremendous challenge (21).

Figure 1. Markers for 4 P Healthy aging Medicine

Figure 1: Clinical markers (such as those derived from ICOPE) should be assessed at first to determine the category (robust, pre-frail, frail and dependent) of individuals of various chronological ages and what WHO function(s) is (are) altered. This should help to facilitate the search for relevant biological markers taking into account both WHO functions, hallmarks of aging as well as markers based on the SIM-INSPIRE strategy. By using artificial intelligence, we believe that relevant mixed-markers of biological ages will be made available.


A large list of potential biomarkers is already available and discussed in several reviews, but these biomarkers are most often unique, studied and validated as such for specific pathologies or derived from preclinical studies that have taken changes in disease specific physiology or lifespan, rather than health expectancy. A great step forward to identify transverse elements has been taken with the development of geroscience that considers biological aging as the main risk factors of chronic diseases (22). To go beyond these concepts and in the framework of WHO perspective, we propose to implement this analysis with a physiological transverse analysis. While the current approach to biology of aging is mainly focused on parenchymal cells that are specific to each organ consistent with the current organ/disease centered view, we propose that age-related loss of function is largely determined by dysfunctions of key generic components of tissue repair processes associated with tissue renewal and particularly the defective support of parenchyma cell environment (23, 24). As such, we propose to focus much of the investigations on the stroma, the immune system and metabolism (SIM), three elements considered as the general ground that constitutes the soil supporting the emergence and the functioning of the different specific parenchyma. Inflammation and immunity both represent a warning signal and the housekeeping guardians of tissue integrity. Mesenchymal stem/stroma cells (MSC) are supportive cells of the stroma that create and maintain the functional macro-architecture of tissues, in close connexion with vascular cells. Finally, metabolism is a central component that controls any cell decision and fate. Under challenges and chronic injuries, it is well known that these three inter-related components drive cell turnover and repair outcomes. However, malfunctioning on any of these components leads to i) chronic inflammation and a low-grade immune attrition (25), generation of fibrotic cells with extracellular matrix accumulation and oxidative damage within tissue (26) and ii) accumulation of ectopic adipocytes associated with loss of metabolic flexibility and organ dysfunctions (27). These features are classically described as key elements of the aging process and age-related organ dysfunction. The INSPIRE project proposes to set-up an interdisciplinary approach that gathers clinical and scientific community working on shared bio-resources (28, 29), a fine characterization of these three inter-related components and two types of key parenchyma cells: neurons and muscle cells, corresponding to the WHO-defined functions of locomotion and cognition.


A large panel of bio-resources facilitating the discovery of biomarkers

INSPIRE shared bio-resources

Identifying relevant biomarkers to detect and/or predict human frailty is challenging according to the complexity of the thematic and the human lifespan. These limits can be circumvented by the creation of a panel of biological models and samples that can cover the entire chain from the bench to the bedside and vice-versa. As starting point and at population level, the first personalized assessment to detect frailty to anticipate care dependency can be based on a non-invasive self-evaluation that can be easily deployed to a broad population in its specific environment through a digital application (30). When a problem is flagged, a team of health workers is alerted, and the individual taken care of in a more specific and detailed way guided by further assessment (Figure 1). Beside such e-cohort serving as a pool of participants to further investigations, different cohort can be implemented, including humans, dogs, mice and fish (Fig 2). In longitudinal human cohort, the longitudinal follow-up of the same individuals with a short delay between recurrent evaluations will ensure a reliability of the data and a fine description of any change in IC associated with any relevant life events (28, 30). A computational approach will permit to define predictive and prognostic biomarkers. As many experiments are not easily done in human being due both to experimental procedures, ethics considerations as well as the quite long life-time, complementary animal cohorts should help the quest of putative biomarkers.

INSPIRE aims to build and manage various cohorts in different species including human, rodents, fishes and dogs.


As the simplest but relevant model for studying biological aging, Nothobranchius Furzeri (African or Turquoise killifish) the vertebrate with the shortest known lifespan can be used (31). This animal develops very rapid aging of all its organs (4 to 6 months), with the recapitulation of all the aging hallmarks found in older-adults patients, both at the molecular and functional level (cognitive alterations, reduction of locomotion, sarcopenia onset…) (32, 33) and its health (meaning behavior) can be finely evaluated. It represents a unique animal model to quickly test and validate new hypotheses within the framework of the INSPIRE project. A murine cohort as small mammals and gold-standard physiopathological models has been designed as a mirror to the human cohort and fully described in (34). Finally, a dog cohort from adulthood to elderly represents an unvaluable tool because dog is a mammal displaying rapid aging compared to humans but with a very similar physiology including aging features and pathologies, as well as an evolution towards biological aging with IC reduction. Furthermore, it is possible to evaluate the biological and physical parameters of dogs in a manner almost identical to that of humans (physical and clinical examinations, blood and urine analyzes, medical imaging, functional tests, etc.). In addition, dogs share with their owners their environment (air, water, food, exposure to potential contaminants, …) and their lifestyle (active or sedentary) (35, 36).
In all of these animal cohorts, individuals has to be widely phenotyped and generate a large biobanking in order not only to identify putative biomarkers that can be tested and validated in human cohort but also to better investigate those mechanisms identified in the human cohorts in terms of changes in metabolism, immune and regenerative capacities.

The challenges to identify resilience markers

Prefrail state can be determined only if the individual is submitted to a defined challenge or stress which will led to a return, or not, to a “normal” state if resilience is present, or not. Thus, to evaluate functional/physiological resilience as part of a battery of potential biomarkers of aging, we propose to carefully evaluate responses to stressors, either “lifelong challenges” or well-controlled acute stressors. In fact, different physical and psychological life events and injuries can be considered as real-life challenges, including for example both acute or chronic infections (HIV, flu, COVID-19…) but also other life events such as hip fracture or surviving the death of a spouse or a relative, during a divorce or the loss of the job. Since the follow-up of people enrolled in the INSPIRE human cohort is frequent, as soon as such challenge occurs, the stress effect on functions and biological parameters will be informed during the following evaluation. Thus, INSPIRE design can reveal physiological adaptations specific to biological aging in order to unmask biomarkers following a situation of stress.
For controlled stressors, we propose to challenge immune, inflammatory and metabolic homeostasis by an exercise corresponding to the evaluation of muscle tone (isometric measurements) coupled with intense physical exercise. Blood sampling before and just after this VO2 max challenge will allow to a better appraising of physiological reserve, as well as to test the acute response of metabolic, immune and inflammatory blood biomarkers after a maximal effort. To in vivo assess the immune system, we will monitor the magnitude and quality of the antigen-specific immune response as well as the bystander inflammatory response following administration of a vaccine included in standard care (flu, pneumococcal or zoster vaccination). Serum and peripheral blood mononuclear cells will be prospectively collected prior to, and at different time points following the vaccine shots. We will use several ‘omics’ technologies to monitor population and individual-level changes such as cell-subset phenotyping, cytokine response assays, and gene expression profiles, which will allow the longitudinal tracking of multiple immune features. These investigations can be supplemented by ex vivo challenges of precise immune cell types using stimuli targeting either innate immune cells or lymphocyte subsets. With these investigations, we aim at assessing not only the immune reserves but also immune cell global dysfunction, which may contribute by itself to premature aging, as elegantly documented in mice (37). To assess the role of supportive mesenchymal stroma/stem cells (MSC) in repair processes, the definition of a relevant physiological challenge and the way to investigate it is very tricky. The evaluation of the kinetics and quality of wound healing processes after a standardized small incision required for tissue sampling could represent a relevant test. In addition, the ability of MSC isolated from biopsies to participate in repair processes can be also indirectly evaluated through their fine characterization and their repair potential. The quality of the supportive mesenchymal compartment will be determined after an in vitro challenge induced by various inducers mimicking native molecules released after tissue injury. This corresponds to the determination of MSC multipotency including the sensitivity of these cells to differentiate towards myofibroblast as well as their pleiotropic induced paracrine activities. This will be achieved via robotized phenotyping and standardized assays using Advanced Medicinal Therapeutical Product quality control facility (www.ecellfrance.fr).

