Intimate Partner Physical and Sexual Violence and Outcomes of Unintended Pregnancy Among National Samples of Women From Three Former Soviet Union Countries

Abstract

The article examines the relationship between intimate partner violence (IPV) and unintended pregnancy among nationally representative samples of women in three former Soviet Union countries. Women who experienced physical and/or sexual IPV from their current or most recent husband or living together partner demonstrated higher risks of unintended last pregnancy, either terminated through abortion (in Azerbaijan, Moldova, and Ukraine) or resulting in unintended live birth (in Ukraine). IPV prevention components should be integrated into reproductive health programs to reduce the risk of unintended births and abortions among women living with abusive partners in these former Soviet Union countries.

Keywords

women violence domestic violence abortion unintended pregnancy Eastern Europe and the Commonwealth of Independent States

Introduction

Intimate partner violence (IPV)—defined as a “behavior within an intimate relationship that causes physical, sexual or psychological harm” (World Health Organization [WHO], 2002, p. 89)—is a global social issue that infringes on women’s rights, endangers their safety, and affects their overall well-being. In countries with relatively low levels of awareness about IPV, the consequences of violence are perceived to be limited to the woman’s death or severe bodily harm (Fox et al., 2007). Nonetheless, in addition to injuries as a direct result of violence, IPV can have long-lasting adverse effects on a woman’s reproductive health (Coker, 2007; Jordan, 2007; Plichta, 2004).

Empirical studies have demonstrated that IPV is negatively associated with unintended births (Hindin, Kishor, & Ansara, 2008; Pallitto & O’Campo, 2004; Stephenson, Koenig, Acharya, & Roy, 2008), pregnancy loss, and non-live birth (Coker, Sanderson, & Dong, 2004; Hindin et al., 2008; Kuning, McNeil, & Chongsuvivatwong, 2003; Silverman, Gupta, Decker, Kapur, & Raj, 2007). The majority of evidence, however, comes from studies conducted in Western industrialized countries or developing countries of Africa, Latin America, and Asia (Kuning et al., 2003; Pallitto & O’Campo, 2004; Schei, Guthrie, Dennerstein, & Alford, 2006; Stephenson et al., 2008; Stephenson, Koenig, & Ahmed, 2006). In the transitional countries of the former Soviet Union (fSU; also known as the Newly Independent States [NIS]), the relationship between IPV and unintended pregnancy remains unknown.

Previous studies in developing countries demonstrated that women with a history of IPV report higher likelihood of pregnancies that are either unwanted (“wanted no more children”) or mistimed (“wanted children later”; Martin et al., 1999; Pallitto, Campbell, & O’Campo, 2005; Pallitto & O’Campo, 2004). Thus, unintended pregnancies (unwanted or mistimed) were measured only among pregnancies ending in live births. In fSU countries, only about one half to two thirds of pregnancies result in live births (Centers for Disease Control [CDC], 2003). Therefore, a significant portion of unintended pregnancies may not be captured by focusing only on unintended live births. The fSU region is historically known for liberal abortion laws and easy access to safe abortions. The research findings from other parts of the world may not be applicable to countries where abortions are easily available and acceptable and women have a choice of terminating unintended pregnancies through abortions.

Therefore, this article examines the following research question:

Research Question 1: Are women who have experienced IPV more likely to have an unintended pregnancy manifested through unintended live birth or abortion compared with women who have not experienced violence from their partners, in Azerbaijan, Moldova, and Ukraine?

The findings may provide a foundation for violence prevention and reproductive health programs in some countries of the fSU region.

Regional Context

Politically, geographically, and socially, the fSU region represents a common area that deserves special consideration. As part of the Soviet Union for 70 years, Azerbaijan, Moldova, and Ukraine shared similar systems of universal public health care services providing free access to the entire population (Suhrcke, Rocco, & McKee, 2007). The economic crisis of the 1990s, following the collapse of the Soviet Union, resulted in growth in poverty and deterioration of health and social services (Scott, 2000).

Nevertheless, over the last decade, improving women’s reproductive health has become a national priority in all three countries (U.S. International Development Agency [USAID], 2007). After gaining independence, reproductive health programs in Azerbaijan, Ukraine, and Moldova have focused primarily on increasing the knowledge, availability, and use of modern contraception and on improving the quality of family planning services (United Nations Population Fund [UNFPA], 2008). Despite high awareness of modern contraception methods such as birth control pills, condoms, intrauterine devices (IUDs), and other modern methods (95% in Azerbaijan and 99% both in Moldova and Ukraine), the use remains relatively low in the fSU countries (CDC, 2003; Ross & Winfrey, 2002). Studies in the fSU region have identified that young age, low education, rural residence, and low socioeconomic status are associated with poor reproductive health outcomes (CDC, 2003).

Moreover, proponents of gender analysis in health research argue that despite increased information about contraceptives and the accessibility of supplies, even informed women may not be able to utilize protection and their behavior may remain unchanged due to a power imbalance in intimate relationships (Bawah, Akweongo, Simmons, & Phillips, 1999; Blanc, 2001). However, the role of gender factors such as IPV in affecting women’s reproductive health behaviors and decisions has not been examined for this region.

Due to differences in ethnicity and religion, family relationships, health behaviors, and cultural norms vary across these three countries (CDC, 2003). Ukraine and Moldova, with predominantly Christian populations, share relatively more egalitarian gender beliefs, while Azerbaijan, a secular Muslim nation, is characterized by more traditional values and conservative norms with respect to women (Asian Development Bank [ADB], 2006; Heyat, 2006; Manijeh, 1999; Tohidi, 1996). Furthermore, while Azerbaijan and Ukraine exhibit significant economic growth, Moldova remains the poorest country in Eastern Europe (Hensel & Gudim, 2004). Therefore, an examination of cross-country similarities and differences in the relationships between IPV and unintended pregnancy in Azerbaijan, Moldova, and Ukraine may contribute to the understanding of IPV and reproductive health issues in a unique sociocultural context.

Theory

The article is guided by feminist theory and analyzes reproductive health problems through a gender perspective (Dixon-Mueller, 1993; Sen, George, & östlin, 2002; Worth, 1989). IPV and power differentials between men and women pose direct and indirect risks to women’s health by creating significant disparities in access to and use of health information, services, and technologies. In addition to the direct influence of IPV (e.g., unintended pregnancy as a result of forced unprotected sex), partner violence can indirectly affect women’s health by diminishing their functioning and competencies. Living in a situation of chronic violence, intimidation, or fear can impair a woman’s self-esteem, decision-making abilities, and motivation to care for herself (Coker, 2007).

