jfa journal

AND option

OR option



J. Li, P. Liu, Y. Zhang, G. Wang, Y. Zhou, Y. Xing, L. Zhang, Y. Li, L. Ma


Department of Geriatrics, Xuanwu Hospital, Capital Medical University, Capital Medical University National Clinical Research Center for Geriatrics Disorders, Beijing 100053, China. Jiatong Li and Pan Liu are co-first authors.

Corresponding Author: Lina Ma, M.D, Ph.D, Department of Geriatrics, Xuanwu Hospital, Capital Medical University National Clinical Research Center for Geriatrics Disorders, Beijing 100053, China, E-mail: malina0883@126.com

J Frailty Aging 2024;13(2)125-130
Published online March 16, 2024, http://dx.doi.org/10.14283/jfa.2024.24



BACKGROUND: Physical resilience is an emerging concept that describes an individual’s capacity to recover from stressors. However, few instruments are currently available for assessing physical resilience.
OBJECTIVE: To develop a scale to assess physical resilience in older adults.
DESIGN: Development of a clinical scale.
SETTING AND PARTICIPANTS: A total of 172 hospitalized older adults were recruited.
MEASUREMENTS: This study comprised two stages. First, a pool of physical resilience scale items was created through a literature review, and the Delphi method was used to establish an initial scale. Second, the initial physical resilience scale was tested on hospitalized older adults.
RESULTS: Five primary and 19 secondary items were identified after reviewing the literature. After two rounds of expert consultations, three primary and 16 secondary items were determined. The overall Cronbach’s alpha for the scale was 0.760. Except for items N2, N4, N5, N8, and N14, Pearson’s correlation between the scores of the remaining items and the total score ranged from 0.407 to 0.672. Except for items N2, N4, and N5, the corrected item-total correlation results ranged from 0.301 to 0.580, indicating good consistency between each item and the overall scale. Factor analysis showed that except for N7, the factor loadings of the remaining items were between 0.584 and 0.844. After expert discussions, items N2, N4, N7, and N14 were included in the scale, and items N5 and N8 were removed.
CONCLUSION: A 14-item physical resilience scale, CHEES, was developed to assess physical resilience levels in older adults.

Key words: Physical resilience, older adults, assessment scale.



Globally, the proportion of the population aged 60 years and older will continue to increase, accounting for 22% of the total population by 2050 (1). Physical resilience, which plays a crucial role in promoting healthy aging (2), refers to an individual’s capability to recover from challenges related to physical health, such as illnesses, injuries, or age-related decline (3). Older adults with higher physical resilience may have better clinical outcomes. In contrast, those with lower physical resilience are more likely to experience poor physical functions and may even face death when exposed to stressors (4).
There are several approaches to assessing physical resilience, including using recovery phenotypes and expected recovery differentials to define the degree of resilience (5), comparing the effects of dynamic stimulation tests before and after evaluation of individual resilience (6), and monitoring changes in indicators of physical resilience, such as activities of daily living, Short Form Health Survey (SF-36), Short Physical Performance Battery (SPPB), and deficit accumulation index (7-10). However, these methods usually require long-term dynamic observations. Some scholars have designed resilience questionnaires, for example, Resilience Scale (RS), Resilience Scale for Older Adults (RSOA), and Physical Resilience Instrument for Older Adults (PRIFOR) (11-14). The clinical application of the physical resilience scale is still in its early stages. Some scales are unable to reflect all characteristics of physical resilience. In addition, reliable physical resilience scales specifically designed for Chinese older adults are lacking.
Therefore, we aimed to develop an easy-to-use physical resilience scale, named the Clinical pHysical rEsilience assEssment Scale (CHEES), using the Delphi method to assess physical resilience in older adults.



Delphi method

The Delphi method is suitable for healthcare research as it allows for the systematic and various collection of opinions from panel experts (15). Investigators solicit several rounds of opinions from selected experts anonymously. After each round, an organized opinion material is sent back to each expert. Experts provide new argumentative opinions. This process is repeated until a consensus is reached, resulting in a more consistent and reliable conclusion (16).
19 experts from 13 hospitals in seven provinces in China were selected as consultants in this study. The Delphi method was employed in two rounds of consultations. All experts had a minimum of 5 years of work experience and had expertise in geriatrics, general medicine, nursing, internal medicine, epidemiology, and other fields. A consultation questionnaire containing the purpose of the consultation, explanations of the meaning of each item, experts’ scoring of the importance and feasibility of the items, and self-assessment of experts’ familiarity with the items was emailed to each expert (17, 18).