Implementation and perspectives

An important aspect is the rapid sharing of bio-resources and data to a large multi-disciplinary community. Pending sufficient data from the living longitudinal human cohort, cross-sectional studies will be carried out to develop first lists of putative biomarker profiles with the corresponding relevant thresholds for their uses in clinical practice. Since the healthy aging field is inherently very complex, the irruption of artificial intelligence in precision medicine should be of a great help to build expert systems suitable to deploy 4P Medicine approaches (38, 39). This should largely change the field of biomarkers, moving from a single or a limited profile of biomarkers with associated relevant thresholds towards a profile of multi-modal parameters including multi-Data driven algorithms.
A complete biomarker kit to address the issue of healthy aging should contain biomarkers of pre-frailty and frailty capable of predicting individual health trajectory and the risk of care dependency. Initial biomarker sets will be tested as composite biomarkers on the INSPIRE longitudinal cohort but also will reveal putative targets for innovative therapeutic strategies that could be investigated and tested in animal models. From this first round of development and evaluation, sequential rounds will be conducted to refine and adapt these initial sets with definite validation or invalidation, inclusion of new biomarkers and a better definition of the relevant thresholds (Figure 3). Thus, the first Icope version is particularly appropriate for a large screen of individuals in primary care, but it is primarily adapted to individuals of more than 60 years old. So, if the goal is to be able to predict the risk for subsequent abnormal declines in capacity or frailty at an early age, the present evaluation should be adapted for young people prior to overt losses of capacity, it is imperative that such future measurements include parameters of resilience to challenges and stresses. For the future we can reasonably envision other types of approaches, such as artificial intelligence based image analysis methodologies. Indeed, facial recognition and locomotion gait analyses represent promising bio-visual biomarker tools, because of their accuracy and ease of implementation at the whole population level, using different digital applications.

Figure 3. INSPIRE strategy

INSPIRE will used the shared bio-resources (see figure 2) and mainly focalize on two WHO functions namely cognition (C) and mobility (M).


The innovative paradigm of mixing the clinical perspective provided by the WHO innovative vision of tissue homeostasis with the geroscience perspective and modern digital tools would lead to provide a composite score to determine the biological age of each individual. Artificial intelligence will allow to achieve this objective and play a transformative role in diagnosis and therapeutic strategy. Indeed, data analysis in complex datasets comprising multiple clinical and biological variables repeated over time is extremely difficult for humans. The development of artificial intelligence in the medical field, through machine learning methodologies adapted to the need to take into account temporal data, makes it possible to envisage tools that may in the near future help the clinician to identify relevant biomarkers and to make outcome predictions. One of the keys to success in achieving this objective consists in the development of data sharing with standardized high quality data sets. This is made possible by a major scientific and financial investment by both research institutions and local authorities in INSPIRE project. It is the beginning of a long effort absolutely required to face one of the major societal challenges of our time.


Authors contribution: All authors of the paper “The INSPIRE research initiative: a program for GeroScience and healthy aging research going from animal models to humans and the healthcare system” participated to the writing of the manuscript. LC, LP and IA supervised the writing and the final version of the manuscript.
Acknowledgments: The Inspire Program is supported by grants from the Region Occitanie/Pyrénées-Méditerranée (Reference number: 1901175), the European Regional Development Fund (ERDF) (Project number: MP0022856).

Conflict of interest: All authors of the paper “The INSPIRE research initiative: a program for GeroScience and healthy aging research going from animal models to humans and the healthcare system” declare no Conflicts of Interest related to this manuscript.

Open Access: This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, duplication, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.



