Abstract
Objective
To assess associations between marital violence and type of contraception among women in South Asia.
Methods
Cross-sectional analyses were conducted using marital violence data collected during the most recent Demographic and Health Surveys from Bangladesh (n = 3665), India (n = 56 357), and Nepal (n = 3037). Data were pooled to assess associations of marital violence (physical or sexual) with modern contraception use (current spacing or sterilization).
Results
Sexual marital violence was associated with both modern spacing contraception (adjusted odds ratio [AOR] 1.30; 95% confidence interval [CI], 1.13–1.49) and sterilization (AOR 0.79; 95% CI, 0.70–0.88). Sexual violence was reported more often by pill users (9.8% vs 5.5% for non-users) but less often by condom users (4.5% vs 5.8% for non-users).
Conclusion
Sexual marital violence might increase use of contraception that need not require husband involvement (pill) but decrease use of methods that require his cooperation (condom) or support for mobility, funds, or time (sterilization).
Keywords: Contraception use; Domestic violence; Intimate partner violence; Sexual violence, Family planning; South Asia
1. Introduction
A study published in 2010 estimated that 40% of women in South Asia had experienced physical and/or sexual marital violence [1]. Global evidence suggests that such violence affects women’s reproductive control and contraceptive practices [2–4]. Actions taken by male partners to prevent women from implementing family planning measures have been implicated in increased risk of unplanned and unwanted pregnancies and induced abortion among women in South Asia who report marital violence [5–10]. Complicating this picture are findings from India that demonstrate associations between marital violence and non-use of contraception [5,6,11], and contrasting findings from Bangladesh that document an association between marital violence and increased use of contraception [12]. Research from Nepal found no appreciable association between marital violence and contraception, possibly owing to an inadequate sample size [13].
These dissimilar findings, which were recorded at the national level, might be attributable to the different forms of contraception that predominate in each country; namely, injections in Nepal, the contraceptive pill in Bangladesh, and female sterilization in India [14–165]. Regional, rather than country-specific, analysis could, therefore, offer some insight, given the overlap of key predictors of contraceptive use across South Asia. Such predictors include gendered risks (e.g. early marriage of girls and a preference for sons) and social inequities (e.g. rural residence, poverty, and low education) [14–16]. Such pooled regional analysis would also allow for large samples through which to explore differences in associations of marital violence with spacing contraception (modern contraceptives that allow a woman to delay or space pregnancy; for example, the pill, IUD) versus limiting contraception (or permanent contraception, which prevents further pregnancies from occurring; for example, female sterilization).
Analysis of associations with spacing versus limiting contraception is currently lacking; however, this aspect is important to consider because motivations differ for these forms of contraception. Multi-country analyses, including research conducted in South Asia, suggest that women who report spousal violence are also more likely to report high parity [2,14–17], which suggests that they might be less likely to use limiting forms of contraception. Research from East Africa found that men who held an accepting attitude toward marital violence also desired a large number of children [18]. Such attitudes in the context of marital violence might affect women’s acquisition of limiting forms of contraception (e.g. sterilization). Simultaneously, however, the high rates of unwanted pregnancy [19] and induced abortion [20] observed among women experiencing spousal violence suggest that such women might actually wish to avert a pregnancy. In this context, female-controlled contraceptive methods, such as the pill, might offer greater opportunity for covert use by women contending with marital violence, while still allowing them the opportunity to achieve their husband’s expected fertility goals.
The aim of the present study was to provide insight into the potentially different associations between marital violence and spacing versus limiting forms of contraception in South Asia, through the analysis of data collected from married women not currently pregnant at the time of survey in Bangladesh, India, and Nepal.
2. Methods
Cross-sectional analyses were conducted using the most recent Demographic and Health Survey (DHS) data from Bangladesh (2007), India (2005–2006), and Nepal (2011) [14–16]. The analysis was conducted at the University of California, San Diego, between November 2013 and November 2014. Ethical approval for the DHS design and implementation was provided by ICF International (Fairfax, VA, USA) and the respective host country. All DHS participants provided informed consent prior to their participation. The institutional review board of the University of California, San Diego, USA, approved the present study.
The DHS data from Bangladesh, India, and Nepal were used for present study because they were the only countries in South Asia for which both physical and sexual marital violence measures were available [14–16]. The DHS are nationally representative, two-stage, stratified sample surveys on population health and fertility conducted among women of reproductive age [21]. Response rates were greater than 94% in all three countries included in the present study. The sample for analysis was restricted to currently married women who had completed the domestic violence module; were not pregnant at the time of interview; and had provided responses for all dependent and independent variables assessed. The total cohort size was 63 059; the breakdown by country was Bangladesh (n = 3665), India (n = 56 357), and Nepal (n = 3037).
The primary independent variables assessed in the present study were any occurrence of physical marital violence (assessed by seven items) or sexual marital violence (assessed by two items), ever in the current marital relationship. Other items assessed physical and sexual marital violence during the past year; these data were provided descriptively. Further details are available in the relevant DHS reports [14–16]. Current contraceptive use was set as the primary dependent variable. This variable was categorized as none and/or not modern (including traditional and folkloric methods);modern spacing (contraceptive pill, intrauterine device [IUD], injections, diaphragm, condom, implant, female condom, foam, and jelly); and sterilization (male and female).
Covariates were social equity indicators (respondent age, respondent and husband education, household wealth quintile, and urban vs rural residence) and gender equity indicators (parental marital violence [father’s abuse of mother], female child marriage, position in the household, preference for a son, and actual number of living sons and daughters). These covariates were selected on the basis of previous research documenting their associations with marital violence and contraception in South Asia [5–12,22–24].
2.1. Data analysis
Data were pooled across countries, and multinomial regressions were used to assess relationships between independent variables and the contraception outcome. Models included both physical and sexual marital violence as primary independent variables. Final adjusted models included primary independent variables, fixed effects by country, and all covariates that were statistically significant (P < 0.05). No co-linearity for covariates was indicated for the model, based on a tolerance cutoff of 0.30. All analyses were weighted using individual weights that adjusted for country population sizes and complex survey design using SAS version 9.3 (SAS Institute, Cary, NC, USA). Given the much larger sample size for India, sensitivity analyses were conducted to determine whether the observed effects for the pooled model held true at the national level. The sensitivity analyses involved examination of multivariate models stratified by country and by comparing multivariate models with and without India. In addition, descriptive analyses of specific types of contraceptive use (e.g. pill, condom, or IUD) by physical and sexual marital violence were conducted both for the total pooled sample and by country.
