Introduction

With about 3.8% depressed population worldwide, a common mental health issue, depression, is slowly sneaking into the lives of people of all ages [1]. According to the World Health Organization, depression results from persistent sadness and a loss of interest in one’s usual activities, accompanied by an inability to carry out daily activities for at least two weeks [1]. Even though depression is preventable, studies show that more than 75% of individuals do not receive treatment in low-and-middle-income countries [2]. Such negligence is mostly due to the misconception that depression is a part of the ageing process, making the 60 years and above population more vulnerable [3].

India, one of the lower-middle-income countries, is no exception [3]. According to the National Mental Health Survey 2015–16, one in every 20 Indians suffers from depression [4]. Although the most affected age group were those aged 15–49 years, the continuous Indian demographic transition has generated health concerns for the aged people in recent years. According to the Longitudinal Aging Study in India, about 8.3% of older adults aged 60 years and above experienced major depression based on the Short Form Composite International Diagnostic Interview (CIDI-SF) scale [5]. Depression triggers many health problems like obesity, diabetes, heart disease, physical activeness, and vice-versa [6,7,8,9]. A complex interaction of social, psychological, and biological factors further triggers depression in an individual [10,11,12].

In India, inequities in health and healthcare utilization between advantaged and marginalized groups persisted throughout the years [13]. Extant research has demonstrated social status as a significant predictor of health disparities in India [14,15,16,17]. A study from the World Health Organization’s Survey of Global Ageing and Adult Health (WHO-SAGE) indicated higher mental health issues among Scheduled Castes and Muslims than among Higher caste Hindus [18]. Ample evidence shows the depression vulnerability of the tribal population in the Indian context [18,19,20,21]. Changing lifestyle, beliefs, and community living, moving to urban spaces, and triggering adverse life events like morbidity, unemployment, bereavement, stress, and psychological trauma are the emerging factors predisposing to mental health issues like depression [22].

Several studies have associated depression with lower socioeconomic status; however, a study indicated that emotional well-being (associated with depression) and life evaluation (associated with life satisfaction) have different correlates, which are often considered the same mistakenly [23]. For instance, there is a long-run question of whether money buys happiness or not. A study from the US shows that emotional well-being is highly associated with the positive and negative emotions created by circumstances like caring for a sick relative and spending time with family and friends. In contrast, life evaluation was highly associated with socioeconomic status [23]. The study concludes that a high income can bring life satisfaction but not happiness. According to these findings, the indigenous older population living with their families and friends might experience lesser depressive symptoms due to better emotional well-being. However, the long negligence of mental health issues in the tribal population also raises concern on their current situation.

Thus, the present study aimed to investigate differentials in the prevalence of depression among the tribal and non-tribal populations in India, primarily focusing on the vulnerable older population. Moreover, we explored the contribution of socio-demographic, health-related, and household factors to depression disparity across tribal and non-tribal older adults in India.

Materials and methods

Data source

The present study utilized the data from the baseline wave (wave 1) of the Longitudinal Aging Study in India (LASI), conducted during 2017–18. The LASI is a biennial panel survey representative of the older and elderly population age 45 and above for India and its states and union territories. The LASI wave 1 interviewed 42,949 households and 72,250 individuals aged 45 + years and their spouses across India’s 30 states and six union territories. It adopted a multistage stratified area probability cluster sampling design: a three-stage sampling design in rural areas and a four-stage sampling design in urban areas.

The LASI is a multi-topic, nationally representative, a large-scale survey conducted jointly by the International Institute for Population Sciences (IIPS), the Harvard T.H. Chan School of Public Health (HSPH), and the University of Southern California (USC). It is the Indian version of the Health and Retirement Studies (HRS). It provides vital information on chronic health conditions, symptom-based health conditions, demography, functional and mental health, household economic status, healthcare utilization, health insurance, work, employment and retirement, and life expectations of 45 years and above participants with their spouses, irrespective of age. Further details regarding the sample design, survey instruments, fieldwork, data collection and processing, and response rates are publicly available in the report [5].

Sample and participants

The Indian constitution divides the social groups into three broad categories- Scheduled Tribes [ST], Scheduled Caste [SC], and Other Backward Classes [OBC]. Those who do not fall in these groups are categorized as “Others”. Both SC and ST include mostly socially backward groups. OBC is socially and economically backward but has a much better situation than SC/ST. The ST group, which has a predominantly tribal population, comprises 8.6% of India’s population. Traditionally, these tribal groups reside away from mainstream Indian society and consume nature-based resources for daily health requirements [24, 25]. We have recoded the social group variable into ST (tribal or indigenous) and Non-ST groups in the present study.