A total of 172 hospitalized older adults were enrolled in the study. This study was approved by the Ethics Committee of Xuanwu Hospital, Capital Medical University (number: LYS-2022154). All participants provided written informed consent before participation.


The primary CHEES includes 19 items scored on a five-point Likert scale (1 = strongly disagree, 2 = disagree, 3 = agree, 4 = very agree, and 5 = strongly agree). The total score ranged from 19 to 95, with higher scores indicating better physical resilience.

Statistical analysis

Data were entered into EpiData version 3.1 (EpiData Association, 2008) and analyzed using SPSS version 22 (IBM SPSS Statistics, 2013). A positive coefficient refers to the response rate of the expert consultations. In general, a positive coefficient of at least > 50% is considered acceptable. The familiarity coefficient (Cs) among experts was divided into five levels: unfamiliar, moderately familiar, familiar, quite familiar, and very familiar, with corresponding scores of 0.2, 0.4, 0.6, 0.8, and 1.0. The judgment coefficient (Ca) was assigned by the experts using different values based on their judgment of the impact of each item: major, moderate, and minor for theoretical analysis (0.50, 0.40, and 0.30), practical experience (0.30, 0.20, and 0.10), peer understanding (0.10, 0.08, and 0.05), and intuitive feeling (0.10, 0.07, and 0.05) (19). The authority coefficient (Cr) = (Ca + Cs)/2, with a range of 0–1; Cr ≥ 70% is considered reliable. Kendall’s W represents the degree of agreement and consistency among the experts for each item and ranges from 0 to 1, with higher values indicating better agreement and consistency among the experts. Statistical significance was set at P < 0.05.
The Pearson’s correlation coefficient between each item score and the total score was calculated, with a value < 0.4 indicating a low correlation. Cronbach’s α was used to calculate the corrected item-total correlation (CITC), with a value of < 0.3 indicating a low correlation. If deleting an item increased the Cronbach’s α, that item was removed. Factor analysis was conducted to calculate the common factor variance; generally, deletion was considered when the factor loading was < 0.45 and the communality value was < 0.2. Among the above three analysis methods, those items failed to pass two of them were deleted. For items that failed to pass one, further discussions are required with experts (20).



Literature search and item pool construction

A literature search of English and Chinese databases was conducted to identify questionnaires assessing physical resilience. The search strategy combining keywords such as “physical resilience”, “questionnaires”, “aged”, or “older adults” was employed. After the exclusion of duplicate and irrelevant studies, 13 articles were included. An item pool consisting of five primary and 19 secondary items was established.

Expert reliability evaluation and coordination coefficient

The response rates for the experts’ questionnaires in the first and second rounds were 100% and 86.6%, respectively, suggesting the experts’ active engagement and strong interest. The Cr was 0.88 in the first round and 0.85 in the second round, indicating high authority and reliability.
In the first round, Kendall’s W for the importance of the primary items was 0.08 and for feasibility was 0.04. Kendall’s W for the importance of secondary indicators was 0.10 and for feasibility was 0.16. In the second round, Kendall’s W for the importance of primary items was 0.28 and for feasibility was 0.30. Kendall’s W for the importance of secondary items was 0.18 and for feasibility was 0.16.

Items modification after two rounds of consultations

Based on predefined criteria for the importance and feasibility of item selection, no items were deleted (Table 1). Following the expert opinions, we removed “positive thinking”, “life goal”, and “social environment”, and incorporated “intrinsic ability” in the primary items. The classification and order of the secondary items under the primary items were adjusted. Secondary items Q4, Q9, Q11, and Q18 were deleted, and two additional secondary items were added: “I am forgetting where things were placed.” and “My hearing and vision problems hinder my daily life.” Furthermore, items Q12 and Q13 were merged into “There is someone there to encourage me and help me out when I am sick or encounter difficulties.” Then, the 16-item questionnaire was used in the second round of expert consultations. According to the predefined criteria for the importance and feasibility of item selection, there was no item deleted (Table 2). Thus, an initial scale consisting of three primary items and 16 secondary items was established.