3. Results
The descriptive characteristics of the present study group and observed associations are outlined in Table 1. A history of physical or sexual marital violence ever was reported by 37.2% (unweighted n = 20 225) of the cohort; 23.3% (unweighted n = 12 966) had experienced physical or sexual marital violence during the past year. The occurrence of both physical and sexual marital violence ever was 7.8% (unweighted n = 4192), and in the past year was reported by 4.6% (unweighted n = 2543). In all, 34.9% (unweighted n = 19 051) reported physical marital violence at any time, and 10.1% (unweighted n = 5366) reported sexual marital violence at any time. Modern spacing contraceptive use was reported by 14.5% (unweighted n = 10 923) and sterilization by 37.4% (unweighted n = 22 578; male sterilization, 1.1%, unweighted n = 948).
Table 1.
Descriptive characteristics and associations between marital violence and current contraceptive use among women in South Asia (n = 63059).a,b,c
| Characteristic | Total sample | Current contraceptive use Frequenciesd |
Association analyses | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| No current modern contraceptione |
Modern spacing contraception |
Sterilization | Modern spacing contraception |
Modern spacing contraception |
Sterilization | Sterilization | ||||
| Unweighted sample size |
Prevalence (95% CI) |
Distribution | Distribution | Distribution | OR (95% CI)f |
AOR (95% CI)g |
OR (95% CI)f |
AOR (95% CI)g |
||
| Study cohort overview |
63 059 | 48.1 (47.4–48.9) | 14.5 (14.0–15.0) | 37.4 (36.7–38.1) | - | - | - | - | ||
| Physical marital violence | ||||||||||
| Never | 44 008 | 65.1 (64.3–65.9) | 48.6 (47.8–49.5) | 14.9 (14.3–15.5) | 36.5 (35.6–37.3) | 1.00 | 1.00 | 1.00 | 1.00 | |
| Ever | 19 051 | 34.9 (34.1–35.7) | 47.2 (46.0–48.4) | 13.7 (12.9–14.5) | 39.0 (37.8–40.3) |
0.81 (0.75–0.88) |
1.04 (0.94–1.14) |
1.17 (1.10–1.25) |
0.99 (0.92–1.07) |
|
| Sexual marital violence | ||||||||||
| Never | 57 693 | 89.9 (89.4–90.3) | 47.7 (47.0–48.5) | 14.1 (13.7–14.6) | 38.2 (37.4–38.9) | 1.00 | 1.00 | 1.00 | 1.00 | |
| Ever | 5366 | 10.1 (9.7–10.6) | 51.9 (49.9–53.8) | 17.6 (16.1–19.1) | 30.5 (28.7–32.4) | 0.98 (0.87–1.10) |
1.30 (1.13–1.49) |
0.79 (0.72–0.87) |
0.79 (0.70–0.88) |
|
| Age, y | ||||||||||
| 15–19 | 2727 | 7.0 (6.7–7.4) | 85.6 (83.7–87.4) | 13.2 (11.4–15.0) | 1.2 (0.7–1.7) | 1.00 | 1.00 | 1.00 | 1.00 | |
| 20–24 | 9387 | 16.9 (16.4–17.3) | 66.4 (65.0–67.8) | 19.3 (18.1–20.4) | 14.4 (13.4–15.4) |
2.14 (1.81–2.53) |
1.14 (0.95–1.37) |
14.43 (9.27–22.46) |
7.67 (4.94–11.89) |
|
| 25–29 | 13 746 | 19.8 (19.4–20.3) | 46.6 (45.2–47.9) | 19.7 (18.6–20.8) | 33.7 (32.4–35.0) |
3.17 (2.69–3.74) |
1.26 (1.04–1.52) |
47.90 (31.05–73.88) |
18.48 (11.97–28.53) |
|
| 30–39 | 24 330 | 34.1 (33.6–34.6) | 36.4 (35.4–37.4) | 15.1 (14.4–15.9) | 48.5 (47.4–49.6) |
3.00 (2.55–3.53) |
1.02 (0.84–1.24) |
89.29 (57.79–137.97) |
27.88 (17.99–43.20) |
|
| 40–49 | 12 869 | 22.2 (21.6–22.7) | 41.8 (40.4–43.2) | 5.6 (5.0–6.2) | 52.6 (51.2–54.0) | 0.91 (0.75–1.09) |
0.27 (0.21–0.33) |
86.44 (55.95–133.53) |
23.08 (14.85–35.87) |
|
| Education | ||||||||||
| None | 23 789 | 44.9 (44.0–45.8) | 51.4 (50.3–52.6) | 7.7 (7.1–8.2) | 40.9 (39.7–42.0) | 1.00 | 1.00 | 1.00 | 1.00 | |
| Primary | 10 315 | 16.7 (16.2–17.1) | 44.5 (43.1–46.0) | 14.3 (13.4–15.3) | 41.1 (39.7–42.6) |
1.73 (1.54–1.94) |
1.50 (1.33–1.70) |
1.28 (1.19–1.39) |
1.57 (1.44–1.72) |
|
| Secondary or higher | 28 955 | 38.4 (37.6–39.3) | 45.8 (44.8–46.8) | 22.5 (21.7–23.3) | 31.7 (30.7–32.6) |
3.42 (3.13–3.73) |
2.07 (1.83–2.33) |
0.86 (0.80–0.92) |
1.36 (1.24–1.49) |
|
| Spousal education | ||||||||||
| None | 13 956 | 26.8 (26.0–27.5) | 52.3 (50.9–53.7) | 9.3 (8.6–10.0) | 38.4 (37.0–39.8) | 1.00 | 1.00 | 1.00 | 1.00 | |
| Primary | 10 462 | 17.2 (16.7–17.7) | 45.3 (43.7–46.8) | 13.0 (12.1–14.0) | 41.7 (40.2–43.2) |
1.53 (1.36–1.73) |
1.18 (1.04–1.35) |
1.29 (1.18–1.40) |
1.22 (1.11–1.34) |
|
| Secondary or higher | 38 641 | 56.0 (55.1–56.9) | 47.0 (46.2–47.9) | 17.4 (16.8–18.1) | 35.6 (34.7–36.4) |
2.66 (2.41–2.93) |
1.15 (1.02–1.30) |
0.95 (0.88–1.02) |
0.93 (0.85–1.03) |
|
| Wealth quintile | ||||||||||
| Poorest | 8619 | 18.2 (17.5–19.0) | 61.2 (59.6–62.8) | 8.3 (7.4–9.1) | 30.5 (29.0–32.1) | 1.00 | 1.00 | 1.00 | 1.00 | |
| Poorer | 9833 | 19.3 (18.7–19.8) | 52.6 (51.1–54.0) | 9.9 (9.1–10.8) | 37.5 (36.0–38.9) |
1.40 (1.22–1.60) |
1.30 (1.13–1.51) |
1.43 (1.31–1.57) |
1.54 (1.38–1.72) |
|
| Middle | 12 001 | 19.8 (19.2–20.4) | 47.4 (45.9–48.9) | 10.8 (9.9–11.6) | 41.8 (40.4–43.3) |
