Comparison of four frailty scales in predicting adverse outcomes in Chinese community-dwelling older adults: a longitudinal cohort study

Abstract

BACKGROUND: Frailty is related to numerous adverse outcomes and can be operationalized in a variety of ways. However, data on which frailty scales are most suitable for estimating risk remain limited. Herein we examined and compared four validated frailty scales in predicting adverse outcomes in a large population-based cohort of older adults.

METHODS: A total of 5402 subjects aged 50 and older from the WHO Study on global AGEing and adult health (SAGE) in Shanghai were studied. Frailty was measured using a 35-item frailty index (FI), the frailty phenotype (FP), FRAIL, and Tilburg Frailty Indicator (TFI). Logistic regression models were performed unadjusted/adjusted to evaluate the association between frailty and outcomes including 4-year disability, hospitalization, and 4- and 7-year all-cause mortality. The accuracy for predicting these outcomes was determined by assessing the area under the curve (AUC) as well as sensitivity and specificity.

RESULTS: Prevalence of frailty ranged from 4.2% (FRAIL) to 16.9% (FI). Frailty, however defined, was associated with an increased risk of any outcome. The aforementioned associations, except for those of FP with disability and hospitalization, remained significant after controlling for potential covariates. FI, followed by TFI and FRAIL, showed acceptable predictive ability for disability and mortality (AUC: 0.65-0.78). While specificity estimates (85.3-97.3%) for each scale were higher and similar across all outcomes, sensitivity estimates (6.3-56.8%) varied considerably within a lower range.

CONCLUSIONS: All four scales did have the potential to identify Chinese older adults at high risk of adverse outcomes; however, their predictive accuracy varied.

Background

Frailty describes a non-specific state reflecting cumulative declines in multiple physiological systems with aging, leading to decreased resilience to stressors ^[1]. There have been calls for frailty to be routinely screened for among older adults ^[2]; however, no uniformly accepted operational definition for frailty is currently available ^{[1, 3]}.

Most commonly, frailty has been operationalized as the frailty phenotype (FP) based on the biologic syndrome model proposed by Fried and colleagues ^{[4, 5]}. In comparison, the frailty index (FI), developed as a scale of deficit accumulation model, measures the cumulative burden of, for example, diseases, symptoms, and conditions ^[6]. The FRAIL, a hybrid measure containing elements from both the FI and FP models, and the multidimensional Tilburg Frailty Indicator (TFI) with interview-based questions are also frequently used ^[7].

Many scales including all four above, irrespective of the frailty definition used, have been shown to predict a wide range of adverse outcomes ^[8–10], while in practice choosing a scale is sometimes arbitrary, for example, solely based on available data, yet how frailty is conceptualized affects aging research. For instance, since multiple frailty-related incidents, such as disability and hospitalization, can affect lots of people, it is crucial to determine whether one frailty scale has advantages over others in identifying people at risk. To this end, comparisons between frailty scales have been performed but the results remain controversial ^[11–14], partially attributed to differences in study populations, settings, and even the criteria selected to operationalize frailty. In China, the largest developing country with a rapidly ageing population, it remains unclear which frailty scale should be used as an outcome measure, as few studies have constructed multiple frailty scales within the same timeframe ^{[15, 16]}, and even to date, no longitudinal study has compared the four aforementioned frailty scales in the Chinese population.

To dress the above-mentioned gap, this study seeks to examine and compare the accuracy of the four commonly used scales in predicting disability, hospitalization, and mortality in a large longitudinal population-based cohort of Chinese adults aged 50 and older.

Methods

Frailty Scales

The four frailty scales were briefly described below. An overview of all items constructed in each scale can be found in Additional file 2. In particular, to maximize the use of available data, a scale was included in subsequent analyses if no more than 20% of all items were missing ^[13]; meanwhile, missing items for FP, FRAIL, and TFI were imputed with 0, whereas no substitution procedure was required for FI because of its distinctive derivation method used in this study.

Frailty index (FI). Following a standard procedure ^[18], we created a 35-item FI, which differed from our previous publication ^[19] due to the use of different data. Even so, it has been suggested that an index with 30–40 variables is sufficiently accurate for predicting adverse outcomes ^{[18, 20]}. The included variables were converted as a certain proportion of the deficit. For each participant, these deficits were summed up and then divided by the total possible deficit to derive the FI, with a value of 0.20 or greater defined as frail.

