High prevalence of STIs among men engaged in transactional sex and alcohol use in western Kenya: Important implications for STI prevention interventions

Abstract

Objectives:

Better data on etiologic prevalence of sexually transmitted infections (STIs) among African men could greatly strengthen STI prevention efforts and convey benefits to women as well. In an ongoing study among men in Kenya, we analyzed baseline STI prevalence and individual characteristics associated with STI infection.

Methods:

In Siaya County, Kenya, we recruited men aged 18–39 years who self-reported engagement in transactional sex and alcohol use. We administered a baseline questionnaire to participants and conducted testing for HIV, HSV-2, C. trachomatis (CT) and N. gonorrhoeae (NG) infection. Characteristics associated with bacterial STIs were analyzed using logistic regression, and we estimated the positive and negative predictive values (PPV/NPV) of syndromic management of these infections.

Results:

We enrolled 1,500 participants from July 2022-March 2023. Participant mean age was 27.9 years, 62.2% were married/cohabitating, and 53.5% were heavy alcohol users (AUDIT-C≥4). Participants reported a mean of 4.2 sexual partners and 3.5 transactional sex partners in the past 3 months. HIV prevalence was 9.5%, HSV-2 was 38.7%, CT was 14.3% and NG was 2.5%. Combined CT and/or NG infection was detected in 16.1% of participants.

Compared to participants uninfected, those testing positive for CT and/or NG were younger (p=0.001), had more sexual partners (p=0.027) and transactional sex partners (p=0.039), were less likely to have used a condom at last sex (p=0.015), and were more likely to self-report having an STI besides HIV in the past 12 months (p=0.002). The PPV and NPV for currently experiencing CT and/or NG symptoms was 33.3% and 84.4%, respectively.

Conclusions:

Among Kenyan men engaged in transactional sex and alcohol use, STI prevalence was high. These data fill an important gap about STI prevalence and risk factors in African men, highlighting the risk of ongoing transmission, and the need for targeted prevention programs and expanded access to testing and treatment.

BACKGROUND

The high global burden of sexually transmitted infections (STIs) is a significant public health concern,(1) and an estimated 75–85% of annual global STI cases occur in low- and middle-income countries (LMICs), including sub-Saharan Africa (SSA).(2) Timely diagnosis and management of STIs prevents continued transmission and adverse sequelae. Laboratory testing is the most accurate method of STI diagnosis but is resource intensive, with limited availability in LMICs. Therefore, syndromic management remains the main strategy in LMICs to diagnose and manage STIs in spite of low diagnostic accuracy,(3, 4) under-diagnosis of asymptomatic infections, and unnecessary treatment of non-specific symptoms contributing to antimicrobial resistance.(5)

The management of STIs is particularly inadequate among men, who have lower clinic attendance than women and commonly have asymptomatic infections. Laboratory-based STI data among men in SSA are limited and existing data are largely among symptomatic men or those living with HIV, rather than from the general population or in other higher risk groups.

In an ongoing study among men in western Kenya (NCT04013295), we analyzed baseline STI prevalence and characteristics associated with STI infection. Current data on etiologic STI prevalence and risk factors among men in Kenya would help prioritize STI resources for Kenyan men, and men in Africa more broadly, and would also benefit women by identifying opportunities to reduce STI transmission.

METHODS

The parent study, Akiba, is a randomized trial of a savings intervention designed to motivate men to reduce their spending on risky behaviors including transactional sex and alcohol use, and save more of their income for the future.(6) Outcomes include STI incidence over 24 months, savings, and expenditures. Questionnaire and STI diagnostic data gathered at baseline were used for analyses presented here.

Recruitment occurred July 2022-March 2023 in Siaya County, Kenya, from beach landing sites, transportation hubs, and other locations where men with steady incomes congregate. Study staff provided general information about the study to men and referred them to nearby study offices to be screened for eligibility. To be eligible men needed to be 18–39 years of age, report a steady source of income, have used alcohol or drugs in the past month, and have engaged in transactional sex in the past 3 months. They also needed to reside in the study community, own a mobile phone, and have a national identification card (required for opening a bank account).