1.69 (1.46–1.94) |
1.47 (1.26–1.71) |
1.77 (1.61–1.95) |
2.06 (1.82–2.33) |
|
| Richer | 14 409 | 20.7 (20.0–21.4) | 42.4 (41.0–43.9) | 15.2 (14.2–16.3) | 42.4 (40.9–43.8) |
2.71 (2.35–3.11) |
1.90 (1.62–2.24) |
1.99 (1.80–2.20) |
2.38 (2.08–2.71) |
|
| Richest | 18 197 | 22.0 (21.2–22.8) | 39.5 (38.2–40.7) | 26.3 (25.2–27.4) | 34.2 (32.9–35.6) |
5.47 (4.83–6.20) |
3.22 (2.69–3.84) |
1.68 (1.52–1.86) |
2.01 (1.72–2.36) |
|
| Residence | ||||||||||
| Rural | 35 472 | 69.3 (68.6–69.9) | 51.3 (50.4–52.2) | 11.5 (11.0–12.1) | 37.2 (36.3–38.1) | 1.00 | 1.00 | 1.00 | 1.00 | |
| Urban | 27 587 | 30.7 (30.1–31.4) | 41.0 (39.8–42.2) | 21.1 (20.2–22.1) | 37.9 (36.7–39.1) |
2.64 (2.43–2.87) |
1.42 (1.28–1.58) |
1.21 (1.13–1.30) |
0.95 (0.86–1.05) |
|
| Parental marital violence | ||||||||||
| No | 51 565 | 79.6 (78.9–80.3) | 48.6 (47.8–49.4) | 14.8 (14.3–15.3) | 36.6 (35.8–37.4) | 1.00 | 1.00 | 1.00 | 1.00 | |
| Yes | 11 494 | 20.4 (19.7–21.1) | 46.4 (44.9–47.9) | 13.2 (12.3–14.1) | 40.4 (38.9–41.8) |
0.85 (0.78–0.93) |
1.01 (0.92–1.12) |
1.20 (1.12–1.29) |
1.36 (1.25–1.48) |
|
| Age at marriage, y | ||||||||||
| <15 | 11 303 | 23.1 (22.5–23.7) | 45.0 (43.6–46.4) | 13.2 (12.4–14.1) | 41.7 (40.4–43.1) | 1.00 | 1.00 | 1.00 | 1.00 | |
| 15–17 | 21 251 | 38.2 (37.6–38.8) | 48.4 (47.3–49.5) | 12.2 (11.6–12.9) | 39.4 (38.3–40.5) |
1.17 (1.06–1.30) |
1.02 (0.91–1.14) |
0.77 (0.72–0.83) |
0.80 (0.73–0.87) |
|
| ≥18 | 30 505 | 38.7 (38.0–39.5) | 49.8 (48.8–50.7) | 17.5 (16.7–18.2) | 32.8 (31.8–33.7) |
1.96 (1.77–2.17) |
1.24 (1.10–1.40) |
0.59 (0.55–0.64) |
0.55 (0.51–0.61) |
|
| Relationship to household head | ||||||||||
| Daughter-in-law | 9789 | 18.4 (17.8–18.9) | 58.0 (56.5–59.6) | 20.0 (18.8–21.2) | 21.9 (20.6–23.3) | 1.00 | 1.00 | 1.00 | 1.00 | |
| Head | 3310 | 4.8 (4.5–5.1) | 61.6 (58.8–64.4) | 5.9 (4.8–7.0) | 32.5 (29.7–35.3) |
0.22 (0.18–0.28) |
0.24 (0.19–0.32) |
1.53 (1.32–1.78) |
0.65 (0.55–0.77) |
|
| Wife | 46 046 | 66.4 (65.7–67.1) | 41.4 (40.6–42.2) | 13.7 (13.2–14.2) | 44.9 (44.0–45.8) |
0.81 (0.74–0.88) |
1.00 (0.90–1.10) |
3.06 (2.81–3.32) |
1.50 (1.36–1.64) |
|
| Daughter | 2353 | 6.3 (5.9–6.6) | 72.4 (69.9–74.9) | 12.3 (10.5–14.1) | 15.3 (13.3–17.2) |
0.42 (0.34–0.51) |
0.48 (0.39–0.59) |
0.59 (0.50–0.71) |
0.77 (0.64–0.93) |
|
| Other | 1561 | 4.2 (3.9–4.5) | 59.8 (56.2–63.4) | 15.7 (12.8–18.5) | 24.5 (21.5–27.5) |
0.66 (0.52–0.84) |
0.71 (0.55–0.91) |
1.14 (0.95–1.37) |
0.87 (0.71–1.07) |
|
| Son preference | ||||||||||
| Yes | 13 655 | 23.9 (23.3–24.5) | 53.1 (51.7–54.5) | 10.3 (9.5–11.0) | 36.6 (35.2–38.0) | 1.00 | 1.00 | 1.00 | 1.00 | |
| No | 49 404 | 76.1 (75.5–76.7) | 46.6 (45.8–47.3) | 15.8 (15.3–16.4) | 37.6 (36.9–38.4) |
1.58 (1.44–1.73) |
1.23 (1.11–1.37) |
1.22 (1.14–1.31) |
1.65 (1.53–1.78) |
|
| Living sons | ||||||||||
| 0 | 14 222 | 23.5 (23.0–24.1) | 75.4 (74.3–76.5) | 13.8 (12.9–14.6) | 10.9 (10.1–11.6) | 1.00 | 1.00 | 1.00 | 1.00 | |
| 1 | 24 042 | 36.8 (36.2–37.4) | 44.8 (43.8–45.9) | 18.6 (17.8–19.4) | 36.6 (35.5–37.6) |
2.25 (2.06–2.46) |
2.54 (2.31–2.79) |
5.69 (5.22–6.21) |
3.85 (3.53–4.21) |
|
| ≥2 | 24 795 | 39.7 (39.0–40.3) | 35.0 (33.9–36.2) | 11.1 (10.5–11.7) | 53.9 (52.7–55.1) |
1.62 (1.47–1.79) |
3.22 (2.87–3.62) |
11.02 (10.03–12.10) |
7.52 (6.79–8.33) |
|
| Living daughters | ||||||||||
| 0 | 19 922 | 30.5 (30.0–31.1) | 59.9 (58.8–60.9) | 14.5 (13.8–15.3) | 25.6 (24.7–26.5) | 1.00 | 1.00 | 1.00 | 1.00 | |
| 1 | 24 275 | 36.7 (36.2–37.3) | 41.4 (40.4–42.5) | 16.3 (15.6–17.0) | 42.3 (41.2–43.3) |
1.58 (1.45–1.71) |
1.77 (1.62–1.95) |
2.41 (2.27–2.56) |
1.99 (1.86–2.13) |
|
| ≥2 | 18 862 | 32.8 (32.2–33.3) | 44.7 (43.5–45.9) | 12.4 (11.7–13.2) | 42.9 (41.6–44.1) | 1.09 (1.00–1.20) |
1.89 (1.70–2.11) |
2.28 (2.13–2.45) |
1.58 (1.44–1.72) |
|
Abbreviations: AOR, adjusted odds ratio; CI, confidence interval; OR, odds ratio.
Values given as number or percentage (95% CI), unless otherwise indicated.
The present analysis included married women aged 15–49 years who were not pregnant at the time of completing the Demographic and Health Survey. The breakdown by country was as follows: Bangladesh (n = 3665), India (n = 56 357), and Nepal (n = 3037).
Multinomial regression analyses were used to conduct ORs and AORs, with 95% confidence intervals to determine significance. Significant effects are noted by bolding in text.
Row percentages presented.
Reference category.
Adjusted for country fixed effects.
Adjusted for country fixed effects and all variables shown in the column.
