Pattern of contraceptive use, determinants and fertility intentions among tribal women in Kerala, India: a cross-sectional study.

OBJECTIVES: To assess the pattern of contraceptive use and its determinants, knowledge regarding contraceptives including oral contraceptive pills and fertility intentions among tribal women in the reproductive age group.
DESIGN: Community-based cross-sectional study.
SETTING: Community development blocks in a predominantly tribal district of Wayanad in Kerala, India. PARTICIPANTS: Women in the reproductive age group (15-49 years) from the tribal groups in the district numbering 2495. OUTCOME MEASURES: PRIMARY: Prevalence of contraceptive use and its determinants. SECONDARY: Knowledge regarding contraceptives in general, oral contraceptives and fertility intentions.
RESULTS: The mean age of the study participants was 30.8 years (SD=9.8) and belonged to various tribal groups such as Paniya (59.2%), Kurichiyar (13.6%) and Adiya (10.9%). Current use of contraceptive was reported by about a fourth, 658 (26.4%) (95% CI 27.9 to 24.9) of women. Following logistic regression, belonging to Paniya tribe (adjusted OR (aOR) 2.67, 95% CI 1.49 to 4.77; p<0.001) and age at menarche >13 years (aOR 1.69, 95% CI 1.14 to 2.52; p<0.009) had significantly higher use of contraceptives whereas social vulnerability as indicated by staying in a kutcha house had a lesser likelihood of use of contraceptive (aOR 0.55, 95% CI 0.31 to 0.95; p<0.03). Oral contraceptive use was low (4.8%) among this population and no abuse was observed.Less than half (47%) of the respondents had an above average knowledge on contraception. Multivariable logistic regression indicated that above average knowledge was 2.2 times more likely with higher education (95% CI 1.2 to 3.9), lesser among those who desired more than two children (aOR 0.59; 95% CI 0.38 to 0.94; p<0.02).Two children per family was the preferred choice for 1060 (42.5%) women. No gender bias in favour of the male child was observed.
CONCLUSION: Awareness and use of contraceptives are poor though the fertility is not commensurately high. Along with developing targeted responses to contraceptive use among Indigenous people with indigenous data, awareness also requires attention. Ethnographic studies are also necessary to determine the differences in contraceptive use including traditional methods among the various Indigenous groups. © Author(s) (or their employer(s)) 2022. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

It is a community-based study.

Limited studies have been conducted on contraceptive use among Indigenous tribal women.

Spatial or temporal association could not be studied.

Introduction

India is home to one of the largest population of Indigenous persons at 104 million second to China’s 106.4 million.1 Global data is replete with evidence of poorer health and social outcomes for Indigenous peoples than for non-Indigenous populations. However, this is not uniformly the case, and the size of the rate difference varies.2

Tribal people in independent countries are those ‘Whose social, cultural and economic conditions distinguish them from other sections of the national community and whose status is regulated wholly or partly by their own customs or traditions or by special laws or regulations; and peoples in independent countries who are regarded as Indigenous because of their descent from the populations who inhabited the country, or a geographical region to which the country belongs, at the time of conquest or colonisation or the establishment of present state boundaries and who, irrespective of their legal status, retain some or all of their own social, economic, cultural and political institutions’.3

Among the various social groups in India, the tribes are the most socioeconomically deprived groups, with low literacy and poor economic and living conditions.4 In India, scheduled tribes (ST) refer to various aboriginal ethnic minorities, who are concentrated in hilly lands in the Himalayas, Eastern Ghats, Western Ghats, Deccan Plateau and other hilly regions.5 The STs, also referred to as Adivasis not only have distinct cultural and linguistic characteristics but are also spatially isolated which is also cited as a reason for deprivation.6 In addition fertility indices are found to be high. Though, countries such as India, China and Bangladesh, resist the use of the term Indigenous groups due to contentions in the definition, they have adopted affirmative action and special measures for the ‘tribal’ or ‘Adivasi’.7 Several constitutional benefits have been instituted for their upliftment, including various welfare programmes, reserved positions in administration, government bodies and in educational institutions and so on, but these initiatives appear not to have benefitted the entire community.