In this study, we utilized the sample of 31,464 Indian older adults aged 60 years and above (5,173 belonging to the ST and 26,291 belonging to the non-ST groups). 827 older adults whose depression information was not available in the data were excluded from this study. Therefore, the analytical sample was 30,637 older adults, among whom 5,025 and 25,612 belonged to the ST and Non-ST social groups, respectively.

Outcome variable

The main outcome variable was major depression among older adults categorized as “Depressed” and “Not depressed”.

Major depression was assessed in LASI using the successfully tested and widely-utilized Composite International Diagnostic Interview short-form (CIDI-SF) scale that assesses psychiatric depression [5, 26]. There are ten questions and the response for all the items (except items “2” and “3”) were in binary format (i.e., “No” coded as 0, and “Yes” coded as 1). Again, those individuals who felt sad, blue or depressed “all day long” or “most of the day” were coded as “Yes”, else they were coded as “No”. Equivalently, individuals who felt sad, blue or depressed “every day” or “almost every day” were coded as “Yes”, else, they were coded as “No”. The scale had a Cronbach’s α reliability coefficient of more than 0.6. Next, we summated the ten items to obtain the CIDI-SF depression scale, with scores ranging from 0 to 10. Based on extant literature, Indian older adults with a score of 5 and more were categorized as “Depressed”, and those with a score of 4 and less were classified as “Not depressed” [27, 28].

Explanatory variables

Based on existing research, we included the following socio-demographic, health-related, and household characteristics in this study. The socio-demographic characteristics of older adults include the following –

  1. 1)

    Age group was coded as (oldest-old, old-old, young-old). The young-old, old-old and oldest-old comprise older adults aged 60 to 69 years, 70 to 79 years, and 80 years or more, respectively.

  2. 2)

    Gender was coded as (male, female).

  3. 3)

    Marital status was coded as (currently married, currently not married).

  4. 4)

    Status of work was coded as (never worked, currently not working, currently working, retired).

  5. 5)

    The level of education was coded as (no formal education, upto primary, secondary and above).

  6. 6)

    Social participation was coded as (socially active, socially inactive). LASI collected information on whether a person goes out of the house for eating, goes outdoors for relaxing, plays indoor games, plays outdoor games or exercises, visits relatives, attends cultural events, attends religious functions, attends group meetings, reads books, newspaper and magazines, watches television or listens to the radio, use a computer for electronic communication. Older adults who undertook any of the above social activities were classified as “socially active” and otherwise were classified as “socially inactive”.

  7. 7)

    The importance of religion was coded as (not important, very important).

  8. 8)

    Living arrangement satisfaction was coded as (satisfied, neutral, not satisfied).

  9. 9)

    Received ill-treatment was coded as (no, yes) within one year from the interview.

  10. 10)

    Faces discrimination in life was coded as (no, yes).

    Next, the health-related characteristics of the older adults include –

  11. 11)

    Self-rated health was coded as (good, average, poor).

  12. 12)

    Chronic morbidity status was coded as (no condition, single condition, multiple conditions). LASI collected information on whether an older adult was ever diagnosed with hypertension or high blood pressure, diabetes or high blood pressure, cancer or malignant tumour, chronic lung diseases, chronic heart diseases, stroke, bone or joint diseases, any neurological or psychological problems, high cholesterol. Individuals having no diseases, any one of the diseases and two or more diseases were categorized into “no condition”, “single condition”, and “multiple conditions”.

  13. 13)

    Physical activity status was coded as (physically inactive, physically active). Physical activity status was assessed based on WHO guidelines on physical activity for 18 years and above [29]. Older adults who performed at least 75 min of vigorous-intensity physical activity or at least 150 min of moderate-intensity physical activity in a day or a combination of both were classified as “Physically active” and otherwise were classified as “Physically inactive”.

  14. 14)

    Difficulty in Activities of Daily Living (ADL) was coded as (no difficulty, faces difficulty). Information was collected on whether a person had difficulty in dressing, walking across a room, eating, going in or out of bed, bathing and using the toilet, respectively. During the interview, older adults who faced difficulty in any one of the activities for more than three months were categorized as “faces difficulty”. Those who did not have difficulty in any activities were included in the “no difficulty” category.

  15. 15)

    Difficulty in Instrumental Activities of Daily Living (IADL) was coded as (no difficulty, faces difficulty). LASI obtained information on whether an individual had difficulty in the following seven instrumental activities of daily living: preparing a hot meal, shopping for groceries, making telephone calls, taking medications, doing work around the house or garden, managing household finance and getting around or reaching an unfamiliar place. During the interview, those who faced difficulty in any one of the activities for more than three months were categorized as “faced difficulty”. Else they were included in the “no difficulty” category.