An interviewer-administered questionnaire was completed at baseline to collect demographic, socio-economic and behavioral data. We assessed alcohol use using the Alcohol Use Disorders Identification Test (AUDIT-C) scale,(7) and screened for depression using the Patient Health Questionnaire-2 (PHQ-2).(8) A whole blood sample was collected for human immunodeficiency virus (HIV) and herpes simplex virus type 2 (HSV-2) testing. HIV testing was conducted per the Kenya national testing algorithm using a single rapid test (Determine HIV-1/2 Ag/Ab Combo, Abbott Diagnostic). Reactive or indeterminate results were confirmed using a second rapid test (First Response HIV 1–2.0, Premier Medical Corporation). HSV-2 testing was conducted using the Kalon HSV-2 IgG ELISA (Kalon Biologicals Ltd.) test. A first void urine sample was provided for Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (NG) testing using the Cepheid GeneXpert CT/NG PCR test (Cepheid Inc.). Participants who tested HIV-positive were referred to the nearest health facility, where free treatment is available. Participants positive for CT or NG infection were contacted by study staff and given free treatment based on local protocols.(9)

Statistical analyses were conducted using Stata 17 (StataCorp). Participant characteristics were analyzed descriptively. Infection with CT and NG were collapsed into a single variable. Characteristics associated with bacterial STIs were analyzed using univariable logistic regression. Characteristics with a p-value≤0.1 were then analyzed in multivariate logistic regression, controlling for age, education, and marital status as these variables have been shown in prior literature to be potentially confounding.(10–12) We also estimated the positive and negative predictive value (PPV and NPV) of syndromic management of CT and NG, i.e. the proportions of positive and negative per syndromic diagnosis that are true positive and true negative per laboratory results.

Ethics

Informed consent was obtained from all participants prior to initiating study procedures. The study was approved by the University of Pennsylvania Institutional Review Board (849897) and the Maseno University Ethical Review Committee (MSU/DRPI/MUSERC/01022/21).

RESULTS

A total of 2214 men were assessed for eligibility and 1,500 participants were enrolled. Of the 714 who were not enrolled, 29% did not meet inclusion criteria and 3% declined participation. Among participants, mean age was 27.9 years (range 18–39 years), 62.2% were married and/or cohabitating, and 60.7% had some secondary education (Table 1). Half of participants (50.7%) worked in the transportation sector. Hazardous drinking (AUDIT-C ≥4) was reported by 53.5% of participants. Participants reported a mean of 4.2 sexual partners and 3.5 transactional sex partners in the past 3 months.

Table 1.

Summary of participant characteristics and their association with C. trachomatis and/or N. gonorrhea infection