In India, the southern state of Kerala is home to 0.4% of the total tribal population. The Government of Kerala has identified about 36 scheduled tribes and a majority of them reside in Wayanad district. Until early 1900, they formed the majority in the district. With the roads built by the British, settlers from the other parts of the state reached the area and displaced the original inhabitants further into forests.8 They also introduced tobacco and alcohol to the population which they then became dependent on.9 10

In other Latin American and Caribbean countries, coverage with modern contraception (median coverage ratio 0.82, IQR 0.66–0.92), was lower among Indigenous women than in the reference group. These differences persisted after adjustment for wealth, education and residence.11 Due to all of the above reasons and more, the tribes have a very low contraceptive usage and high unmet need for family planning than the other groups.5 A microlevel study conducted among Adiya and Kattunayakan tribes in Kerala found low contraceptive usage and high unmet need for family planning at 68% and 62%, respectively.12 Furthermore, understanding fertility, fertility intentions and contraceptive use can help in the development of population policies. Based on these findings, educational and interventional measures can be carried out to address the prevalent family planning and fertility issues among women.

In the light of sparse knowledge regarding contraceptive use among the tribal groups, this study was initiated to assess contraceptive use and its determinants, knowledge regarding contraceptives in general, oral contraceptives and fertility intentions.

Methods

The study was conducted among women belonging to tribal groups in the reproductive age group in two of the three randomly selected community development (CD) blocks of Wayanad in Kerala, India, namely Vythiri/Kalpetta and Mananthavady (figures 1 and 2).

Selection of sample and sample size. PPS, probabilityproportionate to size.

From a previous study12 the sample size was calculated to be 1200 per block (details of sample calculation in online supplemental file 1). Informed video consent was obtained from the study participants.

Considering the cluster size to be a minimum of 30, a total of 40 clusters were selected from the two blocks based on the Probability proportionate to size sampling. According to the 2011 census, the total population of tribal origin in the two CD blocks was 46 598. The sampling interval was 1165 and the first number randomly selected was 264 which was ward 1 of Thirunelly Grama Panchayat of Mananthavady CD block. Then 40 clusters were selected by adding the sampling interval.

The women in the 15–49 year age group were interviewed with a semi-structured questionnaire by trained volunteers from the district who underwent a 2-day training on contraception, communication skills, rapport building and questionnaire administration. A total of 2495 women were interviewed; with 1161 from Kalpetta and 1334 from Mananthavady (figure 2). The qualitative variables are expressed as proportion, percentage and the quantitative as mean and SD. The knowledge scores were then categorised into above average and below average. The missing values were in the range of 3%–8%. The occupation variable had a larger missing value at 18% and was interpreted with caution.

Figure 2

Map of the area.

Settings

A CD block is a rural area administratively earmarked for planning and development.13 Pucca House refers to dwellings with walls made up of burnt bricks, stones, cement concrete, etc, and the roof made of tiles, asbestos cement sheets, reinforced brick cement, reinforced concrete cement, timber, etc. In contrast, a Kutcha House, is one with its walls and/or roof made of material other than those mentioned earlier and may include the use of unburnt bricks, bamboos, mud, grass, thatch, loosely packed stones, etc. The particularly vulnerable tribal group (PVTG) have been classified by the Government of India and are tribal communities who are characterised by (a) forest-based livelihoods, (b) pre-agriculture level of existence, (c) stagnant or declining population, (d) extremely low literacy and (e) subsistence economy. They consist of Cholanaikayan (a section of Kattunaickans), Kadar, Kattunayakan, Kurumbas and Koraga.14

The healthcare services that women have used to access contraception include district hospitals, health centres, health workers and medical shops (over-the-counter). Other than the medical shops all services are free of cost. All medical care is free of cost to the persons of tribal origin including transport services which is directly reimbursed to the vehicle owner by the Integrated Tribal Development Office (as reported from Tribal office and tribal health promoters).

