Abstract
Objective
To assess associations of intimate partner violence (IPV) with pregnancy intendedness and pre-pregnancy contraceptive use among pregnant women in South Asia.
Study Design
Cross-sectional analyses were conducted using most recent Demographic and Health Surveys (DHS) from Bangladesh, India and Nepal for married, pregnant women aged 15–49 years who responded to IPV assessments specific to current marriage (N=4738). Adjusted logistic and multinomial regression analyses were conducted with pooled data to assess associations of IPV ever (sexual only, physical only, sexual plus physical, none) with the outcomes of pregnancy intendedness (wanted, mistimed or unwanted) and pre-pregnancy contraceptive use (no, traditional, or modern), for the current pregnancy.
Results
IPV was not associated with a mistimed or unwanted pregnancy. Sexual IPV was associated with pre-pregnancy modern contraceptive use (aOR=2.32, 95% CI=1.24, 4.36); sexual plus physical IPV was associated with pre-pregnancy traditional contraceptive use (aOR=1.85, 95% CI=1.12, 3.07). Post-hoc analysis of reasons for pre-pregnancy contraceptive discontinuation revealed that women with a history of IPV, particularly sexual IPV, had higher prevalence of contraceptive failure (sexual only: 37.3%, sexual plus physical: 30.9%, physical only: 22.6%, no IPV: 13.6%).
Conclusion
Pregnant women who experienced sexual IPV from husbands were more likely to use contraceptives pre-pregnancy but had no reduced risk unintended pregnancy, possibly due to higher rates of pre-pregnancy contraceptive failure among those with this history. These findings suggest that victims of sexual IPV are able to acquire and use family planning services, but require more support to sustain effective contraceptive use.
Implications
Family planning services are reaching women affected by sexual IPV, and programs should be sensitive to this concern and the heightened vulnerability to contraceptive failure these women face. Long-acting reversible contraception could be beneficial by allowing women to have greater reproductive control in situations of compromised sexual autonomy.
Keywords: intimate partner violence, unintended pregnancy, contraceptive use, South Asia
Introduction
South Asia, which includes the sub-Himalayan and neighboring nations of Bangladesh, Bhutan, India, Nepal, Pakistan and Sri Lanka, has among the highest rates of intimate partner violence (IPV) in the world, with one in three women reporting sexual and/or physical IPV, predominantly from a husband [1]. Studies from the region also document that such IPV is associated with increased risk for unintended pregnancy [2–7], which in turn compromises maternal, infant and child health [8]. Much of this work is cross-sectional, limiting assumptions of causality. Nonetheless, it is presumed that poorer reproductive control among women contending with IPV impedes pregnancy prevention in the form of contraceptive use, thereby increasing risk for unintended pregnancy [1, 7, 9]. Studies from diverse regions have found that women reporting IPV are less likely to use contraception, supporting this hypothesis [2, 6, 7, 9, 10, 11]. However, there is conflicting evidence on this point, with recent research from South Asia indicating that IPV is associated with greater likelihood of contraceptive use [12, 13]. Little research has quantitatively examined how IPV affects contraception and subsequent unwanted pregnancy; such work would help support considerations of causality.
The purpose of this study is to assess associations of IPV with pregnancy intendedness and pre-pregnancy contraception, specific to the current pregnancy, among a representative sample of pregnant women from South Asian nations with population-level data available on the issue: Bangladesh, India and Nepal, These three countries of focus all have public health systems that include reproductive health and family planning services, as well as free contraceptives; legal access to safe abortion is also available. Contraceptive prevalence is comparable across the three countries, with only approximately half of the women of childbearing age in Bangladesh (52%), India (49%), and Nepal (43%) reporting modern contraceptive use; the majority of these contraceptors used public health services to acquire their contraceptives, at no or low cost [14–16]. Findings from this work can help clarify prior conflicting research on this topic in the context of South Asia, and guide family planning programs on approaches that attend to IPV.
Methods
Sample
This study uses data from the Demographic and Health Surveys, which are nationally representative household surveys assessing various measures of population health and nutrition. Data were limited to the most recent standard DHS from countries in South Asia that included a domestic violence module assessing both physical and sexual violence, and was thus inclusive of Bangladesh (2007) [14], India (2005–06) [15] and Nepal (2011) [16]. Pakistan’s 2012–13 DHS [17] could not be used as it did not collect data on sexual partner violence. Eligible women response rates were above 94% in all three surveys [14–16]. The sample was further restricted to currently married women who were pregnant at the time of survey, had completed the domestic violence module, and had responses for all dependent and independent variables (n=4,738).
Measures
The primary dependent variables were intendedness of current pregnancy and pre-pregnancy contraceptive use, specific to the current pregnancy. Intendedness of current pregnancy was categorized as wanted (intended at current time) vs. mistimed (wanted later) vs. unwanted (not wanted at all). Pre-pregnancy contraceptive use, derived from the reproductive calendar, was defined as the use of modern contraceptives (pill, IUD, injections, diaphragm, condom, implant, female condom, foam/jelly), traditional contraception (lactational amenorrhea, periodic abstinence, withdrawal, abstinence or other traditional methods) or no contraception. Usage had to be prior to the most recent pregnancy and subsequent to the penultimate pregnancy, or 5 years before to the interview, whichever window was shorter. Modern pre-pregnancy contraception was also broken down by method for descriptive analyses using data from India and Nepal; these data were not available for Bangladesh.
The primary independent variable was ever having experienced sexual and/or physical IPV in current marital relationship (sexual only/physical only/sexual plus physical/none). Sexual IPV was defined as having ever been physically forced to have sex or to perform sexual acts against one’s will (2 items, summated and dichotomized as yes/no). Physical IPV was defined as having ever been pushed, shaken or had something thrown at you, slapped, had one’s arm twisted or hair pulled, punched, kicked, dragged or beaten up, choked or burned, or threatened with a weapon (6 items, summated and dichotomized as yes/no). Sexual and physical IPV variables were then used to create the outcome variable as a combination score.