Adjusted multinomial analyses indicated that history of sexual marital violence was associated with increased likelihood of current modern spacing contraceptive use but reduced likelihood of sterilization (Table 1). The adjusted odds ratios (AORs) were 1.30 (95% confidence interval [CI], 1.13–1.49) and 0.79 (95% CI, 0.70–0.88), respectively. Physical marital violence was not associated with either parameter. Sensitivity analyses—including Bangladesh-specific and Nepal-specific models, and the pooled multivariate model without India—did not yield similar findings to the overall model; in these analyses, neither physical nor sexual marital violence were appreciably associated with the contraception outcomes (data not shown). Small cell sizes for sexual marital violence might have affected these estimates. The results of the India-specific model were comparable to the pooled model.
Covariates in the total pooled model revealed important social equity indicators associated with the contraceptive outcomes (Table 1). Well-educated women with a high wealth index were more likely to report both spacing contraception and sterilization than poorly educated women with a low wealth index. In addition, women with well-educated husbands and those who were urban residents tended to report the use of spacing contraception. Women in the oldest age category (40–49 years) were more likely to report sterilization and less likely to report spacing contraception than women in the youngest age category (15–19 years). Son preference ideologies were associated with increased likelihood of both spacing contraception and sterilization. Although high boy and high girl parity were both associated with the use of spacing contraception and sterilization, having two or more boys demonstrated markedly greater effect sizes for these associations than were detected for two or more girls. The greatest difference was seen for sterilization: women with two or more boys were 7.5-times more likely than those with no boys to report sterilization; by contrast, women with two or more girls were only 1.6-times more likely than those with no girls to use this method of contraception. Heads of household and daughters of the heads of household were less likely than daughters-in-law to report spacing contraception or sterilization, although wives of the heads of household were more likely than daughters-in-law to report sterilization. Early marriage was also associated with sterilization.
To offer further insight, descriptive data on the type of contraceptives used by history of sexual marital violence were reviewed (Table 2). Women who had experienced sexual marital violence were more likely to report pill use (9.8% [507/5366] vs 5.5% [3332/57 693] for non-users) but less likely to report condom use (4.5% [279/5366] vs 5.8% [4335/57 693] for non-users). Similarly, use of injection was more likely, but use of an IUD was less likely, among women who had experienced sexual marital violence; however, the prevalence of these types of birth control was low (1.7%, [1604/63 059]) in the present cohort. Similar findings were documented for physical marital violence. For traditional contraceptive methods, withdrawal was most likely to be used by women who reported sexual marital violence than those who reported no sexual marital violence (4.2% [274/5366] vs 2.7% [2118/57 693]). Country-specific analyses showed similar associations between marital violence and condom use; however, associations between marital violence and other forms of contraception were not consistent across countries. A greater likelihood of pill use among women reporting sexual marital violence was seen for Bangladesh and India, but not for Nepal. A lower likelihood of sterilization among women reporting sexual marital violence was seen for India, but not for Bangladesh or Nepal, with Nepal showing higher likelihood of sterilization among women with such a history. These divergent findings by country might in part be attributable to variation in cell sizes, as predominant forms of contraception were not the same in each country. In Bangladesh, the predominant forms of contraception were the pill (30.0% [1086/3665]), injections (7.7% [295/3665]), and female sterilization (5.9% [200/3665]). In India, they were female sterilization (39.8% [21 031/56 357]) and condoms (5.8% 4233/56 357). In Nepal, they were female sterilization (14.5% [399/3037]), injections (10.7% [357/3037]), male sterilization (9.5% 290/3037), and condoms (5.1% 175/3037).
