Trends In Frailty And Its Associated Factors In Community Dwelling Elderly Indian Population During COVID-19 Pandemic- A Prospective Analytical Study.

Background There is a scarcity of quality literature on the prevalence of frailty among community dwelling elderly in India. This study was originally planned to analyze the longitudinal trends in frailty status of community dwelling elderly in an Indian population as well to identify factors associated with frailty in the Indian context. However the recruitment phase of this study coincided with one of the largest lockdowns in history, associated with the COVID-19 Pandemic, and this gave us a unique opportunity to study the effects this pandemic enforced, absolutely necessary restrictions, had on the frailty status as well the factors affecting frailty in elderly. activity(PASE), gait walk test), nutritional status(MNAT), body composition(BIA), and strength(dynamometry), were measured at baseline and on follow-up exactly after 3 months. The change occurring in these variables over the 3 month period was analyzed and the change in frailty was independently correlated with changes in each of the other outcomes.


Introduction
Frailty derived from Latin word 'Fradilita' meaning Brittleness, is an important and emerging term in Geriatric medicine.(1) There is no de nition which is internationally recognized to be valid, but it is usually associated with adverse outcomes developed as a consequence of increased venerability because of decline in physiological systems with increasing age, triggered by any minor stressor which collectively leads to sudden changes in state of health. (2)Frailty is geriatric syndrome which is multidimensional in nature. World Health Organization (WHO) de ned frailty in 2017 as "a clinically recognizable state in which the ability of older people to cope with everyday or acute stressors is compromised by an increased vulnerability brought by age-associated declines in physiological reserve and function across multiple organ systems". (3) Page 3/13 900 million people around the world comes under elderly age group according to WHO, out of which in India there are 104 million elderly (> 60 years age). It is also estimated that India will hold largest geriatric population around the globe by the year 2050. (4) With the advancement in medical sciences there is decrease in mortality rates, the life expectancy is increased and so is the frailty among the elderly. (5) It is estimated that 4-10% of elderly population dwelling in United state are frail, also 8.1% elderly are observed frail in United kingdom, 6.5% and 7% in Italy and France (respectively). (6) A large and compressive study by WHO showed that among middle and low income countries (South Africa, China, Russia, Ghana, India and Mexico) India has the highest prevalence of frailty (i.e. 56.9%), more number of women are frail then men (47% of elderly men and 67% of elderly women). (7,8) The evaluation of frailty is di cult because of lack of any standardized tool. There exists 67 tools for quantifying Frailty, out of which only 9 of the screening tools are highly cited (more than 200 citations).
(9)Phenotype of frailty (given by Fried)(10) and Frailty index (given by Mintnitski) (11) are the validated and most widely used screening tools.(9) Fried Phenotype of frailty model de nes a person frail when 3 or more than 3 physiological de cits out of 5 are present (10), whereas Frailty index model expresses frailty as a "ratio of existing de cits to the total probable de cits there could be". Wide ranges of diseases, disabilities, signs and symptoms are de ned as these de cits. (11) Ageing leads to numerous changes in physiological systems of our body which are fundamental to the growth of frailty, speci cally the immunological system, the neuromuscular system and neuroendocrine system. (12)These changes in body interact progressively and adversely, leading to loss of physiological function and reserve (state of compromised homeostasis). (12) The risk factors for frailty are varied and have been found to have multiple linear and non linear interactions. For example consequence of normal ageing leads to loss of muscle mass and strength. (13) Muscle loss can also be accelerated due to chronic illness, poor nutrition, decrease in growth hormone production, and reduced physical activities.
(13) All these factors are inter-related to each other through complex interactions and ultimately leads to frailty. Socio economic and demographic variables like availability of disposable income/ nances, level of education, nutritional status, and general living conditions have been found to be confounders of frailty. (7) Several studies have identi ed factors like sarcopenia, loss of muscle strength, functional mobility and gait velocity changes, loss of weight, reduced physical activity and easy exhaustibility to be strong independent confounders of frailty. The most closely associated biological parameters of frailty have been identi ed, as in ammatory markers, dyslipidemic markers, endocrinological markers, insulin resistance and state of glycemia. (12) Available literature on the feasibility of predicting frailty state that changes in functional as well as biological parameters could be used as well quali ed candidates to estimate and quantify frailty. Out of these risk factors, it is purported that functionality could be the strongest predictor or measurer of frailty.
Sarcopenia, connective tissue remodeling, and in ammatory markers mediated physiological and functional changes and their interaction among themselves and other confounders of frailty need to be studied to develop a predictive model of frailty. (14) A recent large scale review on the state of frailty related research in India had stated that there is a rather alarming lack of conceptualization or epidemiological data regarding frailty in Indian scenario, and there is a dire need to identify the key confounders for frailty syndrome among Indian elderly. (7)There is a scarcity of quality literature on the prevalence of frailty among community dwelling elderly in India. This study was originally planned to analyze the longitudinal trends in frailty status of community dwelling elderly in an Indian population as well to identify factors associated with frailty in the Indian context. However the recruitment phase of this study coincided with one of the largest lockdowns in history, mandated to minimize the spread of COVID 19 in India. Restrictions were put in place to minimize the outdoor movement of population in general and speci cally the elderly, who were recognized to be the most vulnerable group with respect to the pandemic. This gave us a unique opportunity to study the effects this pandemic enforced, absolutely necessary restrictions, had on the frailty status as well the factors affecting frailty in elderly.