Covariates included: number of living children (0/1/2/3 or more), months elapsed since the commencement of the index pregnancy, age at interview (15–19/20–24/25–29/30–49), education and spousal education (none/primary/secondary or higher), age at marriage (<15/15–17/≥18), relationship to household head (head/wife/daughter/daughter-in-law/other), household wealth quintile and urban/rural residence. Household wealth quintile is a relative measure computed based on a principal components analysis of household assets and housing characteristics [18].
Data Analysis
Descriptive frequencies were calculated for all covariates. Unadjusted and adjusted logistic regressions were modeled for each dependent variable. All models included country fixed effects. Adjusted models also included all variables that were significant at p<0.20 in unadjusted regressions, as well as contraception for the unintended pregnancy model. No multicollinearity was detected in the final model using a tolerance cutoff of 0.30. All analyses accounted for complex survey design and were weighted using denormalized individual weights to account for variation in country size. Analyses were conducted using SAS 9.3. An exploratory analysis of the reasons for contraceptive discontinuation was conducted among women who provided these data in India and Nepal (n=1,032); Bangladesh did not collect this information.
Results
One-fourth (24.6%) of women reported a history of only physical IPV; 2.8% reported a history of only sexual IPV, and 8.6% of women reported both sexual and physical IPV. (Table 1.) One-fourth (25.7%) reported their current pregnancy as unintended; 15.8% reported wanting the pregnancy later, and 9.9% reported not wanting the pregnancy at all. Most participants (76.0%) reported no contraceptive use prior to their current pregnancy; 16.8% reported a modern method, and 7.2% reported a traditional method. Pill and condom were the most common forms of modern contraception reported (8.5% and 7.4%, respectively).
Table 1.
Descriptive characteristics of currently pregnant women in Bangladesh (2007), India (2005–06) and Nepal (2011) (n=4,738).
| Total | Current pregnancy wanted | Current pregnancy mistimed | Current pregnancy unwanted | |||||
|---|---|---|---|---|---|---|---|---|
| Unwted N | %(95% CI) | Unwted N | % (95% CI) | Unwted N | % (95% CI) | Unwted N | % (95% CI) | |
| Total | 4,738 | - | 3,513 | 74.3% (72.5, 76.2) | 724 | 15.8% (14.2, 17.4) | 501 | 9.9% (8.6, 11.1) |
| Country | ||||||||
| Bangladesh | 292 | 9.5%(8.2, 10.7) | 181 | 8.1% (6.7, 9.5) | 64 | 14.2% (10.2, 18.2) | 47 | 12.1% (8.4, 15.9) |
| India | 4,234 | 89.5%(88.3, 90.8) | 3,190 | 91.0% (89.6, 92.4) | 629 | 84.7% (80.6, 88.7) | 415 | 86.2% (82.4, 90.0) |
| Nepal | 212 | 1.0%(0.8, 1.2) | 142 | 0.9% (0.7, 1.1) | 31 | 1.1% (0.5, 1.7) | 39 | 1.7% (1.0, 2.3) |
| Pre-pregnancy contraceptive use | ||||||||
| None | 3,479 | 76.0% (74.1, 77.8) | 2,666 | 79.2% (77.2, 81.1) | 486 | 67.9% (62.6, 73.1) | 327 | 64.9% (58.8, 71.1) |
| Traditional | 339 | 7.2% (6.1, 8.3) | 246 | 6.7% (5.5, 8.0) | 58 | 9.5% (6.1, 12.9) | 35 | 7.1% (4.0, 10.1) |
| Modern | 920 | 16.8% (15.3, 18.3) | 601 | 14.1% (12.5, 15.6) | 180 | 22.7% (18.0, 27.3) | 139 | 28.0% (22.2, 33.8) |
| Pre-pregnancy modern contraceptive use by method* | ||||||||
| Pill | 424 | 8.5% (7.3, 9.6) | 248 | 6.4% (5.3, 7.6) | 101 | 14.2% (10.3, 18.1) | 75 | 14.8% (10.8, 18.7) |
| IUD | 107 | 1.3% (0.9, 1.7) | 84 | 1.3% (0.9, 1.7) | 9 | 0.9% (0.0, 1.7) | 14 | 2.4% (0.8, 4.1) |
| Injections | 87 | 1.6% (1.2, 2.1) | 48 | 1.1% (0.6, 1.5) | 12 | 2.0% (0.5, 3.4) | 27 | 5.4% (2.9, 7.9) |
| Condom | 401 | 7.4% (6.4, 8.5) | 275 | 6.7% (5.6, 7.9) | 87 | 9.9% (6.8, 13.0) | 39 | 8.9% (4.1, 13.7) |
| Implants | 2 | 0.0% (0.0, 0.0) | 2 | 0.0% (0.0, 0.0) | 0 | - | 0 | - |
| Ever IPV | ||||||||
| None | 3,274 | 63.9% (61.8, 66.0) | 2,525 | 66.4% (64.0, 68.8) | 487 | 61.4% (56.0, 66.7) | 262 | 49.3% (43.2, 55.4) |
| Sexual only | 112 | 2.8% (2.1, 3.6) | 73 | 2.8% (1.8, 3.7) | 22 | 3.7% (1.8, 5.6) | 17 | 2.0% (0.8, 3.3) |