  16. 16)

    Covered by any health insurance was coded as (yes, no).

    Further, the household-related characteristics are –

  17. 17)

    Monthly per capita consumption expenditure was coded as (MPCE) quintile of household (poorest, poorer, middle, richer, richest).

  18. 18)

    Religion was coded as (Hinduism, Islam, Others).

  19. 19)

    The place of residence was coded as (Urban, Rural).

  20. 20)

    The country region that a household is situated (Southern, Northern, Central, Western, Eastern, North-eastern). We constructed the country regions by including the erstwhile 29 states and six union territories of India into six categories based on administrative similarity. The northern part includes Chandigarh, Delhi, Haryana, Himachal Pradesh, erstwhile Jammu and Kashmir, Punjab, Uttaranchal and Rajasthan. The north-eastern region includes Arunachal Pradesh, Assam, Manipur, Meghalaya, Mizoram, Nagaland, Sikkim and Tripura. The central area consists of Madhya Pradesh and Chhattisgarh. The eastern zone consists of Bihar, Jharkhand, Odisha and West Bengal. The western region comprises Dadra and Nagar Haveli, Daman and Diu, Goa, Gujarat and Maharashtra. The southern part includes Andhra Pradesh, Karnataka, Kerala, Pondicherry and Tamil Nadu.

Statistical methods

The present study applied descriptive statistics, bivariate and multivariable regression and decomposition analyses to find out the results. Using the chi-square test for association, the bivariate analysis looked at the unadjusted link between major depression and health-related, socio-demographic, and household variables. The multivariate association was presented using average marginal effects (AME). Multivariable models did not violate the assumptions of multicollinearity.

We also applied the nonlinear multivariate decomposition method to determine the role of explanatory factors in the social divide in major depression among older adults in India. The ST-non-ST disparity in major depression is decomposed in both an overall and specific. The divide in major depression is decomposed into an endowments or features (E) component and a coefficients or effects (C) component in the overall decomposition. The contribution of each health-, and socio-demographic-, and household-related explanatory factor is dissected in-depth for each endowment and coefficient component of the difference in major depression among tribal and non-tribal groups. The endowments component (E) shows how difference in major depression would change if the distribution of older adults from the ST and non-ST groups for each category of the explanatory variable were similar. Suppose older adults from ST and non-ST groups were equally likely to fall into one of the explanatory variable’s categories. In that case, the coefficients component (C) indicates the expected change in the disparity. Decomposition models may be used to get various estimates for different reference categories of categorical data. We provide normalized decomposition estimates to solve this “identifying difficulty.” STATA software version 13 was used for all the statistical calculations in this study.

All methods were carried out in accordance with relevant guidelines and regulations.

Results

Descriptive statistics

Table 1 represents the distribution of ST and Non-ST older adults by socio-demographic, health-related and household characteristics in India. About 4.8% and 8.9% of older adults from ST and non-ST social groups had major depression. About 14.2% of older adults from ST category and 27.5% of older adults from non-ST category never worked. About 73.6% and 55.1% of older adults from ST and non-ST categories had no formal education, respectively. Nearly 29.7% and 20.5% of older adults from ST and non-ST categories did not give importance to religion, respectively. About 4.4% of older adults from ST category and 5.2% of older adults from non-ST category were not satisfied with their living arrangements. Almost 17.6% and 23.6% of the older adults from ST and non-ST categories had poor self-rated health, respectively. About 10.7% of older adults from ST category and 25.0% of older adults from non-ST category had multiple chronic conditions. Almost 38.5% and 26.7% of older adults from ST and non-ST categories were physically activity, respectively. Almost 19.5% & 43.6% of older adults from ST category and 23.2% & 48.0% of older adults from non-ST category had difficulty in ADL and IADL, respectively.

Table 1 Distribution of ST and Non-ST older adults by socio-demographic, health-related and household characteristics in India 2017–18
Full size table

Figure 1 reveals the prevalence of depression among older adults by social groups in India. The highest prevalence of major depression was found among older adults from SC (9.9%) followed by OBC (9.0), Others (7.8%) and ST (4.8%) groups.

Fig. 1
figure 1

Major depression status of older adults by social groups in India 2017–18

Full size image

Bivariate analysis

Table 2 depicts the bivariate association between depression and socio-demographic, health-related, and household characteristics of ST and Non-ST older adults in India.

