Examining fine-scale tribal health inequalities around three forested sites in India: Results of a cross-sectional survey

doi:10.21203/rs.2.21892/v1

PDF

Research

Examining fine-scale tribal health inequalities around three forested sites in India: Results of a cross-sectional survey

https://doi.org/10.21203/rs.2.21892/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted 27 Jan, 2020

You are reading this latest preprint version

Background

The data available for the health of Scheduled Tribes (ST) in India are often coarse-scale snapshots of health status and healthcare access showing poorer indicators when compared to others but do not allow fine-scale analysis. In this paper, we examined health inequalities between ST and non-ST populations in two forested sites and compare healthcare parameters for ST populations across three forested sites in India.

Methods

We conducted a cross-sectional household survey in sites in and around Biligiri Ranganathaswamy Temple (BRT), Kanha and Pakke tiger reserves in the south, central and northeast regions of India respectively. In each site, multi-stage sampling and cluster analysis provided a representative sample of households across villages of 859 ST and non-ST households. We examined sociodemographic and health-related information including self-reported illnesses and healthcare utilisation; from these we explored within-site health inequality patterns for two sites and inter-site differences among ST households in the three sites.

Results

In BRT, ST-non-ST differences favoured the latter for socio-economic characteristics with no difference in self-reported illness/injuries or healthcare utilisation. In Kanha, both groups were similar for socio-economic characteristics and healthcare utilisation. Pakke ST households reported highest healthcare utilisation, while Kanha ST households reported lowest care seeking at hospitals and relied on home networks and health workers. High tobacco consumption prevalence was noted among ST households in all sites.

Conclusions

ST-non-ST inequality patterns at a fine-scale were different between BRT and Kanha. Absence of health inequalities between both groups in Kanha indicates uniform socio-geographical disadvantage while poor healthcare utilisation by ST in BRT indicates health inequities within the state. Pakke ST households reported highest utilisation while Kanha reported the lowest. Thus, programmes addressing health inequalities of STs need to consider site-specific assessments in socio-geographical and health system factors.

Health Policy

Forests

health inequalities

healthcare access

tribal health

indigenous health

tiger reserves

India

India’s tribal population of 104 million people (the second highest in the world) have poor health indicators as compared to other social categories.[1–4] Mortality among children under five (tribal population: 57.2 deaths under-5 per 1000 live births, others: 38.5), malnutrition (tribal population: 42.3% stunting among children under 5 years, others: 33.9%) and many other health indicators related to healthcare access and immunisation (tribal population: 55.7% full immunisation, others: 71.6%) are worse-off among tribal populations in India.[3, 5–7] In fact, these nation-wide patterns mirror findings from the largest global study of indigenous populations where tribal populations had worse-off indicators such as Infant Mortality Rate (IMR, 4.5 times higher), maternal mortality ratio (3.6 times higher) and proportion of children under-five with stunting (2.4 times higher).[1] While acknowledging the importance of the question of why tribal communities are “behind everyone, everywhere” with respect to health status and healthcare access, Anderson and colleagues (2016) call for a more granular analysis of health of indigenous communities.[1, 8]

The term Scheduled Tribes (hereafter ST; derived from Article 342 of the Indian Constitution) is a common identifier for a heterogeneous group of tribal communities in India.[9–10] This constitutional specification of STs aggregates distinct and diverse genetic, ethnic, cultural and social populations.[11–12] This includes 705 notified ST communities, with each Indian State notifying its own ST list. In the general landscape where our study was conducted, south Indian states had lowest percentages of STs (7% in Karnataka) compared with Central India (21% in Madhya Pradesh) while STs were a majority in Northeast India (69% in Arunachal Pradesh).[13]

The main source of information about the health status of ST population is from the National Family Health Surveys (NFHS), nationwide surveys across a representative sample of households. The four rounds of NFHS surveys till date are at a relatively coarse-scale that allow for comparison across ST and other broad social categories.[5, 12] The recent NFHS survey (2015-16) found that STs had the lowest institutional delivery rates (ST: 68%, national average: 80%), lowest full immunisation rates (ST: 56%, national average: 62%), highest stunting (ST: 44%; national average: 38%) and wasting (ST: 45%, national average: 36%) when compared to others.[7] Even in these surveys, there are design constraints for assessing population health of particular ST communities within a district or state. Furthermore, comparisons across tribal communities in different states/regions of the country are not possible.[5, 12] Other sources of information on health of STs are national surveys conducted for specific health problems and reports released by the Indian government’s Ministry of Health & Family Welfare (Rural Health Statistics for instance) or Ministry of Tribal Affairs (Statistical Profile of Scheduled Tribes 2010 for instance). These provide snapshots that allow state-level disaggregation of various health related parameters with limited or nil local-, district- or tribe-level information.[5, 10, 12, 14] The paucity of disaggregated data on health status and health care utilisation of tribal populations in India is repeatedly stated in different tribal health reports, while noting that even when available, disaggregated data is limited in information and dated.[5, 12]