| Physical only | 1,030 | 24.6% (22.7, 26.5) | 714 | 23.1% (20.9, 25.3) | 152 | 24.8% (20.0, 29.6) | 164 | 35.6% (30.0, 41.3) |
| Sexual and physical | 322 | 8.6% (7.4, 9.9) | 201 | 7.7% (6.3, 9.1) | 63 | 10.2% (6.7, 13.6) | 58 | 13.0% (8.6, 17.4) |
| Number of living children | ||||||||
| 0 | 1,584 | 36.4% (34.3, 38.5) | 1,377 | 42.1% (39.6, 44.6) | 193 | 30.2% (25.0, 35.3) | 14 | 2.9% (0.5, 5.3) |
| 1 | 1,427 | 29.2% (27.3, 31.1) | 1,092 | 29.4% (27.2, 31.6) | 282 | 40.8% (35.6, 46.0) | 53 | 9.3% (6.0, 12.6) |
| 2 | 831 | 16.3% (14.8, 17.9) | 567 | 15.4% (13.6, 17.2) | 135 | 15.0% (11.6, 18.4) | 129 | 25.4% (19.6, 31.3) |
| 3 or more | 896 | 18.1% (16.6, 19.6) | 477 | 13.1% (11.5, 14.6) | 114 | 14.0% (10.6, 17.4) | 305 | 62.4% (56.1, 68.7) |
| Months since first month of index pregnancy | 4,738 | 5.4 (5.3, 5.5)1 | 3,513 | 5.5 (5.3, 5.6) 1 | 724 | 5.1 (4.8, 5.4) 1 | 501 | 5.6 (5.3, 5.9) |
| Age at interview | ||||||||
| 15–19 | 729 | 21.5% (19.6, 23.5) | 561 | 22.6% (20.4, 24.9) | 157 | 27.7% (22.7, 32.7) | 11 | 3.6% (0.9, 6.3) |
| 20–24 | 1,879 | 42.8% (40.8, 44.8) | 1,447 | 44.1% (41.7, 46.4) | 319 | 46.4% (41.0, 51.7) | 113 | 27.4% (21.8, 33.1) |
| 25–29 | 1,325 | 23.2% (21.5, 24.9) | 965 | 22.6% (20.6, 24.5) | 175 | 20.1% (16.2, 24.1) | 185 | 33.1% (27.6, 38.5) |
| 30–49 | 805 | 12.4% (11.2, 13.7) | 540 | 10.7% (9.4, 12.1) | 73 | 5.8% (3.7, 8.0) | 192 | 35.9% (30.2, 41.7) |
| Education level | ||||||||
| None | 1,796 | 44.3% (42.1, 46.6) | 1,296 | 44.7% (42.0, 47.4) | 240 | 35.4% (30.3, 40.5) | 260 | 48.2% (42.8, 53.7) |
| Primary | 718 | 15.3% (13.8, 16.8) | 503 | 14.7% (12.9, 16.4) | 117 | 16.4% (12.3, 20.4) | 98 | 18.4% (13.7, 23.0) |
| Secondary or higher | 2,224 | 40.4% (38.2, 42.6) | 1,714 | 40.7% (38.1, 43.2) | 367 | 48.2% (42.8, 53.7) | 143 | 25.7% (20.2, 31.2) |
| Spousal education level | ||||||||
| None | 1,137 | 27.5% (25.4, 29.6) | 815 | 26.9% (24.7, 29.2) | 157 | 24.6% (19.8, 29.4) | 165 | 36.7% (30.6, 42.8) |
| Primary | 792 | 16.6% (15.1, 18.2) | 530 | 15.4% (13.7, 17.2) | 155 | 21.4% (17.0, 25.8) | 107 | 17.9% (13.3, 22.5) |
| Secondary or higher | 2,809 | 55.9% (53.7, 58.1) | 2,168 | 57.6% (55.2, 60.1) | 412 | 54.0% (48.5, 59.5) | 229 | 45.4% (39.1, 51.7) |
| Age at marriage | ||||||||
| <15 | 697 | 19.0% (17.2, 20.7) | 484 | 18.3% (16.4, 20.3) | 90 | 13.1% (9.8, 16.4) | 123 | 33.5% (27.2, 39.8) |
| 15–17 | 1,572 | 39.3% (37.3, 41.4) | 1,133 | 39.0% (36.6, 41.4) | 247 | 41.8% (36.2, 47.4) | 192 | 37.5% (31.7, 43.2) |
| ≥18 | 2,469 | 41.7% (39.6, 43.8) | 1,896 | 42.6% (40.2, 45.1) | 387 | 45.1% (39.6, 50.6) | 186 | 29.1% (23.3, 34.8) |
| Relationship to household head | ||||||||
| Head | 160 | 3.5% (2.8, 4.3) | 106 | 3.2% (2.4, 4.0) | 22 | 2.4% (1.0, 3.8) | 32 | 8.1% (4.6, 11.5) |
| Wife | 2,722 | 44.3% (42.1, 46.4) | 1,974 | 42.3% (40.0, 44.7) | 378 | 40.0% (35.0, 44.9) | 370 | 65.7% (59.1, 72.3) |
| Daughter | 332 | 12.2% (10.5, 13.8) | 262 | 13.1% (11.2, 15.1) | 57 | 12.3% (8.3, 16.4) | 13 | 4.4% (1.5, 7.2) |
| Daughter-in-law | 1,328 | 33.6% (31.4, 35.8) | 1,026 | 35.4% (33.0, 37.9) | 228 | 35.3% (30.1, 40.6) | 74 | 17.0% (11.7, 22.2) |
| Other | 196 | 6.4% (5.2, 7.7) | 145 | 5.9% (4.5, 7.3) | 39 | 10.0% (6.4, 13.6) | 12 | 4.9% (1.1, 8.8) |
| Wealth quintile | ||||||||
| Poorest | 893 | 24.6% (22.6, 26.6) | 634 | 25.1% (22.8, 27.4) | 117 | 19.4% (15.4, 23.4) | 142 | 29.4% (23.9, 34.8) |
| Poorer | 885 | 22.2% (20.3, 24.1) | 623 | 21.4% (19.2, 23.6) | 152 | 23.9% (19.1, 28.7) | 110 | 25.5% (19.8, 31.1) |
| Middle | 966 | 19.8% (18.0, 21.6) | 688 | 18.7% (16.8, 20.7) | 169 | 23.1% (18.5, 27.8) | 109 | 22.3% (16.8, 27.8) |
| Richer | 1,051 | 19.3% (17.5, 21.0) | 792 | 19.5% (17.5, 21.6) | 173 | 21.1% (16.4, 25.8) | 86 | 14.4% (10.3, 18.5) |
| Richest | 943 | 14.1% (12.7, 15.5) | 776 | 15.2% (13.5, 16.9) | 113 | 12.5% (9.2, 15.8) | 54 | 8.4% (4.6, 12.3) |
| Residence | ||||||||
| Rural | 2,931 | 75.4% (73.6, 77.2) | 2,160 | 76.0% (74.0, 78.1) | 444 | 74.0% (69.6, 78.5) | 327 | 73.3% (67.4, 79.1) |
| Urban | 1,807 | 24.6% (22.8, 26.4) | 1,353 | 24.0% (21.9, 26.0) | 280 | 26.0% (21.5, 30.4) | 174 | 26.7% (20.9, 32.6) |
Note: Percentages are weighted.
