Lifetime experiences of gender-based violence,depression and condom use among female sex workers in Cameroon

Abstract

Background:

In general populations, consistent data highlight the relationships among violence, HIV risk behavior and depression; however, these patterns are not well understood among female sex workers (FSWs). We examined the relationship between FSWs’ experiences with sexual violence and consistent condom use as a key HIV risk behavior and explored mental health as a potential mediator.

Methods:

In total, 2,165 FSWs were recruited via respondent-driven sampling in Cameroon in 2016. The women answered questions about violence, condom use and mental health.

Results:

Inconsistent condom use with clients was reported by 23.5% of participants (508/2,165). Lifetime sexual violence was prevalent with 33.0% (713/2,163) of participants. Almost 50% (1,067/2,143) of respondents had some level of depression. Sexual violence was significantly associated with inconsistent condom use (adjusted risk ratio (aRR) 1.4, 95% confidence interval (CI) (1.2–1.6)). Of FSWs with no depression, 24.9% (267/1,071) reported sexual violence, versus 56.1% (32/57) of respondents with severe depression (p < .01). Severe depression significantly increased risk of condomless sex (aRR 1.8, 95% CI (1.3–2.6)); in mediation analysis, both sexual violence and severe depression remained significant predictors of condomless sex (aRR 1.4, 95% CI (1.2, 1.6) and aRR 1.7, 95% CI (1.2–2.4), respectively). Depression did not mediate the relationship between sexual violence and condom use.

Conclusion:

Sexual violence and depression are prevalent and independently associated with condom nonuse with clients among FSWs in Cameroon. Results highlight the need for interventions to address mental health as well as gender-based violence for FSWs.

Keywords

HIV female sex work gender-based violence depression

Introduction

Gender-based violence (GBV), defined as ‘a harmful threat directed at any group or individual based on actual or perceived sex, gender identity and/or expression, sexual orientation, and or lack of adherence to varying socially constructed norms’, is pervasive globally (U.S. Agency for International Development (USAID), 2012). GBV includes physical and sexual violence, economic deprivation, threats, blackmail and psychological abuse (U.S. USAID & U.S. Department of State, 2016). Several decades of research on GBV have enabled an understanding of how GBV fuels health inequalities and perpetuates power dynamics that leave women marginalized (Shannon, Goldenberg, Deering, & Strathdee, 2014).

Disparate power dynamics leave certain groups of women at higher risk of GBV. Female sex workers (FSWs), in particular, report significantly higher rates of GBV than other women, for reasons including social norms, increased exposure to violent partners and lack of support services (Coetzee, Gray, & Jewkes, 2017). GBV is a symptom of gender inequities and is associated with heightened health risks and vulnerabilities among FSWs who report high rates of GBV by clients, partners and state actors (Deering et al., 2013; Lyons et al., 2017). In addition to being a human rights violation, experiences of GBV have been shown to increase risk and transmission of HIV (Lim et al., 2015). Despite declines in HIV rates in the countries most affected by the HIV pandemic, HIV remains concentrated among FSWs (Baral et al., 2012; Lancaster, Cernigliaro, Zulliger, & Fleming, 2016; MacAllister et al., 2015). A 2012 study determined that FSWs in low- and middle-income countries have a 13.5-times increased odds of HIV compared to all women of reproductive age (Baral et al., 2012). This increased HIV burden as both a cause and symptom of the inequalities inherent in the lives of FSWs and is perpetuated by individual, network and structural factors (Decker et al., 2015; Decker et al., 2013). Condom use has been shown to consistently impact rates of HIV prevalence and incidence among both FSW and non-FSW populations (Williams et al., 2014). While condom use may be a direct barrier to HIV transmission, the decision or ability to practice consistent condom use is understood as a decision made in context of larger structures, which often include economic necessity (Decker et al., 2015; Deering et al., 2013). Many sex workers enter the field due to financial vulnerability or lack of other economic opportunities, which can limit access to quality care (Lancaster et al., 2016). As sex workers, financial need limits agency in condom negotiation for FSWs, who may need to accept more money in exchange for condomless sex (Goldenberg, Duff, & Krusi, 2015). Coerced sex is often condomless, and sexual violence can create injuries and therefore increase biological susceptibility of sexually transmitted infections (STI) and HIV (Decker et al., 2013; Swain, Saggurti, Battala, Verma, & Jain, 2011). Client-perpetrated violence has been shown to significantly increase the odds of being offered and accepting condomless sex, showing a lasting impact beyond the incidence of violence (Deering et al., 2013). Finally, GBV can create mental health morbidities which can impact a woman’s ability to negotiate condom use, practice protective behaviors and access HIV health services (Devries et al., 2013; Hutton, Lyketsos, Zenilman, Thompson, & Erbelding, 2004; Jewkes, Dunkle, Nduna, & Shai, 2010).

Structurally, GBV risk is increased for FSWs through the marginalization caused by criminalization and punitive laws. Policies that criminalize the practice of sex work can encourage FSWs to practice in secrecy (Decker et al., 2015). This can impact the ability of FSWs to reach legal and medical services designed to support victims of violence and reach harm-reduction services (Shannon et al., 2015). Behavioral outcomes of GBV include associations with HIV risk behaviors including drug and alcohol use, high-risk sexual behavior and practicing condomless sex (Deering et al., 2014). A study among FSWs in Russia found violence to be significantly associated with HIV, with past-year experiences of client violence increasing the HIV risk over threefold (Decker et al., 2012). Partner violence compromises HIV care trajectories for FSWs, and a study from the Dominican Republic among FSWs living with HIV found that intimate partner violence was associated with decreased care seeking and retention for HIV (Mendoza et al., 2017). Specifically, depression has been shown to be associated with inconsistent condom use; having sex while using drugs or alcohol; reducing positive behavior change; limiting both cognitive and physical functioning; and potentially interfering with HIV prevention, treatment and health care (Hill, Maman, Kilonzo, & Kajula, 2017; Hutton et al., 2004). The pathways between GBV and depression, and depression and health behaviors may play a significant role among FSWs. This dynamic between depression and health behaviors, as it has been evaluated among FSWs, has shown some association between GBV, depression and condom use although merits more research (Patel, Ganju, Prabhakar, & Adhikary, 2016; Sherwood et al., 2015; Swain et al., 2011).

