Prevalence of and factors associated with urethral and rectal STIs among a cross-sectional sample men who have sex with men

Abstract

Background:

We examined prevalence and factors associated with STIs among MSM in Kenya.

Methods:

A cross-sectional analysis was conducted using baseline data from the Mbili Pamoja cohort study, which enrolled 501 MSM (250 in Kisumu and 251 in Nairobi) aged 18–39 years between June and November 2024. Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (NG) were detected in urine and clinician-collected rectal swabs. Multivariable logistic regression was used to identify factors associated with urethral and rectal CT/NG.

Results:

The prevalence of urethral CT and/or NG infections was 9.2% in Kisumu and 3.7% in Nairobi; rectal CT and/or NG prevalence was 10.4% in Kisumu and 18.7% in Nairobi. In multivariable analysis, urethral CT/NG was more common among MSM aged 25–29 years vs. 18–24 years, last sex partner being a cisgender woman, and identifying as transgender or non-binary. Hazardous drug use was inversely associated with urethral STIs. Rectal STIs were increased for participants reporting receptive or versatile sexual role, group sex, being transgender/non-binary, and elevated depressive symptoms. Hazardous alcohol use, food insecurity, and recent antibiotic use were inversely associated with rectal infections. Self-reported dysuria and penile discharge showed low sensitivity (38.6% and 30.6%, respectively) and high specificity (>80%) for urethral infections. Rectal symptoms were infrequently reported but showed moderate sensitivity and high specificity.

Conclusions:

MSM in Kisumu and Nairobi have substantial burden of frequently asymptomatic urethral and rectal CT and NG infections. While some modifiable sexual exposures were identified, results also highlight the need for integrated mental health and substance use services.

Summary:

Kenyan MSM had high prevalence of asymptomatic urethral and rectal chlamydia or gonorrhea; prevalence varied with sexual practices and psychosocial factors, highlighting need for integrated services.

Introduction

Men who have sex with men (MSM) in this region face higher risks of HIV and sexually transmitted infections (STIs) than the general population due to limited access to healthcare services, stigma, and discriminatory policies that criminalize same-sex relationships.¹ STIs such as rectal gonorrhea and chlamydia are often underdiagnosed in MSM due to inadequate screening and syndromic management approaches.² In Uganda, a study of 295 MSM found that 47.3% reported any genital or anal STI symptoms in the last 6 months.³ Among the participants, 13.5% were found to have at least one STI, including syphilis, rectal or urethral chlamydia or gonorrhea. A study from Tanzania found that among 409 MSM aged 18–60 years, 17.4% were HIV-positive, 38.5% had herpes simplex virus 2 (HSV-2), and 0.2% tested positive for syphilis.⁴

Studies from Kenya have observed high STI prevalence among MSM in both urban and peri-urban areas.⁵ Kenya has made progress in addressing STIs among key populations, including MSM, through targeted interventions such as HIV pre-exposure prophylaxis (PrEP), condom distribution, and peer-led outreach programs.⁶ However, despite these efforts, barriers such as stigma, discrimination, and legal challenges continue to hinder MSM from accessing appropriate STI prevention and treatment services. In a cross-sectional study using a snowball sampling approach, 70 MSM in Nairobi and Mombasa reported that concerns about disclosure and past negative encounters in healthcare facilities contributed to delays in seeking treatment for STIs.⁷ The reluctance to access care was primarily linked to expected stigma and discrimination from medical professionals.

Factors such as condomless anal sex, multiple sexual partners, engagement in transactional sex, and substance use contribute to higher rates of urethral and rectal STIs.^{8, 9} Socio-economic challenges such as unemployment, poverty, and housing insecurity further limit MSM in accessing consistent healthcare services. Stigma and discrimination, both within healthcare settings and the broader community, remain significant deterrents to seeking timely STI diagnosis and treatment. Our study aimed to examine the factors associated with urethral and rectal STIs among MSM in two sites in Kenya, to support development of tailored interventions with improved relevance to target populations.