Mean (95% CI)
Assessment of forms of pre-pregnancy contraception also included options of diaphragm, female condom and foam/jelly; no respondents reported use of these forms.
History of sexual and/or physical IPV from spouse were not associated with mistimed or unwanted pregnancy in adjusted analyses. Pre-pregnancy modern contraceptive use, however, was positively associated with both mistimed pregnancy (aOR=1.42, 95% CI=1.01, 2.00) and unwanted pregnancy (aOR=2.20, 95% CI=1.48, 3.26) in adjusted analyses (Table 2). Additional covariates associated with mistimed pregnancy were: greater number of children, younger age, secondary or higher relative to no education, older age at marriage, and having an “other” position in the household (e.g., not being head of household, or wife, daughter or daughter-in-law of head of household). Additional covariates associated with unwanted pregnancy were: greater number of children, education, and having a household role of head, wife of head or “other” (relative to being the daughter in law of the household).
Table 2.
Unadjusted and adjusted models assessing odds of mistimed and unwanted pregnancies (compared to wanted pregnancies) among currently pregnant women living in Bangladesh, India and Nepal (n=4,738).
| Current pregnancy mistimed | Current pregnancy unwanted | Current pregnancy mistimed | Current pregnancy unwanted | |
|---|---|---|---|---|
| OR1 (95% CI) | OR1 (95% CI) | aOR2 (95% CI) | aOR2 (95% CI) | |
| Ever IPV | ||||
| None | 1.00 | 1.00 | 1.00 | 1.00 |
| Sexual only | 1.40 (0.71, 2.76) | 0.97 (0.47, 2.02) | 1.37 (0.72, 2.63) | 1.20 (0.54, 2.66) |
| Physical only | 1.15 (0.87, 1.53) | 2.08 (1.55, 2.79) | 1.17 (0.86, 1.58) | 1.30 (0.94, 1.81) |
| Sexual and Physical | 1.38 (0.91, 2.09) | 2.22 (1.40, 3.54) | 1.41 (0.89, 2.22) | 1.38 (0.87, 2.20) |
| Pre-Pregnancy Contraception | ||||
| None | 1.00 | 1.00 | 1.00 | 1.00 |
| Traditional | 1.64 (1.05, 2.56) | 1.28 (0.76, 2.14) | 1.53 (0.96, 2.43) | 1.18 (0.67, 2.06) |
| Modern | 1.67 (1.21, 2.30) | 2.35 (1.69, 3.29) | 1.42 (1.01, 2.00) | 2.20 (1.48, 3.26) |
| Number of living children | ||||
| 0 | 1.00 | 1.00 | 1.00 | 1.00 |
| 1 | 1.95 (1.45, 2.62) | 4.65 (1.82, 11.87) | 3.11 (2.20, 4.38) | 4.14 (1.67, 10.26) |
| 2 | 1.36 (0.95, 1.94) | 24.20 (9.63, 60.82) | 3.52 (2.23, 5.54) | 25.97 (10.62, 63.52) |
| 3 or more | 1.56 (1.08, 2.25) | 73.25 (30.35, 176.83) | 7.28 (4.27, 12.40) | 93.22 (37.45, 232.07) |
| Time since first month of index pregnancy (months) | 0.95 (0.90, 0.997) | 1.03 (0.97, 1.10) | 0.96 (0.91, 1.01) | 1.00 (0.93, 1.07) |
| Age at interview | ||||
| 15–19 | 1.00 | 1.00 | 1.00 | 1.00 |
| 20–24 | 0.90 (0.67, 1.21) | 4.12 (1.80, 9.42) | 0.43 (0.30, 0.62) | 0.87 (0.39, 1.93) |
| 25–29 | 0.76 (0.54, 1.06) | 9.68 (4.22, 22.18) | 0.26 (0.16, 0.41) | 0.68 (0.29, 1.60) |
| 30–49 | 0.46 (0.29, 0.74) | 22.03 (9.72, 49.97) | 0.13 (0.07, 0.24) | 0.90 (0.37, 2.18) |
| Education level | ||||
| None | 1.00 | 1.00 | 1.00 | 1.00 |
| Primary | 1.28 (0.89, 1.85) | 0.92 (0.64, 1.32) | 1.31 (0.90, 1.92) | 1.85 (1.21, 2.81) |
| Secondary or higher | 1.43 (1.10, 1.87) | 0.49 (0.35, 0.67) | 1.82 (1.24, 2.67) | 1.78 (1.14, 2.78) |
| Spousal education level | ||||
| None | 1.00 | 1.00 | 1.00 | 1.00 |
| Primary | 1.47 (1.03, 2.10) | 0.83 (0.57, 1.21) | 1.22 (0.84, 1.78) | 0.93 (0.60, 1.44) |
| Secondary or higher | 1.06 (0.79, 1.42) | 0.59 (0.44, 0.79) | 0.80 (0.55, 1.15) | 1.07 (0.74, 1.55) |
| Age at marriage | ||||
| <15 | 1.00 | 1.00 | 1.00 | 1.00 |
| 15–17 | 1.59 (1.11, 2.27) | 0.53 (0.39, 0.74) | 1.73 (1.18, 2.54) | 0.83 (0.58, 1.18) |
| ≥18 | 1.64 (1.17, 2.31) | 0.38 (0.27, 0.56) | 3.14 (2.01, 4.90) | 0.96 (0.63, 1.46) |
| Relationship to household head | ||||
| Daughter-in-law | 1.00 | 1.00 | 1.00 | 1.00 |
| Head | 0.73 (0.38, 1.42) | 5.18 (2.85, 9.40) | 0.83 (0.41, 1.71) | 1.99 (1.01, 3.90) |
| Wife | 0.89 (0.69, 1.15) | 3.17 (2.15, 4.68) | 0.89 (0.67, 1.17) | 1.59 (1.03, 2.45) |
| Daughter | 0.90 (0.58, 1.42) | 0.68 (0.31, 1.48) | 1.10 (0.69, 1.74) | 1.25 (0.58, 2.68) |