1. Thuault, S., Reflections on aging research from within the National Institute on Aging. Nature aging, 2021: p. 14-18.
2. Beard, J.R., et al., The World report on ageing and health: a policy framework for healthy ageing. Lancet, 2016. 387(10033): p. 2145-2154.
3. Clegg, A., et al., Frailty in elderly people. Lancet, 2013. 381(9868): p. 752-62.
4. Dent, E., et al., Management of frailty: opportunities, challenges, and future directions. Lancet, 2019. 394(10206): p. 1376-1386.
5. WHO, World report on ageing and health. Global strategy and action plan on ageing and health. J Frailty Aging, 2015.
6. Cesari, M., et al., Evidence for the Domains Supporting the Construct of Intrinsic Capacity. J Gerontol A Biol Sci Med Sci, 2018. 73(12): p. 1653-1660.
7. At, J., et al., Frailty and the prediction of dependence and mortality in low- and middle-income countries: a 10/66 population-based cohort study. BMC Med, 2015. 13: p. 138.
8. Van der Vorst, A., et al., Correction: Limitations in Activities of Daily Living in Community-Dwelling People Aged 75 and Over: A Systematic Literature Review of Risk and Protective Factors. PLoS One, 2017. 12(1): p. e0170849.
9. Beard, J.R., et al., The structure and predictive value of intrinsic capacity in a longitudinal study of ageing. BMJ Open, 2019. 9(11): p. e026119.
10. Ferrucci, L., et al., Measuring biological aging in humans: A quest. Aging Cell, 2020. 19(2): p. e13080.
11. Khosla, S., et al., The role of cellular senescence in ageing and endocrine disease. Nat Rev Endocrinol, 2020. 16(5): p. 263-275.
12. Tabue-Teguo, M., et al., Frailty in elderly: a brief review. Geriatr Psychol Neuropsychiatr Vieil, 2017. 15(2): p. 127-137.
13. Kirkwood, T.B., A systematic look at an old problem. Nature, 2008. 451(7179): p. 644-7.
14. Belsky, D.W., et al., Quantification of biological aging in young adults. Proc Natl Acad Sci U S A, 2015. 112(30): p. E4104-10.
15. Ahadi, S., et al., Personal aging markers and ageotypes revealed by deep longitudinal profiling. Nat Med, 2020. 26(1): p. 83-90.
16. Weston, A.D. and L. Hood, Systems biology, proteomics, and the future of health care: toward predictive, preventative, and personalized medicine. J Proteome Res, 2004. 3(2): p. 179-96.
17. Hood, L., et al., Systems biology and new technologies enable predictive and preventative medicine. Science, 2004. 306(5696): p. 640-3.
18. Hood, L., R. Balling, and C. Auffray, Revolutionizing medicine in the 21st century through systems approaches. Biotechnol J, 2012. 7(8): p. 992-1001.
19. Justice, J.N., et al., A framework for selection of blood-based biomarkers for geroscience-guided clinical trials: report from the TAME Biomarkers Workgroup. Geroscience, 2018. 40(5-6): p. 419-436.
20. Li, Q., et al., Homeostatic dysregulation proceeds in parallel in multiple physiological systems. Aging Cell, 2015. 14(6): p. 1103-12.
21. Guerville, F., et al., Revisiting the Hallmarks of Aging to Identify Markers of Biological Age. J Prev Alzheimers Dis, 2020. 7(1): p. 56-64.
22. Sierra, F., The Emergence of Geroscience as an Interdisciplinary Approach to the Enhancement of Health Span and Life Span. Cold Spring Harb Perspect Med, 2016. 6(4): p. a025163.
23. Lopez-Otin, C., et al., Metabolic Control of Longevity. Cell, 2016. 166(4): p. 802-821.
24. Lopez-Otin, C., et al., The hallmarks of aging. Cell, 2013. 153(6): p. 1194-217.
25. Keenan, C.R. and R.S. Allan, Epigenomic drivers of immune dysfunction in aging. Aging Cell, 2019. 18(1): p. e12878.
26. Lu, L., et al., Cardiac fibrosis in the ageing heart: Contributors and mechanisms. Clin Exp Pharmacol Physiol, 2017. 44 Suppl 1: p. 55-63.
27. Palmer, A.K. and J.L. Kirkland, Aging and adipose tissue: potential interventions for diabetes and regenerative medicine. Exp Gerontol, 2016. 86: p. 97-105.
28. P. De Souto Barreto, S.G., I. Ader, S. Andrieu, L. Casteilla,N. Davezac, C. Dray, N. Fazilleau, P. Gourdy, R. Liblau, A. Parini, P. Payoux,L. Pénicaud, C. Rampon, Y. Rolland, P. Valet, N. Vergnolle, B. Vellas, The Inspire Research Initiative: A Program for Geroscience and Healthy Aging Research going from Animal Models to Humans and the Healthcare System. The Journal of Frailty & Aging, 2020.
29. S Guyonnet, Y.R., C Takeda, P-J Ousset, I Ader, N.Davezac, C Dray, N Fazilleau, P Gourdy, R Liblau, A Parini, P Payoux, L Pénicaud, C Rampon, P Valet, N Vergnolle, S Andrieu,P de Souto Barreto, L Casteilla, B Vellas., The INSPIRE Bio-resource Research Platform for Healthy Aging and Geroscience: Focus on the Human Translational Research Cohort (The INSPIRE-T Cohort). The Journal of Frailty & Aging, 2020.
30. N. Tavassoli, A.P., C. Berbon, J. de kerimel, C. Lafont, P. De souto Barreto, S. Guyonnet, C. Takeda, I. Carrie, D. Angioni, F. Paris, C. Mathieu, P.J. Ousset, L. Balardy, T. Voisin, S. Sourdet, J. Delrieu, V. Bezombes,S. Andrieu, F. Nouhashemi, Y. Rolland, M.E. Soto, J. Beard, Y. Sumi, I. Araujo carvalho, B. Vellas, Framework implementation of the inspire icope-care program in collaboration with the world health organization (who) in the occitania region. The Journal of Frailty & Aging, 2020.
31. Singh, P.P., et al., The Genetics of Aging: A Vertebrate Perspective. Cell, 2019. 177(1): p. 200-220.
32. Valenzano, D.R., et al., The African Turquoise Killifish Genome Provides Insights into Evolution and Genetic Architecture of Lifespan. Cell, 2015. 163(6): p. 1539-54.
33. Poeschla, M. and D.R. Valenzano, The turquoise killifish: a genetically tractable model for the study of aging. J Exp Biol, 2020. 223(Pt Suppl 1).
34. Y Santin, S.L., I Ader, S Andrieu, N Blanchard, A Carrière, L Casteilla, B Cousin, N Davezac, P De Souto Barreto, C Dray, N Fazilleau, D Gonzalez-Dunia, P Gourdy, S Guyonnet, N Jabrane-Ferrat, O Kunduzova, F Lezoualc’h, R Liblau, L O. Martinez, C Moro, P Payoux, L Pénicaud, V Planat-Bénard, C Rampon, Y Rolland, Joost-P Schanstra, F Sierra, P Valet, A Varin, N Vergnolle, B Vellas, J Viña, B P. Guiard, A Parini, Towards a large-scale assessment of relationship between biological and chronological aging: The INSPIRE Mouse Cohort. The Journal of Frailty & Aging, 2020.
35. Creevy, K.E., et al., The Companion Dog as a Model for the Longevity Dividend. Cold Spring Harb Perspect Med, 2016. 6(1): p. a026633.
36. Kaeberlein, M., K.E. Creevy, and D.E. Promislow, The dog aging project: translational geroscience in companion animals. Mamm Genome, 2016. 27(7-8): p. 279-88.
37. Desdin-Mico, G., et al., T cells with dysfunctional mitochondria induce multimorbidity and premature senescence. Science, 2020. 368(6497): p. 1371-1376.
38. Goecks, J., et al., How Machine Learning Will Transform Biomedicine. Cell, 2020. 181(1): p. 92-101.
39. Zhavoronkov, A., et al., Artificial intelligence for aging and longevity research: Recent advances and perspectives. Ageing Res Rev, 2019. 49: p. 49-66.



C.P. Launay1,2,3, L. Cooper-Brown2,3, V. Ivensky2,4, O. Beauchet1,2,4,5,6,7

1. Department of Medicine, Division of Geriatric Medicine, Sir Mortimer B. Davis – Jewish General Hospital and Lady Davis Institute for Medical Research, McGill University, Montreal, Quebec, Canada; 2. Centre of Excellence on Longevity of McGill Integrated University Health and Social Services Network, Quebec, Canada; 3. Faculty of Medicine, McGill University, Montreal, Quebec, Canada; 4. Faculty of Medicine, University of Montreal, Montreal, Quebec, Canada; 5. Departments of Medicine and Geriatrics, University of Montreal, Montreal, Quebec, Canada; 6. Research Center of the Geriatric University institute of Montreal, Montreal, Quebec, Canada; 7. Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore.

Corresponding Author: Cyrille Launay, MD, PhD; Department of Medicine, Division of Geriatric Medicine, Sir Mortimer B. Davis – Jewish General Hospital, McGill University, 3755 chemin de la Côte-Sainte-Catherine, Montréal, QC H3T 1E2, Canada; E-mail: cyrille.launay@mcgill.ca; Phone: (+1) 514-340-8222, #23837; Fax: (+1) 514-340-7547

J Frailty Aging 2021;10(4)361-362
Published online March 18, 2021, http://dx.doi.org/10.14283/jfa.2021.9


Key words: Aged, epidemiology, community dwelling, frailty, COVID-19.