IPV and Unintended Pregnancy

Having limited control over their bodies and decisions related to sexual activities, such as the time of sex and use of contraceptive protection, women are more likely to engage in unwanted and/or unprotected sex, which can lead to unintended pregnancies (Coker, 2007; Pallitto et al., 2005). Lifetime physical and/or sexual IPV was found to be associated with higher odds of unintended pregnancy in Demographic and Health Surveys (DHS) studies in other developing countries (Pallitto & O’Campo, 2004; Silverman et al., 2007; Stephenson et al., 2008). In the DHS questionnaire, however, questions about unintended pregnancy are available only for pregnancies resulting in live births in the past 5 years and for current pregnancies. The question about unintended pregnancy may not solicit accurate information in countries of the former Soviet Union, where the culture of family planning is almost absent and women do not refer to unplanned pregnancies as unwanted or undesired, especially after the child is born. Hence, retrospective measures may underestimate the reports of unintended pregnancy.

According to the nationally representative DHS, the percentage of unintended live births in the 5 years preceding the survey was 17% in Azerbaijan, 21% in Moldova, and 14% in Ukraine (NCPM & ORC Macro, 2006; State Statistical Committee & Macro International, 2008; Ukrainian Center for Social Reforms [UCSR], State Statistical Committee, Ministry of Health, & Macro International, 2008). However, the same surveys found that only about half of pregnancies end in a live birth in Azerbaijan (46%) and Moldova (55%), and about two thirds in Ukraine (68%; NCPM & ORC Macro, 2006; State Statistical Committee & Macro International, 2008; UCSR et al., 2008).

An estimated 49% of pregnancies in Azerbaijan, 34% in Moldova, and 25% in Ukraine are terminated through abortion (NCPM & ORC Macro, 2006; State Statistical Committee & Macro International, 2008; UCSR et al., 2008). Thus, for these fSU countries, the unintended live birth measure in the DHS does not capture pregnancies that were terminated through abortions and provides an incomplete measure of unintended pregnancies.

IPV and Abortions

Women may use abortions in situations when a fetus endangers the woman’s health. However, women may also deliberately terminate a pregnancy if it was unintended (Singh, Prada, & Kestler, 2006). The fSU countries have some of the world’s most liberal abortion laws, and for decades abortion was safe, free, and became a form of birth control (CDC, 2003). A 2002 study found that about 57% of pregnancies in Azerbaijan were unintended and 87% of unintended pregnancies resulted in abortion (Goldberg & Serbanescu, 2002).

In the late 1990s, a woman in the fSU region would have, on average, at least 1 abortion during her lifetime—1.3 abortions per woman in Moldova, 1.6 in Ukraine, and 3.2 abortions per woman in Azerbaijan, one of the highest abortion rates in the world (CDC, 2003; Hyde, 2008). During the Soviet period, there were few legal restrictions constraining woman from having an abortion; abortions were provided free of charge at women’s health clinics, while oral contraceptive methods without significant side effects were not easily available (Serbanescu, Goldberg, & Morris, 2005). After the collapse of the Soviet Union, the practice of treating abortion as a main method of birth control began to change, starting with the introduction of family planning programs.

In this study, we examine the relationship between IPV and unintended pregnancy using a measure of unintended pregnancy more suited to the fSU countries that includes both unintended pregnancies that resulted in live births and pregnancies that were terminated through induced abortion.

Hypothesis

The article examines the relationships between IPV and unintended pregnancy using population-based samples from three fSU countries—Azerbaijan, Moldova, and Ukraine. More specifically, the article examines the associations between a history of physical or sexual partner violence from the most recent intimate partner and unintended pregnancy within the last 5 years, while controlling for socio-demographic characteristics (woman’s age, marital status, urban/rural residence, education, and wealth status) and important reproductive health covariates (having children prior to last pregnancy, previous use of abortion, and lifetime use of contraceptives).

We hypothesize that women who have ever experienced physical or sexual IPV from their recent partner will demonstrate higher likelihood of unintended last pregnancy compared with women who have never experienced physical or sexual IPV, after controlling for key covariates.

Method

Data Source

This study is an analysis of secondary data collected through the DHS in Azerbaijan (2006), Moldova (2005), and Ukraine (2007). The DHS, conducted in more than 80 developing countries, recently included the fSU countries, and for the first time provided nationally representative data on socio-demographic, household, women’s status, and health variables for this region. The DHS uses a unified methodology that allows for cross-country comparisons. The standard DHS survey consists of Household, Women’s and Child’s Health Questionnaires that cover a number of topics including socio-demographic characteristics, women’s status, fertility, family planning, and child survival. The standard DHS questionnaire was fielded in 8 of 15 fSU countries. The standard DHS questionnaire without the Domestic Violence Module was conducted in Armenia (2000, 2005, 2010), Kazakhstan (1995, 1999), Kyrgyz Republic (1997), Turkmenistan (2000), and Uzbekistan (1996, 2002). However, only three countries—Azerbaijan, Moldova, Ukraine—included the optional Domestic Violence Module. Therefore, our analysis is limited to these three countries from the fSU region. The DHS questionnaire was administered by trained local interviewers.

Sampling Scheme

Demographic and Health Surveys are nationally representative household surveys that use a stratified multistage sampling strategy. Using the most recent national census, the list of regional clusters (or PSUs, primary sampling units) was developed in each country. Clusters were randomly selected using a probability-proportionate-to-size sampling procedure within each strata, meaning that more populated clusters had a higher probability of being selected. Households were then randomly selected within each selected cluster. The Domestic Violence module was administered to one randomly selected woman of reproductive age (15-49 years old) within each selected household. In Ukraine, women were not eligible for the Domestic Violence Module in households in which men were administered the Domestic Violence Module (one half of all selected households). The detailed sampling methodology is presented in the DHS sampling manual (Macro International, 1996) and DHS Country reports (NCPM & ORC Macro, 2006; State Statistical Committee & Macro International, 2008; UCSR et al., 2008).

In the Moldova DHS, the clusters were drawn separately for urban and rural areas (strata). In Azerbaijan and Ukraine, the stratification was achieved by separating economic regions into rural and urban areas. Due to administrative difficulties and a blockade, no respondents were included from the following areas: Transnistria (a disputed region of Moldova), Nakhichevan (an autonomous republic in Azerbaijan), the Kalbajar–Lachin region and the four districts of Yukhari Garabakh (occupied territories in Azerbaijan), and a region in Ukraine uninhabited since the Chernobyl nuclear disaster.

The original samples (N = 8,444 in Azerbaijan; N = 7,440 in Moldova; and N = 6,841 in Ukraine) were reduced to include ever-married (or cohabitating) women of reproductive age (15-49) who were eligible and completed the Domestic Violence Module and who had a pregnancy in the last 5 years preceding the survey. Almost half of the respondents from the domestic violence sample in Azerbaijan were pregnant in the last 5 years compared with one third of respondents in Moldova and one quarter in Ukraine. Eligible ever-married women who did not complete the Domestic Violence Module (n = 499 in Azerbaijan, n = 197 in Moldova, n = 676 in Ukraine) or who could not be interviewed because of unachieved privacy during their interviews (n = 41 in Azerbaijan, n = 14 in Moldova, and n = 13 in Ukraine) were excluded from the analysis. To capture the relationship between IPV and women’s decision to sustain an unintended pregnancy or to terminate it through abortion, we excluded from the final analysis last pregnancies resulting in stillbirth or miscarriage (69 cases in Azerbaijan, 123 in Moldova, and 28 in Ukraine). To obtain nationally representative estimates, weights from the Domestic Violence Module were used as sampling weights. The final weighted samples included 1,620 women in Azerbaijan; 1,377 women in Moldova; and 545 women in Ukraine.