| Other | 1.64 (0.996, 2.69) | 1.74 (0.71, 4.23) | 1.70 (1.04, 2.77) | 2.67 (1.05, 6.78) |
| Wealth quintile | ||||
| Poorest | 1.00 | 1.00 | 1.00 | 1.00 |
| Poorer | 1.43 (0.995, 2.06) | 1.01 (0.70, 1.46) | 1.31 (0.90, 1.89) | 1.16 (0.79, 1.70) |
| Middle | 1.59 (1.13, 2.26) | 1.01 (0.70, 1.48) | 1.30 (0.87, 1.94) | 1.54 (1.00, 2.36) |
| Richer | 1.40 (0.98, 2.01) | 0.63 (0.43, 0.94) | 1.08 (0.70, 1.66) | 1.10 (0.64, 1.88) |
| Richest | 1.03 (0.70, 1.52) | 0.47 (0.27, 0.81) | 0.80 (0.49, 1.32) | 1.14 (0.62, 2.10) |
| Residence | ||||
| Rural | 1.00 | 1.00 | - | - |
| Urban | 1.14 (0.88, 1.47) | 1.18 (0.86, 1.62) | - | - |
Bold indicates p<0.05. Reference category is wanted pregnancies.
Adjusted for country fixed effects.
Adjusted for country fixed effects and all variables shown excluding urban/rural residence.
Relative to women reporting no IPV, women with a history of sexual IPV alone (no physical IPV) were more likely to report use of pre-pregnancy modern contraceptive use (aOR=2.32, 95% CI=1.24, 4.36), and those reporting sexual plus physical IPV were more likely to report use of pre-pregnancy traditional contraceptive use (aOR=1.85, 95% CI=1.12, 3.07) (Table 3). Covariates associated with modern relative to no contraceptive use included having more children and more education, being the daughter rather than the daughter-in-law of the household, and being in the richest wealth quintile. Covariates associated with traditional relative to no contraceptive use included having more children, younger gestational age but older maternal age at interview, and having secondary or greater education level.
Table 3.
Unadjusted and adjusted models assessing odds of pre-pregnancy contraceptive use among currently pregnant women living in Bangladesh, India and Nepal (n=4,738).
| Traditional | Modern | Traditional | Modern | |
|---|---|---|---|---|
| OR1 (95% CI) | OR1 (95% CI) | aOR2 (95% CI) | aOR2 (95% CI) | |
| Ever IPV | ||||
| None | 1.00 | 1.00 | 1.00 | 1.00 |
| Sexual only | 1.60 (0.63, 4.03) | 1.37 (0.73, 2.56) | 1.97 (0.74, 5.23) | 2.32 (1.24, 4.36) |
| Physical only | 0.95 (0.65, 1.39) | 0.85 (0.64, 1.12) | 0.85 (0.56, 1.28) | 0.95 (0.69, 1.31) |
| Sexual and physical | 2.10 (1.33, 3.32) | 1.03 (0.66, 1.61) | 1.85 (1.12, 3.07) | 1.62 (0.98, 2.66) |
| Number of living children | ||||
| 0 | 1.00 | 1.00 | 1.00 | 1.00 |
| 1 | 3.37 (2.17, 5.26) | 4.07 (2.98, 5.55) | 3.32 (2.03, 5.41) | 4.75 (3.31, 6.81) |
| 2 | 2.55 (1.57, 4.14) | 2.94 (2.04, 4.25) | 2.42 (1.37, 4.25) | 4.89 (3.01, 7.96) |
| 3 or more | 1.82 (1.09, 3.07) | 2.18 (1.52, 3.13) | 1.66 (0.86, 3.22) | 5.20 (3.10, 8.71) |
| Time since first month of index pregnancy (months) | 0.92 (0.86, 0.99) | 1.00 (0.95, 1.06) | 0.90 (0.83, 0.97) | 1.00 (0.95, 1.06) |
| Age at interview | ||||
| 15–19 | 1.00 | 1.00 | 1.00 | 1.00 |
| 20–24 | 2.45 (1.50, 4.00) | 2.48 (1.70, 3.64) | 1.91 (1.10, 3.33) | 1.16 (0.76, 1.75) |
| 25–29 | 1.92 (1.15, 3.20) | 2.96 (1.99, 4.40) | 1.57 (0.87, 2.87) | 1.15 (0.72, 1.82) |
| 30–49 | 2.82 (1.60, 4.99) | 3.07 (1.98, 4.76) | 2.83 (1.34, 5.95) | 1.13 (0.65, 1.97) |
| Education level | ||||
| None | 1.00 | 1.00 | 1.00 | 1.00 |
| Primary | 1.02 (0.63, 1.66) | 1.58 (1.08, 2.31) | 1.22 (0.73, 2.07) | 1.67 (1.10, 2.52) |
| Secondary or higher | 1.20 (0.83, 1.73) | 3.29 (2.54, 4.26) | 1.69 (1.02, 2.80) | 2.97 (2.08, 4.23) |
| Spousal education level | ||||
| None | 1.00 | 1.00 | 1.00 | 1.00 |
| Primary | 0.88 (0.53, 1.48) | 1.69 (1.16, 2.48) | 0.95 (0.56, 1.61) | 1.33 (0.87, 2.03) |
| Secondary or higher | 1.01 (0.67, 1.53) | 2.77 (2.02, 3.81) | 0.98 (0.58, 1.66) | 1.45 (0.97, 2.18) |
| Age at marriage | ||||
| <15 | 1.00 | 1.00 | 1.00 | 1.00 |
| 15–17 | 1.11 (0.71, 1.75) | 1.34 (0.96, 1.86) | 1.08 (0.68, 1.73) | 1.19 (0.83, 1.71) |
| ≥18 | 1.10 (0.71, 1.69) | 2.03 (1.47, 2.79) | 1.08 (0.64, 1.81) | 1.42 (0.96, 2.10) |
| Relationship to household head | ||||
| Daughter-in-law | 1.00 | 1.00 | 1.00 | 1.00 |
| Head | 0.67 (0.25, 1.82) | 0.46 (0.23, 0.95) | 0.50 (0.17, 1.48) | 0.58 (0.27, 1.26) |