The conceptual framework for the development of the questionnaire (figure 3) consisted of domains of contraceptive use in line with the objectives such as contraceptive use, knowledge, attitude regarding contraception and its determinants. The questionnaire has been provided as online supplemental file 2. Attitude to reproductive rights and contraception is not discussed here. Questions on knowledge regarding contraceptives in general and on oral contraceptives, emergency contraception in particular were assessed with a total score of 2.5 for general knowledge questions, 3.75 for oral contraceptives and 1.5 for emergency contraceptive. Those who scored above mean scores, 0.8 for general knowledge on contraceptives, 2.2 for oral contraceptives and 0.2 for emergency contraception were considered to have above average knowledge. The independent variables for above average knowledge were considered to be age, marital status, education, occupation, type of house, tribe, number of children, tobacco use, alcohol use, age at menarche, current contraceptive use, knowledge regarding emergency contraception, oral contraception and concept of ideal family size. The menstrual history was also assessed along with some associated gynaecological symptoms in order to determine the relation, if any, with contraceptive use. The independent variables of contraceptive use were age, marital status, education, occupation, type of house, tribe, number of children, tobacco use, alcohol use, age at menarche, knowledge in general regarding contraceptives, knowledge regarding emergency contraception and oral contraception. Multivariate logistic regression model was developed using the variables age at menarche, occupation, type of tribe, type of house, marital status, number of children, knowledge about contraceptives in general and emergency contraception.

Figure 3

Conceptual framework for the development of questionnaire. OCPs, oral contraceptive pills.

Patient and public involvement

Public participation and involvement of tribal persons was brought about by training and involving the members of the indigenous groups in data collection. Before the training, discussions were carried out with the participants representing the tribal groups and some changes were made in the questionnaire. Discussions and necessary permission were obtained from tribal officers and the study was informed by their experience. Dissemination activities were also carried out.

In order to address the bias due to poor rapport, self-reports of contraceptive use, a repeat study was conducted in a subset of 298) to assess the contraceptive prevalence and to determine the presence of menstrual huts which was reported during the initial study, which has not been discussed here. The study was conducted from October 2019 to October 2020. A follow-up qualitative study was conducted to understand the barriers and facilitators to contraceptive use, fertility intentions and menstrual practices which is beyond the scope of this paper.

Results

The mean age of the study participants was 30.8 (SD=9.8 years). The women belonged to Paniya, Kurichiyar and Adiya tribal groups at 59.2%, 13.6% and 10.9%, respectively. The PVTG constituted 10% of the study population. About 395 (17.2%) persons were illiterate, a third (33.2%) had 12 years of schooling and about a fifth (19.7%) lived in kutcha houses (table 1). The mean per capita income was $10.66±25.72 from the available responses since many respondents did not reveal their income. The mean age at menarche was 13.21±1.5 years and the mean family size was 4.52±1.62.

Table 1

Sociodemographic distribution of the study population

Variables	Frequency	Percentage
Type of house
Kutcha	491	20.4
Pucca	1690	70.4
Others (*94 missing)	220	9.2
Tribe
Paniya	1478	59.2
Kurichiyar	339	13.6
Adiya	273	10.9
Others	405	29.8
Age (in years)
<31	1303	54
>31 (*81 missing)	1111	46
Education
Illiterate	395	17.2
Literate (*202 missing)	1898	82.8
Occupation
Employed	1419	69.6
Unemployed (*456 missing)	620	30.4
Marital status
Currently married	1695	70
Others (*75 missing)	725	30
Number of children
<3	1232	85.4
>3	211	14.6
Tobacco use
Yes	1436	57.6
No	1059	42.4
Alcohol use
Yes	92	3.7
No	2403	96.3
Regular periods
Yes	2075	83.2
No	420	16.8
Currently pregnant
Yes	83	3.3
No	2412	96.7
Age at menarche (in years)
<13	1428	61.7
>13 (*181 missing)	886	38.3
Ideal family size
≤2	1109	47.6
>2 (*164 missing)	1222	52.4
Desire for a third child if the first two children are of the same sex
Yes	1519	68.4
No (*273 missing)	703	31.6
Knowledge—general
Below average	666	53
Above average	590	47
Knowledge—oral contraceptive pill (OCP)
Below average	28	41.2
Above average	40	58.8
Knowledge—emergency contraception
Below average	899	78.4
Above average	247	21.6
Ever use of contraceptive
Yes	658	26.4
No	1837	73.6
Type of contraceptive used
Permanent	420	63.8
Hormonal (OCP)	57	8.7
Condoms	46	7
Copper T	107	16.3
Others including traditional methods	14	2.1