Measures

Intimate partner violence: Physical and/or sexual IPV is a binary variable defined as having occurred if a woman reported any abusive acts ever committed by her current or most recent husband or living together partner. The binary variable was constructed based on seven items measuring physical violence—husband/partner ever (a) pushed, shoved or threw something at you; (b) slapped you; (c) twisted your arm or pulled your hair; (d) punched with his fist or with something that could hurt you; (e) kicked you, dragged you, or beat you up; (f) tried to choke you or burn you on purpose; or (g) threatened or attacked you with a knife, gun, or other type of weapon—and three items measuring sexual violence—husband/partner ever (a) physically forced you to have sexual intercourse with him, when you did not want to; (b) forced you to perform any sexual acts you did not want to; and (c) respondent reported forced first sexual intercourse with this partner. Almost all women who reported experiencing sexual IPV also reported ever experiencing physical IPV; this measure was combined into physical and/or sexual IPV.

Unintended pregnancy: Only in the fSU countries did the DHS questionnaire include a Pregnancy Calendar and collected detailed information about women’s pregnancy history while collecting birth history in other countries. Women who had a live birth within the 5 years preceding the survey were asked whether they wanted the child then, later, or wanted no more children at the time of the pregnancy. To reduce recall bias, this question was asked only if the woman reported being pregnant in the 5 years preceding the survey. Thus, the measure of unintended pregnancy is available only for pregnancies resulting in live births. In the fSU region, where women historically had an option of terminating an unintended pregnancy through an abortion, this question may not capture a large portion of unintended pregnancies. Therefore, to measure unintended pregnancy, we constructed a new variable that captured unintended pregnancies that resulted in live births and pregnancies that were terminated through abortion. This new measure of unintended pregnancy was based on outcome of last pregnancy within the last 5 years preceding the survey and included three categories: 0 = live birth wanted then, 1 = unintended live birth (live birth wanted later or not wanted at all), and 2 = abortion.

Covariates: The analysis was adjusted for socio-demographic covariates—rural residence, low levels of education, and poverty—often associated with poor reproductive health outcomes (Remez, 2003). Residence was measured as a binary variable (rural/urban areas). Age demonstrated a non-linear relationship with the outcome variable, and regression models included age in a categories variable (15-24, 25-34, and 35-44 years of age). Marital status is a binary variable representing (0) currently married (or current cohabitating) women or (1) formerly married (or formerly cohabitating) women that includes divorced, widowed, or separated women.

To measure wealth status, the DHS uses a wealth index that is computed using principal component analysis and includes five quintiles (from 1, lowest wealth, to 5, highest) based on ownership of durable assets (e.g., TV, mobile phone, washing machine, photo and video cameras, computer, air conditioner), means of transpiration (e.g., car), and ownership of land and farm animals. The index has been tested in many countries and has been demonstrated to be an accurate measure of household economic status in countries with irregular income data (Rutstein & Kiersten, 2004). In this article, the wealth index was dichotomized into two categories: the lowest two quintiles (poor and poorest) were coded as 1 to indicate poor wealth status and the middle and two highest quintiles (rich and richest) were coded as 0 to indicate middle or high wealth status. A continuous variable measuring total years of schooling was skewed and recoded as an education variable that included two categories: 1 = secondary education or below and 0 = above secondary education. Due to the policy of mandatory secondary education in the fSU (11 years, an equivalent of U.S. high school), the number of individuals who have no education or primary education was minimal (<0.5 in Ukraine, <1% in Moldova, and <2% in Azerbaijan). The above secondary education category included institutions of higher education (university-level undergraduate or graduate education) as well as secondary specialized education (e.g., tekhnicums).

Lifetime use of contraceptives included three categories: (0 = no method; 1 = only ever used traditional methods of contraception such as rhythm, withdrawal, and folkloric methods; and 2 = ever used modern methods of contraception). Modern methods of contraception include birth control pills, intrauterine devices (IUDs), diaphragm/cap, condoms, sterilization, injectables, implants, spermicides/foam/jelly, ring, lactational amenorrhea method (LAM), and emergency contraception. Finally, to estimate the association between IPV and decision regarding most recent pregnancy, the study adjusted for reproductive health covariates commonly associated with pregnancy outcomes: number of living children prior to last pregnancy and number of abortions prior to last pregnancy. Both variables were positively skewed and the regression models included recoded binary variables (had at least one child prior to last pregnancy and had at least one abortion prior to last pregnancy).

Data Analysis

The statistical analysis was performed in Stata 12 that handles complex survey data. To account for DHS’s stratified multistage cluster sampling design, we used Stata’s survey command (-svy-) for descriptive statistics and regression models. Regional clusters were specified as primarily sampling units (PSU), and weights from the Domestic Violence Module were used as sampling weights. To adjust for stratification, we used urban/rural area for the DHS Moldova and a newly created variable representing rural and urban areas within each economic region for the data set from Azerbaijan and Ukraine. Without specifying sampling design, the analysis may underestimate the standard errors and produce more statistically significant results, running the risk of Type I Error (Chambers & Skinner, 2003). Descriptive statistics with 95% confidence intervals provide nationally representative estimates and were obtained using Stata’s survey and subpopulation (-subpop-) commands.

To examine associations between IPV and outcomes of last pregnancy, we performed a multinomial survey logistic regression adjusted for the DHS’s sampling design. The regression coefficients (B), standard errors (SE), and 95% confidence intervals (CI) for unadjusted and adjusted regression models (Model 1 and Model 2, respectively) are reported in the article. Models were adjusted for key socio-demographic characteristics (woman’s age, marital status, urban/rural residence, education, and wealth status) and reproductive health covariates (lifetime use of contraceptives, having living children prior to last pregnancy, and previous use of abortions). To examine the association between IPV and unintended pregnancy as a reproductive health outcome that usually differs by age, the final model included an interaction term by age. The models were tested for multicollinearity, and no evidence of high collinearity was observed (Tabachnick & Fidell, 2006). The statistical analysis was conducted individually for each country and the results were compared.

Results

Descriptive Results

Weighted percentages with 95% CI, presented in Table 1, account for the sampling design and sampling weights and are more accurate and representative estimates of socio-demographic characteristics, prevalence of IPV, and women’s reproductive health outcomes at the national level.

Table 1.