Frailty phenotype (FP). The FP was constructed using an adapted phenotypic definition based on the criteria of five components proposed by Fried et al. ^[5]: slowness, weight loss, low grip strength, exhaustion, and low physical activity. It has been previously operationalized in SAGE ^{[21, 22]}, and the same criteria were applied in this study. Likewise, participants were classified as frail if 3 or more criteria were present.

FRAIL scale. We used an adaption of the FRAIL scale ^[23], which considers deficits accumulated in five domains: fatigue, resistance, ambulation, illness, and loss of weight. FRAIL has not been explored in SAGE before. Briefly, fatigue, resistance, and ambulation were assessed using separate self-reported questions. Participants were classified as illnesses if they had 5 or more out of 9 self-reported chronic diseases. The weight loss criterion was ascertained based on the lowest quintile of BMI. Individuals with 3 or more deficits were recognized as frail.

Tilburg Frailty Indicator (TFI). The TFI, developed as an integral conceptual model of frailty, comprises two subscales ^[24]. One subscale addresses the determinants of frailty such as socio-demographics, the latter addresses the level of frailty across physical (8 items), psychological (4 items) and social domains (3 items), and is used in this study, yet it has not previously been explored in SAGE. Items were assessed using self-reported questions as well as several functional tests. Theoretical scores of the TFI range from 0 to 15, with a score of 5 or greater defining frailty ^[25].

Outcome Measures

Outcome measures were disability, hospitalization at 4 years, and 4- and 7-year all-cause mortality.

Disability was assessed both during 2014/15 and 2018/19 using eight activities of daily living (ADL) tasks (moving around, bathing, dressing, maintaining appearance, getting up from lying down, eating, toileting and controlling urine) ^[26]. For each ADL task, participants were asked, “Do you have difficulty in” performing the task in the preceding 30 days? The response was in a Likert scale format ranging from “none” to “extreme/cannot”. Respondents were considered as disability if they reported severe or extreme difficulties in performing at least one of the eight tasks listed above; then, the onset of new disability was defined as a newly identified disability during 2018/19. Spending overnight in hospital at least once, ascertained by asking “have you been hospitalized in the prior four years?” during 2018/2019, was regarded as hospitalization. 4- and 7-year all-cause mortality was determined by linking data to the Shanghai Death Registry during 2018/2019 and on December 31, 2021, respectively.

Covariates

Using the literature on disability, hospitalization, and mortality in older adults as a guide, commonly cited risk factors were selected as potential covariates and then identified in the dataset. Hence, covariates include age, sex, marital status (partnered [married/cohabiting], not partnered [separated/divorced/widowed or never married]), educational level achieved (no education, less than primary, primary, secondary or higher), smoking status (never smoked, current smoker or former smoker) and body mass index (BMI). Measured height and weight were used to calculate a standard BMI (calculated as weight in kilograms divided by height in meters squared).

Results

Baseline characteristics of the cohort are described in Table 1. The subjects ranged in age from 50 to 97 years, with a mean age of 66.3 (SD: 9.6) years. The prevalence of frailty varied between scales: FRAIL, 4.2%; TFI, 7.3%; FP, 12.6%; FI, 16.9%, although between 103 (1.9%) up and 244 (4.5%) participants were unable to be assessed by the four scales due to missing data (> 20% items) (Supplement Table S3). After 4 years of follow-up, 325 (6.0%) of 5402 participants died; of the 3485 responders, 125 (3.6%) developed disability and 720 (20.7%) reported hospitalization. Additionally, after a longer 7-year follow-up, a total of 516 participants died, resulting in a greater mortality rate of 9.6%. Compared with their counterparts, those with all the adverse outcomes were older, less educated, and frailer at baseline.

As shown in Table 2, frail participants as classified by FI, FRAIL, and TFI had a higher risk of adverse outcomes, with adjusted ORs of 3.50, 5.56, 1.93 for 4-year disability, 1.44, 1.64, 1.71 for 4-year hospitalization, 2.22, 1.91, 1.95 for 4-year mortality, and 2.88, 2.29, 1.85 for 7-year mortality respectively. Being frail as defined by FP was also associated with a higher risk of adverse outcomes in unadjusted analyses, whereas after controlling for potential covariates it only independently predicted 4- and 7-year mortality, with adjusted ORs of 1.54 and 2.16, respectively.