Majority of the women 2075 (83.2%) had regular menstrual periods and 97.2% had a menstrual cycle duration of less than 45 days. Abdominal pain associated with periods was experienced by 37.5% of the participants. The majority (86%) did not experience any bleeding or spotting between periods. The mean duration of menstrual periods was 4.07±1.33. Only, 4.7% reported that they had bleeding in between periods and 0.4% said that they had bleeding after intercourse. A small proportion, 3.3% of the women in the reproductive age group were currently pregnant. The mean number of children born to the respondents was 2.34+1.18. About a tenth, 11.6% gave a history of abortion and the mean number of abortions was 1.41±1.08 (table 1).

Initially, only about 41.2% (1027/2495) reported having heard of contraceptives and on further probing this increased to 50.3% (1256/2495). Only more than a third (36.7%) had above average knowledge of contraceptives. Less than a fifth, 433 (17.4%) reported having heard of oral contraceptives. Only 68 answered further knowledge questions on OCP of which about a third (33.8%) of the women had above average knowledge. The mean knowledge scores about contraceptives in general, oral contraceptives and about emergency contraception were uniformly low at 0.8 (maximum 2.5), 2.2 (maximum 3.5) and 0.2 (maximum 1.5). Only 206 (8.3%) had heard of emergency contraception and the knowledge regarding emergency contraception was above average among only a fifth, 21.5% (247/1146).

More than a fourth 658 (26.4%) (95% CI 27.9 to 24.9) of the women reported current use of contraceptive. This was validated by a second study among a subset of 298 reproductive age women where the prevalence of contraceptive use was found to be almost the same at 29%.

Among those who used contraceptives, a majority 420/658 (63.8%) had used permanent methods and a similar percentage of women had used hormonal contraceptive 57/658 (8.7%) and condoms 46/658 (7%). Intrauterine contraceptive device such as Copper T (CuT) was reported to have been used by 107/658 (16.3 %) and 14/658 (2.1%) reported using other methods including traditional methods. Nobody reported the use of an injectable hormonal contraceptive. Most of them relied on the Government health system for their contraceptive needs and obtained it from the primary health centres, Accredited Social Health Activists, district hospital and a few from medical shops. The most recent use of contraceptives was reportedly for family planning more than half of the time 58.6% (237/404), 7.1% (29/404) for religious reasons and 4.9% (20/404) used it at the partner’s insistence. Despite the current use of contraceptive being only 26%, less than 1% (23/2495) women reported accidental pregnancy. About 39 women (1.6%) reported that they had used emergency contraception.

Among those who reported not to have used a contraceptive, worry about the side effects (13.5%) and partner not wanting to use a method (11.8%) were the main reasons. About 4.8% also mentioned that they did not know how to use the contraceptive method and 7.2% did not need to use a contraceptive method as they were below 18. About a fourth, 26.4% mentioned a variety of reasons for not using such as husband’s illness, have decided not to use, have not felt its need, wanted pregnancy, divorced, self-control, need one more child, recently married and not pregnant, natural methods, planning to conceive and planning to do tubectomy, staying away from husband, widow.

Regarding oral contraceptives which was a particular focus of this study, about 61.5% had never used oral contraceptives. Most of the users of oral contraceptives (27/58) had only used it sometimes. About a half 28/58 (48.2%) used OCP for attending religious rituals for not more than a week and 24/58 (41.3%) for contraceptive use. The most reported side effects were irregularity in menstrual cycle (24.64%) and vomiting (15.94%). Overuse or continuous use of contraceptive for non-contraceptive use was not found, although present in other neighbouring tribal communities15 16 and this was also corroborated by the tribal officers of the area.