| Wife | 1.04 (0.73, 1.49) | 1.11 (0.85, 1.43) | 0.83 (0.56, 1.24) | 1.12 (0.84, 1.50) |
| Daughter | 1.05 (0.54, 2.06) | 0.40 (0.24, 0.67) | 1.25 (0.63, 2.49) | 0.44 (0.25, 0.75) |
| Other | 0.32 (0.11, 0.95) | 1.10 (0.59, 2.05) | 0.35 (0.12, 1.07) | 1.12 (0.62, 2.01) |
| Wealth quintile | ||||
| Poorest | 1.00 | 1.00 | 1.00 | 1.00 |
| Poorer | 1.06 (0.67, 1.68) | 1.54 (1.06, 2.23) | 1.00 (0.61, 1.62) | 1.32 (0.88, 1.99) |
| Middle | 0.88 (0.51, 1.51) | 1.64 (1.10, 2.44) | 0.70 (0.36, 1.35) | 1.15 (0.73, 1.80) |
| Richer | 0.85 (0.51, 1.42) | 2.82 (1.97, 4.05) | 0.58 (0.30, 1.11) | 1.58 (0.98, 2.53) |
| Richest | 0.96 (0.56, 1.64) | 4.94 (3.35, 7.30) | 0.57 (0.26, 1.28) | 2.41 (1.42, 4.09) |
| Residence | ||||
| Rural | 1.00 | 1.00 | 1.00 | 1.00 |
| Urban | 1.07 (0.73, 1.55) | 2.29 (1.79, 2.92) | 1.19 (0.77, 1.85) | 1.28 (0.95, 1.72) |
Bold indicates p<0.05. Reference category is no pre-pregnancy contraceptive use.
Adjusted for country fixed effects.
Adjusted for country fixed effects and all variables shown.
To understand why pre-pregnancy modern contraception was more likely in cases of sexual IPV alone (i.e., absent physical IPV) and unintended (mistimed and unwanted) pregnancy, we conducted a descriptive analysis of reasons for discontinuation of modern contraception prior to current pregnancy among participants from India and Nepal (n=1032), where such data were available. Analyses were conducted for the total subsample, as well as by IPV and pregnancy intendedness categories. For the total subsample of contraceptors, most (61.9%) discontinued contraception prior to the current pregnancy due to a desire to become pregnant; 18.1% reported method failure, and 11.9% reported concerns with methods (e.g., side effects, inconvenience). (Table 4.) Only 4.2% reported husband disapproval as the reason for discontinuation. Analysis of IPV revealed that those who reported sexual IPV, alone or with physical IPV, were more likely to report contraceptive failure than those reporting no IPV (37.3% and 30.9%, respectively versus 13.6%); overlapping confidence intervals suggest a trend rather than significant differences in these prevalence estimates. Analysis of pregnancy intendedness revealed that those reporting their current pregnancy as mistimed or unwanted were significantly (based on non-overlapping confidence intervals) more likely to report contraceptive failure than did those reporting a wanted pregnancy (39.6% and 40.3%, respectively, versus 7.9%).
Table 4.
Descriptive analysis of reasons for pre-pregnancy contraceptive discontinuation among currently pregnant women living in India and Nepal (n=1,032).
| Total | Ever IPV | Intendedness of current pregnancy | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Unwted N | % (95% CI) | None % (95% CI) | Sexual only % (95% CI) | Physical only % (95% CI) | Sexual and physical % (95% CI) | Wanted % (95% CI) | Mistimed % (95% CI) | Unwanted % (95% CI) | |
| Method failure | 157 | 18.1% (14.1, 22.1) | 13.6% (9.0, 18.2) | 37.3% (12.0, 62.5) | 22.6% (14.9, 30.2) | 30.9% (17.8, 44.0) | 7.9% (4.7, 11.1) | 39.6% (28.4, 50.8) | 40.3% (26.1, 54.6) |
| Wanted to become pregnant | 664 | 61.9% (57.0, 66.8) | 67.8% (61.9, 73.6) | 45.1% (19.4, 70.8) | 51.8% (41.4, 62.2) | 50.4% (36.3, 64.6) | 75.8% (70.9, 80.8) | 35.5% (24.5, 46.5) | 27.2% (13.3, 41.1) |
| Husband disapproved | 31 | 4.2% (2.1, 6.2) | 3.1% (1.0, 5.2) | 5.1% (0.0, 14.9) | 7.5% (1.0, 14.0) | 4.2% (0.0, 10.2) | 2.5% (0.5, 4.4) | 10.2% (3.0, 17.4) | 3.8% (0.0, 7.9) |
| Method concerns1 | 138 | 11.9% (9.0, 14.7) | 12.6% (8.9, 16.4) | 12.5% (0.0, 29.5) | 11.5% (5.6, 17.5) | 7.4% (0.5, 14.4) | 10.4% (7.0, 13.9) | 10.4% (4.5, 16.2),, | 22.1% (11.8, 32.4) |
| Access barriers2 | 9 | 0.9% (0.1, 1.7) | 0.0% (0.0, 0.1) | - | 2.9% (0.0, 5.7) | 2.7% (0.0, 7.8) | 0.5% (0.0, 1.1) | - | 4.7% (0.0, 10.1) |
| Other3 | 33 | 3.1% (1.4, 4.7) | 2.8% (0.7, 5.0) | - | 3.7% (0.4, 7.0) | 4.3% (0.0, 10.4) | 2.9% (0.7, 5.1) | 4.3% (0.6, 7.9) | 1.8% (0.0, 4.4) |
Note: Percentages are weighted.