Methodology
After obtaining necessary permission to recruit subjects, the study recruitment was planned to commence in March 2020. A total of 28 samples were screened between rst and third week of March of whom 22 subjects were found to be eligible for study. The criteria for inclusion were that the age must be greater than 65 years, MOCA score greater than 26 at the time of rst evaluation. Subjects with known diagnosis of any progressive disorder, as well as those with, cardiovascular, musculoskeletal, neurological or systemic illness which could potentially interfere with data collection were excluded.
Subjects' demographics as well as medical history were recorded using appropriate tools following which evaluation for frailty index questionnaire was administered to identify and quantify frailty among them.
Lower extremity muscle strength was evaluated using baseline hand held dynamometer. A Tanita® Segmental Body Composition analyzer was used to determine the body composition variables of muscle mass, visceral fat and total body fat percentage for each subject. Subjects were then made to do a 10 m walk test to analyze the gait velocity following which the nutritional status and socioeconomic status were evaluated using MNAT-NESTLE® and BG prasad scale respectively. Physical activity level was recorded using PASE scale following which a timed up and go test was then performed to analyze the functional mobility status following the collection of outcome measures each subject was given a date exactly 3 month from date of 1st evaluation for the follow up assessment.
Unexpectedly India went into complete lockdown in the third week of March 2020 following the rise of COVID-19 cases which in its strict form lasted for approximately 70 days following which there was a phase wise gradual unlock. Our follow up data collection coincided with phase two of unlock but still there was a general advisory for elderly subjects to be home bound to minimize chances of exposure. Of the 22 subjects recruited only 19 subjects returned for the timely follow up evaluation. All the outcomes were again collected using same tool and in the same order at the end of third month.

Data Analysis
The data was collected and analyzed using JAMOVI version 1.6.14. Statistical software. Normality of continuous variables was tested using Shapiro-Wilk test. Demographic variables were expressed in terms of descriptive statistics. Differences in Gait velocity, strength, body composition and functional mobility at baseline and at 3 months of follow up was analyzed using Wilcoxon sign rank test/Students pared sample T-test. Changed scores of independent variables were individually associated with changes scores in "Evaluation of Frailty index for physical activity questionnaire" using Spearman's Correlation test.

Results
A total of 28 subjects were screened out of which 19 samples after ful lling the inclusion and exclusion criteria were recruited. The demographic data of all participants are represented in