The need for granular ST data is because of social and cultural heterogeneity across ST groups as well as the varying geographical landscapes that they live and depend on.[5, 12] A close association with forests is seen across most ST populations. For instance, the Forest Survey of India 2017 revealed that 215 districts with a relatively higher tribal population had an average of 37% forested area when compared with the nationwide forest cover average of 21%.[15] In Northeast India with large areas under forest cover, STs are a majority in six of the seven states (75–90% in some states).[10] Thus, it is important to understand healthcare access of STs in relation to the forested landscape they live in, in addition to other social determinants of health.[16] Furthermore, the history of several ST communities is closely aligned with struggles for land rights and access to forest produce to sustain livelihoods, and their overarching social effects are linked with overall socio-economic and political disadvantage.[5, 17, 18] Access to forests for livelihood and secure land tenure differ from one area to another in India, and forest regimes and ST identity in northeast India are well known to be different from that in central and south Indian forested areas.[19, 20] The National Health Policy 2017 also acknowledges the challenges faced by ST communities as being geographical and infrastructural and calls for situation-specific reforms in health service delivery although it does not convert this into any specific strategy or reform.[21]

In this study, we explored the health of select households in forested landscapes in three states, each in a different region of the country, namely Arunachal Pradesh, Madhya Pradesh and Karnataka. These states significantly vary with respect to their ST populations as evident in Table 1. Arunachal Pradesh has the highest proportion of ST population with relatively better-off indicators with an IMR comparable to the national average; while Karnataka on the other end of the spectrum has the lowest proportion of ST population with the lowest IMR but relatively poorer social indicators among its ST population (Table 1). State-level disaggregated data reveals that health inequalities between ST and non-ST varying from state to state with a significant gap in mortality rates among children under-five in Madhya Pradesh as compared to Karnataka (Fig. 1).

Table 1

A profile of ST population indicators in the three study states (Arunachal Pradesh, Karnataka and Madhya Pradesh) compared with national averages [5, 12]
States	ST population proportion	Education till secondary school	Households using clean cooking fuel^	Infant Mortality Rate^^	Sources for treatment seeking
States	ST population proportion	Education till secondary school	Households using clean cooking fuel^	Infant Mortality Rate^^	Public	Private
Arunachal Pradesh	69%	34.5%	23.9%	67.6	96.5	2.5
Karnataka	7%	24.5%	14.3%	45.8	47.2	51.8
Madhya Pradesh	21%	14.2%	3.6%	95.6	59.9	38.8
India	9%	21.9%	9.5%	62.1	77.3	20.2
Source	Census 2011	NSSO 68th 2011-12	Census 2011	NFHS 3 2005-06	DLHS 2007-08
^Clean cooking fuel includes PNG/LPG, electricity, biogas ^^infant deaths per 1000 live births

In this paper, we examine health inequalities between ST and non-ST population living in the same area in forested landscapes in south and central India. Further we compared various healthcare parameters for ST populations in three forest areas in three regions of India based on data obtained from a larger study that examined current and future correlates of forest dependence in four Indian forest areas, and also focused on collecting select health-related parameters at household level across ST and non-ST communities.[22]

Study setting and design

The study was conducted among communities living around four Indian tiger reserves (a class of Protected Areas with the highest degree of restrictions on human activities) namely Kanha (Madhya Pradesh), Pakke (Arunachal Pradesh), Biligiri Ranganathaswamy Temple (BRT, Karnataka) and Corbett (Uttarakhand).[22] In this paper, we examined a subset of data on healthcare parameters from households identified using multi-stage cluster sampling in three of its sites excluding Corbett (Fig. 2). The ST communities living around these three sites are mainly the Gonds and Baigas around Kanha, the Nyishis, Akas and Puroiks around Pakke (six other STs live in the area but in smaller numbers) and Soligas in and around BRT. The non-ST communities in these sites were Pawar, Marar, Lodhi and Yadavs around Kanha, temporary migrant populations from outside the state in Pakke albeit in small numbers, and Dalits, Upparas, Lingayats, and Brahmins around BRT.