AUC comparisons showed that the four scales had the distinctive predictive ability (Fig. 2). FI yielded the highest AUC for predicting disability and mortality (0.76–0.78), followed by TFI (0.71) or FRAIL (0.65–0.72), which performed better than FP (0.57–0.59). By contrast, all scales showed poor ability to predict hospitalization (AUC: 0.53–0.57).

The associated diagnostic values were described in Table 3. Each scale showed higher and similar levels of specificity for all outcomes (FI: 85.3–87.7%, FP: 88.2–89.0%, TFI: 93.8–95.0%, FRAIL: 96.7–97.3%). In contrast, sensitivity estimates varied within lower ranges: FI was more sensitive to disability and mortality (51.2–56.8%) than the others (19.5–34.1%); the lowest sensitivity estimates were found regarding hospitalization (FI 21.0%, FP 15.2%, TFI 10.0%, FRAIL 6.3%). Furthermore, the sensitivity and specificity for each scale were found to vary considerably when higher or lower cutoffs were applied.

Discussion

To date, this large-scale prospective cohort study has been the first attempt to simultaneously identify and compare the four validated frailty scales for their utility in predicting multiple adverse outcomes in the Chinese community dwellers. Our findings corroborate the argument that frailty is a valid state that can be defined in different ways, which may capture different groups of older adults and provide different estimates of frailty prevalence. As mentioned earlier ^{[15, 29, 30]}, multiple frailty scales can be used as predictors for health outcomes, exhibiting different predictive properties. Although comparisons are hampered by huge differences in characteristics of the study subjects, our results were substantially in line with this view.

The reported frailty prevalence among community-dwelling older adults in LIMICs varied from 3.9–51.4%, depending on the study population and the frailty scales. In China, previous SAGE studies ^{[19, 31, 32]} showed lower prevalence ranged 8.6% (FP) to 14.7% (FI) in the younger population. In this study, the frailty prevalence, estimated using different scales, ranged from 4.2% (FRAIL) to 16.9% (FI), similar to those found in other Chinese populations ^{[35, 36]}.

Previous research has validated the studied scales longitudinally in European population. However, these results may not be generalizable as they have shown a different exposure pattern and disease spectrum in the western populations compared to those with Chinese backgrounds. The FI, FRAIL and TFI demonstrated independent predictive validity against all study outcomes in this study, suggesting that they could identify high-risk Chinese older adults. This finding is consistent with those of other studies ^{[12, 13, 15, 33]}, although most of them focused on mortality. Furthermore, while these three scales were comparably associated with hospitalization and mortality, their strengths of association with disability were different; FRAIL conferred the greatest risk of disability, followed by FI and TFI. Notably, there was no evidence of an independent association between FP and disability or hospitalization on multivariate analysis, which contrasted with previously published data ^{[29, 33]}. This discrepancy may be attributable to the partially modified components as well as different covariates adjustments used in our study. Nevertheless, we found that FP was independently associated with mortality, even allowing for different follow-up periods.

Good scales should have high predictive accuracy. Regarding disability and mortality, AUCs for FI, FRAIL and TFI were acceptable and slightly higher than those of other population-based studies ^{[16, 29, 30, 33, 34]}; all four scales, however, were least able to discriminate hospitalization, which was consistent with those studies. For example, the FI, FP and TFI were investigated in a 2-year follow-up study in Dutch community-dwelling older people, and the findings showed AUCs of 0.64, 0.60 and 0.64 respectively, for disabilities, 0.63, 0.59 and 0.61 for hospitalization and 0.64, 0.65 and 0.62 for mortality ^[14]. We demonstrated that the FI, FRAIL, and TFI are useful tools for predicting disability and all-cause mortality, whereas none of the four scales should be used as the sole tool for screening for risk of hospitalization.

Differences were also found in sensitivity and specificity. Our high and similar specificity indicated that all scales can be comparably useful in identifying non-frail participants in those without adverse outcomes. Nevertheless, sensitivity estimates, which were below the acceptable threshold, were in lower ranges compared with values reported elsewhere ^{[14, 29]}. There are several possible explanations. First, we must consider that frailty is a dynamic process rather than a well-defined cutpoint, which increases the difficulty of capturing changes in frailty over time using certain scales. Further, this study used the cutoffs proposed by the original authors for FP, FRAIL and TFI, however, these cutoffs may not be sensitive enough to detect small changes in frailty status when applied to the Chinese population ^[35–37]. Last, different components of the scales and definitions of the outcomes may also have contributed to this result. Our findings showed that all scales failed to correctly identify an adequate number of those who were frail due to low sensitivity.