Table 2.
Descriptive characteristics on contraceptive method for total sample and by history of physical and sexual marital violence, among currently married, not currently pregnant women aged 15–49 years in South Asia (n = 63 059; Bangladesh n = 3665, India n = 56 357, Nepal n = 3037).a,b
| Contraceptive method | Total sample | No physical marital violence | Physical marital violence | No sexual marital violence | Sexual marital violence |
|---|---|---|---|---|---|
| None | 39.5 (38.8–40.2) | 39.7 (38.8–40.5) | 39.2 (38.1–40.4) | 39.3 (38.5–40.0) | 41.7 (39.6–43.4) |
| Pill | 5.9 (5.8–6.2) | 5.3 (5.0–5.7) | 6.9 (6.4–7.5) | 5.5 (5.1–5.8) | 9.8 (8.6–11.0) |
| IUD | 1.9 (1.7–2.0) | 2.2 (2.0–2.4) | 1.3 (1.0–1.6) | 1.9 (1.8–2.1) | 1.2 (0.8–1.6) |
| Injections | 0.9 (0.8–1.1) | 0.7 (0.6–0.8) | 1.4 (1.2–1.6) | 0.9 (0.7–1.0) | 1.8 (1.3–2.2) |
| Condom | 5.7 (5.4–5.9) | 6.6 (6.2–6.9) | 3.9 (3.5–4.3) | 5.8 (5.5–6.1) | 4.5 (3.7–5.3) |
| Female Sterilization | 36.3 (35.6–37.0) | 35.5 (34.6–36.3) | 37.9 (36.7–39.1) | 37.1 (36.3–37.8) | 29.5 (27.7–31.3) |
| Male sterilization | 1.1 (0.9–1.2) | 1.0 (0.9–1.1) | 1.1 (0.9–1.4) | 1.1 (0.9–1.2) | 1.1 (0.7–1.4) |
| Periodic abstinence | 5.4 (5.1–5.7) | 5.6 (5.2–6.0) | 5.0 (4.5–5.5) | 5.4 (5.1–5.7) | 5.3 (4.4–6.3) |
| Withdrawal | 2.9 (2.6–3.1) | 3.1 (2.8–3.3) | 2.5 (2.1–2.8) | 2.7 (2.5–2.9) | 4.2 (3.4–5.0) |
| Bangladesh | |||||
| None | 39.6 (37.5–41.7) | 43.7 (40.8–46.6) | 35.2 (32.3–38.0) | 40.1 (37.8–42.4) | 36.9 (31.9–41.8) |
| Pill | 30.0 (27.9–32,0) | 28.5 (25.8–31.2) | 31.5 (28.5–34.4) | 28.9 (26.8–31.0) | 34.7 (30.4–39.1) |
| IUD | 1.1 (0.60–1.7) | 0.5 (0.2–0.8) | 1.8 (0.8–2.8) | 1.1 (0.5–1.7) | 1.2 (0.1–2.3) |
| Injections | 7.7 (6.5–8.9) | 6.3 (4.8–7.7) | 9.2 (7.7–10.7) | 7.7 (6.4–8.9) | 7.8 (5.3–10.3) |
| Condom | 4.7 (3.9–5.5) | 6.3 (5.0–7.5) | 3.0 (2.1–3.9) | 5.0 (4.1–5.8) | 3.4 (1.8–5.1) |
| Female Sterilization | 5.9 (4.8–7.0) | 4.2 (3.1–5.3) | 7.6 (5.6–9.7) | 5.9 (4.7–7.2) | 5.7 (3.8–7.5) |
| Male sterilization | 0.6 (0.3–0.9) | 0.3 (0.0–0.6) | 1.0 (0.5–1.5) | 0.6 (0.2–0.9) | 1.0 (0.2–1.9) |
| Periodic abstinence | 5.7 (4.8–6.6) | 5.3 (4.1–6.5) | 6.9 (4.8–7.3) | 5.9 (4.9–7.0) | 4.4 (2.6–6.2) |
| Withdrawal | 3.4 (2.6–4.1) | 3.6 (2.6–4.7) | 3.1 (2.1–4.1) | 3.4 (2.6–4.2) | 3.2 (1.6–4.8) |
| India | |||||
| None | 39.4 (38.7–40.2) | 39.2 (38.4–40.1) | 39.8 (38.6–41.0) | 39.1 (38.3–39.9) | 42.7 (40.4–44.9) |
| Pill | 3.4 (3.1–3.6) | 3.5 (3.2–3.7) | 3.3 (2.9–3.7) | 3.2 (3.0–3.5) | 4.9 (4.0–5.7) |
| IUD | 2.0 (1.8–2.1) | 2.3 (2.1–2.5) | 1.2 (0.9–1.5) | 2.0 (1.8–2.2) | 1.2 (0.8–1.6) |
| Injections | 0.1 (0.1–0.2) | 0.1 (0.1–0.2) | 0.1 (0.0–0.2) | 0.1 (0.1–0.1) | 0.4 (0.1–0.7) |
| Condom | 5.8 (5.5–6.1) | 6.6 (6.2–7.0) | 4.1 (3.7–4.5) | 5.9 (5.5–6.2) | 4.8 (3.8–5.7) |
| Female sterilization | 39.8 (39.0–40.6) | 38.3 (37.4–39.2) | 42.6 (41.3–44.0) | 40.3 (39.5–41.1) | 34.3 (32.3–36.4) |
| Male sterilization | 1.0 (0.9–1.2) | 1.0 (0.8–1.1) | 1.1 (0.8–1.4) | 1.0 (0.9–1.2) | 1.0 (0.6–1.4) |
| Periodic abstinence | 5.4 (5.1–5.8) | 5.7 (5.3–6.1) | 4.8 (4.3–5.4) | 5.4 (5.1–5.7) | 5.6 (4.5–6.7) |
| Withdrawal | 2.8 (2.6–3.0) | 3.0 (2.7–3.2) | 2.4 (2.0–2.7) | 2.6 (2.4–2.8) | 4.4 (3.5–5.3) |
| Nepal | |||||
| None | 46.7 (43.8–49.6) | 46.9 (43.6–50.1) | 46.1 (40.2–52.0) | 47.8 (44.7–50.9) | 39.8 (33.7–45.9) |
| Pill | 4.1 (3.2–5.0) | 4.3 (3.2–5.4) | 3.3 (1.9–4.7) | 4.3 (3.2–5.3) | 3.0 (1.3–4.6) |
| IUD | 1.5 (0.9–2.2) | 1.5 (0.9–2.2) | 1.5 (0.4–2.7) | 1.4 (0.7–2.1) | 2.2 (0.6–3.8) |
| Injections | 10.7 (9.1–12.2) | 10.7 (9.0–12.3) | 10.6 (7.7–13.5) | 10.6 (9.0–12.3) | 10.9 (6.9–14.8) |
| Condom | 5.1 (4.2–6.0) | 6.1(5.0–7.2) | 1.6 (0.7–2.4) | 5.4 (4.3–6.4) | 3.3 (1.6–4.9) |
| Female sterilization | 14.5 (12.2–16.8) | 12.3 (10.1–14.5) | 21.7 (17.3–26.2) | 12.6 (10.5–14.8) | 25.9 (19.8–32.1) |
| Male sterilization | 9.5 (7.7–11.3) | 9.8 (7.7–11.8) | 8.7 (5.4–12.1) | 10.0 (8.0–12.0) | 6.5 (3.6–9.4) |
| Periodic abstinence | 1.4 (0.9–2.0) | 1.6 (0.9–2.4) | 0.7 (0.1–1.4) | 1.5 (0.9–2.2) | 0.8 (0.0–1.7) |
| Withdrawal | 5.4 (4.2–6.5) | 5.7 (4.4–7.0) | 4.3 (2.2–6.4) | 5.2 (4.0–6.4) | 6.3 (2.9–9.6) |
Abbreviation IUD, intrauterine device.
Values given as percentage (95% confidence interval).
Methods reported by <1% across all assessed countries were not included in the present table. These methods included diaphragm, foam, implant, female condom, other modern methods, and other methods.
4. Discussion