Dear Editor,

The COVID-19 pandemic had severe consequences for older adults. First, COVID-19 was associated with more severe medical complications and an increased mortality rate in older compared to younger adults (1). Second, home confinement, an intervention that reduces the spread of COVID-19, was associated with adverse consequences for the older community-dwelling population (2). It broke down social networks and the continuum of primary care, resulting in medication or food delivery issues, psychological fallout and increasing frailty risks (3). Frailty assessment provides insight into the degree of older community dwellers’ health status vulnerability, social isolation and adverse health event risks, and it should be assessed before interventions are proposed (3). We designed a short assessment tool known as “Evaluation SOcio-GERiatrique” (ESOGER) for Montreal’s homebound community-dwelling older adults (3). In a phone call, ESOGER briefly assessed frailty and social isolation and provided recommendations, facilitating contact with health or social care providers who initiate appropriate health and social care plans (3). This study aims to examine the longitudinal effects of ESOGER on frailty and social isolation in Montreal’s homebound community-dwelling older adults.
Adopting a pre-post intervention, single arm, prospective and longitudinal design, this experimental study enrolled 119 community-dwelling older adults (70.0% female) for a mean follow-up period of 23.2±13.6 days. Selection criteria were being age 70 and over, being homebound, living in Montreal, not being involved in an experimental study and agreeing to participate. The assessments (i.e., baseline and follow-up) were performed using ESOGER. ESOGER assessed frailty using the 6-item brief geriatric assessment (BGA), with scores ranging from 0 (no frailty) to 14 (severe frailty) (4), as well as social isolation through accessibility to essential services (i.e., medication and food delivery, home care) and contact with individuals (family, neighbours, friends, healthcare or social professionals) over the phone or in person. When services and/or social contact were absent, ESOGER provided recommendations addressing social isolation and/or care disruption. Evaluated outcomes were: 1) the difference between baseline and follow-up BGA scores (/14) calculated as (follow-up score – baseline score) / ((follow-up score + baseline score) /2); 2) new-onset moderate to severe frailty staging defined by scores ≥5/14 at follow-up, but not baseline, assessment; and 3) unsuccessful or successful implementation of ESOGER recommendations, using two proxies: a) new-onset social isolation and care disruption (i.e., no social isolation and no care disruption at baseline assessment with social isolation and care disruption at follow-up assessment) when ESOGER recommendations were not implemented; or b) addressed social isolation and care disruption (i.e., social isolation and care disruption at baseline assessment with no social isolation and no care disruption at follow-up assessment) when ESOGER recommendations were implemented. Linear and logistic regressions were used to examine the association between frailty (mean BGA score and moderate-to-severe frailty stage used as dependent variables in separate models) and ESOGER recommendations (used as independent variables) adjusted to participants’ baseline characteristics (Age≥ 85, sex, polypharmacy, length of follow-up). P-values less than 0.05 were considered statistically significant. All statistics were performed using SPSS (version 24.0; SPSS, Inc., Chicago, IL). The protocol received Jewish General Hospital (McGill University, Quebec, Canada) Research Ethics Committee approval.
Baseline characteristics of participants are shown in the Table 2. In our study, social isolation decreased significantly from baseline to follow-up assessments (62.2% versus 38.7% with P≤0.001), whereas frailty increased significantly (BGA score 4.1±3.2 and moderate-to-severe frailty 47.9% versus BGA score 5.1±3.5 and moderate-to-severe frailty 58.0% with P≤0.001). As shown in Table 1, the implementation of ESOGER recommendations was not associated with significant variations in frailty (P>0.166), whereas the absence of the implementation of ESOGER recommendations was associated with a significant increase in frailty (P≤0.038).

Table 1. Multiple linear and regression models showing the association between frailty (mean BGA score and moderate-to-severe frailty stage used as separated dependent variables) and ESOGER recommendations (used as independent variables), adjusted for participants’ baseline characteristics (n=119)

ESOGER: “Evaluation SOcio-GERiatrique”; ß: Coefficient or regression beta; CI: Confidence interval; *: Difference between baseline and follow-up score ranged from 0 (no frailty) to 14 (severe frailty) and calculated from the formula: ((follow-up score – baseline score) / ((follow-up score + baseline score) /2); †: Brief geriatric assessment score ≥5/14; significant P-values (i.e., ≤0.05) in bold.

Table 2. Baseline characteristics of participants (n=119)

BGA: brief geriatric assessment; SD: standard deviation; CI: Confidence interval; *: no formal (i.e., health and/or social professional) or informal (i.e., family and/or friends) help; †: inability to give the current year and/or month; ‡: Number of different medications taken daily ≥5; §: Score 6-item BGA >6 with score ranging from 0 (no frailty) to 14 (severe frailty)


Home confinement was associated with increased frailty in the studied sample of older community dwellers, and this change in frailty was dependent on ESOGER recommendation implementation. There was a significant increase in frailty when ESOGER recommendations were not implemented, whereas no change was observed when they were implemented. The negative impact of social isolation on older adults’ health condition has previously been reported (5,6). This effect could explain the increased frailty observed among participants in this study for whom ESOGER recommendations were not implemented. The absence of significant changes to frailty when ESOGER recommendations were implemented shows that telemedicine may be an effective approach to sustaining the continuum of care in vulnerable homebound patients during crises like COVID-19. The main limitation of our study is the small sample size of participants, and thus studies recruiting greater number of participants are needed to confirm the results of our pilot study. In addition, further developing such interventions may help to remotely stabilize patients and to avoid seeing them in consultation.


Conflicts of Interest: None declared by the authors.



1. Du RH, Liang LR, Yang CQ, Wang W, Cao TZ, Li M, Guo GY, Du J, Zheng CL, Zhu Q, Hu M, Li XY, Peng P, Shi HZ. Predictors of mortality for patients with COVID-19 pneumonia caused by SARS-CoV-2: a prospective cohort study. Eur Respir J. 2020;55(5):2000524.
2. https://blogs.bmj.com/bmj/2020/06/15/covid-19-will-be-followed-by-a-deconditioning-pandemic/ (July 07, 2020 date last accessed)
3. Beauchet O, Cooper-Brown L, Ivensky V, Launay CP. Telemedicine for housebound older persons during the Covid-19 pandemic. Maturitas 2020 [Epub ahead of print]
4. Launay CP, de Decker L, Kabeshova A, Annweiler C, Beauchet O. Screening for older emergency department inpatients at risk of prolonged hospital stay: the brief geriatric assessment tool. PLoS One. 2014;9(10): e110135.
5. Brooke J, Jackson D. Older people and COVID-19: Isolation, risk and ageism. J Clin Nurs. 2020;29(13‑14): 2044‑2046
6. Kf B, Jl H, Ta D. Preventing Frailty Progression during the COVID-19 Pandemic. J Frailty Aging. 2020;9(3):130‑131



Q. Roquebert1,4, J. Sicsic1, B. Santos-Eggimann2, N. Sirven1, T. Rapp1,3


1. LIRAES (EA4470), Université de Paris; Paris, France; 2. Centre universitaire de médecine générale et santé publique (Unisanté), Université de Lausanne, Switzerland; 3. LIEPP, Sciences Po, France; 4. Université de Strasbourg, Université de Lorraine, CNRS, BETA, 67000 Strasbourg, France
Corresponding author: Jonathan Sicsic (PhD), LIRAES, 45 rue des Saints-Pères, Université de Paris 75006 Paris, Jonathan.sicsic@u-paris.fr. +33(0)177219361

J Frailty Aging 2021;10(3)272-280
Published online March 15, 2021, http://dx.doi.org/10.14283/jfa.2021.7



This systematic literature review documents the link between frailty or sarcopenia, conceptualized as dimensions of physical health, and the use of long-term care services by older individuals. Long-term care services include formal and informal care provided at home as well as in institutions. A systematic review was performed according to PRISMA requirements using the following databases: PubMed-Medline, Embase, CINAHL, Web of Science, and Academic Search Premier. We included all quantitative studies published in English between January 2000 and December 2018 focusing on individuals aged 50 or more, using a relevant measurement of sarcopenia or physical frailty and a long-term care related outcome. A quality assessment was carried out using the questionnaire established by the Good Practice Task Force Report of the International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Five subsets of long-term care outcome were considered: 1/ nursing home placement (NHP), 2/ nursing home short stay (NHSS), 3/ formal personal care (FPC), 4/ formal home help (FHH), 5/ informal care (IC). Out of 1943 studies, 17 were finally included in the review. With some studies covering several LTC outcomes, frailty and / or sarcopenia were associated with increased LTC use in 17 out of 26 cases (NHP: 5/6, NHSS: 3/4, FPC: 5/7, FHH: 1/4, IC: 3/5) The association was not consistent in 5 cases (NHP: 1/6, NHSS: 1/4, FPC: 2/7, FHH: 0/4, IC: 1/5) and the association was either not significant or the results inconclusive in the remaining 9 cases. Overall, while results on sarcopenia are scarce, evidence support a positive association between frailty and LTC use. The evidence is stronger for the association of physical frailty with nursing home placement / short stay as well as on FPC. There is less (more heterogeneous) evidence regarding the correlation between physical frailty and FHH or IC use. Results need to be confirmed by more advanced statistical methods or design based on longitudinal data.