Overall, our results suggested that all scales were valid predictors for adverse outcomes, while FP independently predicted fewer outcomes (i.e., 4- and 7-year mortality). Differences were evident, however, between unadjusted logistic models where the predictive ability (estimated using AUC) by FI was significantly higher than that of either FRAIL or TFI, all of which offered an advantage over FP. This is perhaps unsurprising, as multidimensional geriatric measures may provide better identification of frailty-related outcomes than a unidimensional index exclusively focused on muscular fitness. Moreover, given none of our scales had both acceptable sensitivity and specificity, nor when the cutoffs were increased or decreased, the choice of scale used will greatly depend on the purpose and setting for frailty assessment ^[38]. Specifically, the FI, TFI, and FRAIL are useful predictors for disability and mortality in intervention programs such as being inclusion criteria for clinical trials, in which higher specificity is preferred over sensitivity, as it is preferable to correctly identify frail individuals, although some frail individuals will be missed. When screening for geriatric conditions in primary care, higher sensitivity is preferred, as it is better to identify as many frail individuals as possible, rather than to miss those who are actually frail; considering our low sensitivity for all scales, a strategy for maximizing the feasibility of frailty screening would be to conduct a stepwise process of increasingly more detailed assessment, that is, a combination of these scales, which may result in a comprehensive geriatric assessment. Another consideration when choosing a scale is the time required to complete it. The FP and FRAIL have the advantage of being easy and quick to administer, although for the former, the relatively restrictive set of criteria may not apply to all individuals. Conversely, while FI and TFI provide broader coverage of deficits and allow for better identification of high-risk individuals, they are more time-consuming and may be less practical to apply in clinical settings and epidemiologic studies. Nevertheless, the increasing use of electronic medical and health records (e.g., hospitals or general practices) enables ready access to health measures across multiple domains. They can then be easily used as screening scales.

Strengths And Limitations

Strengths of this research include the longitudinal cohort design, a large, well-defined population-based sample, a wide range of baseline age, and a repeated comprehensive set of health-related assessments. These enable the operationalization and comparison of these four scales in the same Chinese population within the same timeframe.

Our study also has several potential limitations. A potential limitation is the reliance on self-reported questionnaires. We cannot rule out recall bias (e.g., regarding hospitalization in the last 4 years). Furthermore, the scales used here were adapted from the original definitions to utilize the data available from SAGE in shanghai, some important ageing-specific variables, therefore, were not included, which may have influenced how each scale predicted outcomes. In particular, we used the lowest BMI for self-reported weight loss, which may have modified the scale characteristics, although this modification has been used previously in many studies ^{[21, 22, 32]}. The modified measures, however, may be advantageous. For instance, our measure of memory performance (assessed using a verbal recall test instead of a single self-reported question) can predict functional decline ^[39] and, unlike the self-reported memory used in the original scale, relies on objective assessments. A third limitation is that those who were excluded from the analyses of a scale because of missing data had a higher proportion of 4- and 7-year mortality (Additional file 3). This may slightly underestimate the ability of scales to predict all-cause mortality. Future studies could focus on verifying the usefulness of our operational approaches to frailty by replicating and extending our findings in other populations and settings.

Conclusion

In conclusion, all four scales did have the potential to identify older Chinese at high risk of adverse outcomes; however, only FI, FRAIL and TFI had acceptable predictive ability. Each scale, while performing well at ruling out high-risk groups through high specificity, was likely to miss large numbers of frail individuals as measured by adverse outcomes due to low sensitivity.

Abbreviations

Declarations

Ethics approval and consent to participate

All methods were performed in accordance with the relevant guidelines and regulations, according to the Declaration of Helsinki. Ethical approval for the study was obtained in the Shanghai Center for Disease Control and Prevention Ethical Review Committee (approval notice 200601), with all participants providing written informed consent.

Consent for publication

Not applicable.

Availability of data and materials

The datasets used and/or analyzed during the current study are not publicly available due to the need of approval from each local investigator before sharing, but are available from the corresponding author on reasonable request.

Competing interests

The authors declare no conflict of interests.