Table 1. Item importance and feasibility scores and coefficients of variation for the first round

CV: Coefficient of variation.

Table 2. Item importance and feasibility scores and coefficients of variation for the second round

CV: Coefficient of variation.


Item selection analysis results

Pearson’s correlation coefficients were calculated for each item’s score and the total score. Items N2, N4, N5, N8, and N14 had low correlations with the total score (all r < 0.4). The remaining items showed a high correlation with the total score (r = 0.407–0.672) (Table 3).

Table 3. Correlation analysis and Cronbach’s α in the scale

CITC: Corrected Item-Total Correlation; CAID: Cronbach’s Alpha if Item Deleted.


The overall Cronbach’s α of the scale was 0.760. Items N2, N4, and N5 had a CITC of 0.289, 0.037, and 0.056, respectively. Deleting N5 increased the overall Cronbach’s α, indicating that N5 lowered the internal consistency of the scale (Table 3).
The Kaiser–Meyer–Olkin measure (KMO) test and Bartlett’s test of sphericity were also performed, yielding a KMO value of 0.718 and a significant Bartlett’s test result (χ2 = 686.077, P < 0.001), indicating suitability for factor analysis. Factor analysis was used to extract the communalities of each item. The communalities of all the items were greater than 0.20, indicating that no item needed to be removed (Table 4). When extracting the fifth common factor, the explained cumulative percentage of the scale’s variables was 59.644%. Furthermore, based on the analysis in conjunction with the scree plot (Figure 1), five common factors were extracted to calculate the factor loadings of each item. The factor loading for N7 was 0.345, which is less than 0.45 and does not meet the screening criteria, indicating that it should be deleted or modified (Table 4).

Table 4. Analysis results of common factor variance method and factor loading in the scale

Figure 1. Scree plot in factor analysis


Item N5 was deleted due to failing to pass two kinds of test methods. Items N2, N4, N7, N8, and N14 did not pass one kind of test method, requiring further discussion by experts to decide whether these items should be retained. After detailed discussions, N8 was suggested to be removed, while the other items remained since they are more essential for resilience assessment. Therefore, the final CHEES consisted of 14 secondary items.
Furthermore, we explored the relationship between the CHEES and physical functions in older adults. The results showed that the CHEES scores were positively correlated with SPPB scores (r=0.264, p<0.001).



A 14-item physical resilience scale (CHEES) was established by using the Delphi method in this study. The initial pool of items was determined through a literature review, followed by two rounds of expert consultations. Physical resilience, the ability to withstand external stressors, may be influenced by age, physiological reserves, psychological state, nutritional status, cognition, and genetics (21-23). Intrinsic capacity emphasizes the positive attributes of physical status in older adults and may be a pivotal determinant of physical resilience (24). Research using the Psychological Resilience Scale showed the importance of psychosocial factors in determining resilience after exposure to stressors (25). The primary level of CHEES encompasses the definition of resilience and potential determinants, which were categorized into “positive thinking”, “adapt to change”, “life goal”, “external support”, and “social environment”. After the first round of expert consultations, the primary items were reduced to the following three parts: “intrinsic capacity”, “adapt to change”, and “external support”.
In the secondary items, some items were drawn from the PRIFOR scale (14), SF-36 scale (10), the RS (11), RSOA (12), Physical Resilience Scale (PRS) (26), and Connor-Davidson Resilience scale (CD-RISC) (27). Although inconsistencies among experts existed in the first round (Kendall’s W was 0.10), a consensus by experts was reached in the second round (Kendall’s W was 0.18 and feasibility coordination coefficient was 0.16). A preliminary physical resilience scale consisting of three primary and 16 secondary items was formulated.
Correlation coefficients, Cronbach’s α, and factor analysis were used to ensure the selection of items with robust psychometric properties. Item N5 was deleted due to not meeting the above two standards. While items N2, N4, N7, N8, and N14 need further discussions. Item N2 was related to the current disease conditions and nutritional status, and good appetite can help maintain physical health and prevent malnutrition. Moreover, cognition and mood are important foundations for participating in social life and accessing external information. Items N4 and N7 were associated with the mental health of cognitive functions and depressive functions, respectively. Income, as an essential material and economic foundation for people, varies in investment in the field of health due to different income levels. Evidence showed that the lower a person’s income level, the less investment in health, and the more their health will develop in a negative direction (28). N14 was an important component of social support. In contrast, item N8 was included in item N9. After the full discussions among the clinical expert panel, items N2, N4, N7, and N14 remained, while item N8 was removed. Thus, the final CHEES consisted of three primary items and 14 secondary items.
The physical resilience model promoted by Whitson et al was a theoretical framework including physical, psychological, and social aspects based on a systematic review. However, the CHEES was developed based on a literature review, two rounds of expert consultations, and tested on hospitalized older adults to establish a formal clinical scale. Commonly, the above two scales both pay much attention to the psychological and social support factors. More research is needed to test the CHEES’ reliability and validity and explore its clinical application soon.
There are several limitations in the study. Firstly, as the developed CHEES is a self-reported assessment tool, the objective physical assessments were restricted. But we also found the CHEES scores were positively correlated with physical functions assessed by SPPB in this study. Secondly, further studies are needed to validate the CHEES’s reliability and validity and compared it with those established comprehensive assessment scales for older adults, which can provide insights into the scale’s performance and effectiveness across different population groups. Thirdly, all the participants were from a single center, and future research on older adults from different settings and multi-centers should be conducted to ensure applicability and generalizability.
In conclusion, this study developed a simple scale, the 14-item CHEES, to evaluate the level of physical resilience in older adults in response to health-related stressors in clinical settings, aimed to prevent function decline. Future validation studies are required to be conducted with larger sample sizes and different settings.