Materials and Methods

Study Setting, Design, and Participants

Mbili Pamoja (“Two Together” in Swahili), is a prospective study conducted in Nairobi (the capital of Kenya, centrally located) and Kisumu (Kenya’s third largest city, and located in the west, adjacent to Lake Victoria). According to the Kenya Demographic and Health Survey (KDHS),¹⁰ Kisumu and Nairobi counties had a higher prevalence of men who reported to have sex with persons who were neither their partners nor lived with them (37.6% and 43.6%) respectively, as compared to the national prevalence of 34.7%. The prevalence of HIV among adults in Kisumu County is 17.4% and 3.3% in Nairobi,¹¹ in comparison to the national prevalence of 3.7%.¹²

Recruitment occurred between June 7 and November 18, 2024, with a target of 250 participants in each study location. In Kisumu, 250 participants were recruited from a MSM friendly facility called Drop-in Centre (DICE). At DICE, peer educators informed their peers of the study; contact was either in person, virtual, or through phone calls. Potential participants were also informed of the study during their routine visits and during weekly programmatic activities. Potential participants then self-selected and came for screening, consenting and enrolment, with some as walk-in and others with an appointment. Those who met eligibility criteria and were willing to participate underwent informed consent and were enrolled in the study. In Nairobi recruitment occurred between June 7 and November 14, 2024. Participants were recruited from existing HIV prevention and care programs across 10 clinics. Potential participants were invited to the study clinic for screening, consenting, and enrolment, with all being given appointments (i.e., no walk-ins) to facilitate logistics and reduce waiting time.

Mbili Pamoja is designed to follow participants for one year with an enrolment visit, followed by 6- and 12- month visits, and seeks to examine penile and rectal microbiome, mucosal and systemic immune features in relation to incident infection, and changes after antibiotic treatment. To be eligible, participants had to be aged 18–39 years, assigned male sex at birth, reporting anal insertive or receptive sex with another man in the past 3 months, and providing informed consent. Having had diarrhea or vomiting in the past 14 days was a temporary ineligibility; after symptoms had been resolved for 2 weeks, potential participants could come for rescreening and consenting while enrolment remained open. For eligible participants, informed consent was obtained in their preferred language (English, DhoLuo, Swahili). The current analysis makes use of baseline data.

Data collection

Data were collected through ACASI (Audio Computer Assisted Self-Interview) and CAPI (Computer Assisted Personal-Interview). Variables included sociodemographic factors, sexual practices, alcohol and substance use, depressive symptoms, household food surveys and social networks. The Patient Health Questionnaire 9 (PHQ-9) assessed depressive symptoms, dichotomized at ≥5 for mildly elevated or greater symptoms.¹³ Alcohol use was assessed using the Alcohol Use Disorders Identification Test (AUDIT) tool dichotomized at ≥8 for potentially hazardous or harmful alcohol use.¹⁴ Drug Abuse Screening Test 10 (DAST) was used to measure drug dependency, with responses dichotomized at a threshold of ≥3 for potentially hazardous or harmful drug use.¹⁵ An 11-question assessment was was used to assess the level of stigma related to same sex practices and identity, dichotomized at the median for higher vs. lower perceived stigma.^{16, 17} We assessed the association of PHQ-9, AUDIT, DAST, stigma, and food insecurity (as an indicator of economic strain) in relation to STI risk because ours and others’ studies have found them to be associated with sexual risk taking.^18–20

STI and HIV Testing: Laboratory Methods and Treatment

All participants underwent a medical examination which included genital and rectal examination for signs of STIs every 6 months, with results recorded using a closed-ended standardized form. Urine and clinician-collected rectal swab specimens were collected at baseline, and months 6 and 12. Infection with CT and NG was evaluated by GeneXpert^® CT/NG (Cepheid, Sunnyvale, CA) qualitative PCR. Participants were provided antibiotic treatment for NG (cefixime 400mg orally or ceftriaxone 125mg intramuscular, AND azithromycin 1g orally or doxycycline 100mg orally, twice a day for 7 days) and CT (azithromycin 1g orally or doxycycline 100mg orally, twice a day for 7 days) according to Kenyan treatment guidelines.²¹

In the current analysis, HIV status was analyzed as self-reported for participants reporting themselves as positive, and results of testing for those who reported themselves as HIV negative. HIV testing was conducted with serial rapid tests, following Kenyan guidelines.²² In our prior analysis of MSM and non-binary individuals in Kenya,²³ the prevalence of oropharyngeal CT/NG infection was 2.8%. Given the low prevalence, costs of testing, and overall objective and aims of the study, we did not conduct oropharyngeal testing.