This relationship is particularly important within Cameroon, where HIV prevalence among FSWs is significantly elevated at 26%, compared to 4.5% among reproductive-aged adults (Decker et al., 2016). With high rates of GBV documented among FSWs in Cameroon, understanding how GBV affects overall health is central to addressing the country’s HIV epidemic (Decker et al., 2013). Mental illness and depression has been under-researched and evaluated for FSWs in Cameroon, potentially in light of the focus on HIV transmission and less on the context in which it occurs (Baral et al., 2012; Patel et al., 2016). While there is limited data, several previous studies have shown high rates of depression among adult Cameroonians (Kanmogne et al., 2017).

Within this body of work, depression and other mental health issues have received relatively limited attention. Greater clarity is needed as depression and mental health have long been associated with GBV and may potentially play a major role in impeding protective health behaviors. Mental health issues and depression must be understood in context of the root causes in which they exist and the comorbidities that impact condom use. The objective of this article is to evaluate the relationships between lifetime experiences of sexual violence, depression and inconsistent condom use with clients among FSWs in Cameroon.

Methods

Study population

In tandem with a large-scale, multi-dimensional programmatic intervention for HIV/AIDS among FSWs and men who have sex with men called Continuum of Prevention, Care and Treatment (CoPCT) of HIV/AIDS with Most-at-risk-Populations in Cameroon (CHAMP), an integrated behavioral and biological survey (IBBS) was conducted with FSWs at baseline. This cross-sectional study was conducted across five cities in Cameroon: Yaoundé, Douala, Bertoua, Kribi and Bamenda (Project Search: Research to Prevention, 2004, 2014).

Participants were recruited through respondent-driven sampling (RDS), which is a strategy for reaching stigmatized or hidden communities (Greiner et al., 2014; Heckathorn, 1997). Recruitment is done through coupons, which allow members of the population to recruit their communities into the study (Schonlau & Liebau, 2012). Women were successfully enrolled in the study if they had a coupon, reported practicing sex work within the past 3 months, were 18 years old or over, were capable of giving informed consent, lived in the study site/region, reported sex work as their principal source of income, had not participated in a similar study in the past 6 months and understood English or French.

Ethical approval was granted by Cameroon’s National Ethics Committee, CNERSH (Comite National d’Ethique sur la Recherce en Sante Humaine), study number 016/06/782/CE/CNERSH/SP. This analysis was submitted to the Johns Hopkins Bloomberg School of Public Health and was determined to be not research.

Procedures and methods

The survey was an interviewer-administered tablet-based structured questionnaire and was distributed in English, French and Pidgin. The questionnaires took around 60 minutes to administer and were conducted in private rooms within the study sites which were themselves strategically located to ensure confidentiality and comfort. Interviewers were trained in confidentiality procedures and on interview best practices. Respondents received a financial incentive upon completion of the study (2,000 XAF, ~$3 USD) and an additional incentive for each of their three referrals who completed the study (1,000 XAF per referral, ~$2 USD), though respondent’s answers or results did not impact if incentive was received. All women were administered an HIV test at site.

Measures

Condom use with clients was our main dependent variable, sexual violence experience was the main independent variable of interest and depression was the potential mediator of interest. Variables were chosen for analysis based on an evidence-based conceptual framework, which are considered to be associated with the main outcome of condom use for vaginal sex with clients among the literature.

Demographics

Demographics included study city, age of respondents, education levels, marital status, employment outside of sex work, experience living abroad in the past 10 years, having biological children, from an urban/rural community and living situation.

Risk determinants

The survey included questions about potential associated determinants including arrest history and drug/alcohol use. Alcohol use was defined as alcohol use in the past year, and drug use was defined as non-injectable illicit drug use, and illicit injectable drug use. Sex work–related characteristics included location of sex with clients (brothel, hotel/guest houses) and number of clients in the past month. The number of casual clients in the past month was dichotomized into over and under 20 clients to estimate the impact of high client loads. All measures were self-reported. Finally, HIV status was determined by blood samples through the lab within the IBBS. The tests were Alere Determine TM HIV – ½ and OraQuick HIV – ½.

Condom use

The main outcome, consistent condom use for vaginal sex with clients, was captured through questions on condom use with both regular and casual clients. Participants were first asked whether they had ever had regular clients or ever had casual clients. They were then asked about their lifetime experiences of condom use with each type of client. Inconsistent condom use was determined for this study as reporting anything other than ‘always’ using condoms with either type of client (MEASURE Evaluation, n.d.). For those who reported having regular and casual clients, consistent condom use was determined by reporting ‘always’ using a condom with both their regular and casual clients. If participants reported only having either casual or regular clients, consistent condom use was determined by ‘always’ using a condom for just the type of client they reported. No significant differences were found in condom use in vaginal sex with regular and casual clients for the covariates examined, so they were combined into one outcome indicator: condom use for vaginal sex with clients, dichotomized into outcomes of consistent or inconsistent use.

Violence

Sexual violence was defined as having someone ever force you to have sex when you did not want to (oral, anal, vaginal or with an object). Responses were ‘no’, ‘yes, in the last six months’, or ‘yes, but not in the last six months’. Responses were dichotomized into yes and no to determine proportions of ever experiencing violence, and GBV refers specifically to this type of sexual violence. For this article, we only looked at sexual violence.

Depression

Depression was assessed via the Patient Questionnaire 9 (PHQ9) (Cholera et al., 2014). Composite scores were developed and then categorized into levels of depression. Scores ranged from 0 to 26 and were classified to ensure maximum sensitivity into groups of no depression (0–4), mild depression (5–9), moderate depression (10–14), moderately severe depression (15–19) and severe depression (20–27) (Adewuya, Ola, & Afolabi, 2006; Pence et al., 2012).

Stigma

Stigma was categorized into three different scales to determine types of stigma experienced: anticipated, enacted (non-violent) and perceived. Anticipated stigma included afraid to go to health center, avoided going to health center and scared to be in public spaces. Enacted stigma included overheard health care workers gossip, faced verbal harassment or had been blackmailed. Perceived stigma was comprised of excluded from family, family made discriminatory remarks, rejected by friends, not treated well in a health center and police refused to protect, all due to status as a FSW (Stahlman et al., 2016).