Table 2 Bivariate association between depression and socio-demographic, health-related, and household characteristics of ST and Non-ST older adults in India 2017–18
Full size table

Figure 2 represents the weighted percentage of the total, Non-ST and ST older adults with major depression in India. It was found that in all the age groups, older adults from non-ST groups had higher prevalence of major depression.

Fig. 2
figure 2

Weighted percentage of the total, Non-ST and ST older adults with major depression in India 2017–18

Full size image

Multivariate analysis

Table 3 represents the multivariate association between depression and socio-demographic, health-related and household characteristics of ST and Non-ST older adults in India. The odds of depression were higher among young-old in compairison to the odds of depression among oldest old in non-ST population (OR: 1.51; CI: 1.27–1.80). Among the ST (OR: 1.97; CI: 1.36—2.87) and non-ST (OR: 1.26; CI: 11.13—1.41) population, the odds of depression among currently non-married older adults were higher in reference to the odds to depression among currently married older adults. The likelihood of depression was high among currently working older adults in reference to older adults who never worked among non-ST population (OR: 1.29; CI: 1.10—1.52). The likelihood of depression was high among socially inactive older adilts in comparison to older adults who were socially active (OR: 1.45; CI: 1.24—1.70) in non-ST population. In the non-ST population the odds of depression were higher among older adults for whom religion was very important in comparison to the odds of depression among older adults for whom religion was not important (OR: 1.34; CI: 1.17—1.53). The likelihood of depression was high among older adults who were not satisfied with their current living arrangement (OR: 2.55; CI: 2.17—3.01), received ill treatment (OR: 2.02; CI: 1.70—2.40) and faced discrimination in life (OR: 1.68; CI: 1.49—1.90) in reference to their counterparts in both ST and non-ST population. Among the ST and non-ST population the likelihood of depression was high among older adults with poor SRH, multiple morbidity status and difficulty in ADL. The older adults who were physically active had higher odds of depression in reference to the odds of depression among older adults who were physically inactive in ST (OR: 1.86; CI: 1.24—2.79) and non-ST population (OR: 1.20; CI: 1.06—1.35).

Table 3 Multivariate association between depression and socio-demographic, health-related and household characteristics of ST and Non-ST older adults in India 2017–18
Full size table

Decomposing the caste differentials in depression

Table 4 reveals the overall decomposition of differential in depression among older adults from ST and non-ST groups in India. It was found that differences in effects (C) account for 56.2% of the observed social group differences in the prevalence of major depression. However, 43.8% of the differentials in major depression among ST and non-ST groups was explained by differences in compositional characteristics (E).

Table 4 Overall decomposition of differential in depression among ST and non-ST older adults in India 2017–18
Full size table

Table 5 represents the detailed decomposition (with normalized coefficients) of ST-non-ST differential in depression among older adults in India. Differences due to characteristics (E) is explained at first. It was revealed that if the regional differences were minimized, it would decrease the gap in major depression among ST and non-ST groups by about 19.6%. Similarly, an equal level of self-rated health and chronic conditions among ST and non-ST would result in reduced gap in major depression by almost 9.6% and 7.9%, respectively. Additionally, equal IADL and ADL difficulty among ST and non-ST older adults would result in reduced gap in major depression by about 3.8% and 3%, respectively. Moreover, if health insurance coverage is equalized among older adults from ST and non-ST populations, it would decrease the gap in major depression by about 3.4%.

Table 5 Detailed decomposition (with normalized coefficients) of the differential in depression among older adults in India 2017–18
Full size table

Also, findings from differences due to coefficients (C) revealed that currently married status (8.6%) and being socially active (10.3%) significantly contributed to the ST-non-ST inequality in major depression in this study. Similarly, it was found that if older adults from non-ST population had the same status of ADL as of older adults from ST population, it would decrease the gap in major depression by about 11.8%. Similarly, if older adults from ST group had the same coverage of health insurance as of non-ST group, this would decrease the gap in major depression by almost 7.9%. Additionally, if older adults from non-ST population had same level of physical activity as of older adults from ST population, it would decrease the gap in major depression by about 5.7%.

Discussion

The present study investigated the differentials in the prevalence and determinants of major depression among ST and non-ST populations and decomposed the key socio-demographic, health, and household factors that accounted for the ST/non-ST-based inequalities in major depression among older adults in India by utilizing large-scale nationally representative survey data. The prevalence of major depression in the ST population was lower than the non-ST population (4.8% vs. 8.9%). Multiple studies based on indigenous tribal populations in India reported a lower prevalence of severe depression than in our study [30, 31]. For instance, a study conducted in the tribal community living in the Phek district of Nagaland, India, found that nearly 1.2% of older adults were experiencing depression [30]. Another recent study based on the tribal population of the Mysuru district in Karnataka, India, found a prevalence of severe depression of 3.5% among older adults [31]. In contrast, several other small-scale studies have reported a greater prevalence of depression in tribal older adults in India [32, 33]. The inconsistencies in the prevalence of depression may be partially due to variations in geographical areas, sample measures, and measurement tools.