Determinants of knowledge of contraceptive use

After multivariate logistic regression, the educated were found to have a 2.2 times greater likelihood of above average knowledge (95% CI 1.2 to 3.9) about contraceptives. Compared with other Indigenous groups Paniya’s had 67% poorer knowledge (adjusted OR (aOR) 0.33 95% CI 0.17 to 0.65). Though the current use of contraceptives was 1.2 times more likely with above average knowledge, this was not significant. Knowledge of emergency contraception was an important determinant of general knowledge of contraceptives (aOR 6.9, 95% CI 3.86 to 12.36; p<0.001).

Though a clustering of risk factors for below average knowledge of contraceptives such as tobacco, alcohol use and illiteracywas observed (table 2A) only education was found to be an independent variable for above average knowledge (aOR 2.2; 95% CI 1.2 to 3.9). Those who desired more than two children irrespective of whether they were boys or girls had significantly lower knowledge (aOR 0.59; 95% CI 0.38 to 0.94; p<0.025) (table 2B). Others such as age >31 years, employed, married women, more than three children having significant association with knowledge disappeared with regression analysis (tables 2A, B). More than half of the women 198 (55.6%) in the reproductive age group with above average knowledge said that another child is not required if existing two children belong to the same sex compared with 360 (44.8%) in those who said yes (p<0.001) indicating no bias for male children.

Table 2A

Determinants of knowledge of contraceptives: univariate analysis

	Above average (%)	Total	P value
Age
≤31 years	279 (44.9)	622	0.12
>31 years	294 (49.3)	596
Education
Illiterate	53 (36.1)	147	0.001
Literate	501 (51)	983
Occupation
Employed	319 (45.8)	26	0.02
Unemployed	188 (53.9)	349
Tribe
Paniya	66 (34.7)	190	0.00
Kattunaickan	254 (41.6)	610
Others	270 (59.2)	456
Type of house
Kutcha	113 (42)	269	0.09
Pucca	413 (49.5)	835
Others	50 (50)	100
Marital status
Currently married	484 (51.3)	943	0.00
Others	94 (33.6)	280
Number of children
≤3	368 (53.6)	687	0.04
>3	49 (43)	114
Current use of contraceptive
No	457 (49.7)	919	0.001
Yes	133 (39.5)	337
Tobacco use
Yes	269 (40.1)	670	0.00
No	321 (54.8)	586
Alcohol use
Yes	9 (18.8)	48	0.002
No	581 (48.1)	1208
Knowledge regarding emergency contraceptives
Below average	323 (37.6)	859	0.00
Above average	185 (77.7)	238
Knowledge regarding oral contraceptives
Below average	12 (42.9)	28	0.61
Above average	17 (50)	34
Attitude
Below average	127 (28)	520	0.00
Above average	463 (57.7)	736

Table 2B

Independent determinants of above average knowledge of contraceptives

	Crude OR	95% CI	P value	Adjusted OR	95% CI	P value
Education
Literate	1.84	1.28 to 2.64	0.001	2.22	1.23 to 3.98	0.01
Illiterate	1			1
Tribe
Paniya	0.36	0.26 to 0.52	0.001	0.33	0.17 to 0.65	0.001
Kattunaickan	0.49	0.38 to 0.62	0.01	0.7	0.41 to 1.17	0.17
Others	1			1
Occupation
Employed	0.72	0.55 to 0.93	0.01	1	0.65 to 1.55	0.98
Unemployed	1			1
Attitude
Good	3.49	2.72 to 4.48	0.01	4.88	3.11 to 7.65	0.001
Poor	1			1
Marital status
Ever married	2.08	1.57 to 2.75	0.01	0.65	0.32 to 1.32	0.24
Others	1			1
Tobacco Use
Yes	0.55	0.44 to 0.69	0.01	0.67	0.41 to 1.08	0.1
No	1			1
Alcohol use
Yes	0.25	0.12 to 0.52	0.01	0.56	0.11 to 3.05	0.51
No	1			1
Ever use of contraceptive
Yes	0.66	0.51 to 0.85	0.01	1.24	0.75 to 2.06	0.4
No	1			1
Knowledge regarding emergency contraceptive
Above average	5.79	4.14 to 8.09	0.001	6.9	3.86 to 12.36	0.001
Below average	1			1
Number of children
>3	0.65	0.44 to 0.97	0.04	1.07	0.58 to 1.96	0.82
≤3	1			1
Fertility intention
Yes	0.65	0.50 to 0.83	0.001	0.59	0.38 to 0.94	0.03
No	1			1