Discussion
Present study was undertaken primarily to analyze the trends in frailty status of a cohort of community dwelling elderly, residing in the Dakshina Kannada district of Karnataka state in India over a period of 3 months as well as to analyze the strength of association between change in frailty score and cognition, nutritional status, gait velocity, functional mobility, body mass, and strength The recruitment period of the study coincided with the beginning of ongoing SARS COVID 19 which proved to be a major hindrance in approaching, screening and evaluating elderly subjects. Over the period of study, a total of 28 subjects were screened, of which 22 ful lled the criteria of inclusion in the study. However of the 22 subjects, the follow up evaluation could be done only for a total of 19 subjects and hence the goals of the study were realigned to investigate the in uence of pandemic induced lockdown and the associated reduction in physical activity on the outcome variables. In the current study we found that there is an observable change in frailty status over a period of 3 months but it was not statically signi cant.
Frailty is an umbrella term and there are many tools to measure Frailty. EPIF scale was used in the current study because it covers all domains of frailty (like physical, psychological, social functioning and general health), and has been proven to have good reliability and validity. (15) The data collection involved administering 4 questionnaires (EPIF, MOCA, MNAT NESTLE®) which on an average took 45 min-1 hour to complete. Objective measures of strength, functional mobility, gait velocity, and body composition analysis would take an additional hour to complete. This made the entire data collection process a time consuming one thereby adversely affecting the number of subjects recruited in a day. However other than the afore mentioned 3 subjects who chose to forgo the follow-up evaluation because of the pandemic situation, there were no additional dropouts in the span of study and no reported discomfort or adverse event pertaining to data collection.
The primary objective of the study was to detect any association between changes in frailty status and other outcome variables. It must be noted that there was only very minimal difference (over a period of 3 months) in Frailty score (mean difference 0.625) and ndings were not statically signi cant. We could not nd any statically signi cant relationship between changes in frailty score and the changes in strength, muscle mass, cognition, nutritional status, gait velocity, or functional mobility.
However it must be emphasized that, when the independent variables were compared at baseline and 3 month of follow-up there was a statistically signi cant difference found in the scores of MOCA, TUG, visceral fat, PASE and muscle mass. The muscle mass and gait velocity showed a marginal but statically signi cant reduction, whereas total body fat as well as visceral fat content showed an increment. Cognitive functions as measured by MOCA and gait velocity (implied by an increase in time taken to complete 10M walk test) showed a decline in the above mentioned period, whereas the time taken to complete TUG had marginally increased. The observed differences in MOCA scores though were never su cient to imply a cognitive decline. It can be inferred from these nding that a short span of 3 months has brought about measurable differences in variables which have been previously associated with frailty.
Previous research corroborated our ndings in that there is a de nitive decline in muscle mass ranging between 2 to 4% annually in older men and women of all ethnicity. There is also a concurrent increase in body fat content averaging about 0.8% within the same time span. (16) Factors that in uence body composition, especially muscle mass include genetic variables, metabolic variables, endocrinological variables, co-morbidities, diet, alcoholism, smoking, as well as gender and ethnicity. It must be emphasized however that physical activity as an independent variable is a strong predictor for loss of muscle mass and changes in body composition in elderly. (17)The data collection of present study coincided with the period of pandemic enforced restriction and all of the recruited subjects had reported a considerable decline in the amount of physical activity they indulged in the same period.
For measuring physical activity we had used Physical activity Scale for elderly (PASE) and we found a highly signi cant reduction in the physical activity (Mean Difference = 43, p < 0.05) over the period of 3 months. For the population which we had studied, the major source of physical activity used to be walking in public places like parks or attending organized social gatherings like yoga and group exercise sessions. Since most of these activities were deemed to be unsafe, especially in elderly population, there was virtually a complete absence of these activities in the lockdown period.
Our data analysis shows there is a statically signi cant decline in functional mobility as measured by TUG with ageing, but it must be emphasized that this decline was barely consequential and for all intentions and purpose it is safe to assume there was no decline in functional mobility of the studied cohort. Gait velocity showed a statically signi cant but minuscule difference when compared over a period of 3 months.
In all major muscle groups of lower extremity, there was a signi cant difference noted in strength which ranged from a difference of 0.7 kg to 1.5 kg. One of the key associated nding was that the decline in strength of bilateral hip and knee musculature (hip abductors, hip adductors, knee extension of right side and knee extensors, hip exors, hip extensors and hip adductors on the left side) showed statically signi cant moderate correlation with decline in muscle mass. Previous studies have shown that there is insu cient evidence of a liner relationship between the loss of muscle strength and muscle mass in aging, though both have been individually established as de nitive outcomes of ageing. Other factors affecting muscle strength have been identi ed as impaired reciprocal inhibition, alteration in rate coding of motor unit activation, as well as changes in metabolic characteristic of muscle bers. These changes can happen independent of the changes in muscle mass.(18)The changes in muscle strength could then be attributed to the de nitive decline in physical activity levels as previously stated, which would have precipitated a deconditioning/reversal effect on muscle strength. In our study cohort we observed neither statically signi cant nor any amount of change in the nutritional status of study population as measured by MNAT®.

Conclusion
Two key ndings of this study are that, 1) there has been a de nitive decline in physical activity of elderly within the lockdown period, and, 2) there is absolutely no signi cant change in the frailty status of community dwelling elderly, even in a time period characterized by physical activity restrictions due to the COVID 19 induced lockdown, though some of the independent determinants of frailty showed a decline in the same period. Present study failed to establish any association between frailty and changes in cognitive, functional mobility, body composition, strength, or nutritional factors, during a relatively short span of 3 months. The study was approved by the Scienti c and Institutional Ethics committee of KMC Mangalore (IEC KMC MLR 11-19/590). All stages of the study were conducted in strict adherence to the principles of "Helsinki Declaration" for research on human subjects. All subjects were explained the objectives of the study in a language which they fully comprehended, and were recruited for the study only after signing a written informed consent.