Statistical analyses

Data from the five sites were combined into one total sample (N = 2,165). All analyses were conducted in Stata 14 (College Station, TX) (StataCorp LP, 2015). Table 1 reports demographic variables in crude numbers and proportions overall and by consistent condom use as the main outcome, with comparisons evaluated using chi-square p values. RDS-adjusted estimates were determined for demographic characteristic proportions as well. While crude numbers and proportions represent actual reported numbers in the study sample, RDS estimates attempt to determine the estimates of the covariates within the FSW population of Cameroon (Lyons et al., 2017). Crude estimates were used in the regressions in place of the RDS estimates, because there is no conclusion reached on how RDS should be handled within regressions (Schonlau & Liebau, 2012). Table 2 examined structural and sex work–related characteristics for the sample overall and in relation to condom use as the key outcome, again with chi-square analysis to evaluate comparisons.

Table 1.

Sociodemographic characteristics of female sex workers (N = 2,165).

Condom use, vaginal, any client
	Total		RDS adjusted proportion		Inconsistent condom use		Chi-square p value	RDS adjusted
	Total		RDS adjusted proportion		Inconsistent condom use		Chi-square p value	Proportion
	n	%	%	95% CI	n	%		%	95% CI
Number of participants	2,165	100.0			508	23.5
Demographics
Study city
Yaounde	566	26.1			116	20.5	.000
Douala	438	20.2			41	9.4
Bertoua	279	12.9			96	34.4
Bamenda	336	15.5			128	38.1
Kribi	546	25.2			127	23.4
Age, years
18–23	573	26.5	32.8	29.5–36.3	137	23.9	.649	23.4	18.3–29.5
24–30	702	32.4	32.0	28.8–35.4	171	24.2		20.4	16.2–25.2
>30	890	41.1	35.2	31.9–38.6	200	22.5		23.0	20.2–26.1
Education (n = 2,164)
None	50	2.3	1.9	1.3–2.8	12	24.0	.413	32.4	15.6–55.4
Some primary	286	13.2	12.3	10.2–14.8	70	24.5		25.9	18.6–34.8
Completed primary	334	15.4	12.8	10.6–15.3	91	27.3		24.6	17.9–32.7
Some secondary	1,307	60.4	64.7	61.3–68.0	287	22.0		21.5	17.9–25.6
Completed secondary	103	4.8	4.0	2.9–5.4	27	26.2		28.7	16.0–45.9
Technical training/University	84	3.9	4.3	3.1–6.0	21	25.0		23.7	12.0–41.2
Marital status (n = 2,164)
Never married	1,672	77.3	80.3	77.5–82.9	403	24.1	.248	23.7	20.4–27.3
Married/stable partner	231	10.7	10.0	8.2–12.1	53	23.0		19.6	13.3–27.9
Widowed	117	5.4	4.0	3.0–5.4	28	23.9		22.2	12.2–37.1
Separated/divorced/other	144	6.7	5.7	4.3–7.5	24	16.7		20.0	10.4–35.0
Employed/student outside of sex work
None/looking	1,183	54.6	47.3	43.8–50.9	280	23.7	.070	22.8	19.1–22.0
Training/student	174	8.0	11.5	9.2–14.1	31	17.8		20.4	12.4–31.7
Short-term contract/intern	93	4.3	2.8	1.9–4.1	31	33.3		31.4	15.9–52.8
Indefinite contract	28	1.3	1.2	0.7–2.1	8	28.6		32.7	12.4–62.5
Independent worker/other	687	31.7	37.2	33.8–40.8	158	23.0		23.2	18.4–28.7
Lived abroad in past 10 years
No	1,999	92.3	91.7	89.6–93.5	474	23.7	.345	23.4	20.23–26.7
Yes	166	7.7	8.3	6.6–10.5	34	20.5		19.2	12.4–28.5
Biological children
No	343	15.8	17.0	14.6–19.7	69	20.1	.111	19.8	14.3–26.7
Yes	1,822	84.2	83.0	80.3–85.4	439	24.1		23.7	20.5–27.2
From urban/rural community (n = 2,164)
Urban	1,635	75.6	77.2	74.0–80.0	356	21.8	.001	21.4	18.2–25.0
Rural	529	24.5	22.8	20.0–26.0	152	28.7		28.7	22.7–35.5
Living situation
Live alone	967	44.7	41.6	38.1–45.2	204	21.1	.017	19.2	15.4–23.8
Live with family	951	43.9	47.4	43.8–50.9	254	26.7		27.3	23.0–32.1
Live with roommates/friends/brothel	111	5.1	4.5	3.3–6.0	21	18.9		31.4	17.4–49.9
Other	136	6.3	6.6	4.8–8.9	29	21.3		10.5	4.8–21.2

RDS: respondent-driven sampling; CI: confidence interval.

Table 2.

Violence, HIV and sexual history characteristics among FSWs.