Statistical Analysis

We employed logistic regression to identify factors associated with STIs as a composite of infection with chlamydia and/or gonorrhea. Urethral and rectal infections were modelled separately. We analysed CT and NG as a composite outcome because they have the same routes of infection, and because there would have been insufficient power to analyse them separately by anatomic site. As missing data did not exceed 5% in any covariate or outcome variable, we proceeded with complete case analysis. Explanatory variables of interest were selected a priori as those having previously demonstrated relation to increased STI risk, and included demographic factors such as age, education level, and employment status, as well as sexual practices and psychosocial factors (depressive symptoms, drug and alcohol use). Univariate logistic regression was performed to assess crude associations between each independent variable and the STI outcome. Variables with a p-value <0.2 in the univariate analysis were then included in the multivariable logistic regression model to control for confounding factors. Minimized Akaike’s Information Criterion (AIC) was used to for model selection, with additional checking for whether removal of variables indicated confounding via change in coefficients >10%.²⁴ Adjusted odds ratios (AORs) with corresponding 95% confidence intervals (CIs) were reported to quantify the strength and direction of associations. Variance was estimated with clustering based on study site. All analyses were conducted using Stata/SE version 17.

Results

A total of 501 participants were enrolled (250 in Kisumu, 251 in Nairobi). Participants were median age 25 years, 88% cis-male and 12% trans or non-binary, 81.6% having secondary education or higher, 65.5% employed, and 19.4% HIV positive. There were significant differences in the distribution of participants’ demographics and behavioral practices between study sites.²⁵ Briefly, participants in Kisumu were older, while participants in Nairobi more commonly had secondary education or higher and were more likely to report being non-cis-male.

The prevalence of urethral CT/NG was 9.2% in Kisumu and 10.4% in Nairobi. The prevalence of rectal CT/NG was 3.7% in Kisumu and 18.7% in Nairobi (Figure 1). At both locations, urethral and/or rectal CT was more common than urethral and/or rectal NG (15.8% vs. 10.1%). Most infections (N=95) were single-aetiology (N=57 CT infections; N=38 NG infections) and single anatomic site (n=72 infections rectal only, n=38 urethral infections only) rather than dual aetiologies (n=21) or anatomic sites (n=11). There were N=37 rectal NG infections in the absence of rectal CT (n=3 Kisumu, N=34 Nairobi), N=33 rectal CT infections in the absence of rectal NG (n=4 Kisumu, n=29 Nairobi), N=11 urethral NG infections in the absence of urethral CT (n=3 Kisumu, n=8 Nairobi), and N=35 urethral CT infections in the absence of urethral NG (n=18 Kisumu, n=17 Nairobi). Antibiotic treatment was documented for 80.0% of urethral NG, 90.0% of rectal NG, 84.2% of urethral CT, and 87.2% of rectal CT.

Figure 1. Distribution of urethral and rectal infections among study participants.

Figure 1 shows the intersections of urethral and rectal CT and NG, with the rows of the matrix corresponding to the status (indicated with the various shapes), and the columns to the intersections between these statuses. The number of participants observed in the intersection (vertical drop lines between circles) is shown in the grey bar with the number of participants labelled (y-axis and each bar). Thus, there were 29 participants with urethral CT (black circle) in the absence of urethral NG or rectal CT/NG; 5 participants with urethral CT who also had rectal CT (black triangle); 3 participants with urethral CT who also had urethral NG (black square); and 1 participant with urethral CT who also had rectal CT (black diamond) and rectal NG (black triangle).

