Male victims of sexual violence in rural Malawi: The overlooked association with HIV infection

Abstract

In sub-Saharan Africa, research on intimate partner violence (IPV) has largely failed to consider men’s experiences as victims by female perpetrators—particularly within ongoing heterosexual relationships such as marriage. The objectives of this study were to document the prevalence of sexual coercion among men, to describe the characteristics of male victims, and to test for an association between sexual coercion and HIV positivity. In 2010, cross-sectional data on HIV risk behaviors, HIV status, and IPV was collected from a sample of 684 mostly-married men in rural Malawi. Bivariate analyses were used to examine differences in HIV risk characteristics between victims and non-victims of sexual coercion. Multivariate logistic regression was used to determine the association between sexual coercion and HIV positivity. Over one-tenth (10.4%) of men reported being a victim of sexual coercion. Male victims of sexual coercion were more likely to be married (p<0.05), older than 24 years (p<0.05), physically abused by a female partner (p<0.001), believed their partners were at higher risk for HIV (p<0.05), and had consumed alcohol in the past month (p<0.01). After controlling for potential confounders, the odds of being HIV positive were 7.2 times higher among men who had experienced sexual coercion (p<0.000). In sub-Saharan Africa, research on men’s experience of violence as victims is long overdue. More formative research is needed to understand the mechanisms through which men experience violence and how to appropriately measure IPV among male victims. While the data are cross-sectional and cannot evaluate causality, the strength of the association with HIV positivity merits further attention.

Introduction

In many parts of sub-Saharan Africa, intimate partner violence (IPV) is a pervasive public health issue that places women at increased risk for HIV infection (Dunkle et al., 2004; Jewkes et al., 2006; Jewkes et al., 2010; Zablotska et al., 2009). One important form of IPV related to HIV risk is sexual coercion, which refers to a range of experiences that compel a person to have sex against their will, including the use of violence, threats, verbal insistence or pressure, deception, and cultural expectations that have social or physical consequences if the person refuses (Heise et al., 1995). Given the gendered nature of sexual violence, most research from developing countries has centered on female victims or male perpetrators (Palermo et al., 2013). Consequently, little research has examined heterosexual men’s experiences of sexual coercion by female perpetrators. Of the few studies from sub-Saharan Africa, sexual coercion was reported by 11–70% of men (Erulkar, 2004; Hoffman et al., 2006; Ybarra et al., 2012).

Research on men’s experiences of sexual coercion is limited to studies on young men in premarital or casual relationships (Erulkar, 2004; Hoffman et al., 2006; Sikweyiya & Jewkes, 2009; Ybarra et al., 2012), which suggest that age discrepancies between partners plays an important role. For example, qualitative research highlights cases of ‘assertive seduction’ committed by older (and often wealthier) women against younger men who feel compelled to participate (Dunkle et al., 2007; Sikweyiya & Jewkes, 2009; Simpson, 2007). In a quantitative study, having an older partner was the strongest predictor of sexual coercion after controlling for HIV risk behaviors (Erulkar, 2004). While informative, these findings are unlikely to represent everyday relationships such as marriage: a quasi-universal demographic of Malawi (MDHS, 2011) and the setting where most new HIV infections occur (Dunkle et al., 2008).

Using a sample of mostly-married men from Malawi, we sought to document the prevalence of sexual coercion and describe the characteristics of male victims of sexual coercion. Among women, sexual coercion is associated with having more sexual partners, more alcohol use, more physical abuse, and less condom use (Erulkar, 2004; King et al., 2004; Pettifor et al., 2004; van der Straten et al., 1998). Thus, we assessed whether similar HIV risk characteristics were more prevalent among male victims, thereby giving insight into potential HIV transmission routes. However, our primary objective was to test for an association between sexual coercion and HIV positivity. In Malawi, women are 1.5 times more likely to be HIV positive (MDHS, 2011) and more likely to enter the marriage HIV-infected than men (Dunkle et al., 2008). Accordingly, we hypothesized that men who experienced unprotected, coercive sex with their wives would be more likely to be HIV positive.