	Total			RDS adjusted proportion		Inconsistent condom use		Chi-square p value	RR	95% CI	aRR^a	95% CI
	n	n	%	%	95% CI	n	%
Violence
Physical violence (ever) (N = 2,161)
No	1,617	2,161	74.8	80.5	77.8	370	22.9	.273	REF		REF
Yes	544	2,161	25.2	19.5	17.1	137	25.2		1.1	0.9–1.3	1.1	0.9–1.3
Forced to have sex against will (ever) (N = 2,163)
No	1,450	2,163	67.0	70.5	67.2–73.6	297	20.5	.000	REF		REF
Yes	713	2,163	33.0	29.5	26.4–32.8	210	29.5		1.4***	1.2–1.7	1.4***	1.2–1.7
Justice
Ever arrested
No	1,236	2,165	57.1	66.1	62.8–69.2	340	27.5	.000	REF		REF
Yes	929	2,165	42.9	33.9	30.1–37.2	168	18.1		0.7***	0.6–0.8	0.7***	0.6–0.8
Alcohol in last year (N = 2,163)
No	582	2,163	26.9	26.1	23.1–29.4	114	19.6	.011	REF		REF
Yes	1,581	2,163	73.1	73.9	70.6–76.9	392	24.8		1.3*	1.1–1.5	1.3*	1.1–1.5
Non-injectable illicit drug use (N = 2,162)
No	1,957	2,162	90.5	92.8	91.1–94.3	453	23.2	.304	REF		REF
Yes	205	2,162	9.5	7.1	5.7–9.0	54	26.3		1.1	0.9–1.5	1.2	0.9–1.5
Illegal injection drug use (N = 2,163)
No	2,135	2,163	98.7	99.4	98.9–99.7	501	23.5	.849	REF		REF
Yes	28	2,163	1.3	0.6	0.4–1.1	7	25.0		1.1	0.6–2.0	1.1	0.5–2.0
Sexual risk
Casual clients in the last month (N = 2,064)
<20	957	2,064	46.4	57.6	54.2–61.1	284	29.7	.000	REF		REF
20+	1,107	2,064	53.6	42.4	39.0–45.8	198	17.9		0.6***	0.5–0.7	0.6***	0.5–0.7
Meet clients for sex in brothel
No	1,301	2,165	60.1	64.2	60.6–67.6	268	20.6	.000	REF		REF
Yes	864	2,165	39.9	34.9	32.4–39.4	240	27.8		1.3***	1.2–1.6	1.4***	1.2–1.6
Meet clients for sex in hotel/guest house
No	330	2,165	15.2	22.5	19.3–26.0	82	24.9	.519	REF		REF
Yes	1,835	2,165	84.8	77.5	74.0–80.7	426	23.2		0.9	0.8–1.4	0.9	0.8–1.1
Stigma
Perceived Stigma Index (N = 2,164)
0	1,629	2,164	75.3	82.7	80.3–84.9	369	22.7	.115	REF		REF
1+	535	2,164	24.7	17.3	15.1–19.7	139	26.0		1.1	1.0–1.4	1.1	0.9–1.3
Anticipated Stigma Index (N = 2,163)
0	1,979	2,163	91.5	93.8	92.4–95.0	447	22.6	.002	REF		REF
1+	184	2,163	8.5	6.2	5.0–7.6	60	33.0		1.4**	1.2–1.8	1.4**	1.1–1.8
Enacted (Non-Violent) Stigma Index (N = 2,163)
0	790	2,163	36.5	47.4	43.8–51.0	216	27.3	.000	REF		REF
1	537	2,163	24.8	19.0	16.3–21.9	133	24.8		0.9	0.8–1.0	0.9	0.8–1.1
2+	836	2,163	38.7	33.6	30.6–36.7	158	18.9		0.7***	0.6–0.8	0.7***	0.6–0.8
HIV status
HIV result (N = 2,158)
Negative	1,632	2,158	75.6	79.1	76.1–81.8	376	23.0	.380	REF		REF
Positive	526	2,158	24.4	20.9	18.2–23.9	131	24.9		1.1	0.9–1.3	1.0	0.9–1.2

FSW: female sex worker; RDS: respondent-driven sampling; RR: risk ratio; CI: confidence interval; aRR: adjusted risk ratio.

Adjusted for urban/rural and living situation.

p < .05. **p < .01. ***p < .00.

Poisson regressions with robust error variance were used for all regressions (Zou, 2004). Bivariate models were used to determine risk ratios (RRs) and 95% confidence intervals (CIs) of all characteristics. Multivariate Poisson regressions were used to determine adjusted risk ratios (aRRs) of the sex work characteristics, justice-related variables, stigma experiences and HIV status covariates included in Table 2 and were adjusted for the demographic variables that were significantly associated with the outcome (chi-square p values < .05).

The final models used were to conduct a cross-sectional mediation analysis (Baron & Kenny, 1986). Mediation analysis was done by first describing the proportions of inconsistent condom use with clients for each level of depression and the associated chi-square p value. A bivariate model was used to assess the crude association of depression with inconsistent condom use. Second, we explored proportions of depression levels among those exposed and those unexposed to sexual violence and associated chi-square p value to examine the relationship between the potential mediator (depression) and the independent variable of interest (sexual violence). Three final models were used to assess association of inconsistent condom use with clients, sexual violence and depression using aRRs, and evaluated if depression was mediating the effect of GBV on inconsistent condom use with clients, as evidenced by significant changes in the aRR. Depression was chosen as a mediation factor by its significance in the adjusted model, as well as the possibility of its role in impacting condom use. Model 1 examined the impact of sexual violence on inconsistent condom use without controlling for depression, model 2 included depression but not sexual violence in the model, and model 3 included both depression and reported sexual violence. All final models included demographic and risk characteristics that were found to be significantly associated with inconsistent condom use in chi-square analyses at p < .05. All the covariates included in the final model were assessed for collinearity (Table 3).

Table 3.

Distribution of potential mental health mediators over outcome (unprotected sex) and independent variable (sexual violence).

Potential mental health mediator	Inconsistent condom use						SV			Mediation models
Potential mental health mediator	n	n	%	Chi-square p value	RR	95% CI	Among SV unexposed, %	Among SV exposed, %	Chi-square p value		aRR (95% CI)^a	aRR (95% CI)^b
PHQ9 scores: No depression	224	1,074	21.0	.005	REF	REF	75.1	24.9	<.001	No depression	REF	REF
PHQ9: Mild depression	167	680	24.6		1.2	1.0–1.4	60.1	39.9		Mild depression	1.1(0.9–1.3)	1.0(0.9–1.2)
PHQ9: Moderate depression	56	209	26.8		1.3	1.0–1.7	61.7	38.3		Moderate depression	1.1(0.9–1.4)	1.1(0.8–1.4)
PHQ9: Moderately severe depression	29	116	25.0		1.2	0.9–1.7	55.2	44.8		Moderately severe depression	1.1(0.8–1.6)	1.1(0.8–1.5)
PHQ9: Severe depression	23	57	40.4		1.9***	1.4–2.7	43.9	56.1		Severe depression	1.8** (1.3–2.6)	1.7**(1.2–2.4)
−										Sexual violence	1.4*** (1.2–1.6)	1.4***(1.2–1.6)

PHQ: Patient Questionnaire; RR: risk ratio; CI: confidence interval; aRR: adjusted risk ratio; SV: sexual violence.

Inconsistent condom use as a function of independent variable noted in row, adjusted for urban/rural, living situation, arrest history, alcohol in last year, high number of casual clients in last month, brothel use, anticipated stigma and enacted stigma demographics.

Inconsistent condom use as a function of mental health and sexual violence, adjusted for urban/rural, living situation, arrest history, alcohol in last year, high number of casual clients in last month, brothel use, anticipated stigma and enacted stigma demographic.