		C. trachomatis and/or N. gonorrhea
				Univariatea		Multivariateb
Predictor	Overall (N = 1,500)	CT and/or NG positive (N = 241)	CT and NG negative (N = 1,259)	Odds Ratio (95% CI)	p-value	Odds Ratio (95% CI)	p-value
Sociodemographic
Age, mean (range)	27.9 (18–39)	26.9 (19–39)	28.1 (18–39)	0.95 (0.93, 0.98)	0.001	0.95 (0.92, 0.98)	0.002
Married/co-habitatingc	933 (62.2)	141 (58.5)	792 (62.9)	0.83 (0.63, 1.10)	0.20
Education
≤Primary	590 (39.3)	93 (38.6)	497 (39.5)	ref	0.80
≥Some Secondary	910 (60.7)	148 (61.4)	762 (60.5)	1.04 (0.78, 1.38)
Employment
Fishing and related activities	115 (7.7)	22 (9.1)	93 (7.4)	ref	0.72
Transport workers	760 (50.7)	126 (52.3)	634 (50.4)	0.84 (0.51, 1.39)
Informal sector	454 (30.3)	70 (29.1)	384 (30.5)	0.77 (0.45, 1.31)
Traders/shopworkers	153 (10.2)	20 (8.3)	133 (10.6)	0.64 (0.33, 1.23)
Other	18 (1.2)	3 (1.2)	15 (1.2)	0.85 (0.22, 3.18)
Income (in Kenyan Shillings), mean (SD)d	3,875.2 (3,597.1)	4,020.0 (2,669.4)	3,847.5 (3,748.8)	1.01 (0.98, 1.05)	0.50
Depression (PHQ-2 ≥3)	576 (38.4)	94 (39.0)	482 (38.3)	1.03 (0.78, 1.37)	0.83
Behavioral
AUDIT-C ≥4 (hazardous drinking)	803 (53.5)	129 (53.5)	674 (53.5)	1.00 (0.76, 1.32)	1.00
Used tobacco/cigarettes, bhangi, cocaine, heroin, or another drug in the past 6 months	830 (55.3)	128 (53.1)	702 (55.8)	0.90 (0.68, 1.18)	0.44
Gambled in the past 6 months	928 (61.9)	156 (64.7)	772 (61.3)	1.16 (0.87, 1.54)	0.32
Number of sexual partners in the 3 months, mean (SD)	4.2 (2.3)	4.5 (2.9)	4.1 (2.2)	1.06 (1.01, 1.12)	0.027	1.05 (1.00, 1.11)	0.05
Used condom at last sexual encounter in the past 3 months	742 (49.5)	102 (42.3)	640 (50.9)	0.71 (0.53, 0.94)	0.015	0.69 (0.52, 0.91)	0.009
Number of transactional sex encounters in the past 3 months, mean (SD)	10.6 (10.5)	10.9 (11.0)	10.5 (10.4)	1.00 (0.99, 1.02)	0.66
Number of transactional partners in the past 3 months, mean (SD)	3.5 (2.3)	3.8 (2.6)	3.4 (2.2)	1.06 (1.00, 1.12)	0.039	1.05 (0.99, 1.11)	0.12
Spending on transactional sex in the past 3 months (in Kenyan Shillings), mean SDe	7,764.8 (7,856.5)	8,911.0 (10,252.1)	7,545.4 (7,293.7)	1.02 (1.00, 1.03)	0.016	1.02 (1.00, 1.04)	0.013
Most recent transactional encounter: drank/used drugs	926 (61.7)	143 (59.3)	783 (62.2)	0.88 (0.67, 1.18)	0.40
STIs
HIV seropositivity at baseline	143 (9.5)	18 (7.5)	125 (9.9)	0.73 (0.44, 1.23)	0.24
HSV-2 seropositivity at baseline	581 (38.7)	93 (38.6)	488 (38.8)	0.99 (0.75, 1.32)	0.96
Self-reports having had an STI (other than HIV) in the past 12 Months	248 (16.5)	56 (23.2)	192 (15.3)	1.68 (1.20, 2.35)	0.002	1.66 (1.18, 2.33)	0.003
Experienced abnormal discharge from penis in the past 12 months	204 (13.6)	50 (20.8)	154 (12.2)	1.87 (1.32, 2.67)	<0.001	1.84 (1.29, 2.63)	0.001
Had a sore or ulcer near penis in past 12 months	126 (8.4)	26 (10.8)	100 (7.9)	1.40 (0.89, 2.21)	0.15
Still experiencing:f
Dysuria	25 (9.1)	8 (13.1)	17 (8.0)	1.74 (0.71, 4.25)	0.22
Penile sore or ulcer	32 (11.7)	7 (11.5)	25 (11.7)	0.97 (0.40, 2.38)	0.96
Urethral discharge	17 (6.2)	6 (9.8)	11 (5.2)	2.00 (0.71, 5.66)	0.19

^aOdds ratio and p-value from univariate logistic regressions to determine the association between each individual covariate and infection with C. trachomatis and/or N. gonorrhoeae.

^bFor variables with a p-value≤0.1 in univariate regressions, odds ratio and p-value from multivariable logistic regression to determine the association between covariate and infection with C. trachomatis and/or N. gonorrhoeae, conditional on age, education, and whether the participant was married and/or cohabitating.

^cFor marital status, participants were only asked whether or not they were married or living as married.

^dOdd ratio and 95% confidence interval should be interpreted as the odds associated with a 1,000 Kenyan shilling increase total income.