Funding

The work was supported by the WHO and the US National Institute on Aging through Interagency Agreements (OGHA 04034785, YA1323-08-CN-0020, Y1-AG-1005-01) and through a research grant (R01-AG034479). The sponsor had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; in the preparation of the manuscript; or in the review or approval of the manuscript.

Authors' contributions

All authors contributed to the study concept and design, acquisition, analysis, or interpretation of data. FQ did the statistical analyses. FQ conducted the literature search and wrote the first draft of the manuscript. All authors critically revised the manuscript for important intellectual content and approved the final version.

Acknowledgements

We wish to thank all the participants in the study as well as the departments involved in their recruitment.

References

1. Hoogendijk EO, Afilalo J, Ensrud KE, Kowal P, Onder G, Fried LP. Frailty: implications for clinical practice and public health. Lancet. 2019;394(10206):1365–75.
2. Morley JE, Vellas B, van Kan GA, et al. Frailty consensus: a call to action. J Am Med Dir Assoc. 2013;14(6):392–7.
3. Pilotto A, Custodero C, Maggi S, Polidori MC, Veronese N, Ferrucci L. A multidimensional approach to frailty in older people. Ageing Res Rev. 2020;60:101047.
4. Buta BJ, Walston JD, Godino JG, et al. Frailty assessment instruments: Systematic characterization of the uses and contexts of highly-cited instruments. Ageing Res Rev. 2016;26:53–61.
5. Fried LP, Tangen CM, Walston J, et al. Frailty in older adults: Evidence for a phenotype. J Gerontol Ser A-Biol Sci Med Sci. 2001;56(3):M146-M56.
6. Rockwood K, Song X, MacKnight C, et al. A global clinical measure of fitness and frailty in elderly people. CMAJ. 2005;173(5):489–95.
7. Dent E, Kowal P, Hoogendijk EO. Frailty measurement in research and clinical practice: A review. Eur J Intern Med. 2016;31:3–10.
8. Benraad CEM, Haaksma ML, Karlietis MHJ, et al. Frailty as a predictor of mortality in older adults within 5 years of psychiatric admission. Int J Geriatr Psychiatry. 2020;35(6):617–25.
9. Thompson MQ, Theou O, Tucker GR, Adams RJ, Visvanathan R. FRAIL scale: Predictive validity and diagnostic test accuracy. Australas J Ageing. 2020;39(4):e529-e36.
10. Negm AM, Kennedy CC, Thabane L, et al. Management of Frailty: A Systematic Review and Network Meta-analysis of Randomized Controlled Trials. J Am Med Dir Assoc. 2019;20(10):1190–8.
11. Woo J, Leung J, Morley JE. Comparison of Frailty Indicators Based on Clinical Phenotype and the Multiple Deficit Approach in Predicting Mortality and Physical Limitation. J Am Geriatr Soc. 2012;60(8):1478–86.
12. Malmstrom TK, Miller DK, Morley JE. A comparison of four frailty models. J Am Geriatr Soc. 2014;62(4):721–6.
13. Theou O, Brothers TD, Mitnitski A, Rockwood K. Operationalization of frailty using eight commonly used scales and comparison of their ability to predict all-cause mortality. J Am Geriatr Soc. 2013;61(9):1537–51.
14. Veld LPMOh, Beurskens AJHM, de Vet HCW, et al. The ability of four frailty screening instruments to predict mortality, hospitalization and dependency in (instrumental) activities of daily living. European Journal of Ageing. 2019;16(3):387–94.
15. Zhang D, Tang W, Dou L-Y, Luo J, Sun Y. Four different frailty models predict health outcomes in older patients with stable chronic obstructive pulmonary disease. BMC Geriatr. 2022;22(1).
16. Shi G-P, Ma T, Zhu Y-S, et al. Frailty phenotype, frailty index and risk of mortality in Chinese elderly population- Rugao longevity and ageing study. Archives of Gerontology and Geriatrics. 2019;80:115–9.
17. Kowal P, Chatterji S, Naidoo N, et al. Data resource profile: the World Health Organization Study on global AGEing and adult health (SAGE). Int J Epidemiol. 2012;41(6):1639–49.
18. Searle SD, Mitnitski A, Gahbauer EA, Gill TM, Rockwood K. A standard procedure for creating a frailty index. BMC Geriatr. 2008;8:24.