Key words: Frailty, sarcopenia, long-term care, systematic review, nursing home, informal care, formal care.



In a context of aging populations, the demand for long-term care (LTC) is increasing rapidly. Referring to the relatively large range of services needed in the long run to individuals because of their functional limitations, LTC can be provided in the community by relatives (informal care), professional nurses or care workers (formal care), or in specific institutions such as in nursing homes. LTC spending represents an increasing share of Western countries’ Gross Domestic Product (GDP) and most of these expenditures are funded by public mechanism schemes involving taxes and/or social insurance (1, 2). Projections suggest that LTC public expenditures should increase by 69% at least (in a healthy aging scenario) reaching 2.7 % of GDP in the European Union in 2070 (3).
Beyond disability status, increasing attention is being paid to the pre-disability condition as conceptualized by frailty and sarcopenia. While frailty is defined as a vulnerable health status resulting from the reduction of individuals’ physiological reserve, sarcopenia indicates the loss of muscle mass and muscle strength (4). In practice, frailty and sarcopenia are closely related, and it may be difficult to empirically distinguish these concepts. Indeed, frailty and sarcopenia often overlap; most frail older people exhibit sarcopenia, and some older people with sarcopenia are also frail (5). Frailty is relatively easy to measure in the clinical practice. Some readily accessible screening tests also exist for sarcopenia, such as the SARC-F (6). Yet precise measurement of sarcopenia involves more complex instruments such as dual x-ray absorptiometry scan to measure appendicular lean mass (ALM) and classify individuals according to pre-defined cut-points (7, 8). Overall, both are prevalent in old age and associated with adverse health events. Efforts to adapt healthcare and social systems to aging populations might benefit from taking into account the effects of frailty and sarcopenia on health and long-term care services utilization.
A growing literature has been exploring the relationship between frailty or sarcopenia and LTC use since the early 2000’s. However, it is difficult to have a clear perspective of the overall findings and the quality of evidence stemming from this literature. Indeed, publications used various measures of frailty, explored sometimes different LTC outcomes (formal care use, informal care use, nursing home use etc.), applied different statistical modelling techniques, sampling methods etc. Therefore, there is need for a systematic review to assess the quality of the methods, classify findings according to different outcomes, and provide an overall perspective of the relationship between frailty, sarcopenia and LTC use. To the best of our knowledge, such an analysis of the literature has not been done yet.
Three prior reviews of the literature have provided evidence that frailty is associated with a higher risk of hospitalization (9, 10) and nursing home placement (10, 11). These studies have found moderate evidence of the association between frailty and hospitalization or institutionalization. Results on sarcopenia are more limited: it was positively correlated with hospitalization in one study (12) but no study examined its link with nursing home entry.
The present review fills a gap in the literature by focusing on the link between frailty or sarcopenia and LTC, not only including nursing home admission but also formal and informal care provided in the community. Moreover, we evaluated the quality of evidence and summarized the current knowledge on frailty and sarcopenia as correlates of LTC use in populations aged 50 years and over. This article addresses the following question: are frailty and sarcopenia, conceptualized as dimensions of physical health, associated with the use of long-term care services by older individuals? Our systematic review is the first to focus on various LTC outcomes, and to provide an evaluation of both the results of the literature and the quality of the published evidence. The following aspects of LTC utilization are investigated: first, informal care, referring to the unpaid care provided by relatives; second, formal care, referring to the paid care provided by professionals, either nurses or personal care workers; finally, we also consider LTC provided in nursing homes.



The protocol for our systematic review was prepared and registered on the International Prospective Register of Systematic Reviews (PROSPERO; Ref CRD42020137212). The systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA), and with close adherence to the Cochrane Handbook for Systematic Reviews of Interventions. The resulting PRISMA checklist (see Table A.1, Appendix A.1, in Supplemental Materials), and the flowchart (see Figure 1) depict the stages involved in the selection process. The following subsections briefly discuss the methodology.

Figure 1
Flowchart of screening phases


Data Sources and Search Strategy

The search was conducted from April 2019 to June 2019 on the following databases:
PubMed-Medline, Embase, CINAHL (Academic Search Premier/EconLit), Web of Science. A preliminary search was conducted in order to develop a search strategy. The final search strategy is described in Appendix A.2. We employed some methodological components of the Effective Practice and Organisation of Care (EPOC) group’s search strategy, combined with selected MeSH terms and free text terms related to the PICOS (i.e. population, intervention, comparator, outcome and study design) elements. Studies published in English between January 2000 and December 2018 were collected. Also, the reference lists of the retrieved articles were examined to look for further relevant literature (through a ‘pearling procedure’).

Study Selection

Two researchers independently screened the titles and abstracts of papers published in English between 2000-2018. The selection of relevant papers was completed according to predefined eligibility criteria: i) the population of interest was composed of people aged 50 or more; ii) sarcopenia or physical frailty were explicitly measured by an identified tool (phenotype, index, scale, indicator) measuring a physical or biomedical component (thus excluding papers focusing only on psychological or social conceptions of frailty); iii) the comparators were expected to be populations with different levels of frailty or sarcopenia, including the comparison of a frail (respectively sarcopenic) population and a non-frail (respectively non sarcopenic) population; iv) the outcome should be a long-term care outcome (formal care, informal care, nursing home admission); v) we focused on quantitative research papers and excluded case reports, qualitative works or other studies without design (discussion, commentary, workshop report, systematic review).
A pilot phase was conducted to check that these criteria were precise enough. In the selection process, the information regarding the paper (authorship, institutions, journal titles and year of publication) were not blinded: even though inclusion decisions could be affected by these parameters, blind assessment has been shown to be costly and have a limited value in the study selection (13). Differences in screening results were solved by dialogue between the five members of the research team involved in the preparation of the manuscript.

Study design

We searched for cross-sectional, cohort, and pre-post studies, using a validated measure of physical frailty or sarcopenia. There were no restrictions in terms of study setting, or characteristics of sample population (except the age condition).

Typology of LTC outcomes

The main outcomes were categorized as utilization of any type of LTC services, at the intensive (i.e., probability of use) or extensive (i.e., volume of use, volume of costs) margins. We considered five different categories: 1/ Nursing home placement (NHP), regrouping any long or permanent stay in a nursing home or LTC institution, 2/ Nursing home short stay (NHSS), 3/ Formal personal care (FPC), regrouping home care services explicitly stated (e.g., ‘home care’, ‘nursing care’), 4/ Formal home help (FHH), consisting in any type of non-care help (e.g., help with meals of household duties), and 5/ Informal care (IC), provided by spouse, children or relatives.

Data Extraction, Analysis and Synthesis

Two independent reviewers (QR, JS) screened each study to identify the abstract, title, keywords, and concepts reflecting both the study’s contribution and research context. Databases were searched for a combination of three groups of keywords, referring to the variables of interest (frailty and sarcopenia), the outcomes of interest (long-term care, included formal care provided by nurses and care workers, informal care and nursing homes), and a third group complementing the request with a condition relating to utilization (use, utilization, visits, consumption). The complete search strategy is described in Appendix A.2.
Once the search strategy was completed, the full text of the relevant studies was retrieved and assessed by the two review authors with respect to the inclusion/exclusion criteria. When consensus was difficult to reach, a third review author (TR) assessed the study. The data were extracted by one reviewer (QR or JS) and checked by a second reviewer (JS or QR). The raw data from each manuscript included authors’ names, year of publication, title, journal, country and regions (if specified), study setting, study design, study population, participant demographics (age, share of women), details on the physical frailty measure used, type of LTC outcome use, raw and adjusted estimates, and risk of bias assessment details. The collected data were organized manually and tabulated using standardized forms. Given the heterogeneity of studies in terms of type of outcome and reporting (i.e., odds ratio, risk ratio, hazard ratio, marginal effects, etc.) we were unable to pool the results and conduct a meta-analysis. Therefore, we provide a comparative summary of findings for the main outcomes, using measures of the association in the same way as they were reported.