Funding: This work was supported by the National Key R&D Program of China (2020YFC2008606)..

Acknowledgement: We acknowledge all the people who participated in the cohort study.

Conflicts of Interest: All the authors declare no conflicts of interest in the present study.

Author’s contributions: Lina Ma conceived the study design and supervised the study, Jiatong Li, Pan Liu, Yun Li, Yaxin Zhang, Guanzhen Wang, Yaru Zhou, Yiwen Xing, and Li Zhang performed the data collection and analysis, interpreted the results, and drafted the initial manuscript. All authors critically revised the manuscript and approved the final manuscript.

Ethical standards: This study was carried out in accordance with the ethical standard.





1. Ageing and health. World Health Organization. 2022. https://www.who.int/news-room/fact-sheets/detail/ageing-and-health.
2. Friedman SM, Mulhausen P, Cleveland ML, et al. Healthy Aging: American Geriatrics Society White Paper Executive Summary. J Am Geriatr Soc 2019;67(1):17-20. doi: 10.1111/jgs.15644.
3. Whitson HE, Duan-Porter W, Schmader KE, et al. Physical Resilience in Older Adults: Systematic Review and Development of an Emerging Construct. J Gerontol A Biol Sci Med Sci 2016;71(4):489-95. doi: 10.1093/gerona/glv202.
4. Hadley EC, Kuchel GA, Newman AB, Workshop S, Participants. Report: NIA Workshop on Measures of Physiologic Resiliencies in Human Aging. J Gerontol A Biol Sci Med Sci 2017;72(7):980-90. doi: 10.1093/gerona/glx015.
5. Colón-Emeric C, Pieper CF, Schmader KE, et al. Two approaches to classifying and quantifying physical resilience in longitudinal data. J Gerontol A Biol Sci Med Sci 2020;75(4):731-38. doi: 10.1093/gerona/glz097.
6. Varadhan R, Walston JD, Bandeen-Roche K. Can physical resilience and frailty in older adults be linked by the study of dynamical systems? J Am Geriatr Soc 2018;66(8):1455. doi: 10.1111/jgs.15409.
7. Pedone C, Costanzo L, Finamore P, et al. Defining resilience in older people: does a subjective definition of stressor work? J Gerontol A Biol Sci Med Sci 2021;76(8):1480-85. doi: 10.1093/gerona/glaa189.
8. Sotos-Prieto M, Ortola R, Lopez-Garcia E, Rodriguez-Artalejo F, Garcia-Esquinas E. Adherence to the Mediterranean Diet and Physical Resilience in Older Adults: The Seniors-ENRICA Cohort. J Gerontol A Biol Sci Med Sci 2021;76(3):505-12. doi: 10.1093/gerona/glaa277.
9. Kolk D, Melis RJ, MacNeil-Vroomen JL, et al. Physical Resilience in Daily Functioning Among Acutely Ill Hospitalized Older Adults: The Hospital-ADL Study. J Am Med Dir Assoc 2022;23(5):903. e1-e12. doi: 10.1016/j.jamda.2021.08.029.
10. Laskow T, Zhu J, Buta B, et al. Risk Factors for Nonresilient Outcomes in Older Adults After Total Knee Replacement. J Gerontol A Biol Sci Med Sci 2022;77(9):1915-22. doi: 10.1093/gerona/glab257.
11. Wagnild G.. A review of the Resilience Scale.. J Nurs Meas 2009;17(2):105-13. doi: 10.1891/1061-3749.17.2.105..