Includes wanting a more effective method, health concerns/side effects and inconvenience of use.
Includes lack of access/too far, and costing too much.
Includes infrequent sex/husband away, marital dissolution/separation and other.
Discussion
Current findings suggest that for these South Asian nations of focus: Bangladesh, India and Nepal, more than one in four pregnancies is unintended, comprised of 16% mistimed and 10% unwanted pregnancies. While this is less than the global estimate of 41% of pregnancies being unintended [19, 20], it is still far higher than desirable, and may contribute to ongoing high rates of maternal and child health concerns in the region [7, 8]. Contrary to prior research [2–6, 10], however, this study found that IPV does not increase women’s risk for unintended pregnancy. Crude findings of the current study indicate associations between physical IPV (alone and with sexual IPV) and unintended pregnancy, but these results were lost after accounting for covariates such as higher parity, which is also associated with IPV [7, 9, 21]. Assessment of intendedness of current pregnancy may have contributed to the different findings for our study; prior research focused on ever having had an unintended pregnancy [2–6, 10] whereas this study is specific to current pregnancy. Notably, modern contraceptive use prior to the current pregnancy was associated with greater likelihood of current pregnancy being unintended, possibly because 38% of the pre-pregnancy contraceptors reported problems or concerns with their contraception. Similar results have been seen in the United States, where inconsistent and incorrect use of contraception accounts for 43% of unintended pregnancies [22].
While physical and/or sexual IPV were not associated with unintended pregnancy, sexual IPV alone was associated with increased likelihood of pre-pregnancy modern contraceptive use, and sexual plus physical IPV was associated with increased likelihood of pre-pregnancy traditional contraceptive use. Heightened use of contraceptives among women reporting sexual IPV may be due to their attempts to establish reproductive control in the absence of sexual autonomy. Further research is needed to understand why there are differential associations between sexual IPV and the type of contraceptives used based on whether physical IPV is also occurring.
Findings also document that higher rates of pre-pregnancy contraceptive use among sexual IPV-affected women do not garner them greater protection against unintended pregnancy, perhaps due to their higher rates of contraceptive failure. Pill and condom are the most commonly used forms of modern contraception in the region, and they are more easily disrupted than the less frequently used IUD and injection methods. Longer-term contraception, such as IUDs, may be more useful for women contending with sexual IPV from husbands. Prior research documents lower likelihood of condoms and greater likelihood of sabotage of oral contraceptive pill intake among abused women [3, 23]; this reinforces the need for supporting use of long-acting reversible contraceptives (LARC) that are not under male control. A recent study from the United States demonstrates that while discontinuation of contraception is more likely for women with a history of violence, effect of abuse history on LARC methods is less than that seen on non-long-acting methods [24]. Very low use of LARC methods in this sample suggests that greater awareness and availability of these methods are likely needed.
While current findings expand our understanding of whether IPV affects contraception and subsequent unintended pregnancy, they must be considered in light of certain important limitations. Findings are based on use of only three countries in South Asia, limiting generalizability of findings regionally and globally, though representative data sets were used from three of the largest nations in the region. Measures of IPV were limited to those collected across assessed countries, and therefore omitted other important aspects of partner violence, including emotional abuse which has also been linked to unintended pregnancy [25]. Data utilized were the most recent data available that included IPV data for the three countries, but they were not collected within the same timeframe. In addition, data were pooled and results of these pooled analyses may not be consistent with findings based on analyses limited to individual nations included in the current study. Further, disproportionate representation of India in the pooled data set may skew data toward findings specific to India; analyses were weighted to account for variation in country size. Pooled regional analysis was required to explore effects that may be insufficiently powered for analysis at the country-level, particularly those for sexual IPV. These data are based on self-report and are thus are susceptible to social desirability and recall bias. Analyses are cross-sectional in nature; thus, causal relationships between variables cannot be inferred, though use of chronological data regarding pre-pregnancy contraceptive use and current intendedness of pregnancy offers some insight.
Conclusion
The current study documents that marital sexual IPV absent physical IPV is not uncommon and is associated with increased likelihood of modern contraceptive use prior to pregnancy in South Asia, possibly as a means of reproductive control in the absence of sexual autonomy. However, such use does not reduce risk for subsequent unintended pregnancy among sexual IPV-affected women, possibly due to the higher rates of contraceptive failure for these women. Findings highlight the need for further research on vulnerabilities for contraceptive failure among women with a history of IPV as well as the need for more tailored family planning support for women contending with IPV, particularly sexual IPV. Data suggesting growing rates of sexual violence in the region reinforce the need for such an approach [26].
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Contributor Information
Anita Raj, Email: anitaraj@ucsd.edu.
Lotus McDougal, Email: lmcdouga@ucsd.edu.