Risk of bias and quality assessment

The risk of bias was assessed using the Report from the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Task Force which provides a questionnaire to assess the relevance and credibility of observational studies (14). The credibility of each study was assessed by scoring each study across six quality domains: Design (8 items), Data (4 items), Analysis (3 items), Reporting (7 items), Interpretation (8 items), and Conflict of Interest (2 items). Two independent reviewers assessed each study (QR and JS). When a difference persisted between them, a third reviewer (TR) assessed the study and came to an agreement. Details regarding the grid and the coding strategy for each item are provided in Appendix A.3: Table A.3.1 assesses the credibility with respect to the study design domain, Table A.3.2 with respect to the data, Table A.3.3 with respect to the analysis, Table A.3.4 with respect to reporting, Table A.3.5 with respect to interpretation.
The Task Force Report (14) identified two specific criteria, where failure to address these criteria would indicate a “fatal flaw” and require particular caution in interpreting the results. One criterion concerned the Data: “was exposure outcome valid?” and the second concerned the Analysis: “was there a thorough assessment and control of confounding?”. Because the first criterion was already included as a study selection criterion, we only considered the second criteria as a potential fatal flaw. In case no multivariate regression analysis was conducted (adjusting for potential confounders), we classified the study as providing “not enough evidence” to conclude on the relationship being investigated.



Results of screening phases

The literature search yielded 2,534 titles in the preliminary phase. Figure 1 depicts the flow chart associated to the screening procedure. Removal of duplicates resulted in 1,933 potentially relevant papers. Then, at the screening stage, 1,920 studies were removed after applying the exclusion criteria. The pearling procedure allowed us to identify 4 more relevant papers. In the end, our systematic review comprised 17 studies on the relationship between frailty and /or sarcopenia and LTC use (16 studies using frailty, and 1 study using sarcopenia as main independent variable). Main outcomes, descriptions, and statistics (when available) are presented in Table 1. A complete summary of findings is provided in Appendix A.4, including information on the frailty/sarcopenia measurements, outcomes considered and statistical uncertainty around point estimates.

Table 1
Summary of study design and outcomes

a. Time frame only for longitudinal / cohort studies; *: Significant at the 5% level; b. OR calculated by authors based on raw coefficients (log OR); Abbreviations. HR: hazard ratio; OR: odds ratio; RR: risk ratio; ME: marginal effect; raw coefficient. NA: not available. ALM: appendicular lean mass; Note: details regarding measurements and statistical uncertainty around point estimates are provided in Appendix A.3.


As regard settings, 12 of the 17 reviewed studies took place in the general (community-dwelling) population, (15–26). Two studies considered residents of housing facilities or long-stay home care clients (27, 28) and two studies focused on the specific population of adults with intellectual disabilities (29) or cognitive impairments (30). One study focused on patients with diagnostic of Alzheimer disease (AD) (31). Interest in the association between physical frailty, sarcopenia, and LTC use seems to have arisen since 2008, with an increasing trend since. Most of the research was conducted in North America (n=4), Australia (n=2), and central Europe (n=11) (see Table A.5.1 and Figure A.5.1 for more details on the geographical distribution of studies).

Quality assessment

Overall, the mean relative score achievement (quality) was 84% (min= 67%, max= 100%) for the subdomain design, 97% (min= 67%, max= 100%) for the subdomain data, 28% (min= 0%, max= 67%) for the subdomain analyses, 84% (min= 57%, max= 100%) for the subdomain reporting, and 82% (min= 75%, max= 100%) for the subdomain interpretation. Appendix A.3 provides details on the scoring of each paper according to each domain. In the quality assessment, an important issue lies in the way potential confounders are dealt with: few studies used of an appropriate design (Design, Question 5, 2 studies coded 1 out of 17) and they were scarcely discussed (Interpretation, Question 4, 5 studies coded 1 out of 17). Overall, no study had all criteria 100% fulfilled. There were no large discrepancies in terms of quality according to the class of LTC outcome, though studies investigating FHH had lower scores on average, and studies focusing on LTC placements (short or permanent) tended to have higher quality (see Figures A.3.1 in Appendix A.3).
A “fatal flaw” (assessed by the absence of control or assessment of potential confounders) was diagnosed for four studies (19, 25, 26, 29). Two of the studies focused on formal home care (19, 29), among which one study additionally focused on home care (19), and two studies focused on informal care (25, 26).

Association between frailty, sarcopenia and LTC use

Frailty and LTC use

Overall, 10 studies reported a positive association between frailty and LTC use (15–18, 23, 24, 27–29, 31) (Table 2). Conversely, only two studies showed no significant association between frailty and LTC use (21, 22). In three studies, there was not enough evidence to conclude on the relationship between frailty and LTC use (e.g., because no multivariate regression was run) (19, 25, 26). Among the 10 studies reporting positive associations, three found mixed results (17, 24, 31), which varied according to the frailty measure used (see Table 1 for a summary of results according to each frailty measure used).

Table 2
Summary of results per class of outcome

Legend: NS nonsignificant; (+) Outcome increased significantly; NEE: not enough evidence (e.g., no adjusted multivariate regression conducted); (*) inconsistent association (the association varies according to the frailty measure considered); In each study, the LTC outcomes were groups into one of the five categories: 1/ Nursing home placement (NHP), 2/ Nursing home short stay (NHSS), 3/ Formal personal care (FPC), 4/ Formal home care (FHC), 5/ Informal care (IC). Several outcomes per study are possible.


Sarcopenia and LTC use

Only one study focused on the relationship between sarcopenia and LTC use [20]. Using a 3-classes characterization of sarcopenia defined as sarcopenia I (i.e., low appendicular lean mass / ALM alone), sarcopenia II (low ALM with weakness), or sarcopenia III (low ALM with weakness and poor gait speed), the authors found a positive association between sarcopenia (either I, II, or III) and NHP. Moreover, the relationship was stronger when sarcopenia severity was higher (the effect size was higher for sarcopenia III compared to sarcopenia I or sarcopenia II).

Heterogeneity of results according to the LTC outcome

The number of studies having focused on each type of LTC outcome was distributed as follows (one study may have analysed multiple LTC outcomes): six studies on nursing home placement (NHP) (15, 20, 22, 27, 28, 31), four studies on nursing home short stay (NHSS) (16, 17, 21, 23), seven studies on formal personal care (FPC) (16–19, 24, 25, 29), four studies on formal home help (FHH) (16, 19, 25, 26), and five studies on informal care (IC) (17, 24–26, 30). Among the six studies investigating the association between frailty and NHP, 5 studies found a significant positive association, including one with mixed results (i.e., depending on frailty measurement) (see Table 1). Regarding NHSS, three studies confirmed a positive significant association, including one with mixed results. Regarding FPC, five studies confirmed a positive significant association, including two with mixed results. The evidence was less compelling for the two other LTC outcomes. Regarding FHH, only one study found a positive significant association; it was not possible to conclude for the other three studies because no multivariate regression models were estimated. Last, concerning IC, three studies found a positive significant association with frailty, including one with mixed results.