Factors Associated with Urethral Chlamydia and Gonorrhea

Several factors were associated with urethral CT/NG infection (Table 1): aged 25–29 years (vs. 18–24), being trans/non-binary, having female partner at last sex, having a main male sex partner, not practising receptive anal intercourse in past 6 months, not currently using PrEP. Multivariable analyses identified four factors associated with urethral STI (Table 2). Urethral STI was increased among participants aged 25–29 years (vs. 18–24 years; AOR 1.47; 95% CI 1.10 – 1.98) among transgender or non-binary individuals (AOR = 1.12; 95% CI: 1.04 – 1.20), and those whose last sex partner was cis-female (AOR = 1.62; 95% CI: 1.4 – 1.77). Participants with hazardous or harmful drug use (DAST ≥3) had 33% decreased odds of urethral CT/NG than those with lower scores (95% CI: 0.63 – 0.71).

Table 1.

Factors Associated with Urethral Chlamydia and/or Gonorrhea in Kisumu and Nairobi

	Negative, N=451 (90.2%) n (%)	Positive, N=49 (10.8%) n (%)	OR (95% CI)	P-value
Study Site
Kisumu	227 (90.8)	23 (9.2)	Ref
Nairobi	224 (89.6)	26 (10.4)	1.15 (0.63–2.07)	0.652
Age
18–24 years	216 (90.0)	24 (10.0)	Ref
25–29 years	133 (86.4)	21 (13.6)	1.42 (1.03–1.96)	0.033
30 years and above	102 (96.2)	4 (3.8)	0.35 (0.08–1.47)	0.153
Median Age in years (interquartile range)	25 (21–29)	25 (22–26)	0.97 (0.87–1.08)	0.538
Participant Gender
Cis-male	398 (90.5)	42 (9.5)	Ref
Transgender, non-binary	53 (88.3)	7 (11.7)	1.25 (1.09–1.44)	0.002
Worried on whether food would run out in the past 30 days
No	151 (92.1)	13 (7.9)	Ref
Yes	300 (89.3)	36 (10.7)	1.39 (0.79–2.45)	0.248
Antibiotics use past 30 days
No	340 (90.7)	35 (9.3)	Ref
Yes	111 (88.8)	14 (11.2)	1.23 (0.71–2.11)	0.466
HIV Positive
No	362 (89.8)	41 (10.2)	Ref
Yes	89 (91.8)	8 (8.2)	0.79 (0.48–1.31)	0.362
Circumcision status
Uncircumcised	30 (90.9)	3 (9.1)	Ref
Circumcised	421 (90.2)	46 (9.8)	1.08 (0.36 – 3.30)	0.888
Insertive Anal Intercourse Ever
No	75 (91.5)	7 (8.5)	Ref
Yes	376 (90.0	42 (10.0)	1.20 (0.54–2.67)	0.660
Vaginal or anal sex with a cisgender female ever
No	224 (92.2)	19 (7.8)	Ref
Yes	226 (88.3)	30 (11.7)	1.56 (0.86–2.84)	0.140
Refuse to Answer	1	0
Partner gender at last sex
Male	417 (90.5)	44 (9.5)	Ref
Female	34 (87.2)	5 (12.8)	1.39 (1.28–1.52)	<0.001
Usual sexual positioning with male partner
Usually insertive	184 (87.6)	26 (12.4)	Ref
Usually receptive	126 (94.0)	7 (6.0)	0.45 (0.16–1.27)	0.132
Usually versatile	141 (90.4)	15 (9.6)	0.75 (0.40–1.42)	0.383
Currently have a main male sex partner
No	119 (93.7)	8 (6.3)	Ref
Yes	298 (89.2)	36 (10.8)	1.80 (1.03–3.14)	0.039
Missing	33 (86.8)	5 (13.2)	2.25 (1.32–3.86)	0.003
Circumcision status of main Partner
Uncircumcised	13 (81.3)	3 (18.7)	Ref
Circumcised	284 (89.6)	33 (10.4)	0.50 (0.33–0.76)	0.001
No main male sex partner	152 (92.1)	13 (7.8)	0.37 (0.25–0.55)	<0.001
Used condom for insertive anal sex at last sexual encounter
No	95 (88.0)	13 (12.0)	Ref
Yes	221 (91.0)	22 (9.0)	0.73 (0.43–1.22)	0.231
No insertive anal intercourse past 6 months	132 (90.4)	14 (9.6)	0.78 (0.33–1.81)	0.556
Given money or gifts by partner at last sexual encounter
No	256 (89.5)	30 (10.5)	Ref
Yes	194 (91.1)	19 (8.9)	0.85 (0.59–1.47)	0.562
Refuse to answer	1	0
Group sex past 6 months
No	388 (90.0)	43 (10.0)	Ref
Yes	62 (91.2)	6 (8.8)	0.87 (0.09–8.67)	0.908
Refuse to Answer	1	0
Number of different male sex partners in past 6 months 1
0–1	105 (92.1)	9 (7.9)	Ref
2	94 (85.5)	16 (14.5)	1.84 (0.85 – 4.00)	0.122
3–4	121 (91.0)	12 (9.0)	1.14 (0.50 – 2.62)	0.752
5 or more	95 (93.1)	7 (6.9)	0.87 (0.34 – 2.25)	0.773
Missing	33 (86.8)	5 (13.2)	1.67 (0.59 – 4.67)	0.331
Refused or Don’t Know	2	0
Prep Use by HIV status
HIV negative, never taken PrEP	94 (87.0)	14 (13.0)	Ref
HIV negative, taken PrEP in past	139 (89.1)	17 (10.9)	0.84 (0.70 – 1.01)	0.069
HIV negative, currently taking PrEP	129 (92.8)	10 (7.2)	0.55 (0.18 – 1.76)	0.316
HIV positive	89 (91.8)	8 (8.2)	0.64 (0.62 – 0.66)	<0.001
Mean Stigma Score (SD)	25.1 (19.9)	19.4 (17.1)	0.98 (0.95 – 1.01)	0.054
Elevated depressive symptoms
No (PHQ-9 <5)	223 (88.8)	28 (11.2)	Ref
Yes (PHQ-9 ≥5)	228 (91.6)	21 (8.4)	0.73 (0.22–2.43)	0.613
Hazardous or harmful alcohol use
No (AUDIT <8)	294 (89.6)	34 (10.4)	Ref
Yes (AUDIT ≥8)	157 (91.3)	15 (8.7)	0.83 (0.26–2.60)	0.744
Hazardous or harmful drug use
No (DAST <3)	315 (89.5)	37 (10.5)	Ref
Yes (DAST ≥3)	136 (91.9)	12 (8.1)	0.75 (0.62 – 0.92)	0.005