Socio-Demographic Characteristics, Prevalence of Intimate Partner Violence, and Key Reproductive Health Variables Among Ever-Married (or Cohabitating) Women of Reproductive Age in Azerbaijan, Moldova, and Ukraine Who Had a Pregnancy in the Past 5 Years Preceding the Survey.

	Azerbaijan		Moldova		Ukraine
	N = 1,620^a		N = 1,377^a		N = 545^a
Variables	%	95% CI	%	95% CI	%	95% CI
Socio-demographic characteristics
Age in years, M	30.08	[29.61, 30.54]	28.88	[28.47, 29.3]	29.27	[28.58, 29.97]
Age in categories (years)
15-24	22.97	[19.8, 26.48]	28.33	[25.56, 31.26]	26.6	[22.05, 31.71]
25-34	50.57	[47.06, 54.07]	53.35	[50.41, 56.26]	52.35	[47.34, 57.31]
35-45	26.45	[23.86, 29.22]	18.33	[16.02, 20.88]	21.05	[17.38, 25.26]
Residence
Urban	54.39	[49.68, 59.02]	40.09	[37.35, 42.9]	63.44	[59.15, 67.52]
Rural	45.61	[40.98, 50.32]	59.91	[57.1, 62.65]	36.56	[32.48, 40.85]
Marital status
Currently married or in union	96.43	[94.77, 97.58]	92.39	[90.72, 93.78]	88.85	[85.11, 91.74]
Formerly married or in union	3.57	[2.42, 5.23]	7.61	[6.22, 9.28]	11.15	[8.26, 14.89]
Wealth status
Middle and rich	58.86	[54.19, 63.38]	65.45	[61.95, 68.78]	61.99	[56.71, 67.01]
Poor	41.14	[36.62, 45.81]	34.55	[31.22, 38.05]	38.01	[32.99, 43.29]
Years of education, M	10.77	[10.59, 10.96]	11.27	[11.1, 11.44]	13.59	[13.32, 13.9]
Level of education above secondary	26.37	[23.36, 29.62]	40.21	[37.19, 43.32]	63.4	[58.55, 68]
Secondary or below (≤11 years)	73.63	[70.38, 76.64]	59.79	[56.68, 62.81]	36.6	[32, 41.45]
IPV from most current or recent husband or cohabitating partner
Any physical IPV	13.04	[11.16, 15.18]	21.52	[19.19, 24.04]	10.28	[7.57, 13.81]
Any sexual IPV	2.9	[2.05, 4.08]	4.67	[3.52, 6.16]	4.15	[2.27, 7.48]
Any physical or sexual IPV	13.75	[11.88, 15.86]	22.44	[20.09, 24.98]	10.85	[8.02, 14.51]
Reproductive health
Ever used contraception
No method	26.96	[24.16, 29.96]	4.79	[3.59, 6.35]	15.83	[11.9, 20.75]
Traditional methods only	38.52	[35.04, 42.13]	8.99	[7.26, 10.83]	12.62	[9.92, 15.91]
Modern methods	34.51	[31.22, 37.96]	86.33	[83.96, 88.4]	71.55	[66.84, 75.84]
Living children (prior to last pregnancy), median (minimum-maximum)	2	(0-7)	1	(0-9)	1	(0-7)
0	22.2	[19.52, 25.12]	33.25	[30.46, 36.16]	38.87	[34.42, 43.51]
1-2	58.54	[54.95, 62.03]	57.86	[55.02, 60.64]	56.54	[51.74, 61.23]
3 or more	19.27	[16.87, 21.92]	8.89	[6.98, 11.26]	4.59	[3.15, 6.64]
Abortions (prior to last pregnancy), median (minimum-maximum)	0	(0-17)	0	(0-21)	0	(0-9)
0	56.69	[53.7, 59.64]	63	[60.08,65.82]	75.32	[70.72, 79.41]
1-2	30.83	[28.13, 33.67]	29.02	[26.43,31.75]	17.62	[14.18, 21.68]
3 or more	12.48	[10.51, 14.75]	7.99	[6.44, 9.88]	7.06	[4.84, 10.18]
Last pregnancy outcome
Wanted live birth	42.96	[40.06, 45.9]	50.29	[46.88, 53.69]	61.51	[56.87, 65.96]
Unintended live birth	8.51	[6.95, 10.38]	14.29	[12.11, 16.78]	10.87	[8.19, 14.29]
Abortion	46.54	[45.81, 51.27]	35.42	[32.57, 38.39]	27.62	[23.56, 32.09]

Note. Data weighted to account for multistage stratified sampling design and obtain nationally representative estimates. IPV = intimate partner violence.

Total weighted sample.

Socio-demographic characteristics

The mean age of ever-partnered women with a pregnancy in the past 5 years preceding the survey was 29 to 30 years in all three countries. In Ukraine, respondents predominantly lived in urban areas (63.4%), and in Moldova in rural areas (59.9%), with Azerbaijan having a relatively equal proportion of respondents in urban and rural areas (54.3% vs. 45.6%, respectively). The percentage of women with education above secondary school was highest in Ukraine (63.4%), followed by Moldova (40.2%) and Azerbaijan (26.4%). The percentage of formerly married (or cohabitating) women was lowest in Azerbaijan (3.6%), followed by Moldova (7.6%) and Ukraine (11.2%).

Physical and/or sexual intimate partner violence

Table 1 shows that 22.4% of respondents in Moldova, 13.8% in Azerbaijan, and 10.9% in Ukraine reported ever experiencing any act of physical and/or sexual IPV from their most recent partner (being punched, kicked, strangled, burned, or threatened or attacked with a weapon; being physically forced to have sexual intercourse or perform other sexual acts against their will; or having forced first sexual intercourse).

Unintended pregnancy and reproductive health covariates

A number of substantial differences were observed among the three countries in regard to reproductive health indicators. Last pregnancy within the last 5 years preceding the survey resulted in abortion for 46.5% of women in Azerbaijan, 35.4% of women in Moldova, and 27.6% of women in Ukraine. In Azerbaijan, 27% of respondents have never used any method of contraception compared with 4.8% in Moldova and 15.8% in Ukraine. The median number of living children prior to the last pregnancy was one child in Moldova and Ukraine and two children in Azerbaijan. The majority of women in the three countries had no abortions prior to the last pregnancy (median = 0); however, 12.5% in Azerbaijan, 8% in Moldova, and 7% in Ukraine had 3 or more abortions prior to the last pregnancy. The maximum number of abortions per woman prior to last pregnancy was as high as 17 in Azerbaijan, 21 in Moldova, and 9 in Ukraine.

Regression Results

Unadjusted analysis

In all three countries, unadjusted models (Model 1) demonstrated strong associations between ever experiencing physical and/or sexual IPV from most recent partner and having unintended last pregnancy (Table 2). Women with a history of physical and/or sexual IPV demonstrated higher likelihood of last pregnancy resulting in abortion in Azerbaijan (B = 0.67, SE = 0.19, p < .001), Moldova (B = 0.83, SE = 0.15, p < .001), and Ukraine (B = 1.25, SE = 0.36, p < .001). In addition, in Moldova and Ukraine, women with a history of physical and/or sexual IPV were more likely to report unintended last birth (B = 0.7, SE = 0.22, p < .001 and B = 1.67, SE = 0.42, p < .001).