References
- 1.WHO; Department of Reproductive Health and Research, London School of Hygiene and Tropical Medicine, South African Medical Research Council. [Accessed August 31, 2013];Global and regional estimates of violence against women. Prevalence and health effects of intimate partner violence and non-partner sexual violence. 2013 Jun; http://www.who.int/reproductivehealth/publications/violence/9789241564625/en/index.html.
- 2.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]
- 3.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]
- 4.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]
- 5.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]
- 6.Salam A, Alim A, Noguchi T. Spousal abuse against women and its consequences on reproductive health: a study in the urban slums in Bangladesh. Matern Child Health J. 2006;10(1):83–94. doi: 10.1007/s10995-005-0030-6. [DOI] [PubMed] [Google Scholar]
- 7.World Health Organization. WHO Multi-country Study on Women’s Health and Domestic Violence against Women: Initial results on prevalence, health outcomes and women’s responses. Geneva, Switzerland: 2005. [Accessed at: April 21, 2014]. http://www.who.int/gender/violence/who_multicountry_study/summary_report/summary_report_English2.pdf?ua=1. [Google Scholar]
- 8.Gipson JD, Koenig MA, Hindin MJ. The effects of unintended pregnancy on infant, child, and parental health: a review of the literature. Stud Fam Plann. 2008 Mar;39(1):18–38. doi: 10.1111/j.1728-4465.2008.00148.x. [DOI] [PubMed] [Google Scholar]
- 9.McCleary Sills J, McGonagle A, Malhotra A. Women’s demand for reproductive control: understanding and addressing gender barriers. International Center for Research on Women; 2012. [Accessed at:April 21, 2014]. http://www.icrw.org/files/publications/Womens-demand-for-reproductive-control.pdf. [Google Scholar]
- 10.Stephenson R, Jadhav A, Hindin M. Physical domestic violence and subsequent contraceptive adoption among women in rural India. J Interpers Violence. 2013;28(5):1020–39. doi: 10.1177/0886260512459379. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zakar R, Zakar MZ, Mikolajczyk R, Krämer A. Intimate partner violence and its association with women’s reproductive health in Pakistan. Int J Gynaecol Obstet. 2012;117(1):10–4. doi: 10.1016/j.ijgo.2011.10.032. [DOI] [PubMed] [Google Scholar]
- 12.Das S, Bapat U, Shah More N, Alcock G, Joshi W, Pantvaidya S, Osrin D. Intimate partner violence against women during and after pregnancy: a cross-sectional study in Mumbai slums. BMC Public Health. 2013;9(13):817. doi: 10.1186/1471-2458-13-817. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.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]
- 14.National Institute of Population Research and Training (NIPORT), Mitra and Associates, Macro International. Bangladesh Demographic and Health Survey 2007. Dhaka, Bangladesh and Calverton Maryland, USA: National Institute of Population Research and Training, Mitra and Associates, and Macro International; 2009. [Google Scholar]
- 15.International Institute for Population Sciences (IIPS) and Macro International. National Family Health Survey (NFHS-3), 2005–06. I. India:.Mumbai: IIPS; 2007. [Google Scholar]
- 16.Ministry of Health and Population (MOHP) [Nepal], New ERA, and ICF International Inc. Nepal Demographic and Health Survey 2011. Kathmandu, Nepal and Calverton, Maryland: Ministry of Health and Population, New ERA, and ICF International; 2012. [Google Scholar]
- 17.National Institute of Population Studies (NIPS) [Pakistan] and ICF International. Pakistan Demographic and Health Survey 2012–13. Islamabad, Pakistan, and Calverton, Maryland, USA: NIPS and ICF International; 2013. [Google Scholar]
- 18.Rutstein S, Johnson K. The DHS Wealth Index. Calverton, Maryland: ORC Macro; 2004. [Google Scholar]
- 19.Singh S, Sedgh G, Hussain R. Unintended pregnancy: worldwide levels, trends, and outcomes. Stud Fam Plann. 2010;41(4):241–50. doi: 10.1111/j.1728-4465.2010.00250.x. [DOI] [PubMed] [Google Scholar]
- 20.Sedgh G, Singh S, Hussain R. Incidence of unintended pregnancies worldwide in 2012 and trends since 1995. Presented at the 141st APHA Annual Meeting and Exposition; Nov 2–6, 2013; Boston, MA. [Google Scholar]
- 21.Hindin M, Kishor S, Ansara DL. Intimate Partner Violence among couples in 10 DHS Countries: Predictors and Health Outcomes. [Accessed at: April 21, 2014];DHS Analytical Studies. 2008 Dec 8; http://dhsprogram.com/pubs/pdf/AS18/AS18.pdf.
- 22.Gold RB, Sonfield A, Richards CL, Frost J. Next Steps for America’s Family Planning Program: Leveraging the Potential of Medicaid and Title X in an Evolving Health Care System. New York: Guttmacher Institute; 2009. [Accessed at: April 21, 2014]. http://www.guttmacher.org/pubs/NextSteps.pdf. [Google Scholar]
- 23.Silverman JG, Decker MR, McCauley HL, Raj A, Reed E, Miller E. Coercive and deceptive forms of sexual risk and associated violence perpetrated by Male partners of female adolescents. Persp Sex & Reproduct Health. 2011;43(1):60–65. doi: 10.1363/4306011. [DOI] [PubMed] [Google Scholar]
- 24.Allsworth JE, Secura GM, Zhao Q, Madden T, Peipert JF. The impact of emotional, physical, and sexual abuse on contraceptive method selection and discontinuation. Am J Public Health. 2013;103(10):1857–64. doi: 10.2105/AJPH.2013.301371. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Salazar M, San Sebastian M. Violence against women and unintended pregnancies in Nicaragua: a population-based multilevel study. BMC Women’s Health. 2014;14(26) doi: 10.1186/1472-6874-14-26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Raj A, McDougal L. Sexual violence and rape in India. Lancet. 2014;383(9920):865. doi: 10.1016/S0140-6736(14)60435-9. [DOI] [PubMed] [Google Scholar]