These tiger reserves span different management histories and residents use these forests in different ways. Kanha is one of India’s oldest tiger reserves (declared in 1974) while Pakke and BRT were declared as tiger reserves in 2002 and 2011 respectively. The three sites vary in the history and intensity of relocation efforts.[23] ST population in all three study sites rely to varying extent on extraction of non-timber forest products in addition to firewood and livestock grazing albeit with important local differences in access to forests for ST. The Scheduled Tribes and Other Traditional Forest Dwellers (Recognition of Forest Rights) Act, 2006 which recognises land and occupation rights has been implemented to a greater extent in BRT (compared to other sites) where 25 gram sabhas (a unit of local level governance) received community forest rights.[24] These sites have diverse socio-economic and cultural settings, with differing relationships between communities and forests, and policies and practices with respect to forest access and management. A detailed summary for each study site and its implications on conservation efforts is published elsewhere.[22]

The study design was cross-sectional, and we examined the socio-demographic and healthcare-related data for 859 households from the three sites (BRT: 329, Kanha: 322, Pakke: 208). Our sampling frame for STs composed 41% of households in BRT, 70% in Kanha and 95% in Pakke. For each site, we selected villages (from the 2011 Census of India) using a cluster analysis (with distance to tiger reserve boundary, forest, road, town or city) from five clusters (> 75% of the variance was captured by the cluster analysis). After our pilot survey, we added new villages that were not on our initial map and/or we replaced villages from the same cluster that could not be found. Using a random start point in the village we sampled 8 to 12 households per village where we walked in different cardinal directions in order to sample three to four households in each direction. In smaller villages we sampled nearly all households and while this may have resulted in clustering within a village, we expect that within-site across village bias would be minimised with our differing random start point for each village. Our households were represented by various demographic profiles that spanned different tribes, education levels, land-holding sizes, household occupations (Table 2).

Table 2

Table 2. Socio-economic and socio-political characteristics of ST and non-ST households within BRT and Kanha and for ST alone in Pakke (Proportion of households (unless otherwise indicated) with 95% CI indicated in parentheses, significant findings indicated by*with 95% CI not overlapping)
Household characteristics		BRT		Kanha		Pakke
Household characteristics		ST	Non-ST	ST	Non-ST	ST
No. of households studied (within-site proportion)		136 (41%)	193 (59%)	225 (70%)	97 (30%)	208
Average household size		3.9 (3.6–4.2)	4.4 (4-4.7)	5 (4.8–5.3)	5 (4.6–5.3)	6.4 (6.1–6.8)
Average no. of dependents^#		1.3 (1.2–1.5)	1.6 (1.4–1.8)	1.8 (1.6–1.9)	1.9 (1.7–2.2)	4 (3.7–4.3)
Firewood only as cooking fuel		71% (63–78)*	34% (27–40)*	89% (85–93)	82% (74–90)	34% (27–40)
Water-source at or near home		54% (45–62)	68% (61–76)	3% (1–5)	4% (2–7)	76% (71–83)
Toilet present in house		36% (28–44)*	61% (55–68)*	38% (32–44)	47% (38–58)	82% (76–87)
Average asset count per household^		2.9 (2.7–3.3)*	5.1 (4.9–5.5)*	2.6 (2.4–2.8)	2.7 (2.3-3)	5.2 (4.9–5.5)
Substance use in past week by any household member	Tobacco (in any form)	64% (56–72)*	35% (28–41)*	89% (85–93)*	73.% (64–82)*	67% (60–73)
Substance use in past week by any household member	Alcohol	26% (19–33)	26% (20–33)	53% (47–60)	41% (31–51)	38% (31–45)
At least one member migrated for work in past three months		13% (7–19)	8% (5–13)	42% (36–48)	36% (26–46)	61% (54–68)
Spend > 50% income on food		75% (68–82)*	88% (83–92)*	72% (66–78)	68% (59–77)	37% (31–44)
Average no. of income setbacks in past year^^		3.5 (3.2–3.9)	3.1 (2.9–3.3)	5.4 (5.2–5.5)	5 (4.7–5.3)	5.4 (5.1–5.7)
Average no. of schemes utilised in past year^^^		2.5 (2.3–2.7)	2.4 (2.2–2.6)	2.3 (2.2–2.4)	2.2 (2-2.4)	4.1 (4-4.2)
Participation in village or higher politics by any member		42% (34–50)*	18% (12–23)*	3% (1–5)	8% (3–14)	24% (19–30)
^#non-income earning household members ^Sum of 18 different assets for each household including different livestock ^^15 income setbacks were assessed including environmental, social or other relevant setbacks ^^^9 government schemes including ration card, various pensions, Rashtriya Swasthya Bima Yojana [22]

Data collection

Prior to the survey, we conducted field visits to each site to establish field teams and develop the questionnaire through small pilots. The survey was conducted in three languages (Hindi, Assamese and Kannada), each corresponding to the most widely spoken language in that site. Subsequently, we conducted training sessions for data collectors and finally data collection was completed between December 2016 and September 2017. We conducted household surveys using a structured questionnaire with visual aids. The questionnaire collected information on various themes, of which those related to health were: (1) the nature of healthcare utilisation and illnesses: self-reported illnesses, the point of first-care, treatment of minor ailments, hospitalisation, maternal care and preferences for health services; and (2) how these behaviours and preferences are related to socio-demographic characteristics and tribal and non-tribal affiliations. Data for consumption of tobacco and alcohol were collected for the past week, minor ailments for the past month, and hospitalisation for the past year. We interviewed either the household head or any adult who was willing to speak on behalf of the household following verbal consent (51% were females). Our interviews typically lasted between 40–60 minutes. Data was entered into a spreadsheet at each site, and all three data spreadsheets were checked for errors and merged to create a master dataset. Further details of data collection along with the questionnaire with variables are published elsewhere.[22]

