Sexual Violence and Reproductive Health among Youth in Port-au-Prince, Haiti

Abstract

We examine sexual violence and reproductive health outcomes among sexually experienced youth in Port-au-Prince, Haiti, using the Priorities for Local AIDS Control methodology to identify participants in locations where sexual partnerships are formed. Sexual violence is common and is significantly associated with condom use, pregnancy experience and recent STI symptoms.

For young women, the sequalae of sexual violence include increased likelihood of risky sexual behavior and adverse reproductive health outcomes [1-3]. Few studies on sexual violence include young men. In Haiti, where socio-political instability has been accompanied by widespread violence [4], the 2005 Demographic and Health Survey (DHS) data indicate that 18% of 15-24 year old females have ever experienced sexual violence.¹ The present study examines the effect of sexual violence on reproductive health outcomes among male and female youth in Port-au-Prince, Haiti.

Methods

Using data collected in 2006-2007 with the Priorities for Local AIDS Control Efforts (PLACE) method, we examine the association between lifetime experience of sexual violence and reproductive health outcomes and sexual behaviors among sexually-experienced youth. The PLACE method is a novel, rapid assessment methodology aimed at improving HIV/AIDS prevention programming in places transmission is most likely to occur [5]. The method uses a systematic, venue-based sampling approach [6]. In the present study, the method was modified to focus on locations where youth meet sexual partners. The method uses five steps: community identification of a target study area; identification of venues where youth meet sexual partners through interviews with knowledgeable community informants; (n=494 in this study); venue confirmation; survey of about 14 randomly selected youth aged 15-24 in each of 37 randomly selected venues; and community-based results dissemination. In this PLACE study, a total of 504 youth were interviewed. This analysis focuses on sexually-experienced youth (75% of full sample, n=376). Nineteen youth missing data on sexual violence, pregnancy experience, STI symptoms and transactional sex were dropped from the analysis; the analysis sample is 357 youth. The University of North Carolina Institutional Review Board (IRB) reviewed this project and determined that it did not constitute human subjects research as defined in federal regulations, and that further IRB review and approval were therefore not required.

The outcomes of interest are condom use at last sex, pregnancy experience (having ever been pregnant or gotten a partner pregnant), and transactional sex (giving and/or receiving gifts/money in exchange for sex). Sexual violence is a binary variable, measured from questions about whether the respondents was physically forced or raped at first sex and whether the respondent had ever had sex because of threats or physical force. Results are presented for the full sample, as well by gender. Pearson’s chi-squared tests and multivariate logistic regression models are presented, adjusting for factors previously found to be associated with sexual risk-taking including age, gender, education level, residence in the study area, knowledge of someone with HIV/AIDS, and history of meeting a sexual partner at the venue [3, 7].

Results

The mean age among respondents was 20 years. The majority had completed secondary education (72%, not shown). About 18% of the sample ever experienced sexual violence during their lifetimes (Table 1). Fifteen percent of males reported sexual violence during their lifetimes, compared to 22% of females; this difference was marginally significant (p=0.10). For all individuals, those who experienced sexual violence are significantly more likely to have ever been pregnant or gotten a partner pregnant, and to have experienced recent STI symptoms compared with youth who had not experienced sexual violence. While those who experienced sexual violence compared with those who have not experienced violence have lower levels of condom use at last sex and higher levels of transactional sex, these differences do not attain statistical significance. Female and male youth who experienced sexual violence are significantly more likely to report recent STI symptoms compared with females and males with no sexual violence experience. Males who experienced sexual violence are also significantly more likely to report a partner’s pregnancy than males who did not experience sexual violence. While differences on the other outcomes are not statistically significant in genderspecific cross-tabulations, females and males who experienced sexual violence generally exhibit greater levels of sexual risk.

Table 1.

Sexual violence experience and selected reproductive health measures for full sample and by gender, frequency and percentage distributions according to the experience of sexual violence

	Full sample			Females			Males
	Proportion reporting behavior n = 357	Experienced sexual violence n = 65	Did not experience sexual violence n = 292	Proportion reporting behavior n =154	Experienced sexual violence n = 34	Did not experience sexual violence n = 120	Proportion reporting behavior n = 203	Experienced sexual violence n = 31	Did not experience sexual violence n = 172
Sexual violence
Any sexual violence in lifetime	18.2	--	--	22.1	--	--	15.3	--	--
First sex described as physically forced or rape	7.1	--	--	3.5	--	--	11.8	--	--
Ever had sex because of partner’s threats or physical force	12.9	--	--	12.8	--	--	13.0	--	--
Reproductive health
Condom use at last sex1	65.4	60.0	66.5	59.4	53.3	61.1	69.8	68.0	70.1
Ever pregnant/gotten a partner pregnant	24.1	36.9	21.2**	24.0	32.4	21.7	24.1	41.9	20.9*
STI symptoms in past 4 weeks	58.8	76.9	54.8**	70.8	82.4	67.5+	49.8	71.0	45.9+
Ever exchanged money or gifts sex	22.4	26.2	21.6	11.0	17.7	9.2	31.0	35.5	30.2

⁺significant at 10%;

^*significant at 5%;

^**significant at 1%.