**p < .01. ***p < .00.

Results

Demographics

A total of 2,165 FSWs were included in the study sample. The proportion of individuals were dispersed relatively evenly among the cities, with the highest coming from Yaoundé (26.1%, 566/2,165) and the lowest percentage coming from Bertoua (12.9%, 279/2,165). The mean age of the participants was 30.1 (inter-quartile range: 23–36). The majority of participants had never been married (77.3%, 1,672/2,165). Over half of the population was unemployed or looking for work outside of sex work (54.6%, 1,183/2,165). Reported living situation was significantly associated with inconsistent condom use with clients (p < .05). The majority of the sample was split between living alone (44.7%, 967/2,165) and with family (44.0%, 951/2,165), with the rest living with roommates/friends/brothel (5.1%, 111/2,165) or other (6.3%, 136/2,165).

Risk characteristics: violence, justice, sexual risk, stigma and HIV outcomes

Overall, 23.5% (508/2,165) of respondents reported inconsistent condom use for vaginal sex with clients. Of the 33.0% (713/2,163) of FSWs who reported ever facing sexual violence, 29.5% (210/713) had practiced inconsistent condom use with clients (RR 1.4, 95% CI (1.2–1.7); aRR 1.3, (1.1–1.5)). Table 2 presented proportions, RDS adjustments and regressions unadjusted and adjusted for the significant demographic variables (urban/rural divide and living situation).

Arrest was found to be associated with inconsistent condom use. Just under half of the respondents reported being ever arrested (42.9%, 929/2,165), 18.1% of which reported inconsistent condom use with clients. Experience of arrest remained significantly associated with inconsistent condom use even when adjusted for significant demographic variables (RR 0.7, 95% CI (0.6–0.8); aRR 0.7, 95% CI (0.6–0.8)). Of the 2,163 respondents who responded regarding their alcohol use, the majority reported drinking alcohol in the past year (73.1%, 582/2,163). A higher percentage of those who had used alcohol in the past year reported inconsistent condom use with clients (24.8%, 392/1,582) than non-users (19.6%, 114/582), and alcohol use in the past year remained significantly associated with condom use in the adjusted model (RR 1.3, 95% CI (1.1–1.5); aRR 1.3, 95% CI (1.1–1.5)).

In total, 53.6% (1,107/2,064) of respondents reported having over 20 paying casual clients in the past month, which was significantly associated with inconsistent condom use. Inconsistent condom use was prevalent among 29.7% (284/957) of FSWs who had fewer than 20 casual clients in the past month and was significant in both the unadjusted and adjusted models (RR 0.6, 95% CI (0.5–0.7); aRR 0.6, 95% CI (0.5–0.7)). When considering location of meeting with clients, 39.9% (864/2,165) of FSWs reported meeting their clients in brothels. Among FSWs who used brothels, inconsistent condom use was significantly higher than those who used other locations: 27.8% (240/864) to 20.6% (268/1,301), respectively (RR 1.3, 95% CI (1.2–1.6); aRR 1.4, 95% CI (1.2–1.6)).

Anticipated stigma was significantly associated with inconsistent condom use (RR 1.4, 95% CI (1.2–1.8); aRR 1.4, 95% CI (1.1–1.8)), as was the higher level of enacted stigma (RR 0.7, 95% CI (0.6–0.8); aRR 0.7, 95% CI (0.6–0.8)).

Mediation model

Any level of depression was present among 49.8% (n = 1,067) of respondents and was significantly associated with inconsistent condom use (p < .01). Among those with any depression, those with severe depression showed almost twice the risk of inconsistent condom use compared to those without depression (RR 1.9, 95% CI (1.4–2.7)). With increasing depression level, the proportion of those who had experienced sexual violence increased; among respondents with no depression, 24.9% (267/1,071) reported exposure to sexual violence versus 56.1% (32/57) of respondents with severe depression who reported experiencing sexual violence (p < .01).

In the final multivariate mediation models, the RR of inconsistent condom use among FSWs who reported sexual violence did not change when depression was not included in model 1 (aRR 1.4, 95% CI (1.2–1.6)) compared to when it was included in model 3 (aRR 1.4, 95% CI (1.2–1.6)). Severe depression was significantly associated with inconsistent condom use in model 2, which accounted for significant demographic and risk characteristics (aRR 1.8, 95% CI (1.3–2.6)). In model 3, sexual violence and depression were both significantly associated with inconsistent condom use, which suggests independent impact rather than mediation (sexual violence: aRR 1.4, 95% CI (1.2–1.6); severe depression: aRR 1.7, 95% CI (1.2–2.4)).

Discussion

Women face power imbalances during condom negotiation, and structural determinants reinforce these gendered power disparities. This dynamic is profound for FSWs, who negotiate condom use repeatedly in the context of high rates of violence and depression (Decker et al., 2012). As seen in this study, GBV and depression significantly impacted condom use with clients in Cameroon, both independently and in conjunction. Condom use by FSWs is not a behavior that occurs in isolation but in the context of high rates of comorbidities and underlying inequalities. In these results, sexual violence and depression were independent, significant cross-sectional predictors of inconsistent condom use with clients. While sexual violence was associated with depression, depression did not appear to mediate the relationship between sexual violence and condom use. Results illustrate the need to understand the larger contextual factors which may be fueling GBV and depression among FSWs, as both impact inconsistent condom use. Through decreased use of consistent condom use, GBV and depression may ultimately increase vulnerability for HIV transmission and create an environment of risk for FSWs.

FSWs are members of a socially and economically marginalized group with limited access to comprehensive care. FSWs included here reported low levels of education and high rates of unemployment. In part due to such underlying inequalities, the work environment for FSWs offers particular risk of violence, which may fuel depression; the intersection of isolated work locations, punitive policing practices and abusive clients suggests interventions that address structural issues would enable a more effective HIV prevention approach (Lyons et al., 2017). The social, economic, work and political contexts such as poverty and gender inequality may play a larger role in determining condom use among FSWs, and research is needed to further understand depression and violence in this context (Decker et al., 2015; Deering et al., 2014; Papworth et al., 2014). Addressing violence and depression for FSWs requires addressing the power dynamics and larger structures which create and fuel health disparities for women who sell sex (Sagtani et al., 2013).