In each site, the researchers coordinated with local community representatives to inform them of the research and sought permissions to pursue this research. In all three sites, members of local communities were involved in piloting the tools and were the primary respondents of the survey. The data collection teams in the three sites comprised of members from the local communities, who were trained in administering the survey tools.

Data analysis

We used Program R (R Core Team 2017) for cluster analysis and SPSS statistics (version 23) for data analyses. We classified the households into ST and non-ST categories, and analysed household and health-related characteristics across these two categories within each site for BRT and Kanha only (BRT ST: 136 non-ST: 193, Kanha ST: 225 non-ST:97). In Pakke due to the low proportion of non-ST household presence, only the ST household data was included (only 10 non-ST households were identified and not included due to the low number). Comparisons of ST households were conducted across the three sites: BRT, Kanha and Pakke (BRT: 136, Kanha: 225, Pakke: 208). The results of the study are presented in this order. We estimated bootstrapped means and 95% Confidence Intervals (CI, based on 1000 bootstrap iterations). We inferred significance in differences when CIs were non-overlapping.

Health inequalities within the site: BRT and Kanha

The pattern of inequality between ST and non-ST households varied between BRT and Kanha. We report below the specific differences.

In BRT, ST households, (41% of total households in the site), were more dependent on firewood (ST: 71%, non-ST: 34%), less wealthy (ST: 2.9, non-ST: 5.1 average asset count per household), less likely to have toilets in their houses (ST: 36%, non-ST: 61%) and had higher overall tobacco use (ST: 64%, non-ST: 34.7%), compared to non- STs (Table 2; all differences based on non-overlapping CIs). However, relatively lesser BRT ST households reported spending > 50% income on food (ST: 75%, non-ST: 88% households) and reported higher political participation (ST: 42%, non-ST: 18%) when compared to non-STs (Table 2; all differences based on non-overlapping CIs). Overall, the ST-non-ST differences in BRT (unlike in Kanha; see below) were in favour of non-ST communities.

Despite differences in various household characteristics, there were no differences in BRT between ST and non-ST households for self-reporting minor ailments, infectious diseases or serious injuries. Only reporting of Non-Communicable Disease (NCD) was higher among non-STs than for ST households (11% and 3% respectively with non-overlapping 95% CIs)(Table 3). Healthcare utilisation was similar for both groups of households. The first point of care for minor ailments was either the health worker (46% ST 44% non-ST) or hospital (36% ST 43% non-ST), with half (46% ST 60% non-ST) approaching the primary health centre for care eventually (Table 3). Most hospitalisations among ST households (58%) were in public hospitals as compared to those among non-ST households (73%) in private hospitals though this was not statistically significant.

Table 3

Health-related characteristics and healthcare utilisation of ST and non-ST households within BRT and Kanha and for ST alone in Pakke (Proportion of households (unless otherwise indicated) with 95% CI indicated in parentheses, significant findings indicated by*with 95% CI not overlapping)
Health related characteristics		BRT				Pakke
Health related characteristics		ST	Non-ST	ST	Non-ST	ST
Minor ailments reported in past month		51% (42–59)	40% (33–47)	19% (14–24)	10% (5–17)	34% (28–41)
First point of care for minor ailment	Home/friends	17% (9–26)	13% (6–21)	67% (53–81)	40% (9–75)	10% (4–17)
	Health worker	46% (34–58)	44% (33–56)	33% (19–47)	60% (25–91)	0
	Hospital	36% (25–48)	43% (31–53)	0	0	90% (83–97)
Type of hospital eventually visited for minor ailment^$	Primary health centre (public)	46% (34–59)	60% (48–71)	7% (0–16)	20% (0–50)	31% (21–42)
	Other public hospital	4% (0–10)	17% (9–25)	19% (8–30)	50% (17–80)	44% (32–55)
	Private hospital	38% (26–50)	18% (10–28)	7% (0–16)	0	25% (16–36)
	Did not visit hospital	12% (5–19)	4% (0–8)	67% (54–81)	30% (0–63)	0
Infectious diseases reported in past year (malaria/tuberculosis)		9% (4–14)	5% (2–8)	5% (2–8)	6% (2–11)	12% (8–17)
NCDs reported in past year (diabetes mellitus/hypertension/cancer)		3% (1–6)*	11% (7–16)*	0.9% (0–2)	0	6% (3–10)
Serious injury (including snake bite)		1% (0–3)	2% (1–4)	1% (0–3)	0	1% (0–3)
Household with a childbirth in past year		12% (7–17)	6% (3–10)	7% (4–10)	9% (4–16)	7% (4–10)
Hospital admission(s) reported in past year		24% (17–32)	23% (18–29)	10% (6–14)	14% (8–22)	33% (26–39)
Type of hospital utilised for admission^$$	Public health centre	58% (39–74)	27% (14–40)	82% (64–96)	53% (29–80)	54% (42–66)
Type of hospital utilised for admission^$$	Private health centre	42% (26–61)	73% (61–86)	18% (4–36)	47% (20–71)	42% (30–54)
^$Details for 1 non-ST household in BRT not available ^$$Not available for 3 ST households in Pakke