Results from multivariable analysis demonstrated that physically active older adults had significantly higher odds of major depression than their physically inactive counterparts, which is inconsistent with most of the previous studies [34, 35]. This inconsistent relationship may be due to different definitions of physical activeness in other studies. This study did not find any association between various socio-demographic factors such as age and working status and major depression in the tribal population. On the other hand, in the non-tribal group, increasing age and current working status were significantly positively related to major depression. Moreover, other socio-demographic factors such as gender and educational attainments were unrelated to major depressive disorders in tribal and non-tribal populations.

There are possible explanations for the lower depressive symptoms in the ST population than in the non-ST population. Many studies have reported that tribal populations have a lower level of non-communicable chronic conditions [36], more physically active lifestyle [37, 38], healthy diet and lifestyles [39, 40], and belief in spiritual rituals that can play a significant role in healing and wellness [41]. On the other hand, these behaviors like physical activeness, lesser chronic conditions, healthy diet patterns, active lifestyles, and spiritual orientations have been associated with reduced depressive symptoms [42,43,44,45]. Therefore, it could be possible that such behaviors in tribal populations could effectively enhance mental health.

The decomposition analysis showed that the selected explanatory variables explained nearly 43.8% of the ST/non-ST inequality in major depression in the present study. The major contributors to the explained component were regional differences, self-rated health status, and chronic conditions. Around 19.6% of inequalities in major depression were attributed to regional differences; this is due to the dissimilarities in the distribution of ST and non-ST populations across different regions. For instance, as our data showed, the ST population is majorly concentrated in the north-eastern region, while a greater proportion of the non-ST population is concentrated in the northern and western regions of India. Differentials in self-rated health contributed to nearly 10% of the caste inequalities in the prevalence of major depression, resulting from the fact that a greater proportion of older adults in the non-ST group reported poor self-rated health than older adults in the ST group (23.4% vs. 14.9%). These findings are consistent with the other studies suggesting that poor self-rated health [46,47,48] is linked to having major depressive disorders in older persons.

Importantly, age, marital status and social participation contributed significantly to the ST-Non-ST differentials in major depression which may imply the greater buffering effects of increasing age, presence of structural support and participation in social activities on reducing major depression among older adults from ST category than those from non-ST group. Consistent with our findings, numerous studies have reported that widowhood and lack of social support can elevate stress, therefore, leading to higher levels of depression [49,50,51,52].

Nearly 8% of the inequalities in major depression were attributed to the differentials in the prevalence of chronic conditions among ST and non-ST populations; this is because a greater proportion of older adults in the non-ST group reported having multiple chronic conditions than older adults in the ST group (26.8% vs. 13.9%). Although studies related to chronic conditions in the tribal population in India are limited, the available studies reported a severe lack of awareness of chronic diseases among the tribal population in India [53]. However, in comparison to the general population, tribal population are expected to have lesser chances of cardiovascular diseases as they are physically active and are more prone to have plant-based foods. Another important contributor to the ST-non-ST differential in major depression was ADL and IADL functional difficulty. The findings are also in concordance with previous studies on the linkage between multiple chronic conditions [54, 55], functional limitations [56] and major depression.

There are some strengths and limitations of this study worth mentioning. The utilization of a large sample from a national representative survey in deriving the estimates is the potential strength of the study. An additional strength of the study is using the CIDI-SF scale for deriving reliable and valid estimates for major depression. There are potential limitations to the study. Firstly, our explanatory variables, such as self-reporting of the diagnosed chronic conditions, can be under-reported either due to unawareness about the chronic condition among the older adults or by recall bias. Moreover, the responses to the variables like self-rated health may be biased because individuals use response scales in systematically different ways. Finally, our results are based on the cross-sectional data; therefore, establishing a causal direction of the association of explanatory variables with major depression is not possible.

Conclusions

The present study found a greater prevalence of major depression in older adults in the ST group than in the non-ST group. A greater percentage of the ST/non-ST inequality in major depression, advantageous to tribal population in the present study was explained by age, marital status, and social participation, suggesting further research on the coping mechanisms of mental illnesses among the indigenous population in India. The ST-non-ST differentials in the major depression were also attributed to regional differences, self-rated health status, and chronic conditions, which need to be explored in future research.