The present study revealed that sexual marital violence affected one in 10 women in a cohort from South Asia. Such sexual marital violence was associated with an increased likelihood of spacing contraceptive use but a reduced likelihood of sterilization. These findings clarified previous research that documented a positive association between marital violence and contraceptive use (particularly oral contraceptives) [12], but a negative association where sterilization was the predominant means of birth control [5,6,11]. Contrary to prior studies [5–7,11,12], physical marital violence was not associated with contraceptive use in the present study, possibly owing to the model adjusting for sexual marital violence. Previous research from India found that an association of physical marital violence with contraception was apparent only in situations where there was also sexual marital violence; however, the prevalence of sexual marital violence alone was too rare in that study to provide a reliable estimate [11]. The present findings suggested that sexual marital violence could affect contraceptive use to a greater extent than physical marital violence, and that women who experienced sexual marital violence might have an increased tendency to access reproductive health services, perhaps based on greater perceived risk for unintended pregnancy. Such findings are consistent with other research indicating greater likelihood of use of female-controlled contraception [12,25–28] among women experiencing male partner violence.
Findings of a reduced likelihood of sterilization among women with a history of sexual marital violence require further analysis. Spacing contraception might feel a safer or more manageable option for women in the context of marital violence, which has been linked with a desire for a large number of children among men [18]. In the country-specific sensitivity analyses for both Bangladesh and Nepal, no association between marital violence and sterilization was observed. Country-specific considerations of differences in access and motivations for sterilization might offer some insight into this observation. In contrast to the conflicting findings seen for sterilization, the data for condom use in the present study indicated low use of this method in the context of marital violence, across both pooled and country-specific models. These findings reinforced prior research documenting that abusive male partners are less likely to use condoms with their wives [2,29].
The descriptive analyses indicated that sexual marital violence was associated with an increased likelihood of pill utilization among the present study cohort. South Asian women experiencing sexual marital violence might have more reproductive control via contraceptives not easily subject to their husband’s knowledge, approval, or assistance. A similar finding has been reported from Jordan [28]. However, other issues could have been at play given that withdrawal—a traditional form of contraception within the man’s control—was also more likely to be reported by the women with a history of sexual marital violence in the present study. Additional research, including qualitative studies, is needed to understand the contraceptive decision-making and practices of women experiencing sexual violence in marriage. Very low use of the pill (India and Nepal), contraceptive injection (India and Bangladesh) and other forms of effective spacing contraception such as IUDs and implants (all three countries) could potentially obscure full understanding of the observed associations between marital violence and contraception. Nonetheless, there is a clear need for sustainable interventions centered on reproductive health care to identify and assist women experiencing sexual marital violence.
Other gender equity issues were also associated with contraceptive use in the region. Son preference ideology was associated with increased use of contraceptives, and a greater effect of multiple boys versus multiple girls on contraceptive practices was observed, in agreement with other studies from South Asia [30–34]. Such an effect might be contributing to the sex ratio imbalance prevalent in the region [35, 36]. Although supporting women and families to use contraception should be a part of family planning programs, inadvertent reduction of the relative numbers of girls to boys might require more careful consideration in the context of South Asia.
Limitations of the present study included restriction of the regional analysis to just three countries with data not collected within the same timeframe; consequently, the results of pooled analyses might not be consistent with nation-specific findings. Disproportionate representation of India in the pooled data potentially skewed the findings. Sensitivity analyses were conducted to address this issue, but the country-specific multivariate models for Bangladesh and Nepal, as well as the multivariate model without India, offered unstable estimates owing to inadequate power, which inhibited the ability to make valid conclusions at the country level. The India-specific model provided results comparable to the pooled model but, as noted above, this effect might reflect disproportionate representation. Pooled regional analyses did, however, allow adequate power to explore effects that were insufficiently powered for analysis at the country-level, particularly for sexual marital violence. Confirmation of the findings at the national level requires additional data from the region, possibly across multiple years.