In Kanha, the ST households (70% of all households in the site) appeared to be similar to non-ST households in terms of all studied household characteristics including household size, firewood dependence, average asset count per households to name a few, except for tobacco use (Table 2). ST households in Kanha reported significantly higher tobacco use as compared to non-STs (ST: 89%, non-ST: 73% with non-overlapping CIs). The Kanha ST and non-ST households were similar with respect to self-reported minor ailments and hospitalisations in the past year as well. We find that care-seeking at hospitals was never the first choice in case of minor ailments for both ST and non-ST, and eventually both tend to seek care at public hospitals for hospitalisations too. The absence of significant inequalities between ST and non-ST households in Kanha is in contrast to BRT, although absolute health and household parameters indicate a more uniform disadvantage across ST and non-ST households across Kanha.

Variations in healthcare parameters of ST households across the three sites: BRT, Kanha and Pakke

From the literature review, we expected ST households to vary significantly between sites, but the patterns seen in the results are significantly different. Pakke ST households had a larger household size (average 6.4 Pakke vs 3.9 BRT 5 Kanha) with more dependents (average 4 Pakke vs 1.3 BRT 1.8 Kanha) and higher migration (61% Pakke vs 13% BRT 42% Kanha). However, they had better-off indicators for most household characteristics when compared with ST households at BRT and Kanha. For instance, most Pakke ST households reported a toilet at home (82% Pakke vs 36% BRT 38% Kanha), water-source at or near home (76% Pakke vs 54% BRT 3% Kanha) with low firewood dependence (34% Pakke vs 71% BRT and 89% Kanha), high average asset count (5.2 Pakke vs 2.9 BRT 2.6 Kanha) and lowest reports of spending > 50% income on food (37% Pakke vs 75% BRT 72% Kanha) when compared to BRT and Kanha ST households.

Kanha on the other hand reported the highest firewood dependence (mentioned earlier), poorest availability of water-source at or near home with highest reported alcohol (53% Kanha vs 26% BRT 38% Pakke) and tobacco consumption (89% Kanha vs 64% BRT 67% Pakke) when compared to BRT and Pakke ST households (Table 2). BRT ST households were at the middle of this spectrum though for a few household characteristics, they were comparable to Kanha ST households as for instance, the presence of toilet in house (36% BRT 38% Kanha), average asset count (2.9 BRT 2.6 Kanha) and relatively higher proportions of spending > 50% income on food (75% BRT 72% Kanha)

With respect to the healthcare-related characteristics of the ST households in the three sites, differences were noted for self-reporting of minor ailments, hospitalisations and healthcare utilisation (Table 3). The highest self-reporting of minor ailments was seen among BRT ST households (51%) and lowest in Kanha (19%). The first point of care for BRT households was a health worker closely followed by hospital (46% and 36%). In Kanha however, home remedies were the first choice (67%) with none going to a hospital while the opposite was seen in Pakke with 90% of households with minor ailments approaching hospitals immediately. In Kanha 67% with minor ailments never went to any hospital (Table 3, Fig. 3), and for those who did make it to hospitals in the other sites, 75% Pakke and 50% BRT went to public hospitals. Coming to hospitalisations, the lowest was reported among Kanha ST households (10% Kanha vs 24% BRT 33% Pakke) and 82% of them went to public hospitals. More than half of BRT and Pakke hospitalisations were reported at public hospitals too (58% and 54% respectively) (Fig. 3).

There are few studies that examine ST and non-ST inequalities in a given landscape.[25–28] Rather, most studies (such as those based on NFHS data) allow for ST-non-ST comparisons at state or higher levels (up to districts from NFHS-4 onwards).[4, 29] By sampling in particular forested landscapes, our study enabled a finer scale comparison within sites at a local level. This allows us to examine if national- and state-level ST-non-ST differences persist at such scales as well, given that both ST and non-ST communities that we sampled face comparable geographical and social disadvantages associated with living in or around forests. The persistence of ST-non-ST differences in our samples could help deepen our understanding of the drivers of ST inequalities. The picture of inequalities when examined at a finer scale were significantly different than those reported by NFHS surveys, and the contrast between ST-non-ST inequality patterns in BRT and Kanha is striking. In the latter site, geographical and social disadvantages and their healthcare utilisation appeared to be distributed across households irrespective of ST status, indicating factors beyond ST identity driving overall socio-economic and health indicators.