The purpose of this systematic review was to assess the association between physical frailty or sarcopenia and LTC use. Our paper is the first to systematically review the results of these studies and assess their strength and weaknesses using a validated quality assessment grid (14). We found that most of the reviewed studies reported a positive association of physical frailty and LTC outcomes. Moreover, there is evidence showing an association between frailty and LTC use for each class of LTC outcome. Specifically, there seems to be already reasonable evidence to conclude on a positive relationship of physical frailty with nursing home placement / short stay as well as on FPC. However, there is less (more heterogeneous) evidence regarding the association between physical frailty and FHH or IC use. Besides, only one study investigated the association between sarcopenia and LTC use, and found a positive association for NHP. Further studies should thus focus on the particular LTC use outcomes with insufficient evidence, and on sarcopenia.
We found important heterogeneity across studies in terms of i) type of frailty indicator used, ii) statistical analysis, and iii) reporting of results. This heterogeneity made it difficult to compare or transpose the results from one study to another. Overall, 12 different physical frailty measures and three sarcopenia indicators were used (altogether in one study). Almost all studies used multidimensional frailty indexes (most of which based on self-reported data), with a majority using either the original 5 items Fried frailty index (with categorisation into robust, pre-frail, frail) (16, 18, 21–24, 30), the Tilburg frailty indicator (17, 22, 26) or composite frailty indexes counting the number (or percentage) of deficits accumulated (27, 28, 31). In one study, sarcopenia was defined using clinical assessment of appendicular lean mass, muscle strength, and gait speed (20).
Most studies used multivariate regression analyses controlling for several of the following covariates: age, gender, morbidity, disability, and some basic sociodemographic characteristics such as education, living arrangements, or income. However, none of the studies included exactly the same number and types of controls, making the comparing of results difficult. Moreover, in only few instances the rationale for including specific controls was provided. One study performed sensitivity analysis with respect to the model specification. We conclude that future studies should carefully explain and justify their model, and perform sensitivity analyses regarding model specification (inclusion or exclusion of covariates). Despite the fact that a majority of studies relied on cohort data, very few used statistical methods designed to reduce unobserved confounding using longitudinal or panel data models (e.g. fixed effects or correlated random effects models). Specifically, only two studies analysed LTC outcome in relation with frailty transitions (19, 24), among which only one used fixed effect model to control for time invariant confounders (24). Future studies investigating the association between frailty and LTC use should consider using more advanced designs and more robust statistical methods such as the one suggested.
Finally, the studies were highly heterogeneous in terms of reporting of results. Studies using binary outcomes reported the results either in terms of raw estimates, odds ratio, or marginal effects. Studies using continuous outcomes (e.g., costs or time between events) reported either hazard ratio or risk ratio. We thus recommend that future studies report results that can be compared across settings, such as reporting average marginal effects, which can be computed using linear or non-linear (e.g., logit, probit) models.
In conclusion, our article is the first systematic literature review to focus on the relationship between frailty and sarcopenia, operationalized using existing validated measures, on various LTC outcomes among people worldwide. Overall, there is evidence suggesting that frailty is a key correlate of LTC use. But there is still a lack of evidence on the relationship between sarcopenia and LTC use, due an insufficient number of studies that investigated that issue. However, the strength of evidence is heterogeneous and varies according to the type of LTC outcome considered. Overall, the quality of studies was judged weak to moderate. Very few studies were considered as providing strong evidence, in particular because of weaknesses in statistical analyses performed, although many studies relied on cohort data. Future studies should address potential confounding using longitudinal models allowing integrating the effect of within-individual heterogeneity and assessing the effect of transitions into frailty (or sarcopenia) on LTC use. Moreover, the studies should perform sensitivity analyses by varying the control variables or using stratified analyses. In particular, it is of importance to know if the effect of frailty varies according to predefined socio-economic categories (32). This information would be useful for the definition of appropriate LTC policies aiming at ensuring equity of access.
With regards to the implications, our systematic literature review supports three main orientations for aging policies. First, eligibility criteria for public allowances should include frailty measures because evidence shows that regardless of the measure used, frailty is associated with increased LTC use. Second, the challenges associated with frailty among older people are global, since correlations between frailty measures and LTC use are similar across countries. This calls for global aging initiatives, to learn from each country’s experience. Third, frailty prevention in older populations is a policy priority. While many countries (France, Germany, Netherlands etc.) implemented disability reforms, future policies should also target frail populations, and enhance innovations to detect frailty in the 65-75 age group.


Funding and acknowledgements: The research leading to these results has received support from the Innovative Medicines Initiative Joint Undertaking under grant agreement n° 115621, resources of which are composed of financial contribution from the European Union’ Seventh Framework Programme (FP7 / 2007-2013) and European Federation of Pharmaceutical Industries and Associations (EFPIA) companies’ in kind contribution.
Conflicts of Interest: None of the authors report any conflict of interest whatsoever.