Depression was significantly associated with inconsistent condom use throughout all models, indicating the impact of mental health on protective behaviors among FSWs. A 2015 study in Cameroon among HIV-positive patients with depressive symptoms showed decreased PHQ9 scores and increased overall health following 4 months of antidepressant treatment (Gaynes et al., 2015). The study also found that self-efficacy scores increased in the same time period. However, self-efficacy – impacted by GBV and depression – may not be enough to create consistent condom use among FSWs (Jones et al., 2016; Shannon et al., 2009). For FSWs, further research is needed to understand how and why depression and GBV decrease condom use, and how established mental health interventions can be adapted to address the context of their lives. GBV has been studied as a power relationship and representative of gender inequality, often fueled by the types of poverty and inequality which may be the same as those which cause a woman to enter sex work (Papworth et al., 2014; Shannon et al., 2014; Tsai, 2013). Research is beginning to discuss depression in this light as well, and the rates of depression seen in this study make clear the need to understand depression and mental illness as a symptom of larger health disparities. This study represents the impact on FSWs of multiple burdens of GBV, gender inequality and depression, each exaggerated by poverty and lack of economic access seen among these respondents (Beattie et al., 2010; Goldenberg et al., 2015).

The elevated levels of sexual violence suggest that interventions focusing on supporting FSWs should address structural drivers of violence. With health equity paradigms encouraging health systems to think about ‘demand-side’ interventions, documenting the prevalence of depression, violence and their impact on condom use for marginalized populations has potential to make real change (Ostlin et al., 2011). This study attempts to center the inter-connected demands of the FSWs’ complexities to inform interventions that are driven by the users of the health systems (Ostlin et al., 2011).

Limitations

There are several limitations with this study. This was cross-sectional, limiting the ability of the study to determine temporal ordering, causation or directionality of events. Furthermore, there is potential of RDS to create bias in the sample by not reaching FSWs who have limited social networks and may as a result be at even increased risk for marginalization not represented within this study (Lyons et al., 2017). Moreover, the primary sexual violence indicator, forced sex, may have missed sexual experiences that were pressured or coerced (Patel et al., 2016). The sensitive and stigmatized nature of both sexual violence and mental health may have caused some respondents to self-censure, which would result in missed cases, despite the research team training to ensure accurate and comfortable reporting. There is potential for unmeasured or missed confounding, or factors which may have impacted the final indicator but were not included in the analysis.

Conclusion

Inconsistent condom use with clients remains highly prevalent among FSWs in Cameroon. Violence and depression faced by the FSW population impact condom use, yet remain under-researched among FSW populations. Understanding the factors that drive inconsistent condom use with clients is pivotal to reducing HIV incidence among FSWs, improving their overall health and ensuring their ability to seek care. While there is value in increased study of research integrating social inequities, power asymmetrics and social determinants of health, these data already provide guidance on the importance for programs to address underlying inequities driving sexual practices and health outcomes (Ostlin et al., 2011). Mental illness and depression remain poorly researched among FSWs in light of the intersecting epidemics of GBV and HIV. While this study focused on indicators specifically around depression, the overall research indicates a huge need to further understand mental health as a factor in behaviors. Mental health potentially impacts every level of HIV care and must be better considered as a factor impacting the health of FSWs. Taken together, these data suggest that research and programming which focuses on violence and harm reduction for FSWs, and addresses disparities beyond the individual level, have the potential to address health disparities for this highly marginalized population.

Footnotes

Acknowledgements

The authors are grateful to all who supported this work. This study was conducted through Continuum of Prevention, Care and Treatment of HIV/AIDS with Most-at-risk Populations in Cameroon (CHAMP). The program is implemented by CARE International in Cameroon in collaboration with Johns Hopkins University (JHU), Global Viral (GV) and eight local community-based organizations (MotoAction, Alternatives Cameroun, CMWA, Horizons Femmes, Humanity First, ASAD, CAMNAFAW and SWAA Littoral) and in coordination with the National AIDS Control Committee and the Ministry of Public Health.

Conflict of interest

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: This study was made possible by the generous support of the American people through the U.S. Agency for International Development (USAID) and the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR). The contents are the responsibility of Johns Hopkins University and do not necessarily reflect the views of USAID, PEPFAR or the U.S. Government. CHAMP, a 5-year cooperative agreement (AID-624-A-14-00003), is a large-scale HIV prevention, care and treatment program for key populations in Cameroon. CHAMP is led by CARE Cameroon in partnership with Johns Hopkins School of Public Health, Global Viral and MotoAction. S.B.’s effort was funded in part from the Johns Hopkins University Center for AIDS Research, an NIH-funded program (P30AI094189), which is supported by the following NIH Co-Funding and Participating Institutes and Centers: NIAID, NCI, NICHD, NHLBI, NIDA, NIMH, NIA, FIC, NIGMS, NIDDK and OAR. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or AHRQ.

ORCID iD

A Abelson

References

Adewuya

A. O.

Ola

B. A.

Afolabi

O. O.

(2006). Validity of the patient health questionnaire (PHQ-9) as a screening tool for depression amongst Nigerian university students. Journal of Affective Disorders, 96, 89–93. doi:10.1016/j.jad.2006.05.021

Baral

Beyrer

Muessig

Poteat

Wirtz

A. L.

Decker

M. R.

… Kerrigan

(2012). Burden of HIV among female sex workers in low-income and middle-income countries: A systematic review and meta-analysis. The Lancet, 12, 538–549. doi:10.1016/s1473-3099(12)70066-x

Baron

R. M.

Kenny

D. A.

(1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182.

Beattie

T. S.

Bhattacharjee

Ramesh

B. M.

Gurnani

Anthony

Isac

… Moses

(2010). Violence against female sex workers in Karnataka state, south India: Impact on health, and reductions in violence following an intervention program. BMC Public Health, 10, 476. doi:10.1186/1471-2458-10-476

Cholera

Gaynes

B. N.

Pence

B. W.

Bassett

Qangule

Macphail

… Miller

W. C.

(2014). Validity of the Patient Health Questionnaire-9 to screen for depression in a high-HIV burden primary healthcare clinic in Johannesburg, South Africa. Journal of Affective Disorders, 167, 160–166. doi:10.1016/j.jad.2014.06.003

Coetzee

Gray

G. E.

Jewkes

(2017). Prevalence and patterns of victimization and polyvictimization among female sex workers in Soweto, a South African township: A cross-sectional, respondent-driven sampling study. Global Health Action, 10, Article 1403815. doi:10.1080/16549716.2017.1403815

Decker

M. R.

Crago

A.-L.