The inter-site differences in socio-demographic and health-related characteristics of ST households highlight the variation in ST populations in relation to local context and state health systems. In terms of overall state-level ST health indicators, the IMR for Karnataka ST population is better-off (45.8 per 1000 live births), followed by Arunachal Pradesh (67.6 per 1000 live births), with Madhya Pradesh (95.6 per 1000 live births); with a similar pattern for several other health outcomes (Table 1). However, when we compare health services’ infrastructure and human resources in ST areas as per Rural Health Statistics 2014-15, we find a mixed picture.[14] Arunachal Pradesh reports adequate numbers of subcentres and Primary Health Centres (PHCs) with 43% shortage of female health workers in subcentres and only 13% doctor vacancies in PHCs; Madhya Pradesh needs to increase its subcentres by 38% but has female health workers in excess with only 10% doctor vacancies at PHC; and Karnataka however needs to increase its subcentres by 72% numbers with 25% female health workers shortage and 40% doctor vacancies in PHCs.[14] In our study, accessing health services was highest in Pakke and lowest in Kanha despite having comparable services in place as per state-level information (Fig. 3). Access to health services was relatively low for BRT in comparison with Pakke reconfirming that services in tribal areas are lacking in Karnataka despite having a better-off health services distribution among the three states.[30] Hence, with widespread improvements in availability of health services infrastructure through the National Health Mission in states like Karnataka, efforts to address tribal health inequalities at district and sub-district level need to focus on the fine-scale patterns of healthcare access to ST communities in the state.

The high prevalence of tobacco and alcohol consumption among tribal populations is documented in many reports and studies.[5, 10, 12, 31] In our study too, we found this to be significantly higher among ST households in BRT and Kanha when compared to non-ST. It must be noted that in Kanha the prevalence was high for both groups as compared to other sites with the highest prevalence across all sites reported in Kanha ST households (89%). In addition, the relatively high prevalence of alcohol consumption (53% ST 41% non-ST) was noted in the site. Excess consumption of tobacco and alcohol’s contribution to increased mortality is known and at this high prevalence, their contribution to the adverse health in the form of various NCDs and perhaps excess mortality can be inferred though this needs to be studied to recognise its impact on these communities.[31–33] This reflects an urgent and important gap for health services in tribal areas to address from individual-targeted deaddiction services to community-level awareness and health promotion programmes.

The health parameters chosen in the study were a subset of a larger research study and were limited in their scope by covering only self-reported conditions and utilisation of health services, as a proxy for overall healthcare access, and is not expected to be a comprehensive assessment of population health of these communities. Yet, in the lack of fine-scale data on tribal health, this provides an insight into within-group differentials and extent of variation across sites. Increasing the number of indicators, covering more sites and incorporating historical and qualitative inquiries on specific sites could improve the potential for theorizing the drivers of inequalities among populations in and around forest areas.[34]

Overall, our study revealed patterns of inequalities at different sites but there is a need to explain the fine-scale drivers of inequalities taking into account the local socio-economic and health system factors, including the geographical and environmental factors related to living in and around protected forest areas. The contrast between the inequality patterns in Kanha and BRT highlight the importance of studying state and local health system factors in explaining tribal health inequities. For a state like Karnataka, the poor healthcare utilisation of ST in BRT indicates health inequities within the state and/or the contribution of local environmental factors that differentially affect tribal communities. The pattern seen in Kanha though appears to be a function of overall health services and system performance in the state and wider geographical/landscape level factors, that appear to affect both ST and non-ST communities.

BRT – Biligiri Ranganathaswamy Temple

CI – Confidence Interval

IMR – Infant Mortality Rate

NFHS – National Family Health Surveys

NSSO – National Sample Survey Organisation

NCD – Non-Communicable Disease

PHC – Primary Health Centre

ST – Scheduled Tribe

ETHICS APPROVAL

Ethics approval for this study was granted by the Institutional Review Board at Columbia University to NDV [IRB-AAAR2467]

CONSENT FOR PUBLICATION

Not applicable

AVAILABILITY OF DATA AND MATERIALS

Deidentified participant data is available on request from Nandini Velho ([email protected]).

COMPETING INTERESTS

None declared

FUNDING

This work was supported by the Earth Institute Fellows Program at Columbia University [to NDV] and the Wellcome Trust/DBT India Alliance fellowship [grant number IA/CPHI/16/1/502648 to PNS; NSN is a research staff under this fellowship].