Determinants of contraceptive use

A multivariate logistic regression of the determinants of current use of contraceptive showed that age at menarche >13 years (aOR 1.69, 95% CI 1.14 to 2.52; p<0.009) and belonging to Paniya tribe had significantly higher use of contraceptive (aOR 2.67, 95% CI 1.49 to 4.77; p<0.001). Though Kattunaickans, a PVTG had lower contraceptive use this was not significant (table 3A). Those who stayed in a kutcha house were also less likely to have used a contraceptive (aOR 0.55, 95% CI 0.31 to 0.95; p<0.035)(table 3B). Though the employed had a higher use of contraceptive in univariate analysis after regression the contraceptive use was significantly lesser among the employed. The final multivariate regression model also included variables such as more than three children, marital status, poorer knowledge scores though significantly associated in the univariate analysis disappeared in the regression analysis done by the enter method (tables 3A, B). Among the different tribes Kattunaickans, a PVTG had the lowest use of contraceptive at 23.8%, followed by other groups at 26.5% and 39.9% among Paniyas.

Table 3A

Factors affecting current contraceptive use

		Current contraceptive use
		Yes	Total	P value
Age	≤31 years	343 (26.3%)	1303	0.88
	>31 years	289 (26%)	1111
Education	Illiterate	84 (21.3%)	395	0.06
	Others	488 (25.7%)	1898
Tobacco use	Yes	382 (26.6%)	1436	0.78
	No	276 (26.1%)	1059
Occupation	Employed	378 (26.6%)	1419	0.02
	Unemployed	134 (21.6%)	620
Alcohol use	Yes	24(26.1)	92	1
	No	634 (26.4)	2403
Age at menarche	≤13 years	344 (24.1)	1428	0.001
	>13	270 (30.5)	886
Regular periods	Yes	540(26)	2075	0.4
	No	11 (28.1)	420
Knowledge general	Below average	204 (30.6%)	666	0.001
	Above average	133 (22.5%)	590
Knowledge of oral contraceptive pill	Below average	9(32.1%)	28	0.58
	Above average	10(25%)	40
Knowledge of emergency contraceptive	Below average	256 (28.5%)	899	0.02
	Above average	52 (21.1%)	247
Attitude	Poor	422 (29.3%)	1440	0.00
	Good	236 (22.4%)	1055
Tribe	Paniya	109 (39.9%)	273	0.00
	Kattunaickan	352 (23.8%)	1478
	Others	197 (26.5%)	744
Type of house	Kutcha	104 (21.2%)	491	0.004
	Others	554 (27.6)	2004
Marital status	Currently married	415 (24.5%)	1695	0.01
	Others	213 (29.4%)	725
Number of children	≤3	295 (23.9)	1232	0.01
	>3	69 (32.7)	211

Table 3B

Independent determinants of current contraceptive use

	Current contraceptive use
	Crude OR	95% CI	Adjusted OR	95% CI
		P value		P value
Age at menarche (years)
>13	1.38	1.15 to 1.67	1.69	1.14 to 2.52
		0.001		0.01
≤13	1		1
Tribes
Paniya	1.85	1.38 to 2.47	2.46	1.54 to 3.94
		<0.001		<0.001
Kattunaickan	0.87	0.71 to 1.06	0.98	0.67 to 1.43
		0.17		0.24
Others	1		1
Occupation
Employed	1.32	1.05 to 1.65	0.59	0.38 to 0.9
		0.02		0.02
Unemployed	1		1
Type of house
Kutcha	0.7	0.55 to 0.89	0.55	0.315 to 0.959
Others	1			0.04
Marital status
Ever married	0.77	0.64 to 0.94	0.85	0.422 to 1.71
		0.01		0.65
Others	1		1
Number of children
>3	1.54	1.12 to 2.11	1.22	0.69 to 2.16
		0.01		0.48
≤3	1		1
Knowledge—general
Above average	0.65	0.51 to 0.85	0.93	0.58 to 1.47
		0.001		0.75
Below average	1		1
Knowledge—emergency contraceptive
Above average	0.67	0.47 to 0.94	0.86	0.50 to 1.49
Below average	1	0.02	1	0.61