Symbols indicate significant differences in the outcome or behavior between those who experienced sexual violence and those who did not experience sexual violence.

¹30 individuals were missing data on condom use (16 females and 14 males).

In multivariate models (Table 2), individuals who experienced sexual violence are marginally less likely (OR: 0.6, 95% CI 0.3-1.1) to have used a condom at last sex and more likely to have pregnancy experience (3.0, 95% CI 1.4-6.5) and to have recent STI symptoms (2.7, 95% CI 1.3-5.5) compared with youth with no history of sexual violence. In models stratified by gender, sexual violence was a significant risk factor for having ever been pregnant for women (3.2, 95% CI 1.0-10.0) and getting a partner pregnant for men (3.1, 95% CI 1.0-9.5), as well as recent STI symptoms for men (3.1, 95% CI 1.3-7.6). No other significant relationships between sexual violence and outcome measures were found.

Table 2.

Odds ratios and 95% confidence intervals from multivariate logistic regression models examining the effect of sexual violence on selected reproductive health outcomes and behaviors

	Full sample n = 328	Females n = 140	Males n = 188
Condom use at last sex1	0.6 (0.3 - 1.1)+	0.5 (0.2 - 1.3)	0.6 (0.2 - 1.6)
Ever pregnant/gotten a partner pregnant	3.0 (1.4 – 6.5)**	3.2 (1.0 – 10.0)*	3.1 (1.0 – 9.5)*
STI symptoms in past 4 weeks	2.7 (1.3 - 5.5)**	2.1 (0.7 - 6.4)	3.4 (1.3 – 8.6)+
Ever exchanged money or gifts for sex	0.8 (0.3 - 1.8)	2.3 (0.3 - 17.3)	0.7 (0.3 - 1.9)

⁺significant at 10%;

^*significant at 5%;

^**significant at 1%

Models controlled for age, residence in the study area, educational attainment, whether the respondent knows someone with HIV/AIDS and whether the respondent has ever met a sexual partner at the venue.

Ns vary slightly from Table 1 due to missing data on control variables.

¹Ns for condom use models are 301 (full sample), 126 (females) and 175 (males).

Discussion

Sexual violence is significantly associated with condom use, pregnancy experience, and recent STI symptoms among youth socializing at venues where young people meet sexual partners in Port-au-Prince. Among young women in the PLACE sample, the prevalence of sexual violence (22%) is higher than among young women surveyed in the nationally-representative Haiti DHS (18%). While not population-based, PLACE data may be more useful to program planners given that the method identifies locations where sexually-active youth congregate [8]. Additionally, a unique feature of this study is the inclusion of males’ experiences with sexual violence and reproductive health outcomes, since both male and female youth visiting high-risk venues were surveyed.

Limitations exist with the present analysis. First, these findings represent only associations, not causal pathways. While the STI symptom measure represents the previous four weeks, pregnancy is a lifetime measure. Thus, the temporality of sexual violence and reproductive health outcomes is unclear. Additionally, we do not find a significant association between sexual violence and transactional sex in all models, and condom use was not significant in gender-stratified models. While sexual violence may not be related to these behaviors, it is also possible that our small sample size lacks power to detect significant differences in these measures. Finally, sexual violence is challenging to measure and often under-reported [9]. More details about the circumstances, perpetrators, frequency and intensity of sexual violence experiences are needed to inform the development of violence and HIV prevention programs.

This and other studies [1-3] indicate a need for programs for both young men and women to prevent sexual violence and promote protective behaviors among youth who have previously experienced sexual violence. While pregnancy and STI symptoms may be directly associated with sexual violence, victims of sexual violence may also have limited power in negotiating for condom and contraceptive use in future sexual encounters. Sexual violence may also establish an unhealthy pattern of sexual risk-taking throughout life [10]. Screening programs in schools could help to identify victims of sexual violence; these youth could be offered counseling, referrals, and training on sexual negotiation and refusal skills. In addition, interventions could take place in these high-risk venues to reach out-of-school youth, offering condoms to all youth and counseling and referrals for victims of sexual violence. Venue-based interventions have the potential to target high-risk youth and to reduce sexual violence and its deleterious effects among youth in Haiti.