Table 2.

Multinomial Survey Logistic Regression for Physical and/or Sexual IPV From Most Recent Cohabitating Partner and Outcomes of Last Pregnancy Among Ever-Married (or Cohabiting) Women in Azerbaijan, Moldova, and Ukraine Who Had a Pregnancy in the Last 5 Years Preceding the Survey.

	Azerbaijan						Moldova						Ukraine
	Outcome of last pregnancy						Outcome of last pregnancy						Outcome of last pregnancy
	Unintended live birth			Abortion			Unintended live birth			Abortion			Unintended live birth			Abortion
Variables	B	SE	[95% CI]	B	SE	[95% CI]	B	SE	[95% CI]	B	SE	[95% CI]	B	SE	[95% CI]	B	SE	[95% CI]
Model 1 (unadjusted)
Ever experienced any physical or sexual IPV
No	ref			ref			ref			ref			ref			ref
Yes	0.18	0.32	[−0.45, 0.81]	0.67***	0.19	[0.3, 1.03]	0.7***	0.22	[0.27, 1.13]	0.83***	0.15	[0.54, 1.12]	1.67***	0.42	[0.85, 2.49]	1.25***	0.36	[0.54, 1.96]
Constant	−1.64***	0.13	[−1.9, −1.38]	0.03	0.07	[−0.1, 0.17]	−1.41***	0.12	[−1.64, −1.17]	−0.52***	0.08	[−0.69,−0.38]	−1.95***	0.18	[−2.31,− 1.59]	−0.93	0.12	[−1.16, −0.7]
Model 2 (adjusted)
Ever experienced any physical or sexual IPV by age interaction
15-24 years	0.83	0.52	[−0.19, 1.86]	0.49	0.48	[−0.45,1.42]	0.23	0.38	[−0.52, 0.99]	0.14	0.40	[−0.64, 0.92]	1.94*	0.96	[0.05, 3.82]	2.83**	1.04	[0.78, 4.89]
25-34 years	−0.67	0.62	[−1.9, 0.56]	1.2***	0.35	[0.51, 1.9]	−0.41	0.36	[−1.12, 0.3]	−0.15	0.33	[−0.79, 0.5]	0.57	0.85	[−1.11, 2.24]	0.69	0.66	[−.6, 1.99]
35-45 years	0.56	0.66	[−0.73, 1.85]	1.27**	0.48	[0.33, 2.21]	0.52	0.62	[−0.71, 1.75]	0.99*	0.47	[0.07, 1.91]	0.91	0.87	[0.81, 2.62]	−1.81	1.06	[−3.89, 0.28]
Age
15-24	ref			ref			ref			ref			ref			ref
25-34	0.05	0.30	[−0.54, 0.64]	0.16	0.26	[−0.034,0.67]	−0.66*	0.27	[−1.19, −0.13]	−0.52*	0.24	[−1.0, −0.4]	−1.19*	0.49	[−2.16,−0.22]	−0.28	0.43	[−1.12, 0.56]
35-45	0.06	0.47	[−0.86, 0.97]	1.29***	0.33	[0.64, 1.94]	−0.01	0.39	[−0.78, 0.76]	0.06	0.30	[−0.52, 0.65]	−0.83	0.72	[−2.24, 0.59]	0.15	0.50	[−0.83, 1.12]
Residence
Urban	ref			ref			ref			ref			ref			ref
Rural	−0.22	0.37	[−0.95, 0.51]	−0.11	0.20	[−0.5, 0.29]	−0.26	0.25	[−0.75, 0.23]	−0.39*	0.18	[−0.74,−0.04]	1.18**	0.39	[0.42, 1.94]	−0.83*	0.38	[−1.57,−0.08]
Marital status
Currently married	ref			ref			ref			ref			ref			ref
Formerly married	0.25	0.68	[−1.1, 1.59]	0.75	0.68	[−0.59, 2.1]	1.26***	0.37	[0.54, 1.98]	1.78***	0.40	[1.0, 2.56]	0.73	0.64	[−0.52, 1.98]	1.51**	0.48	[0.57, 2.45]
Level of education
Above secondary school	ref			ref			ref			ref			ref			ref
Secondary or below	−0.11	0.30	[−0.7, 0.47]	0.25	0.22	[−0.19, 0.69]	0.11	0.23	[−0.33, 0.56]	−0.12	0.18	[−0.47, 0.23]	0.19	0.40	[−0.59, 0.97]	−0.01	0.33	[−0.65, 0.64]
Wealth status
Middle or high	ref			ref			Ref			ref			ref			ref
Poor	−0.3	0.34	[−0.98, 0.37]	−0.21	0.22	[−0.65, 0.22]	0.58**	0.22	[0.16, 1.0]	−0.33	0.21	[−0.75, 0.09]	−1.02*	0.49	[−1.98, −0.06]	0.17	0.40	[−0.61, 0.95]
Living children prior to last pregnancy
None	ref			ref			ref			ref			ref			ref
One or more	1.79***	0.38	[1.05, 2.53]	3.31***	0.43	[2.45, 4.16]	1.33***	0.22	[0.89, 1.77]	2.8***	0.27	[2.26, 3.33]	0.79*	0.39	[0.03, 1.54]	2.69***	0.45	[1.81, 3.57]
Had an abortion prior to last pregnancy
None	ref			ref			ref			ref			ref			ref
One or more	−0.01	0.26	[−0.52, 0.52]	1.66***	0.18	[1.3, 2.02]	0.12	0.20	[−0.28, 0.51]	0.59***	0.16	[0.27, 0.9]	−0.61	0.48	[−1.55, 0.34]	2.14***	0.27	[1.6, 2.68]
Ever used contraceptives
No methods	ref			ref			ref			ref			ref			ref
Traditional only	−0.26	0.30	[−0.86, 0.33]	0.41	0.22	[−0.02, 0.84]	−0.07	0.60	[−1.26, 1.11]	−0.26	0.46	[−1.16, 0.65]	1.27	0.66	[−0.03, 2.56]	0.69	0.59	[−0.48, 1.85]
Modern methods	−0.21	0.31	[0.82, 0.4]	0.29	0.21	[−0.12, 0.71]	0.14	0.54	[−0.93, 1.2]	−0.17	0.37	[−0.9, 0.56]	0.94	0.48	[−0.01, 1.88]	1.03*	0.48	[0.08, 1.97]
Constant	−2.52*	1.07	[−4.62, −0.42]	−4.92***	0.99	[−6.87,−2.98]	−3.16***	0.84	[−4.81, −1.51]	−3.42***	0.66	[−4.72, −2.11]	−4.65***	1.02	[−6.65, −2.64]	−5.03***	0.87	[−6.74, −3.32]
F(df)	F(26, 276) = 10.73***						F(26, 373) = 9.97***						F (26, 420) = 8.50***
PSU	318						400						498
Strata	17						2						53
Weighted sample	1,620						1,377						544

Note. The multinomial regression simultaneously compares two pregnancy outcomes (“unintended live birth” and “abortion”) against “wanted live birth” as the base outcome category. IPV = intimate partner violence; PSU = primary sampling units.

p < .05. **p < .01. ***p < .001.