A larger proportion of women (42.5%) preferred two children per family, closely followed by (38.2%) women desiring three children. The spacing between children suggested was 3 years by 31.2% and 4 years by 17.8%. There did not appear to be a gender bias in favour of the male child as only 98 (3.9%) of women reported that another child was necessary if both the children were girls. Similarly, only about 102 (4.1%) reported that another child was necessary if the children were both boys.

Discussion

The current use of contraceptives was reported by 26.4% of Indigenous women in the reproductive age group. Knowledge of contraceptives underpins use. The above average knowledge scores were found to be higher among the educated and also aligned with the above average knowledge of emergency contraceptives. Knowledge was lower among those who desired more than two children and among one of the backward tribes, the Paniyas. Contraceptive use was higher among the those belonging to the Paniya tribe and among those with an age of menarche more than 13 years and lesser among those with a kutcha house.

Among the 26.4% contraceptive users, 63.9% used permanent methods, 9% used hormonal contraceptive, 7% used condoms, 16.3% used CuT and 2.1% used other methods including traditional methods. This is much less than that the state average of 53%.17 The contraceptive use varies in other states among the tribals with higher use reported in Assam (36%)18 and Bengal (40%).19 In the state of Maharashtra it varied from 18.7%20 to a spacing contraceptive use of 27.7% (10.8% pill, 13.3% condom, 2.0% IUD and 1.6% withdrawal or rhythm).21 This also underlines the heterogenous character of the tribal groups across the country varying in population size, language and their interaction with the rest of society22 and thus having different contraceptive use rates. In neighbouring Bangladesh the contraceptive use among tribals was very high at 73% which was also much higher than among the general population of that country.23 The primary reason for low contraceptive use may be the low access to contraceptive knowledge and services as observed in Maharashtra and Tamil Nadu.21 24 Other reasons could be early marriage, childbearing observed among tribal groups in Madhya Pradesh and Rajasthan25 which was in turn associated with low contraceptive use11 and lower education.24 Other studies in central India and Nepal substantiated the role of women and husbands’ education, age of women and number of surviving boys in the use of any modern method of contraception.5 26 Nevertheless, the different ethnographic backgrounds of the various tribal groups across different parts of India result in a non-uniform pattern of use of contraceptives. The independent determinants of contraceptive use in this study were belonging to the Paniya group and a lower use among those staying in a kutcha house. Though, India as a whole has maintained a steady decline in fertility,27 studies among tribal communities’ report some of the lowest rates of contraceptive use and an increasing fertility rate.27 With the exception of the scheduled tribes who have a total fertility rate (TFR) of 2.3, the TFR of all population groups in Kerala are below replacement level at 1.6.17 This also corresponds to the mean number of children in this study at 2.34±1.18. The TFR in central Indian state of Madhya Pradesh among the Indigenous tribes was 2.7.28 Though, the fertility levels in this study are higher than the general population it is not as high as the other tribal communities indicating the use of perhaps other traditional methods such as withdrawal method, calendar method, etc, which may not have been revealed fully in this cross-sectional study or the influence of the general population. Traditional method use such as rhythm and withdrawal is the highest in this district at 5.6%.17 Other reasons could be the high prevalence of reproductive tract infections. The few studies among tribal women on reproductive infection in a southern and northern state of Karnataka and Himachal Pradesh found a prevalence of 12%–14%.29 30