The differing time periods for the DHS assessments represented a further limitation of the present study. The measure of marital violence included all experiences of violence from the current spouse, whereas assessments of contraception largely related to current behavior. Consequently, it was not possible to discern how differing timing of marital violence affected contraceptive use. No variable was available to assess current desire to become pregnant, an important predictor of contraceptive use among women. Covariates were designed to adjust for social inequities but indicators of access to health care were not available in the DHS datasets. Multivariate analyses could not be conducted for each form of contraception owing to small cell sizes; however, more robust analyses might be possible when multi-year data with marital violence measures become available for these countries. Finally, DHS data are susceptible to social desirability and recall biases and are cross-sectional in nature; thus, causal relationships could not be inferred in the present study.
5. Conclusion
The present regional analysis found that marital sexual violence affected one in ten women and decreased the likelihood of sterilization and condom use, the leading means of modern contraception in South Asia. By contrast, marital sexual violence was associated with increased likelihood of oral contraceptive pill use, which can be used surreptitiously by affected women. The results support the need to prioritize improved access to contraceptive services and choices in South Asia, and to address marital violence within the context of reproductive health care services.
Acknowledgments
The authors’ time with regard to all aspects of manuscript development and submission was funded by a grant from the David and Lucile Packard Foundation to the principal investigator (A.J; grant #2011-37366).
Footnotes
Conflict of interest
The authors have no conflicts of interests to declare.
References
- 1.Devries KM, Kishor S, Johnson H, Stöckl H, Bacchus LJ, Garcia-Moreno C, et al. Intimate partner violence during pregnancy: analysis of prevalence data from 19 countries. Reprod Health Matters. 2010;18(36):158–170. doi: 10.1016/S0968-8080(10)36533-5. [DOI] [PubMed] [Google Scholar]
- 2.World Health Organization, Department of Reproductive Health and Research, London School of Hygiene and Tropical Medicine, South African Medical Research Council. Global and Regional Estimates of Violence Against Women. [Accessed February 18, 2015];Prevalence and Health Effects of Intimate Partner Violence and Non-Partner Sexual Violence. http://www.who.int/reproductivehealth/publications/violence/9789241564625/en/index.html. Published 2013.
- 3.Miller E, Jordan B, Levenson R, Silverman JG. Reproductive coercion: connecting the dots between partner violence and unintended pregnancy. Contraception. 2010;81(6):457–459. doi: 10.1016/j.contraception.2010.02.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Miller E, Decker MR, McCauley HL, Tancredi DJ, Levenson RR, Waldman J, et al. A family planning clinic partner violence intervention to reduce risk associated with reproductive coercion. Contraception. 2011;83(3):274–280. doi: 10.1016/j.contraception.2010.07.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Stephenson R, Koenig MA, Acharya R, Roy TK. Domestic violence, contraceptive use, and unwanted pregnancy in rural India. Stud Fam Plann. 2008;39(3):177–186. doi: 10.1111/j.1728-4465.2008.165.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Stephenson R, Jadhav A, Hindin M. Physical domestic violence and subsequent contraceptive adoption among women in rural India. J Interpers Violence. 2013;28(5):1020–1039. doi: 10.1177/0886260512459379. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Begum S, Dwivedi SN, Pandey A, Mittal S. Association between domestic violence and unintended pregnancies in India: findings from the National Family Health Survey-2 data. Natl Med J India. 2010;23(4):198–200. [PubMed] [Google Scholar]
- 8.Silverman JG, Gupta J, Decker MR, Kapur N, Raj A. Intimate partner violence and unwanted pregnancy, miscarriage, induced abortion, and stillbirth among a national sample of Bangladeshi women. BJOG. 2007;114(10):1246–1252. doi: 10.1111/j.1471-0528.2007.01481.x. [DOI] [PubMed] [Google Scholar]
- 9.Yoshikawa K, Agrawal NR, Poudel KC, Jimba M. A lifetime experience of violence and adverse reproductive outcomes: findings from population surveys in India. Biosci Trends. 2012;6(3):115–121. doi: 10.5582/bst.2012.v6.3.115. [DOI] [PubMed] [Google Scholar]
- 10.Rahman M, Sasagawa T, Fujii R, Tomizawa H, Makinoda S. Intimate partner violence and unintended pregnancy among Bangladeshi women. J Interpers Violence. 2012;27(15):2999–3015. doi: 10.1177/0886260512441072. [DOI] [PubMed] [Google Scholar]
- 11.Kishor S, Gupta K. Mumbai: International Institute for Population Sciences; Calverton, Maryland, USA: ICF Macro; [Accessed February 23, 2015]. Gender Equality and Women’s Empowerment n India. National Family Health Survey (NFHS-3), India, 2005–06. http://pdf.usaid.gov/pdf_docs/PNADQ633.pdf. [Google Scholar]
- 12.Dalal K, Andrews J, Dawad S. Contraception use and associations with intimate partner violence among women in Bangladesh. J Biosoc Sci. 2012;44(1):83–94. doi: 10.1017/S0021932011000307. [DOI] [PubMed] [Google Scholar]
- 13.National Institute of Population Research and Training, Mitra and Associates, Macro International. Bangladesh Demographic and Health Survey 2007. [Accessed February 18, 2015]; http://dhsprogram.com/pubs/pdf/FR207/FR207%5BApril-10-2009%5D.pdf. Published 2009. [Google Scholar]
- 14.International Institute for Population Sciences, Macro International. National Family Health Survey (NFHS-3) 2005–06 India Volume I. [Accessed February 18, 2015]; https://dhsprogram.com/pubs/pdf/FRIND3/FRIND3-Vol1AndVol2.pdf. Published 2007.