Adjusted analysis

Key covariates associated with outcomes of last pregnancy are presented in Table 2. Compared with 15- to 24-year-old ever-married women, 25- to 34-year-old ever-married women in Moldova and Ukraine had a lower probability of unintended last pregnancy and 35- to 45-year-old ever-married women in Azerbaijan demonstrated a higher probability of last pregnancy ending in abortion.

In all three countries, the reports of unintended last pregnancy (resulting in unintended birth or abortion) were higher among women who already had living children prior to the last pregnancy. In Moldova and Ukraine, the probability of unintended pregnancy (either manifested through abortion and/or unintended birth) was higher among formerly married women, when compared with currently married women.

In Moldova, women of poor wealth status were more likely to keep unintended pregnancy reporting higher probably of unintended last birth (B = 0.58, SE = 0.22, p < .01), whereas in Ukraine the reports of unintended last births were lower among women from poor households (B = −1.02, SE = 0.49, p < .05). Compared with women from urban areas, women from rural areas in Moldova and Ukraine were less likely to use abortion. Furthermore, women from rural areas in Ukraine were more likely to keep unintended last pregnancy, reporting higher probability of unintended last birth (B = 1.18, SE = 0.39, p < .05). In all three countries, having an abortion in the past was highly associated with terminating last pregnancy in abortion as well.

The adjusted analysis demonstrated that the relationship between IPV and unintended pregnancy differed by age groups and final adjusted regression models (Model 2) included IPV-by-age interaction term.

After adjusting for key covariates, the probability of last pregnancy resulting in abortion was significantly higher among 25- to 34-year-olds (B = 1.2, SE = 0.35, p < .001) and 35- to 45-year-old women (B = 1.27, SE = 0.48, p < .001) in Azerbaijan and 35- to 45-year-old women in Moldova (B = 0.99, SE = 0.47, p < .05), who had ever experienced physical and/or sexual IPV from their most recent partner. On the contrary, in Ukraine, younger women (15-24 years of age) with a history of physical or sexual IPV from their most recent partner were more likely to report unintended last birth (B = 1.94, SE = 0.96, p < .05) or terminate last pregnancy through abortion (B = 2.83, SE = 1.04, p < .01), compared with women in this age category who had no history of physical or sexual IPV.

Discussion

The study findings support the hypothesis and show that IPV is associated with unintended pregnancy in three countries of the former Soviet Union. Women with a history of physical and/or sexual IPV from current or most recent partner demonstrate elevated risks of last pregnancy resulting in abortion in all three countries (Azerbaijan, Moldova, and Ukraine) and higher risks of unintended live births in Ukraine. Ever experiencing physical and/or sexual IPV from most recent partner was associated with higher risk of unintended pregnancy among younger women in Ukraine and among older women in Moldova and Azerbaijan.

Outside the fSU region, the DHS uses a retrospective approach to measuring unintended pregnancy; women report the wantedness of a child after giving birth or becoming pregnant. Using this measure, lifetime physical and/or sexual IPV was found to be associated with higher odds of unintended pregnancy in studies using DHS data from Colombia (aOR = 1.4; Pallitto & O’Campo, 2004), India (aOR = 1.3; Stephenson et al., 2008), and Bangladesh (aOR = 1.5; Silverman et al., 2007). The risks of unintended pregnancy were particularly higher when sexual abuse was combined with physical violence (Martin et al., 1999; Pallitto & O’Campo, 2004). A case-control study in Uganda identified a similarly strong relationship between IPV and unwanted pregnancy among women seeking post-abortion services, after adjusting for contraceptive use, pregnancy intentions, and other confounders (Kaye, Mirembe, Bantebya, Johansson, & Ekstrom, 2006). A study in rural India demonstrated that the relationship between IPV and unintended pregnancy is even stronger when prospective measures of unintended pregnancy are used (Stephenson et al., 2008).

Among the three selected countries of the fSU, the relationship between IPV and unintended live birth was observed only in Ukraine, whereas in all three countries women exposed to IPV were more likely to report induced abortions. Focusing only on unintended live birth would underestimate the effect of IPV on unintended pregnancy in countries, where abortions are more culturally acceptable, safe, and accessible and where women have a choice of terminating unintended pregnancy through abortion.

A number of studies in other developing countries have also examined associations between IPV and pregnancy loss (Hindin et al., 2008; Kuning et al., 2003; Silverman et al., 2007). Significant associations have been identified between lifetime exposure to IPV (physical and/or sexual) and a history of pregnancy loss in Bangladesh (aOR = 1.5), Bolivia (aOR = 1.6), Dominican Republic (aOR = 1.8), Malawi (aOR = 1.6), and Zimbabwe (aOR = 1.7), as well as Moldova (aOR = 1.7). These studies, however, used a binary measure of pregnancy loss that combined a history of stillbirth, miscarriage, or abortion without demonstrating a clear relationship between IPV and abortions. One study in Bangladesh examined different types of pregnancy loss as separate outcomes and confirmed significant relationships for abortions and miscarriages but not for stillbirths (Silverman et al., 2007).

Since the collapse of the Soviet Union with the introduction of family planning and reproductive health programs, the most significant reduction in abortion has been observed in Ukraine, where the abortion rate dropped to 0.4 abortions per woman in 2007 (UCSR et al., 2008). In Moldova, however, the abortion rate was dropping at first, but after 2001 stagnated at the rate of 1.1 abortions per woman (NCPM & ORC Macro, 2006). The abortion rate in Azerbaijan fell but remains high, at 2.3 abortions per woman (State Statistical Committee & Macro International, 2008). A recent gap between the number of boys and girls in Azerbaijan demonstrates a preference for sons (Hortacsu, Bastug, & Muhammetberdiev, 2001) and suggests a tendency toward sex-selective abortions (ADB, 2006).

In Moldova and Ukraine, women from rural areas were less likely to terminate last pregnancy through abortion, and in Ukraine, women from rural areas were also more likely to keep unintended last pregnancy reporting high likelihood of unintended live birth. In rural areas, women may have limited access to abortion compared with women from urban centers. There were no significant rural–urban differences observed with regard to the outcomes of unintended pregnancy in Azerbaijan. Moldova and Ukraine have a better system of reproductive and maternal health care services compared with Azerbaijan, where regardless of residence, women may have limited access to and poor utilization of reproductive health services. In Moldova, being the poorest among the three countries, poverty was associated with higher reports of unintended live birth. Women from poor households in Moldova may have limited means to use abortion and are more likely to keep unintended pregnancy. The reports of unintended pregnancy were also higher among formerly married women in Ukraine and Moldova, where divorce and separation are more common and socially accepted compared with that in Azerbaijan.