AUTHORS’ CONTRIBUTIONS

NDV conceptualised the study and conducted the larger research across the different sites. PNS contributed to the design of the health-related section in the research. TS coordinated the study in BRT tiger reserve. NSN initiated the analysis of the health section and wrote the first draft of this paper. TS completed the analysis with support from PNS and NDV and wrote the subsequent versions of the paper with significant contributions from all the co-authors. All authors read and approved the final manuscript.

ACKNOWLEDGEMENTS

We thank Aditi Patil, Arpit Deshmukh, Neyi Jamoh, Prachi Kardam, Santosh Sogal, Jaishree Subrahmaniam for their inputs in planning, implementation, data collection and management efforts. Mahadevi, Shivamma, Periswamy, Mahadevaiah, Shivamallu, Pandegowda, Yamuna Markam, Mahendra Bisen, Raju Khan, Kepu Riba, Masem Tachang, Pahi Tachang, Pema Tacho, Sarsomi Degio, Mize Degio and numerous others provided invaluable field support. We thank the Vivekananda Girijana Kalyana Kendra and the Zilla Budakkattu Girijana Abhivrudhhi Sangha and its taluk counterparts, village leaders, mukhyas, gaon burrahs, Tana Tapi, Sanjay Shukla, Rakesh Shukla, Kime Rambia and other forest department staff and officers . We thank Benjamin Clark, Meghna Agarwala, Rita Banerji, Imrana Khan and Umesh Srinivasan for their support.

1. Anderson I, Robson B, Connolly M et al. Indigenous and tribal peoples’ health: a population study. The Lancet 2016;6736(16): 1–27 doi:10.1016/S0140-6736(16)00345-7.

2. Balarajan Y, Selvaraj S, Subramanian SV. Health care and equity in India. The Lancet 2011;377(9764):505–15 doi:10.1016/S0140-6736(10)61894-6.

3. Das MB, Hall GH, Kapoor S et al. India: The Scheduled Tribes. In Hall GH & Patrinos HA eds, Indigenous peoples, poverty, and development. Cambridge: Cambridge University Press 2012;205-48 doi:10.1017/CBO9781139105729.006.

4. Maity B. Comparing health outcomes across Scheduled Tribes and Castes in India. World Development 2017;96:163-81.

5. Ministry of Tribal Affairs. Report of the high-level committee on socio-economic, health and educational status of tribal communities of India. New Delhi 2014. http://www.indiaenvironmentportal.org.in/files/file/Tribal%20Committee%20Report,%20May-June%202014.pdf. (accessed 24 June 2019).

6. United Nations Children’s Fund (UNICEF). Nourishing India’s tribal children: The nutrition situation of children of India’s scheduled tribes. New Delhi: UNICEF 2014. http://unicef.in/Uploads/Resources/Tribal-low-res-for-view.pdf (accessed 13 September 2019).

7. International Institute for Population Sciences (IIPS) and ICF. National Family Health Survey (NFHS-4), 2015-16: India. Mumbai: IIPS. 2017. International Institute for Population Sciences (IIPS) and ICF. National Family Health Survey (NFHS-4), 2015-16: India. Mumbai: IIPS 2017 (accessed 13 September 2019).

8. Stephens C, Porter J, Willis R, et al. Indigenous people’s health-why are they behind everyone, everywhere? The Lancet 2005;366:10–13 doi:10.1016/S0140-6736(05)66801-8.

9. Chandramouli C. Scheduled Tribes in India as revealed in Census 2011. New Delhi 2013. https://www.tribal.nic.in/ST/3-STinindiaascensus2011_compressed.pdf. (accessed on 24 June 2019).

10. Ministry of Tribal Affairs. Statistical profile of Scheduled Tribes in India. New Delhi 2010. http://14.139.60.153/bitstream/123456789/2563/1/Statistical%20Profile%20of%20Scheduled%20Tribes%20in%20India%202010.pdf (accessed on 13 September 2019).

11. Watkins WS et al. Diversity and divergence among the tribal populations of India. Annals of human genetics 2005;69(6):680–92 doi:10.1046/j.1529-8817.2005.00200.x.

12. Ministry of Health & Family Welfare and Ministry of Tribal Affairs. Report of the expert committee on tribal health: Tribal health in India: Bridging the gap and a roadmap for the future. New Delhi 2018. http://tribalhealthreport.in/full-report/ (accessed on 12 September 2019).

13. Registrar General & Census Commissioner, India. Primary census abstract for total population, Scheduled Castes and Scheduled Tribes 2011. http://censusindia.gov.in/pca/. (accessed on 13 September 2019).