19. Ruan Y, Guo Y, Kowal P, et al. Association between anemia and frailty in 13,175 community-dwelling adults aged 50 years and older in China. BMC Geriatr. 2019;19(1):327.
20. Rockwood K, Mitnitski A, Song X, Steen B, Skoog I. Long-term risks of death and institutionalization of elderly people in relation to deficit accumulation at age 70. J Am Geriatr Soc. 2006;54(6):975–9.
21. Anand A. Categorical Frailty Phenotype and Continuous Frailty Index: A Conceptual Overview of Frailty and Its Components among Older Persons in India. Italian Journal of Public Health. 2019;16(4).
22. Gonzalez-Bautista E, Manrique-Espinoza B, Avila-Funes JA, et al. Social determinants of health and frailty are associated with all-cause mortality in older adults. Salud Publica Mex. 2019;61(5):582–90.
23. Morley JE, Malmstrom TK, Miller DK. A simple frailty questionnaire (FRAIL) predicts outcomes in middle aged African Americans. J Nutr Health Aging. 2012;16(7):601–8.
24. Gobbens RJJ, Luijkx KG, Wijnen-Sponselee MT, Schols J. Towards an integral conceptual model of frailty. J Nutr Health Aging. 2010;14(3):175–81.
25. Gobbens RJJ, van Assen MALM, Luijkx KG, Wijnen-Sponselee MT, Schols JMGA. The Tilburg Frailty Indicator: Psychometric Properties. J Am Med Dir Assoc. 2010;11(5):344–55.
26. Üstün T. Measuring health and disability manual for WHO Disability Assessment Schedule (‎WHODAS 2.0)‎.
27. Royston P, Moons KGM, Altman DG, Vergouwe Y. Prognosis and prognostic research: Developing a prognostic model. Bmj-British Medical Journal. 2009;338.
28. Apostolo J, Cooke R, Bobrowicz-Campos E, et al. Predicting risk and outcomes for frail older adults: a protocol for an umbrella review of available frailty screening tools. JBI database of systematic reviews and implementation reports. 2016;13(12):14–24.
29. Widagdo IS, Pratt N, Russell M, Roughead EE. Predictive performance of four frailty measures in an older Australian population. Age Ageing. 2015;44(6):967–72.
30. Forti P, Rietti E, Pisacane N, Olivelli V, Maltoni B, Ravaglia G. A comparison of frailty indexes for prediction of adverse health outcomes in an elderly cohort. Archives of Gerontology and Geriatrics. 2012;54(1):16–20.
31. Biritwum RB, Minicuci N, Yawson AE, et al. Prevalence of and factors associated with frailty and disability in older adults from China, Ghana, India, Mexico, Russia and South Africa. Maturitas. 2016;91:8–18.
32. Hoogendijk EO, Rijnhart JJM, Kowal P, et al. Socioeconomic inequalities in frailty among older adults in six low- and middle-income countries: Results from the WHO Study on global AGEing and adult health (SAGE). Maturitas. 2018;115:56–63.
33. Ensrud KE, Ewing SK, Taylor BC, et al. Comparison of 2 frailty indexes for prediction of falls, disability, fractures, and death in older women. Arch Intern Med. 2008;168(4):382–9.
34. Daniels R, van Rossum E, Beurskens A, van den Heuvel W, de Witte L. The predictive validity of three self-report screening instruments for identifying frail older people in the community. BMC Public Health. 2012;12:7.
35. Zhu Y, Liu Z, Wang Y, et al. Agreement between the frailty index and phenotype and their associations with falls and overnight hospitalizations. Arch Gerontol Geriatr. 2016;66:161–5.
36. Dong L, Liu N, Tian X, et al. Reliability and validity of the Tilburg Frailty Indicator (TFI) among Chinese community-dwelling older people. Arch Gerontol Geriatr. 2017;73:21–8.
37. Dong L, Qiao X, Tian X, et al. Cross-Cultural Adaptation and Validation of the FRAIL Scale in Chinese Community-Dwelling Older Adults. J Am Med Dir Assoc. 2018;19(1):12–7.
38. Takwoingi Y, Quinn TJ. Review of Diagnostic Test Accuracy (DTA) studies in older people. Age Ageing. 2018;47(3):349–55.
39. Park J-H, Park H, Sohn SW, Kim S, Park KW. Memory performance on the story recall test and prediction of cognitive dysfunction progression in mild cognitive impairment and Alzheimer's dementia. Geriatr Gerontol Int. 2017;17(10):1603–9.