The study has a number of limitations. Despite the strong associations between IPV and pregnancy outcomes found in this article, the use of cross-sectional data precludes us from making causal inferences. Furthermore, the data collection procedure, an interviewer-administered survey, may pose constraints to the disclosure of sensitive information. The presence of an interviewer may create a situation for social desirability bias and decrease reporting of unwanted pregnancies, large numbers of abortions, and IPV, particularly sexual IPV. Despite certain limitations, interviewing is still considered a legitimate data collection method for studies of IPV and reproductive health outcomes, especially if interviewers are trained and administer a standardized instrument, as in the DHS. Another limitation is that the analysis relied on self-reported data about women’s pregnancy outcomes and did not compare their reported health outcomes with health records. Less than 5% of respondents reported ever experiencing sexual IPV, which significantly limited the power to identify significant relationships between pregnancy outcomes and sexual IPV separately from physical IPV.

Despite the limitations, the article has a number of research, policy, and practice implications. The findings lay the groundwork for future studies aiming to understand the complex relationship between IPV and unintended pregnancy. To identify causal relationships between IPV and family planning indicators, future studies accounting for temporal factors—the onset of IPV and decisions related to adoption or discontinuation of contraception—are indispensable. To be especially beneficial in program development, the studies should examine mechanisms linking unintended pregnancy and IPV. Moreover, studies using longitudinal designs and including partner variables (e.g., desire to have children, communication about safe sex practices) are necessary to test the effects of factors mediating the relationship between pregnancy outcomes and IPV and determining their role in preventing and reducing the negative consequences of IPV. Studies examining various mechanisms linking different types of IPV with contraceptive use, unintended pregnancies, and abortions could be extremely useful in developing programs aiming to reduce the effect of IPV on health outcomes among women who experience violence in their relationships.

The findings in this article suggest that the relationship between IPV and women’s health may go far beyond physical injuries and seriously undermine women’s pregnancy outcomes. The results of this study illustrate the need for reproductive health programs in these fSU countries to address IPV as a potential obstacle to preventing unintended pregnancies. Likewise, women receiving or seeking services for IPV from women’s crisis centers or shelters should also be assessed for their family planning needs and risks of unintended pregnancy. Furthermore, women’s health clinics, family planning centers, and hospital units providing abortion services in Azerbaijan, Moldova, and Ukraine could be effective settings for screening women for IPV and providing necessary assistance, referrals, and health information. To further reduce the number of abortions, reproductive health programs in these countries need to take into account the specific risks of women living in abusive relationships, especially in the traditional society of Azerbaijan where divorce is less accepted and women rarely leave their abusive partners. In such situations, a harm-reduction approach could be applied to reproductive health programs to minimize the risk of unintended pregnancies among women who are currently unable or unwilling to leave their abusive partners.

Conclusion

Power dynamics in relationships should be taken into account when understanding the reproductive health behaviors and decisions of women in the fSU region. Strong associations in the fSU region between women’s abusive experiences and unintended pregnancy suggest the importance of designing new programs or modifying current reproductive health programs to address the special circumstances and needs of women who have experienced violence from their partners or who continue living with their abusive partners.

Footnotes

Acknowledgements

Authors would like to thank Dr. Sunita Kishor and Sarah Bradley for their insightful feedback in writing this paper. Special thanks to Isaac Henry Ergas’s family for supporting research in Turkic and Central Asian studies.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study used publicly available data from the Demographic and Health Surveys (DHS ) that have been initiated by the Centers for Disease Control (CDC) in partnership with the Bloomberg School of Public Health at John Hopkins University and implemented with the joint support of the U.S. International Development Agency (USAID), the United Nations Population Fund (UNFPA), the United Nations Children’s Fund (UNICEF), and the National Ministries of Health in respective countries. This article is partially supported by USAID through the DHS Global Health Fellowship at ICF Macro, the Azerbaijan Diplomatic Academy (ADA), and an Isaac Henry Ergas Fellowship at Columbia University. The fellowships were awarded to the first author, Leyla Ismayilova.

Author Biographies

Leyla Ismayilova is an assistant professor at the University of Chicago, School of Social Service Administration. She specializes in the development and cultural adaptation of preventive interventions to improve child and family well-being in developing countries. She obtained her doctorate in clinical social work from Columbia University in 2009. She received her master’s degree in social work from Columbia University School of Social Work, with a concentration in advanced clinical practice, and was among the first Open Society Institute/Soros Foundation fellows from the former Soviet Union at Columbia. She also holds a master’s degree in psychology from Baku State University, Azerbaijan. She was the founder and director of the Center for Psychological Counseling in Azerbaijan, which provided counseling services to women, children, and men affected by violence and abuse.

Nabila El-Bassel is the Willma and Albert Musher Professor at the Columbia University School of Social Work (CUSSW). She is the director of the Social Intervention Group (SIG) at CUSSW, a multidisciplinary research center that works to develop and test effective intervention models to prevent co-occurring problems of intimate partner violence, HIV, drug abuse and trauma. She is also the director of the Columbia University Global Health Research Center of Central Asia (GHRCCA), a multidisciplinary research center committed to advancing solutions to health and social issues in Central Asia.

References

Asian Development Bank. (2006). Gender assessment synthesis report: Mainstreaming gender in poverty reduction strategies in four Central Asian republics. Manila, Philippines: Author.

Bawah

Akweongo

Simmons

Phillips

(1999). Women’s fears and men’s anxieties: The impact of family planning on gender relations in northern Ghana. Studies in Family Planning, 30, 54-66.

Blanc

A. K.

(2001). The effect of power in sexual relationships on sexual and reproductive health: An examination of the evidence. Studies in Family Planning, 32, 189-213.

Centers for Disease Control. (2003). Reproductive, maternal and child health in Eastern Europe and Eurasia: A comparative report. Atlanta, GA: Centers for Disease Control and ORC Macro.

Chambers

R. L.

Skinner

C. J.

(2003). Analysis of survey data. Hoboken, NJ: John Wiley.

Coker

A. L.

(2007). Does physical intimate partner violence affect sexual health? A systematic review. Trauma, Violence, & Abuse, 8, 149-177.

Coker

A. L.

Sanderson

Dong

(2004). Partner violence during pregnancy and risk of adverse pregnancy outcomes. Paediatric and Perinatal Epidemiology, 18, 260-269.

Dixon-Mueller

(1993). The sexuality connection in reproductive health. Studies in Family Planning, 24, 269-282.

Fox

A. M.

Jackson

S. S.

Hansen

N. B.

Gasa

Crewe

Sikkema

K. J.

(2007). In their own voices: A qualitative study of women’s risk for intimate partner violence and HIV in South Africa. Violence Against Women, 13, 583-602.