1. Costa-Font, J., Courbage, C., Swartz, K.: Financing Long-Term Care: ex-ante, ex-post or both? Health Econ. 2015;24, 45–57. https://doi.org/10.1002/hec.3152
2. OCDE: Health at a Glance 2019 : OECD indicators.
3. European Commission, Directorate-General for Economic and Financial Affairs, Economic Policy Committee of the European Communities: The 2018 ageing report: underlying assumptions and projection methodologies. 2017
4. Marcell, T.J.: Sarcopenia: causes, consequences, and preventions. J. Gerontol. A. Biol. Sci. Med. Sci. 2003;58, M911-916
5. Cesari, M., Landi, F., Vellas, B., Bernabei, R., Marzetti, E.: Sarcopenia and Physical Frailty: Two Sides of the Same Coin. Front. Aging Neurosci. 2014;6. https://doi.org/10.3389/fnagi.2014.00192
6. Malmstrom, T.K., Morley, J.E.: SARC-F: a simple questionnaire to rapidly diagnose sarcopenia. J. Am. Med. Dir. Assoc. 2013;14, 531–532. https://doi.org/10.1016/j.jamda.2013.05.018
7. Studenski, S.A., Peters, K.W., Alley, D.E., Cawthon, P.M., McLean, R.R., Harris, T.B., Ferrucci, L., Guralnik, J.M., Fragala, M.S., Kenny, A.M., Kiel, D.P., Kritchevsky, S.B., Shardell, M.D., Dam, T.-T.L., Vassileva, M.T.: The FNIH sarcopenia project: rationale, study description, conference recommendations, and final estimates. J. Gerontol. A. Biol. Sci. Med. Sci. 2014;69, 547–558. https://doi.org/10.1093/gerona/glu010
8. Cawthon, P.M., Peters, K.W., Shardell, M.D., McLean, R.R., Dam, T.-T.L., Kenny, A.M., Fragala, M.S., Harris, T.B., Kiel, D.P., Guralnik, J.M., Ferrucci, L., Kritchevsky, S.B., Vassileva, M.T., Studenski, S.A., Alley, D.E.: Cutpoints for low appendicular lean mass that identify older adults with clinically significant weakness. J. Gerontol. A. Biol. Sci. Med. Sci. 2014;69, 567–575. https://doi.org/10.1093/gerona/glu023
9. Kojima, G.: Frailty as a predictor of hospitalisation among community-dwelling older people: a systematic review and meta-analysis. J. Epidemiol. Community Health. 2016;70, 722–729. https://doi.org/10.1136/jech-2015-206978
10. Wang, S.-Y., Shamliyan, T.A., Talley, K.M.C., Ramakrishnan, R., Kane, R.L.: Not just specific diseases: systematic review of the association of geriatric syndromes with hospitalization or nursing home admission. Arch. Gerontol. Geriatr. 2013;57, 16–26. https://doi.org/10.1016/j.archger.2013.03.007
11. Kojima, G.: Frailty as a Predictor of Nursing Home Placement Among Community-Dwelling Older Adults: A Systematic Review and Meta-analysis. J. Geriatr. Phys. Ther. 2018;2001. 41, 42–48. https://doi.org/10.1519/JPT.0000000000000097
12. Cawthon, P.M., Fox, K.M., Gandra, S.R., Delmonico, M.J., Chiou, C.-F., Anthony, M.S., Sewall, A., Goodpaster, B., Satterfield, S., Cummings, S.R., Harris, T.B.: Do Muscle Mass, Muscle Density, Strength, and Physical Function Similarly Influence Risk of Hospitalization in Older Adults? J. Am. Geriatr. Soc. 2009;57, 1411–1419. https://doi.org/10.1111/j.1532-5415.2009.02366.x
13. Centre for Reviews and Dissemination: Systematic Reviews. CRD’s guidance for undertaking reviews in health care. the Centre for Reviews and Dissemination, York 2009
14. Berger, M.L., Martin, B.C., Husereau, D., Worley, K., Allen, J.D., Yang, W., Quon, N.C., Mullins, C.D., Kahler, K.H., Crown, W.: A Questionnaire to Assess the Relevance and Credibility of Observational Studies to Inform Health Care Decision Making: An ISPOR-AMCP-NPC Good Practice Task Force Report. Value Health. 2014;17, 143–156. https://doi.org/10.1016/j.jval.2013.12.011
15. Rothman, M.D., Leo-Summers, L., Gill, T.M.: Prognostic Significance of Potential Frailty Criteria. J. Am. Geriatr. Soc. 2008;56, 2211–2216. https://doi.org/10.1111/j.1532-5415.2008.02008.x
16. Rochat, S., Cumming, R.G., Blyth, F., Creasey, H., Handelsman, D., Le Couteur, D.G., Naganathan, V., Sambrook, P.N., Seibel, M.J., Waite, L.: Frailty and use of health and community services by community-dwelling older men: the Concord Health and Ageing in Men Project. AGE AGEING. 2010;39, 228–233. https://doi.org/10.1093/ageing/afp257
17. Gobbens, R.J.J., van Assen, M.A.L.M., Luijkx, K.G., Schols, J.M.G.A.: The Predictive Validity of the Tilburg Frailty Indicator: Disability, Health Care Utilization, and Quality of Life in a Population at Risk. The Gerontologist. 2012;52, 619–631. https://doi.org/10.1093/geront/gnr135
18. Hoeck, S., Francois, G., Geerts, J., Van der Heyden, J., Vandewoude, M., Van Hal, G.: Health-care and home-care utilization among frail elderly persons in Belgium. Eur. J. Public Health. 2012;22, 671–677. https://doi.org/10.1093/eurpub/ckr133
19. Bentur N., Sternberg S.A., Shuldiner J.: Frailty transitions in community-dwelling older people. Isr. Med. Assoc. J. 2016;18, 449–453
20. Hirani, V., Blyth, F., Naganathan, V., Le Couteur, D.G., Seibel, M.J., Waite, L.M., Handelsman, D.J., Cumming, R.G.: Sarcopenia Is Associated With Incident Disability, Institutionalization, and Mortality in Community-Dwelling Older Men: The Concord Health and Ageing in Men Project. J. Am. Med. Dir. Assoc. 2015;16, 607–613. https://doi.org/10.1016/j.jamda.2015.02.006
21. Bock, J.-O., König, H.-H., Brenner, H., Haefeli, W.E., Quinzler, R., Matschinger, H., Saum, K.-U., Schöttker, B., Heider, D.: Associations of frailty with health care costs – results of the ESTHER cohort study. BMC Health Serv. Res. 2016;16, 128. https://doi.org/10.1186/s12913-016-1360-3
22. Gonzalez-Colaço Harmand, M., Meillon, C., Bergua, V., Tabue Teguo, M., Dartigues, J.-F., Avila-Funes, J.A., Amieva, H.: Comparing the predictive value of three definitions of frailty: Results from the Three-City study. Arch. Gerontol. Geriatr. 2017;72, 153–163. https://doi.org/10.1016/j.archger.2017.06.005
23. Ferrante, L.E., Pisani, M.A., Murphy, T.E., Gahbauer, E.A., Leo-Summers, L.S., Gill, T.M.: The Association of Frailty With Post-ICU Disability, Nursing Home Admission, and Mortality A Longitudinal Study. Chest. 2018;153, 1378–1386. https://doi.org/10.1016/j.chest.2018.03.007
24. Hajek, A., Bock, J.-O., Saum, K.-U., Matschinger, H., Brenner, H., Holleczek, B., Haefeli, W.E., Heider, D., Koenig, H.-H.: Frailty and healthcare costs-longitudinal results of a prospective cohort study. Age Ageing. 2018;47, 233–241. https://doi.org/10.1093/ageing/afx157
25. Lambotte, D., De Donder, L., Van Regenmortel, S., Fret, B., Dury, S., Smetcoren, A.-S., Dierckx, E., De Witte, N., Verte, D., Kardol, M.J.M., Consortium, D.-S.: Frailty differences in older adults’ use of informal and formal care. Arch. Gerontol. Geriatr. 2018;79, 69–77. https://doi.org/10.1016/j.archger.2018.05.018
26. Verver, D., Merten, H., Robben, P., Wagner, C.: Care and support for older adults in The Netherlands living independently. Health Soc. Care Community. 2018;26, e404–e414. https://doi.org/10.1111/hsc.12539
27. Hogan, D.B., Freiheit, E.A., Strain, L.A., Patten, S.B., Schmaltz, H.N., Rolfson, D., Maxwell, C.J.: Comparing frailty measures in their ability to predict adverse outcome among older residents of assisted living. BMC Geriatr. 2012;12. https://doi.org/10.1186/1471-2318-12-56
28. Campitelli, M.A., Bronskill, S.E., Hogan, D.B., Diong, C., Amuah, J.E., Gill, S., Seitz, D., Thavorn, K., Wodchis, W.P., Maxwell, C.J.: The prevalence and health consequences of frailty in a population-based older home care cohort: a comparison of different measures. BMC Geriatr. 2016;16. https://doi.org/10.1186/s12877-016-0309-z
29. Schoufour, J.D., Evenhuis, H.M., Echteld, M.A.: The impact of frailty on care intensity in older people with intellectual disabilities. Res. Dev. Disabil. 2014;35, 3455–3461. https://doi.org/10.1016/j.ridd.2014.08.006
30. Butler, A., Gallagher, D., Gillespie, P., Crosby, L., Ryan, D., Lacey, L., Coen, R., O’Shea, E., Lawlor, B.: Frailty: a costly phenomenon in caring for elders with cognitive impairment. Int. J. Geriatr. Psychiatry. 2016;31, 161–168. https://doi.org/10.1002/gps.4306
31. Kelaiditi, E., Andrieu, S., Cantet, C., Vellas, B., Cesari, M.: Frailty Index and Incident Mortality, Hospitalization, and Institutionalization in Alzheimer’s Disease: Data From the ICTUS Study. J. Gerontol. Ser. A. 2016;71, 543–548. https://doi.org/10.1093/gerona/glv137
32. Sicsic, J., Ravesteijn, B., Rapp, T.: Are frail elderly people in Europe high-need subjects? First evidence from the SPRINTT data. Health Policy. 2020;124, 865–872. https://doi.org/10.1016/j.healthpol.2020.05.009
33. Berger, M.L., Martin, B.C., Husereau, D., Worley, K., Allen, J.D., Yang, W., Quon, N.C., Mullins, C.D., Kahler, K.H., Crown, W.: A Questionnaire to Assess the Relevance and Credibility of Observational Studies to Inform Health Care Decision Making: An ISPOR-AMCP-NPC Good Practice Task Force Report. Value Health. 2014;17, 143–156. https://doi.org/10.1016/j.jval.2013.12.011