Data
This study utilizes the data from the first wave of Longitudinal Ageing Study in India (LASI) conducted during 2017-18. LASI is a nationally representative survey investigating the health, economic and social determinants and consequences of population ageing in India. The survey used a multistage area probability cluster sampling design to locate the respondents and collect data. A household with at least one adult aged 45 and above has been selected as a sample from all states and union territories of India except Sikkim [17]. The total sample size for the survey covered 65,562 older adults aged 45 and above. The data collection was completed through structured schedules of questionnaires using CAPI and conducted by trained investigators. The data quality and collection process were actively monitored by nodal agency, International Institute for Population Sciences (IIPS) Mumbai [17].
Variable description
Outcome variable
There are two outcome variables used in this study. Both the components were binary response variables.
cognitive health was assessed on five cognitive domains (memory, orientation, arithmetic function, executive function, and object naming). The cognitive domains in the LASI were derived from The University of Michigan Health and Retirement Study (HRS). Memory was assessed using immediate word recall (0–10 points) and delayed word recall (0–10 points); orientation was assessed using the time (0–4 points) and place (0–4 points) measure; arithmetic function was assessed using backward counting (0–2 points), serial seven (0–5 points), and computation method (0–2); executive function was measured through paper folding (0–3) and pentagon drawing method (0–1); and finally, an object naming was conducted (0–2) among the study participants. A composite score of 0–43 for cognitive abilities was computed combining the score from five cognitive domains with a higher score representing better cognitive functioning.
The Centre for Epidemiologic Studies Depression (CES-D) scale was used to assess the depressive symptoms. The CES-D items were selected from a pool of items from previously validated depression scales which included components: depressed mood, feelings of guilt and worthlessness, feelings of helplessness and hopelessness, psychomotor retardation, loss of appetite, and sleep disturbance. A composite score ranging from 0 to 60, with higher scores indicating more symptoms, weighted by frequency of occurrence over the previous week. Cronbach’s alpha indicated that CES-D has excellent internal consistency(α = 0.85). The scale is a useful tool for identifying such high-risk groups and exploring the relationships between depressive symptoms and a diverse range of other variables [18].
Explanatory variables
Social deprivation index (SDI)
We included variables from different dimensions of life satisfaction among the elderly to estimate the SDI.
Table 1 describes the variable included for calculating the index.
Factors
|
|
Includes Variables
|
Social security
|
|
Living alone
|
Socially active
|
Safety from crime
|
Victim of crime
|
Economic
|
|
Wealth quantile
|
Current working status
|
Health Insurance
|
|
|
Loan
|
Household
|
Household condition
|
Crowding
|
Sanitary facilities
|
Ownership of house
|
Housing type
|
Education
|
Number of years of schooling
|
Socio-demographic characteristics
Socio-demographic characteristics included age, sex, marital status, education, working status, caste/tribes, place of residence, and wealth status. The age was categorized into three groups including youngest old (45-59 years), middle old (60-75 years), oldest old (75 and above years). Marital status included three categories i.e., currently married, widowed, and separated/divorced. Wealth status was incorporated as the five quantiles from the wealth indicator ranging from poorest to richest. The caste/tribe variable included scheduled tribes, scheduled caste, other backward castes, and no caste/tribe. Working status was categorized as currently working, retired, and never worked.
Statistical Analyses
The first step in the statistical analyses was to construct hedonic weights [19] for estimating the Social Deprivation Index (SDI). The hedonic weights were calculated using the standardized coefficients from the ordered probit regression by taking self-rated life satisfaction as a dependent variable [19]. The hedonic weights for all the factors contributing to SDI were used to construct the final SDI. The index was categorized into three categories as low (0 < SDI < 0.4), moderate (0.4 < SDI < 0.6), and high (SDI>0.6). The proportion of elderly in each category of SDI was estimated for all the states and socio-demographic characteristics (results are given in the supplementary files).. We calculated the prevalence of cognitive abilities and depressive symptoms across the categories of SDI and socio-demographic characteristics of the elderly. The association of cognitive health and depressive symptoms with SDI and other socio-demographic characteristics was tested using the Chi-square test of association. A p-value less than 0.05 for the test was considered to be significant. Finally, multivariable logistic regression [20] was fitted to examine the adjusted effects of SDI on cognitive health and depressive symptoms among the elderly. The whole analyses were performed by using STATA version 16[21].