^eOdd ratio and 95% confidence interval should be interpreted as the odds associated with a 1,000 Kenyan shilling increase in spending on transactional sex.

^fDenominator for still experiencing symptoms is 274 for the overall population; 61 for CT/NG positive and 213 for CT/NG negative.

HIV prevalence was 9.5%, HSV-2 seroprevalence was 38.7%, and prevalence of CT and NG was 14.3% and 2.5%, respectively. Combined CT and/or NG infection was detected in 16.1% of men. Overall, 16.5% of men self-reported having an STI other than HIV in the past 12 months, 13.6 % reported having urethral discharge in the past 12 months, and among those 20.8% reported currently experiencing urethral discharge and/or penile sore/ulcer and/or dysuria.

Compared to participants without CT and/or NG infection, participants testing positive for CT and/or NG were younger (p=0.001), had more sexual (p=0.027) and transactional sex (p=0.039) partners in the past 3 months, and spent almost 20% more on transactional sex in the past 3 months (7,545.4 vs. 8,911.0 Kenyan Shillings; p=0.029). They were less likely to have used a condom at their last sexual encounter (p=0.015), and were more likely to self-report having an STI besides HIV in the past 12 months (p=0.002).

Participants with CT and/or NG infection had significantly higher reporting of abnormal urethral discharge in the past 12 months (p<0.001), but no difference in current symptoms of dysuria (p=0.22) or urethral discharge (p=0.19) compared to uninfected participants. The PPV and NPV for currently experiencing symptoms of CT and/or NG infection was 33.3% and was 84.4%, respectively (see Supplementary 1).

CONCLUSIONS

This study presents some of the only recent data on laboratory-confirmed bacterial STI prevalence among Kenyan men. Recruitment was based on high alcohol use and engagement in transactional sex and not based on STI symptoms or HIV status. HIV prevalence among study participants was 9.5%, which is on par with the prevalence among the broader population of men in the study region.(13) Similarly, HSV-2 prevalence at 38.7%, was within the range identified in other studies among Kenyan men.(10, 14) Prevalence of CT and NG were 14.3% and 2.5%, respectively, with limited comparable available data. These findings may reflect the higher risk among men who are engaged in transactional sex or have high alcohol use, indicating this is a priority group for STI diagnostic testing and treatment.

National guidelines in Kenya and many other countries encourage syndromic management of STIs other than HIV. However, our data comparing symptomatic CT and NG with laboratory-confirmed infection suggests that syndromic management can lead to a large proportion of untreated infections, continued STI transmission, and overtreatment of uninfected men. Such limitations of syndromic management of STIs have been noted in other studies.(3, 4, 15) Our findings underscore the importance of prioritizing STI diagnostic testing among men who report transactional sex and high alcohol use, particularly as less costly tests and point of care testing become available.

This study also provides insights into demographic and behavioral factors associated with CT and/or NG infection. In the context of high STI prevalence, our findings could provide a roadmap for designing better-targeted testing and treatment strategies that balance the need for controlling the increasing burden of STIs with significant resource constraints by prioritizing men who are at highest risk of infection and transmission.

There were limitations to our research. Eligibility criteria for the study meant that we did not collect data from men who were not engaged in transactional sex or alcohol use, as well as men who did not have a steady income. While STI prevalence is likely to be lower among such men, future research should also evaluate STIs in these populations. Despite these limitations, our study presents important demographic, behavioral and STI data from a large sample of Kenyan men who are engaged in high-risk activities including alcohol use and transactional sex. As such, these data are likely to be generalizable to a large population of men engaging in high-risk activities in Kenya and other African countries.

These findings fill a gap in our understanding of STI prevalence and risk factors among African men and indicate a high prevalence of CT and/or NG among Kenyan men who report alcohol use and transactional sex. They also point to a high risk of ongoing transmission if STIs are untreated. Highly sensitive tests for CT and NG are currently available, and lower cost, point of care tests may soon become available. This improves options for targeted diagnostic testing and treatment among men. Ongoing research in the Akiba study will measure STI incidence in this population over 24 months, which will provide further insight into the need, and potential strategies for, targeted testing and treatment.