Chu

S. K. H.

Sherman

S. G.

Seshu

M. S.

Buthelezi

… Beyrer

(2015). Human rights violations against sex workers: Burden and effect on HIV. The Lancet, 385, 186–199. doi:10.1016/s0140-6736(14)60800-x

Decker

M. R.

Lyons

Billong

S. C.

Njindam

I. M.

Grosso

Nunez

G. T.

… Baral

(2016). Gender-based violence against female sex workers in Cameroon: Prevalence and associations with sexual HIV risk and access to health services and justice. Sexually Transmitted Infections, 92, 599–604. doi:10.1136/sextrans-2015-052463

Decker

M. R.

Wirtz

A. L.

Baral

S. D.

Peryshkina

Mogilnyi

Weber

R. A.

… Beyrer

(2012). Injection drug use, sexual risk, violence and STI/HIV among Moscow female sex workers. Sexually Transmitted Infections, 88, 278–283. doi:10.1136/sextrans-2011-050171

10.

Decker

M. R.

Wirtz

A. L.

Pretorius

Sherman

S. G.

Sweat

M. D.

Baral

S. D.

… Kerrigan

D. L.

(2013). Estimating the impact of reducing violence against female sex workers on HIV epidemics in Kenya and Ukraine: A policy modeling exercise. American Journal of Reproductive Immunology, 69 (Suppl. 1), 122–132. doi:10.1111/aji.12063

11.

Deering

K. N.

Amin

Shoveller

Nesbitt

Garcia-Moreno

Duff

… Shannon

(2014). A systematic review of the correlates of violence against sex workers. American Journal of Public Health, 104, 42–54.

12.

Deering

K. N.

Lyons

Feng

C. X.

Nosyk

Strathdee

S. A.

Montaner

J. S.

Shannon

(2013). Client demands for unsafe sex: The socioeconomic risk environment for HIV among street and off-street sex workers. Journal of Acquired Immune Deficiency Syndromes, 63, 522–531. doi:10.1097/QAI.0b013e3182968d39

13.

Devries

K. M.

Mak

J. Y.

Bacchus

L. J.

Child

J. C.

Falder

Petzold

… Watts

C. H.

(2013). Intimate partner violence and incident depressive symptoms and suicide attempts: A systematic review of longitudinal studies. PLoS Medicine, 10(5), e1001439. doi:10.1371/journal.pmed.1001439

14.

Gaynes

B. N.

Pence

B. W.

Atashili

O’Donnell

J. K.

Njamnshi

A. K.

Tabenyang

M. E.

… Ndumbe

(2015). Changes in HIV outcomes following depression care in a resource-limited setting: Results from a pilot study in Bamenda, Cameroon. PLoS ONE, 10(10), e0140001. doi:10.1371/journal.pone.0140001

15.

Goldenberg

S. M.

Duff

Krusi

(2015). Work environments and HIV prevention: A qualitative review and meta-synthesis of sex worker narratives. BMC Public Health, 15, Article 1241. doi:10.1186/s12889-015-2491-x

16.

Greiner

A. L.

Albutt

Rouhani

S. A.

Scott

Dombrowski

VanRooyen

M. J.

Bartels

S. A.

(2014). Respondent-driven sampling to assess outcomes of sexual violence: A methodological assessment. American Journal of Epidemiology, 180, 536–544. doi:10.1093/aje/kwu149

17.

Heckathorn

(1997). Respondent-driven sampling: A new approach to the study of hidden populations. Social Problems, 44, 174–199.

18.

Hill

L. M.

Maman

Kilonzo

M. N.

Kajula

L. J.

(2017). Anxiety and depression strongly associated with sexual risk behaviors among networks of young men in Dar es Salaam, Tanzania. AIDS Care, 29, 252–258. doi:10.1080/09540121.2016.1210075

19.

Hutton

Lyketsos

Zenilman

Thompson

Erbelding

(2004). Depression and HIV risk behaviors among patients in a sexually transmitted disease clinic. American Journal of Psychiatry, 161, 912–914.

20.

Jewkes

R. K.

Dunkle

Nduna

Shai

(2010). Intimate partner violence, relationship power inequity, and incidence of HIV infection in young women in South Africa: A cohort study. The Lancet, 376, 41–48.

21.

Jones

K. A.

Cornelius

M. D.

Silverman

J. G.

Tancredi

D. J.

Decker

M. R.

Haggerty

C. L.

… Miller

(2016). Abusive experiences and young women’s sexual health outcomes: Is condom negotiation self-efficacy a mediator? Perspectives on Sexual and Reproductive Health, 48, 57–64. doi:10.1363/48e8616

22.

Kanmogne

G. D.

Qiu

Ntone

F. E.

Fonsah

J. Y.

Njamnshi

D. M.

Kuate

C. T.

… Njamnshi

A. K.

(2017). Depressive symptoms in HIV-infected and seronegative control subjects in Cameroon: Effect of age, education and gender. PLoS ONE, 12(2), e0171956. doi:10.1371/journal.pone.0171956

23.

Lancaster

K. E.

Cernigliaro

Zulliger

Fleming

P. F.

(2016). HIV care and treatment experiences among female sex workers living with HIV in sub-Saharan Africa: A systematic review. African Journal of AIDS Research, 15, 377–386. doi:10.2989/16085906.2016.1255652

24.

Lim

Peitzmeier

Cange

Papworth

LeBreton

Tamoufe

… Baral

(2015). Violence against female sex workers in Cameroon: Accounts of violence, harm reduction, and potential solutions. Journal of Acquired Immune Deficiency Syndromes, 68, S241–S247. doi:10.1097/qai.0000000000000440

25.

Lyons

Ketende

Drame

Grosso

Diouf

… Baral

(2017). Physical and sexual violence affecting female sex workers in Abidjan, Cote d’Ivoire: Prevalence, and the relationship between violence, the work environment, HIV and access to health services. Journal of Acquired Immune Deficiency Syndromes, 75, 9–17. doi:10.1097/QAI.0000000000001310

26.