Methods

The data come from Tsologo La Thanzi (TLT)¹, a longitudinal panel study on reproduction and AIDS among young people from southern Malawi. In 2009, approximately 1,500 female and 600 male respondents were randomly selected from a sampling frame of residents living within a seven-kilometer radius of Balaka town. The women were asked to invite their male partners to participate in TLT. The current study only uses data from male respondents (partners of the women and the random sample of men). Men were interviewed every four months for a total of 8 waves. At wave 3, a special module containing questions on IPV was added to the partnership survey. Men were asked the IPV questions with regard to their primary partner (i.e., spouse, if married). In order to minimize the burden on respondents who were subjected to eight lengthy interviews, sexual and physical IPV were captured as two dichotomous indicators.

At wave 3, approximately one-third of the TLT sample was tested for HIV. Our measure of HIV positivity was based on a combination of HIV serostatus (for those tested via TLT) and self-reported HIV status (those not tested via TLT). Sexual coercion was measured by the question, ‘has your partner ever forced you to have sex when you did not want to’ (yes/no). In the local language of Chichewa, ‘forced’ does not translate to physical rape but rather persistent pressure or persuasion to have sex. We also included several HIV risk factors in our multivariate analysis as covariates, including physical violence (‘has your partner ever hurt you by beating you?’; yes/no), consistent condom use with partner (yes/no), sexual concurrency in the past 4 months (yes/no), and alcohol use (‘In the past month, have you drank alcohol?’; yes/no). Finally, we controlled for age, years of education, being married/cohabitating (yes/no), relationship duration (years), and whether HIV positivity was based on serostatus or self-report.

First, we performed chi-squares tests to examine differences in characteristics between victims and non-victims of sexual coercion. Next, we used multivariate logistic regression to assess whether men who had experienced sexual coercion were more likely to be HIV positive after controlling for other factors. Non-normality and multi-collinearity did not influence the results. Since the largest percent of missing data for any variable was less than 2%, we used listwise deletion in our final analysis.

Results

On average, men were 26.1 years old and had been in their relationship for 4.3 years; the majority of men were married or cohabitating (78.6%). Thirteen men had an older female partner (1.9%). Only 10.1% reported using condoms consistently and 4.4% reported recent sexual concurrency. Twelve men (1.8%) reported being physically abused by their partner. Almost 17% consumed alcohol in the past month. One-tenth (10.4%) of men experienced sexual coercion. Table 1 shows that men who experienced sexual coercion were more likely to be married (p<0.05), older (p<0.05), physically abused by a female partner (p<0.001), believed their partners were at higher risk for HIV (p<0.05), and had consumed alcohol in the past month (p<0.01).

Table 1.

Characteristics of male victims of sexual coercion (N=684) at Tsogolo La Thanzi wave 3

Characteristic	Full Sample (N=684)	Male Victims (N=71)	Male Non-Victims (N=613)	P-value
Marital status
Married/cohabitating	78.6	90.1	77.2	0.01
Unmarried	21.4	9.9	22.8
Age
≤24 years	38.9	25.4	40.5	0.01
>24 years	61.1	74.6	59.5
Older partner
Yes	1.9	1.4	2.0	0.75
No	98.1	98.6	98.0
Relationship duration
≤4 years	54.1	48.6	54.8	0.32
>4 years	45.9	51.4	45.2
Consistent condom use
Yes	10.1	4.2	10.8	0.08
No	89.9	95.8	89.2
Physical abuse victim
Yes	1.8	8.5	1.0	<0.001
No	98.2	91.5	99.0
Sexual concurrency past 4 months
Yes	4.4	7.0	4.1	0.25
No	95.6	92.0	95.9
Drank alcohol in the past month
Yes	83.0	28.2	15.7	<0.01
No	17.0	71.8	84.3
Perceived partner infidelity
Yes	10.5	14.1	10.1	0.3
No	89.5	85.9	89.9
Partner’s likelihood of HIV infection
No/low	94.9	89.7	95.5	0.04
High/certain	5.1	10.3	4.5

P-values are based on chi-square tests for categorical variables.