Reduced contraceptive use may also be due to the difficult access to healthcare services. The tribal people rely on the public health system which is in general characterised by low output, low quality and low outcome delivery system.31 An ethnographic study in an adjoining district showed that the hesitancy to access services in spite of being provided free services and a comprehensive financial protection package was due to the failure to provide culturally respectful care and the lack of power on the part of the Indigenous community to negotiate with the health system.32 In India, across 10 states with sizeable tribal population, the percentage surplus/deficit of healthcare providers in tribal areas is found to −33% allopathic doctors at primary health centre, and −84% specialists at community health centre which is indicative of the huge deficiency of specialist doctors and thus quality healthcare in the existing primary healthcare system.31

Knowledge of contraception is almost universal in Kerala. Among all women, 44% knew about emergency contraception.17 However, in the current study, only 36.7% had above average knowledge of contraceptives. This is similar to other studies also with Prusty reporting considerably lower knowledge among tribal women compared with their non-tribal counterparts in the three central Indian states.5 However, tribal women in the northeastern part of India had good knowledge of contraceptives and correspondingly the contraceptive use was also high.18 The awareness levels of contraception in general, was closely aligned with the awareness of emergency contraceptives (aOR 6.9, 95% CI 3.86 to 12.3) and educational levels (aOR 2.2, 95% CI 1.2 to 3.9). Findings suggest that lower spacing contraceptive use among tribals is driven by social vulnerabilities such as low education, housing and higher fertility preferences.21

Gender norms related to fertility is an important issue determining contraceptive use in the tribal community.24 Gender relations among Indian tribes have historically been more balanced and equitable; however there is an increasing trend of gender bias in tribal culture emerging due to the assimilation and modernising process.33 34 The tribes in Kerala seem to be more equitable with a male child preference not indicated with the desire for a third child expressed when the initial two children are of the same sex. In the northeastern state of Meghalaya 86% of the population is tribal and followed a matrilineal system. It had the lowest contraceptive use of 20.2% with a corresponding high total fertility rate of 4.57.35 36 Though, tribal women have more independence and fewer restrictions than women belonging to caste Hindu communities, in some states such as Odisha, their status remained low and decision-making regarding issues like seeking healthcare and large expenditures rested with men.37 Even in neighbouring Bangladesh, no preference of son over daughter was observed among tribal communities,23 whereas women in Kerala are slightly more likely to use contraception if they already have a son.17

Though the Paniya group had lower knowledge scores compared with the other groups, contraceptive use was higher compared with other tribal groups. This may be due to the fact that the Paniyas, a marginalised group among the tribals use more of the public health services38 and the subcentres which provide family planning services do not incur an additional travel cost. This is a cross-sectional study and temporal relationships between knowledge and use cannot be studied. Spatial (or geographical) isolation is cited as the major reason for the exclusion and backwardness of indigenous communities in both China and India.6 The living conditions, an underprivileged background and social vulnerability may contribute to low contraceptive use.

A limitation of the study is that the contraceptive use was self-reported, which inherently carries the risk of under-reporting as the persons of tribal origin are known to be shy and reclusive. However, trained local personnel were employed, who were able to reduce under-reporting. In addition, the prevalence of contraceptive use has been corroborated by a repeat study using a smaller sample. Though, the findings are generalisable to the persons of tribal origin as a whole, the various tribal/indigenous groups have not been covered equally due to difficult terrain and the possibility of human–wildlife conflict. Other ethnographic studies among the various tribal communities are necessary to understand the use of traditional methods of contraception. Data related to some variables are missing though it is negligible, ranging from 3% to 10%.

Thus, at a policy level more changes are necessary. In spite of the numerous schemes and support such as Tribal Health promoter, Oorumithram (friend of the village) there is a huge gap in understanding the tribal persons belonging to different groups. A more culturally sensitive and respectful approach is necessary. Universal access to family planning services, aligned with Sustainable Development Goals 3.7, data suggests that we are far from achieving this goal when it comes to tribal communities.

Conclusion

Awareness and use of contraceptives are poor though the fertility is not commensurately high. Along with developing targeted responses to contraceptive use among Indigenous people with indigenous data,3 awareness also requires attention. This study shows a higher use of contraceptive among Paniyas and ethnographic studies are necessary to determine the differences in contraceptive use including traditional methods among the various Indigenous groups.