MacAllister

Sherwood

Galjour

Robbins

Zhao

Dam

… Baral

(2015). A comprehensive review of available epidemiologic and HIV service data for female sex workers, men who have sex with men, and people who inject drugs in select West and Central African Countries. Journal of Acquired Immune Deficiency Syndromes, 68(Suppl. 2), S83–S90.

27.

MEASURE Evaluation. (n.d.). Consistency of condom use indicator. Retrieved from https://www.measureevaluation.org/prh/rh_indicators/specific/stis-hiv-aids/consistency-of-condom-use

28.

Mendoza

Barrington

Donastorg

Perez

Fleming

P. J.

Decker

M. R.

Kerrigan

(2017). Violence from a sexual partner is significantly associated with poor HIV care and treatment outcomes among female sex workers in the Dominican Republic. Journal of Acquired Immune Deficiency Syndromes, 74, 273–278. doi:10.1097/QAI.0000000000001250

29.

Ostlin

Schrecker

Sadana

Bonnefoy

Gilson

Hertzman

… Vaghri

(2011). Priorities for research on equity and health: Towards an equity-focused health research agenda. PLoS Medicine, 8(11), e1001115.

30.

Papworth

Grosso

Ketende

Wirtz

A. L.

Cange

Kennedy

Baral

(2014). Examining risk factors for HIV and access to services among female sex workers (FSW) and men who have sex with men (MSM) in Burkina Faso, Togo, and Cameroon. Retrieved from https://www.usaid.gov/documents/1860/research-brief-examining-risk-factors-hiv-and-access-services-among-female-sex-2

31.

Patel

S. K.

Ganju

Prabhakar

Adhikary

(2016). Relationship between mobility, violence and major depression among female sex workers: A cross-sectional study in southern India. BMJ Open, 6(9), e011439. doi:10.1136/bmjopen-2016–011439

32.

Pence

B. W.

Gaynes

B. N.

Atashili

O’Donnell

J. K.

Tayong

Kats

… Ndumbe

P. M.

(2012). Validity of an interviewer-administered patient health questionnaire-9 to screen for depression in HIV-infected patients in Cameroon. Journal of Affective Disorders, 143, 208–213. doi:10.1016/j.jad.2012.05.056

33.

Project Search: Research to Prevention. (2004). Examining risk factors for HIV and access to services among female sex workers and men who have sex with men in Cameroon. Washington, DC: U.S. Agency for International Development.

34.

Project Search: Research to Prevention. (2014). Research brief: Examining risk factors for HIV and access to services among female sex workers and men who have sex with men in Cameroon. Retrieved from https://www.jhsph.edu/research/centers-and-institutes/research-to-prevention/publications/cameroon-brief.pdf

35.

Sagtani

R. A.

Bhattarai

Adhikari

B. R.

Baral

Yadav

D. K.

Pokharel

P. K.

(2013). Violence, HIV risk behaviour and depression among female sex workers of eastern Nepal. BMJ Open, 3(6). doi:10.1136/bmjopen-2013-002763

36.

Schonlau

Liebau

(2012). Respondent-driven sampling. The Stata Journal, 12, 72–93.

37.

Shannon

Goldenberg

S. M.

Deering

K. N.

Strathdee

S. A.

(2014). HIV infection among female sex workers in concentrated and high prevalence epidemics: Why a structural determinants framework is needed. Current opinion in HIV and AIDS, 9, 174–182. doi:10.1097/COH.0000000000000042

38.

Shannon

Strathdee

S. A.

Goldenberg

S. M.

Duff

Mwangi

Rusakova

… Boily

M.-C.

(2015). Global epidemiology of HIV among female sex workers: Influence of structural determinants. The Lancet, 385, 55–71. doi:10.1016/S0140-6736(14)60931-4

39.

Shannon

Strathdee

S. A.

Shoveller

Rusch

Kerr

Tyndall

(2009). Structural and environmental barriers to condom use negotiation with clients among female sex workers: Implications for HIV-prevention strategies and policy. American Journal of Public Health, 99, 659–665. doi:10.2105/AJPH.2007.129858

40.

Sherwood

J. A.

Grosso

Decker

M. R.

Peitzmeier

Papworth

Diouf

… Baral

(2015). Sexual violence against female sex workers in the Gambia: A cross-sectional examination of the associations between victimization and reproductive, sexual and mental health. BMC Public Health, 15, Article 270. doi:10.1186/s12889-015-1583-y

41.

Stahlman

Sanchez

T. H.

Sullivan

P. S.

Ketende

Lyons

Charurat

M. E.

… Baral

S. D.

(2016). The prevalence of sexual behavior stigma affecting gay men and other men who have sex with men across Sub-Saharan Africa and in the United States. JMIR Public Health and Surveillance, 2(2), e35. doi:10.2196/publichealth.5824

42.

StataCorp LP. (2015). Stata statistical software (Version release 14). College Station, TX: Author.

43.

Swain

S. N.

Saggurti

Battala

Verma

R. K.

Jain

A. K.

(2011). Experience of violence and adverse reproductive health outcomes, HIV risks among mobile female sex workers in India. BMC Public Health, 11, Article 357. doi:10.1186/1471-2458-11-357

44.

Tsai

A. C.

(2013). Intimate partner violence and population mental health: Why poverty and gender inequities matter. PLoS Medicine, 10(5), e1001440. doi:10.1371/journal.pmed.1001440

45.

U.S. Agency for International Development. (2012). Strategy to prevent and respond to gender-based violence globally: Policy brief. Washington, DC: Author.

46.

U.S. Agency for International Development & U. S. Department of State. (2016). United States Strategy to Prevent and Respond to Gender-Based Violence Globally: 2016 Update. Washington, DC: Author.

47.

Williams

J. R.

Alary

Lowndes

C. M.

Béhanzin

Labbé

A.-C.

Anagonou

… Boily

M.-C.

(2014). Positive impact of increases in condom use among female sex workers and clients in a medium HIV prevalence epidemic: Modelling results from project SIDA1/2/3 in Cotonou, Benin. PLoS ONE, 9(7), e102643. doi:10.1371/journal.pone.0102643

48.

Zou

(2004). A modified Poisson regression approach to prospective studies with binary data. American Journal of Epidemiology, 159, 702–706. doi:10.1093/aje/kwh090