Outcome of older adults at risk of frailty

Abstract Background The integral part of the definition of frailty is the outcome associated with it. Older adults at risk of frailty are in the process of becoming frail. This study looked at the clinical characteristics and outcomes of older adults at risk of frailty. Methodology The study population was selected from outpatient department of the geriatric medicine department in a tertiary care hospital. Older adults identified as at risk of frailty were assessed at baseline and then followed up after 1 year for the composite primary outcome of death, falls, hospitalization, and self‐rated poor quality of life in the follow‐up period. Results The study included 324 older adults who had completed 1‐year follow up. Mean (SD) age was 74.49 (4.58) years, and males were 241 (74.15%). Frail and pre‐frail at baseline among the study population were 31.17% and 61.11%, respectively. The primary outcome occurred in 43 (13.27%) patients. Poor baseline IADL was significantly associated with primary outcome at the end of 1 year. Conclusion An unfavorable outcome in older adults at risk of frailty was significantly higher and independent of their baseline frailty status. Poor baseline IADL value may be considered as a predictor for primary outcome at 1 year of follow up.

recommend that older adults 70 years of age and above or those who had unintentional weight loss of >5% of body weight in the past year be screened routinely for frailty in a clinical setting. 4 The screening criteria can vary with the societies' recommendation, but generally, all older adults 70 years and above may benefit from frailty screening.
Western models used for assessment of frailty in the Indian population may not reflect the actual status. The objective of our study was to screen older adults who are at risk of frailty and follow them prospectively for the outcome of interest. Based on these primary outcomes, we intended to develop a suitable model for the Indian population for identification of frailty.

| ME THODOLOGY
The study was conducted in Geriatric Medicine Outpatient Department (OPD) at the All India Institute of Medical Sciences, New Delhi, from 2017 to 2020. The study was a prospective observational cohort study of older adults 70 years and above who attended Geriatric Medicine OPD were screened for the risk of frailty and inclusion in the frailty registry. The risk of frailty was defined by loss of weight or self-reported decline in physical activity or one or more falls in the past six months. So, older adults 70 years and above with any of the three risk factors were included after written informed consent. Older adults with severe depression, severe cognitive impairment (Hindi Mental State Examination score <18), severe osteoarthritis knee, severe heart failure, decompensated liver or kidney disease, or severe anaemia were excluded from the study.
Detailed baseline data were collected during study recruitment, including demographic data, anthropometry, medical history, physical function, functionality, and mental status. Health-related conditions and medications were identified by reviewing health records and blister packs. Socioeconomic status was assessed by a periodically updated standard Indian scale. 5 A comprehensive geriatric assessment was carried out to find any geriatric syndromes, vision by Snellen or E chart and hearing impairment by Whispering test. 6 A fall was defined as an event that results in a person coming to rest inadvertently on the ground or floor or other lower level. 7 Orthostatic hypotension measured using a digital sphygmomanometer (Omron TM 7310) uses an oscillometric method. 8  Chair raise test done with a standardized armless chair, and the time taken was measured by a stopwatch. The 1 kg arm lift was measured by the standard dumbbell. Gait speed was assessed by the time taken to walk 4 meters at the usual pace. Grip strength measured using a hand dynamometer (JAMAR; Sammons Preston, Rolyon) by Southampton protocol, 18 value of the dominant hand was considered. Mid-arm circumference (MAC) and mid-thigh circumference were measured when the subject was sitting in a relaxed position with arm-by-side and in lying position, respectively. MAC was measured at the midpoint between the acromion and olecranon process, and the thigh circumference was measured between the midpoint between the inguinal crease and the proximal border of the patella (Appendix S1).
The telephonic follow up was done after completion of 1 year.
For those who have not responded to the initial follow-up call, two more calls were made at 1-month interval each. Some participants were followed up when they come for scheduled outpatient visit.
Participants who completed at least one follow-up call or visit after 1 year of recruitment were included in the analysis. The older adults who had not completed follow up were excluded from the study.
The primary outcome was defined as a composite event of history of hospitalization in the 1-year follow-up period or history of fall in the 1-year follow-up period or self-rated poor overall health status during follow-up call/visit or death in the follow-up period.

| Statistical analysis
The data were maintained in an Excel spreadsheet and analyzed using STATA-SE (Version 14.2) (StataCorp). Normality assumption tested using the Kolmogorov-Smirnov test and, accordingly, quantitative data were reported as the mean (standard deviation (SD)).
The qualitative variables of the study participant were reported as numbers with percentages. To find the association between the categorical variables, chi square test or Fisher exact test were used according to the frequency distribution, and for comparing quantitative measures between the two groups, t-test was used. To find out the factors associated with the primary outcome, stepwise multivariable logistic regression analysis approach was used and the results were presented in the form of odds ratio (95% confidence interval).
Multicollinearity and confounders were also explored. Variables that were found to be significant at the level of 25% under crude analysis and/or clinically relevant were considered for stepwise procedure with the entry probability 15% and removal probability 20%. A p value of less than 0.05 was considered as statistically significant.