12. Li YT, Ow YSY. Development of resilience scale for older adults. Aging Ment Health 2022;26(1):159-68. doi: 10.1080/13607863.2020.1861212
13. FW H, CH L, PH L, CY L. Predictive Validity of the Physical Resilience Instrument for Older Adults (PRIFOR). J Nutr Health Aging 2021;25(9):1042–45. doi: doi:10.1016/j.jgo.2019.04.008.
14. Hu FW, Lin CH, Yueh FR, Lo YT, Lin CY. Development and psychometric evaluation of the Physical Resilience Instrument for Older Adults (PRIFOR). BMC Geriatr 2022;22(1):229. doi: 10.1186/s12877-022-02918-7.
15. de Meyrick J. The Delphi method and health research. Health education 2003;103(1):7-16.
16. Kors JA, van Bemmel JH. The Delphi method: a review of its applications in medicine. Medinfo 1989;89:156-60.
17. Nasa P, Jain R, Juneja D. Delphi methodology in healthcare research: how to decide its appropriateness. World Journal of Methodology 2021;11(4):116.
18. Deng X, Ye M, Li W, et al. Development of a humanistic care digital storytelling programme for intensive care unit nursing students: Feasibility and satisfaction analysis. Nurse Education Today 2023:105998.
19. Ruan Y, Song S-j, Yin Z-f, et al. Comprehensive evaluation of military training-induced fatigue among soldiers in China: A Delphi consensus study. Frontiers in Public Health 2022;10:1004910. doi: 10.3389/fpubh.2022.1004910.
20. Dimitrov DM. Statistical methods for validation of assessment scale data in counseling and related fields: John Wiley & Sons 2014.
21. Whitson HE, Cohen HJ, Schmader KE, et al. Physical Resilience: Not Simply the Opposite of Frailty. J Am Geriatr Soc 2018;66(8):1459-61. doi: 10.1111/jgs.15233.
22. Gijzel SMW, Rector J, van Meulen FB, et al. Measurement of Dynamical Resilience Indicators Improves the Prediction of Recovery Following Hospitalization in Older Adults. J Am Med Dir Assoc 2020;21(4):525-30 e4. doi: 10.1016/j.jamda.2019.10.011.
23. Li J, Chhetri JK, Ma L. Physical resilience in older adults: Potential use in promoting healthy aging. Ageing Res Rev 2022;81:101701. doi: 10.1016/j.arr.2022.101701.
24. Chhetri JK, Xue QL, Ma L, Chan P, R. V. Intrinsic Capacity as a Determinant of Physical Resilience in Older Adults. J Nutr Health Aging 2021;25(8):1006-11. doi: 10.1007/s12603-021-1629-z.
25. Gooding PA, Hurst A, Johnson J, Tarrier N. Psychological resilience in young and older adults. Int J Geriatr Psychiatry 2012;27(3):262-70. doi: 10.1002/gps.2712.
26. Resnick B, Galik E, Dorsey S, et al. Reliability and validity testing of the physical resilience measure. Gerontologist 2011;51(5):643-52. doi: 10.1093/geront/gnr016
27. Connor KM, Davidson JR. Development of a new resilience scale: the Connor-Davidson Resilience Scale (CD-RISC). Depress Anxiety 2003;18(2):76-82. doi: 10.1002/da.10113. 28. McFarland MJ, Hill TD, Montez JK. Income Inequality and Population Health: Examining the Role of Social Policy. J Health Soc Behav 2023;64(1):2-20. doi: 10.1177/00221465221109202.

© Serdi 2024