Of the 684 men, 2.5% were HIV positive at wave 3. Of those who were tested at TLT’s wave 3 (N=163), 4.9% were HIV positive. Table 2 shows that the unadjusted odds of being HIV positive were 6.6 times higher for men who had experienced sexual coercion (p<0.000). After controlling for other covariates, the odds of being HIV positive were 7.2 times higher among men who had experienced sexual coercion (p<0.000). The multivariate models also showed that the odds of being HIV positive were 1.1 times higher for each one-year increase in age (p<0.01) and 4.9 times higher for men who reported recent sexual concurrency (p<0.05).

Table 2.

Odds ratios from multivariate logistic regression models predicting HIV positivity among men at Tsogolo La Thanzi (TLT) wave 3.

	OR	95% CI	AOR	95% CI
Primary explanatory variable
Sexual coercion victimization	6.60 ***	[2.43, 17.92]	7.20 ***	[2.42, 21.45]
Secondary explanatory variables
Age (years)	1.10 **	[1.03, 1.19]	1.13 **	[1.04, 1.23]
Education (years)	0.87	[0.75, 1.01]	0.88	[0.75, 1.03]
Married	1.28	[0.36, 4.52]	0.48	[0.08, 2.85]
Relationship duration (years)	1.07	[0.92, 1.25]	1.01	[0.85, 1.20]
Consistent condom use	1.19	[0.26, 5.32]	2.33	[0.34, 15.89]
Sexual concurrency in past four months	5.07 *	[1.37, 18.70]	4.93 *	[1.15, 21.59]
Drank alcohol in past four months	2.09	[0.72, 6.04]	1.56	[0.47, 5.15]

^*p<0.05;

^**p<0.01;

^***p<0.001

Discussion

This study was one of the first to examine men’s experiences of sexual coercion within committed, heterosexual relationships in sub-Saharan Africa. The results revealed that sexual coercion was the strongest predictor of HIV positivity, which warrants attention to the possibility that sexual coercion represents an overlooked risk factor for HIV among men. We also found that sexual coercion was not uncommon among men. While previous research has attributed men’s experience of coercive sex to age discrepancies within casual partnerships (Erulkar, 2004), male victims in our study were mostly married and paired with women of a similar age. Male victims were also more likely to suspect their partner was HIV infected, to experience physical abuse, and to report recent alcohol consumption. While we can only speculate, it is possible that alcohol use may be an important catalyst for physical violence and sexual coercion, as others have noted elsewhere (Zablotska et al., 2009). Male victims were not any more or less likely to use condoms or have multiple sexual partners, which reflects similar findings from Kenya (Erulkar, 2004). However, in our sample, condom use and sexual concurrency were low and potentially underreported, thus limiting our ability to detect differences.

Several limitations are noteworthy. First, sexual coercion was a binary variable and only captured a single dimension of sexual coercion against men. Rates of sexual coercion are likely to be higher with a more sensitive measure. Second, HIV positivity included self-reported HIV status, which may not accurately reflect HIV serostatus (Anglewicz & Kohler, 2009). However, when we restricted the sample to men who tested at wave 3 (N=163), we still found a strong positive association between sexual coercion and HIV positivity. Finally, our analysis was cross-sectional and we cannot make causal statements about directionality. Regardless, the strength of the association with HIV status, which persisted after controlling for potential confounders, merits further attention. Even if sexual coercion does not cause HIV infection but rather the reverse, couple communication and sexual risk reduction with an uninfected partner may represent an important area for intervention.

Despite these limitations, our findings are of considerable importance given the dearth of research and provide a foundation for other researchers to build upon. More formative research is required to understand the mechanisms through which female-to-male violence occurs and how to appropriately measure men’s experiences of violence—which likely differs from women’s. It is conceivable that in high fertility settings, men could be coerced into sex through women’s desires for children, and potentially while intoxicated. Such possibilities could be explored with qualitative research, but also through dyadic studies that collect information from both partners’ perspectives. In conclusion, future studies that are inclusive of men are urgently needed in order to fully understand the causes and consequences of IPV within ongoing heterosexual partnerships in sub-Saharan Africa.