- 15.Ministry of Health and Population, New ERA, ICF International. Nepal Demographic and Health Survey 2011. [Accessed February 18, 2015]; http://dhsprogram.com/pubs/pdf/FR257/FR257%5B13April2012%5D.pdf. Published 2012.
- 16.Wang W, Alva S, Winter R, Burgert C. Contextual Influences of Modern Contraceptive Use Among Rural Women in Rwanda and Nepal. DHS Analytical Studies 41. [Accessed February 18, 2015];USAID Report. 2013 http://pdf.usaid.gov/pdf_docs/pnaec676.pdf; Published 2013.
- 17.Kishor S, Johnson K. Profiling Domestic Violence. [Accessed February 18, 2015];A Multi-Country Study. http://dhsprogram.com/pubs/pdf/OD31/OD31.pdf. Published 2004. [Google Scholar]
- 18.Snow RC, Winter RA, Harlow SD. Gender attitudes and fertility aspirations among young men in five high fertility East African countries. Stud Fam Plann. 2013;44(1):1–24. doi: 10.1111/j.1728-4465.2013.00341.x. [DOI] [PubMed] [Google Scholar]
- 19.Pallitto CC, García-Moreno C, Jansen HA, Heise L, Ellsberg M, Watts C, et al. Intimate partner violence, abortion and unintended pregnancy: results from the WHO Multi-country Study on Women’s Health and Domestic Violence. Int J Gynecol Obstet. 2013;120(1):3–9. doi: 10.1016/j.ijgo.2012.07.003. [DOI] [PubMed] [Google Scholar]
- 20.Hall M, Chappell LC, Parnell BL, Seed PT, Bewley S. Associations between intimate partner violence and termination of pregnancy: a systematic review and meta-analysis. PLoS Med. 2014;11(1):e1001581. doi: 10.1371/journal.pmed.1001581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.ICF International. MEASURE DHS. Calverton, Maryland: ICF International; 2012. Survey Organization Manual for Demographic and Health Surveys. [Google Scholar]
- 22.Hindin MJ, Kishor S, Ansara DL. Intimate Partner Violence Among Couples in 10 DHS Countries: Predictors and Health Outcomes. [Accessed February 18, 2015];DHS Analytical Studies No. 18. http://www.measuredhs.com/pubs/pdf/AS18/AS18.pdf. Published 2008. [Google Scholar]
- 23.Rapp D, Zoch B, Khan MM, Pollmann T, Krämer A. Association between gap in spousal education and domestic violence in India and Bangladesh. BMC Public Health. 2012;12:467. doi: 10.1186/1471-2458-12-467. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kimuna SR, Djamba YK, Ciciurkaite G, Cherukuri S. Domestic violence in India: insights from the 2005–2006 national family health survey. J Interpers Violence. 2013;28(4):773–807. doi: 10.1177/0886260512455867. [DOI] [PubMed] [Google Scholar]
- 25.Roberts TA, Auinger P, Klein JD. Intimate partner abuse and the reproductive health of sexually active female adolescents. J Adolesc Health. 2005;36(5):380–385. doi: 10.1016/j.jadohealth.2004.06.005. [DOI] [PubMed] [Google Scholar]
- 26.Fanslow J, Whitehead A, Silva M, Robinson E. Contraceptive use and associations with intimate partner violence among a population-based sample of New Zealand women. Aust N Z J Obstet Gynaecol. 2008;48(1):83–89. doi: 10.1111/j.1479-828X.2007.00805.x. [DOI] [PubMed] [Google Scholar]
- 27.Alio A, Daley EM, Nana PN, Duan J, Salihu HM. Intimate partner violence and contraception use among women in Sub-Saharan Africa. Int J Gynecol Obstet. 2009;107(1):35–38. doi: 10.1016/j.ijgo.2009.05.002. [DOI] [PubMed] [Google Scholar]
- 28.O'Hara K, Tsai LC, Carlson CE, Haidar YM. Experiences of intimate-partner violence and contraception use among ever-married women in Jordan. East Mediterr Health J. 2013;19(10):876–882. [PubMed] [Google Scholar]
- 29.Silverman JG, Decker MR, Saggurti N, Balaiah D, Raj A. Intimate partner violence and HIV infection among married Indian women. JAMA. 2008;300(6):703–710. doi: 10.1001/jama.300.6.703. [DOI] [PubMed] [Google Scholar]
- 30.Diamond-Smith N, Luke N, McGarvey S. 'Too many girls, too much dowry': son preference and daughter aversion in rural Tamil Nadu, India. Cult Health Sex. 2008;10(7):697–708. doi: 10.1080/13691050802061665. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Rai P, Paudel IS, Ghimire A, Pokharel PK, Rijal R, Niraula SR. Effect of gender preference on fertility: cross-sectional study among women of Tharu community from rural area of eastern region of Nepal. Reprod Health. 2014;11(1):15. doi: 10.1186/1742-4755-11-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Raj A, Vilms RJ, McDougal L, Silverman JG. Association between having no sons and using no contraception among a nationally representative sample of young wives in Nepal. Int J Gynecol Obstet. 2013;121(2):162–165. doi: 10.1016/j.ijgo.2012.12.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Edmeades J, Pande R, Macquarrie K, Falle T, Malhotra A. Two sons and a daughter: sex composition and women's reproductive behaviour in Madhya Pradesh, India. J Biosoc Sci. 2012;44(6):749–764. doi: 10.1017/S0021932012000119. [DOI] [PubMed] [Google Scholar]
- 34.Jayaraman A, Mishra V, Arnold F. The relationship of family size and composition to fertility desires, contraceptive adoption and method choice in South Asia. Int Perspect Sex Reprod Health. 2009;35(1):29–38. doi: 10.1363/ifpp.35.029.09. [DOI] [PubMed] [Google Scholar]
- 35.Guilmoto CZ. Fourth Asia-Pacific Conference on Reproductive and Sexual Health and Rights. [Accessed February 26, 2015];2007 Oct 29–31; http://www.unfpa.org/sites/default/files/resource-pdf/india_0.pdf; [Google Scholar]
- 36.Guilmoto CZ. The sex ratio transition in Asia. Popul Dev Rev. 2009;35:519–549. [Google Scholar]