| RE SULTS
Out of the 550 older adults assessed for eligibility, only 324 had completed at least one follow up and included in the analysis. The baseline characteristics of study participants are given in Table 1.
The mean (SD) age of the study population was 74.49 (4.58) years.
The male population was predominant (74.15%). The primary outcome occurred in 13.27% of the overall study population. Only 10 (3.08%) lived alone; the rest lived with either spouse or children or both. The detailed baseline and follow up information of the characteristics are presented in Table 1.
The study population was categorized into two groups based on outcome as those with no outcome and primary outcome. In anthropometric measures, those without any outcome had significantly higher BMI (23.96 vs. 22.43 kg/m 2 ), and more proportion of them were in the obese category (57.65% vs. 37.21%) than those with the primary outcome. Gait speed was slow, and the Timed-Up-Go test was high in those with the primary outcome ( Table 2). The muscle mass measured by the upper arm and mid-thigh were significantly lower in those with the primary outcome. It was also associated with a significantly lower mini-nutritional assessment score. Both baseline Basic (BADL) and Instrumental (IADL) Activities of Daily living score was significantly lower in those with any negative outcome, and their functional reach was also significantly lower.
In the overall study population, 31.17% were frail, and 61.11% were pre-frail. Older adults who were designated as at risk of frailty and found to be robust were 7.7%. In the population with the primary outcome, 34.88% were frail and 55.81% were pre-frail. In those without any of these outcomes in the follow-up period, 30.60% were frail, and 174 were pre-frail ( were significantly associated with the primary outcome ( Table 3).
The results of the multivariable analysis are presented in Table 4.

| DISCUSS ION
Out of the 324 older adults who completed 1 year of follow-up, the primary outcome occurred in 43 (13.27%). The 4-year incidence of frailty was 7.2% in the phenotype study. 2 The study outcome and measurement were different, but the rate of primary outcome is high in our study. Among the baseline characteristics, there was no difference in age, gender, living status, educational status, or socioeconomic status. There was no association with a history of weight loss, as it was one of the screening criteria for determining those at risk of frailty. However, mean BMI was lower in those who had the primary outcome.
Those with primary outcome had a significant difference in anthropometric measures and physical performance. They all had a significantly lower muscle bulk as measured by mid-arm and midthigh circumference. Along with reduced muscle bulk, the mean grip strength of the group was also poor, though not statistically significant. But when the grip strength was adjusted for age & gender, there was no difference between the two groups. The gait speed was significantly lower in those with the primary outcome. The gait speed as determined for their age and gender was also slow. The Timed-up-Go score was also significantly higher. These findings suggest slowness in gait is an important predictor of poor outcomes in older adults.
Functional Reach Test (FRT), Short Physical Performance Battery (SPPB), and TUG score measure physical performance and predict fall risk and physical frailty. The SPPB score is significantly lower in those with a negative outcome. The lower muscle bulk and poor SPPB score, both a marker of probable sarcopenia, were associated with the primary outcome. In a comparative study by Lim et al., SPPB correlates more with physical frailty, and poor performance in SPPB can predict the at-risk of the frail population. 19  TA B L E 2 Demographic characteristics of the population based on the outcome could be a confounding factor, but we have to also look at how poor physical performance measures might be an effect modifier as it would have led to past and future falls in these study participants.
The BMI was normal-high in the population, and no difference was observed between the groups in the grip strength. In this study, the muscle density and body composition were not examined, and the possibility of sarcopenic obesity cannot be ruled out. The prospective data from the InCHIANTI study showed that though the population had higher BMI, the frailer population had high-fat mass, poor muscle density, and low gait speed, similar to our data. Sarcopenia is part of the vicious frailty cycle, but it would be difficult to determine whether sarcopenia preludes frailty, as the nosology is unknown. The grip strength is insignificant because it only measures the upper body or regional strength.
But the gait speed and the composite physical performance measures examine overall strength and ability to perform daily activities. 22 IADL impairment was one of the most common associations with frailty in the community-dwelling older adults. 23 Frailty is considered as a continuum, and there was a consensus that by improving comorbid conditions, it can be reversible. 24 A prospective study by Zhang et al. identified the frailer population was at risk of disability and falls in the future. 25 A similar association was also observed in the EPIDOS study in community-dwelling older adults using the phenotype model 26 and in the hospitalized patients by Liang et al.
using the cumulative deficit model. 27 The relation between frailty and disability is very close as each of them predisposes to the other.
In our study, IADL impairment is strongly associated with primary outcomes irrespective of baseline frailty status. The study suggests that physical performance measures were poor, and IADL impairment can be used to predict primary outcomes in older adults.
There are few limitations in the study. Though we tried to look at the significant variables to construct a model to predict these outcomes, we had only one variable (IADL) with statistical significance.
Many subjects did not complete at least one follow up and the attrition rate was very high, partially due to the COVID-19-related attrition of health services.

| CON CLUS ION
Older adults who were at risk of frailty had a high incidence of primary outcomes at 1 year of follow up. Poor physical performance measures like slow gait speed, increased TUG score, and lower SPPB score were also associated with the primary outcomes. The primary outcome was independent of the baseline frailty status but associated with their functionality (baseline IADL). Screening for IADL impairment in the atrisk frailty population can be done at primary care level, and rehabilitation can be initiated to prevent falls and poor outcomes.

CO N FLI C T O F I NTE R E S T
None of the authors report any conflict of interest.

AUTH O R CO NTR I B UTI O N S
VG, MSS, and ABD contributed in study concept and design, acquisition of data, andinterpretation of data. VS contributed in analysis of data and interpretation of data. All authors contributed in the preparation of manuscript and final approval.

CO N S E NT FO R PU B LI C ATI O N
Not applicable.