14. Ministry of Health & Family Welfare. Rural health statistics. New Delhi 2015. https://wcd.nic.in/sites/default/files/RHS_1.pdf. (accessed on 13 September 2019)

15. Durlauf SN. Neighbourhood effects. In Henderson JV and Thisse JF eds. Handbook of Regional and Urban Economics. Volume 4. Cities and Geography. Amsterdam 2003:2173–242.

16. Ministry of Health & Family Welfare. National Health Policy. New Delhi 2017. https://mohfw.gov.in/sites/default/files/9147562941489753121.pdf (accessed on 24 June 2019).

17. Mohanty BB. Land distribution among scheduled castes and tribes. Economic and Political Weekly 2001;36(40):3857-68.

18. Bose P, Arts BJM, van Dijk H. 'Forest governmentality': A genealogy of subject-making of forest-dependent 'scheduled tribes' in India. Land Use Policy 2012;29(3):664-673. https://doi.org/10.1016/j.landusepol.2011.11.002 (accessed on 24 June 2019).

19. Kothari A, Pathak-Broom N. Conservation and rights in India: Are we moving towards any kind of harmony? In: Radhakrishna M, ed. First citizens: Studies on Adivasis, tribals and indigenous peoples in India . Washington, DC: Oxford India Studies in Contemporary Society 2016:337-369.

20.Xaxa V. Formation of Adivasi/indigenous peoples' identity in India. In: Radhakrishna M, ed. First citizens: Studies on Adivasis, tribals and indigenous peoples in India . Washington, DC: Oxford India Studies in Contemporary Society 2016:33-52.

21. Forest Survey of India. State of forest report 2017. Dehradun 2017. http://fsi.nic.in/details.php?pgID=sb_64. (accessed on 13 September 2019).

22. Velho N, DeFries RS, Tolonen A et al. Aligning conservation efforts with resource use around protected areas. Ambio 2019;48 (2):160-171 doi:10.1007/s13280-018-1064-5.

23. Lasgorceix A, Kothari A. Displacement and relocation of Protected Areas: A synthesis and analysis of case Studies. Economic & Political Weekly 2009;49:37-47.

24. Tatpati M, Pathak-Broome N. Asserting community forest rights on forest land in India: Emerging paradigms under the Forest Rights Act of India. Chapter 13. Land rights in India: Politics, movements and challenges. Shimla: Indian Institute of Advanced Study 2016.

25. Desai G, Anand A, Modi D et al. Rates, indications, and outcomes of caesarean section deliveries: A comparison of tribal and non-tribal women in Gujarat, India. PLoS ONE 2017;12(12): e0189260. https://doi.org/10.1371/journal.pone.0189260 (accessed 24 June 2019).

26. Nandi S, Sinha D, Joshi D, et al. Evaluation of the Janani Shishu Suraksha Karyakram: Findings on inequity in access from Chhattisgarh, India. BMJ Global Health 2016;1:A4.

27. Chatterjee K, Sinha RK, Kundu AK, Shankar D, Gope R et al. Social determinants of inequities in under-nutrition (weight-for-age) among under-5 children: A cross sectional study in Gumla district of Jharkhand, India. Int. J. Equity Health 2016;15(1):104.

28. Jain Y, Kataria R, Patil S et al. Burden & pattern of illnesses among the tribal communities in central India: A report from a community health programme. Indian J Med Res 2015;141(5):663-72.

29. Das MB, Kapoor S, Nikitin D. A closer look at child mortality among Adivasis in India. World Bank Policy Research Working Paper No. 5231 2010. https://ssrn.com/abstract=1565992 (accessed on 24 June 2019).

30. Niti Ayog. Healthy starts: Progressive India. Report on the rank of states and union territories. 2019 June. http://social.niti.gov.in/uploads/sample/health_index_report.pdf (accessed on 14 Sep 2019)

31. Subramanian SV, Smith GD, Subramanyam M. Indigenous health and socioeconomic status in India. PLOS Medicine 2006;3(10):e421. https://doi.org/10.1371/journal.pmed.0030421 (accessed on 13 September 2019).

32. Mohan, P, Lando HA, & Panneer S. Assessment of tobacco consumption and control in India. Indian Journal of Clinical Medicine 2018;9:1-8. https://doi.org/10.1177/1179916118759289 (accessed on 13 September 2019).

33. Jha P, Jacob B, Gajalakshmi V et al. A nationally representative case-control study of smoking and death in India. N Engl J Med 2008;358(11):1137–47.

34. Ravindran TKS, Seshadri T. A health equity research agenda for India: results of a consultative exercise. Health Res Policy Syst 2018;16(1): 94 doi:10.1186/s12961-018-0367-0

PDF

Version 1

posted 27 Jan, 2020

You are reading this latest preprint version

Examining fine-scale tribal health inequalities around three forested sites in India: Results of a cross-sectional survey

Status:

Version 1

Abstract

Figures

Background

Methods

Results

Conclusions

Abbreviations

Declarations

References

Status:

Version 1