10.

Goldberg

Serbanescu

(2002). Induced abortion in the Caucasus Republics: A detailed analysis. Atlanta, GA: Centers for Disease Control and Prevention.

11.

Hensel

Gudim

(2004). Moldova’s economic transition: Slow and contradictory. The EU and Moldova: On a fault-line of Europe. London, England: Federal Trust.

12.

Heyat

(2006). Globalization and changing gender norms in Azerbaijan. International Feminist Journal of Politics, 8, 394-412.

13.

Hindin

M. J.

Kishor

Ansara

D. L.

(2008). Intimate partner violence among couples in 10 DHS countries: Predictors and health outcomes. Calverton, MD: Macro International.

14.

Hortacsu

Bastug

S. S.

Muhammetberdiev

O. B.

(2001). Desire for children in Turkmenistan and Azerbaijan: Son preference and perceived instrumentality for value satisfaction. Journal of Cross-Cultural Psychology, 32, 309-321.

15.

Hyde

(2008). Abortion and the birth rate in the USSR. Journal of Biosocial Science, 2, 283-292.

16.

Jordan

C. E.

(2007). The health implications of violence against women: Untangling the complexities of acute and chronic effects: A two-part special issue. Trauma, Violence, & Abuse, 8, 87-89.

17.

Kaye

D. K.

Mirembe

F. M.

Bantebya

Johansson

Ekstrom

A. M.

(2006). Domestic violence as risk factor for unwanted pregnancy and induced abortion in Mulago Hospital, Kampala, Uganda. Tropical Medicine & International Health, 11, 90-101.

18.

Kuning

McNeil

Chongsuvivatwong

(2003). Pregnancy loss in the Philippines. Southeast Asian Journal of Tropical Medicine and Public Health, 34, 433-442.

19.

Macro International. (1996). Sampling manual (DHS-III Basic documentation #6). Calverton, MD: Author.

20.

Manijeh

(1999). The impact of economic and political transformation on women: The case of Azerbaijan. Central Asian Survey, 18, 111-120.

21.

Martin

S. L.

Kilgallen

Tsui

A. O.

Maitra

Singh

K. K.

Kupper

L. L.

(1999). Sexual behaviors and reproductive health outcomes: Associations with wife abuse in India. Journal of American Medical Association, 282, 1967-1972.

22.

NCPM & ORC Macro. (2006). Moldova Demographic and Health Survey 2005. Calverton, MD: National Scientific and Applied Center for Preventive Medicine of the Ministry of Health and Social Protection (NCPM, Moldova) and ORC Macro.

23.

Pallitto

C. C.

Campbell

J. C.

O’Campo

(2005). Is intimate partner violence associated with unintended pregnancy? A review of the literature. Trauma, Violence & Abuse, 6, 217-235.

24.

Pallitto

C. C.

O’Campo

(2004). The relationship between intimate partner violence and unintended pregnancy: Analysis of a national sample from Colombia. International Family Planning Perspectives, 30(4), 165-173.

25.

Plichta

S. B.

(2004). Intimate partner violence and physical health consequences: Policy and practice implications. Journal of Interpersonal Violence, 19, 1296-1323.

26.

Remez

(2003). Unintended pregnancy is linked to inadequate prenatal care, but not to unattended delivery or child health. International Family Planning Perspectives, 29(3), 146-147.

27.

Ross

J. A.

Winfrey

W. L.

(2002). Unmet need for contraception in the developing world and the Former Soviet Union: An updated estimate. International Family Planning Perspectives, 28(3), 138-143.

28.

Rutstein

Kiersten

(2004). The DHS Wealth Index. Calverton, MD: ORC Macro.

29.

Schei

Guthrie

J. R.

Dennerstein

Alford

(2006). Intimate partner violence and health outcomes in mid-life women: A population-based cohort study. Archives of Women’s Mental Health, 9, 317-324.

30.

Scott

(2000, April). Recent changes in social conditions of countries in transition (Discussion Paper No. 117). Geneva, Switzerland: United Nations Research Institute for Social Development.

31.

Sen

George

östlin

(2002). Engendering international health: The challenge of equity. Cambridge, MA: MIT Press.

32.

Serbanescu

Goldberg

Morris

(2005). Reproductive health in the transition countries of Europe. In Macura

MacDonald

A. L.

Haug

(Eds.), The new demographic regime: Population challenges and policy responses (pp. 177-198), New York: United Nations Economic Commission for Europe.

33.

Silverman

J. G.

Gupta

Decker

M. R.

Kapur

Raj

(2007). Intimate partner violence and unwanted pregnancy, miscarriage, induced abortion, and stillbirth among a national sample of Bangladeshi women. British Journal of Obstetric and Gynaecology, 114, 1246-1252.

34.

Singh

Prada

Kestler

(2006). Induced abortion and unintended pregnancy in Guatemala. International Family Planning Perspectives, 32(3), 136-145.

35.

State Statistical Committee, & Macro International. (2008). Azerbaijan Demographic and Health Survey 2006. Calverton, MD: Azerbaijan State Statistical Committee (AzSSC) and Macro International.

36.

Stephenson

Koenig

Acharya

Roy

(2008). Domestic violence, contraceptive use, and unwanted pregnancy in rural India. Studies in Family Planning, 39, 177-186.

37.

Stephenson

Koenig

Ahmed

(2006). Domestic violence and contraceptive adoption in Uttar Pradesh, India. Studies in Family Planning, 37, 75-86.

38.

Suhrcke

Rocco

McKee

(2007). Health: A vital investment for economic development in Eastern Europe and Central Asia. Copenhagen, Denmark: World Health Organization.

39.

Tabachnick

Fidell

(2006). Using multivariate statistics (5th ed.). Boston, MA: Allyn & Bacon.

40.

Tohidi

(1996). Soviet in public, Azeri in private: Gender, Islam, and nationality in Soviet and post-Soviet Azerbaijan. Women’s Studies International Forum, 19, 111-123.

41.

Ukrainian Center for Social Reforms, State Statistical Committee, Ministry of Health, & Macro International. (2008). Ukraine Demographic and Health Survey 2007. Calverton, MD: Ukrainian Center for Social Reforms, State Statistical Committee [Ukraine], Ministry of Health [Ukraine], Macro International.

42.

United Nations Population Fund. (2008). UNFPA 2007 annual report. New York: Author.

43.

U.S. International Development Agency. (2007). Family planning situation analysis 2007: Executive summary. Arlington, VA: John Snow/Europe and Eurasia Regional Family Planning Activity for the U.S. Agency for International Development.

44.

World Health Organization. (2002). World report on violence and health. Geneva, Switzerland: Author.

45.

Worth

(1989). Sexual decision-making and AIDS: Why condom Promotion among vulnerable women is likely to fail. Studies in Family Planning, 20, 297-307.