¹For number of different male sex partners in the past 6 months: category 0–1 includes n=9 participants who reported 0 sex partners, though in other questions they did indicate having male sex partners; n=38 responses were missing due to a skip pattern error in survey programming

Table 2.

Results of Univariable and Multivariable Logistic Regression: Factors Associated with Urethral Chlamydia and/or Gonorrhea Infections in Kisumu and Nairobi

	Crude OR (95% CI)	Adjusted OR1 (95% CI) N=496
Age
18–24 years	Ref	Ref
25–29 years	1.42 (1.03 – 1.96)	1.47 (1.10 – 1.98)
30 years and above	0.37 (0.08 – 1.70)	0.36 (0.07 – 1.71)
Gender of last sex partner
Male	Ref	Ref
Cis-Female	1.39 (1.28 – 1.52)	1.62 (1.49 – 1.77)
Participant Gender
Cis-male	Ref	Ref
Transgender, non-binary	1.25 (1.09 – 1.44)	1.12 (1.04 – 1.20)
Hazardous or harmful drug use
No (DAST <3)	Ref	Ref
Yes (DAST ≥3)	0.75 (0.62 – 0.92)	0.67 (0.63 – 0.71)

OR = Odds Ratio; 95% CI = 95% Confidence Interval

¹Model is simultaneously adjusted for all variables presented

Factors Associated with Rectal Chlamydia and Gonorrhea

Several factors were associated with increased prevalence of rectal CT/NG (Table 3): living in Nairobi, having ever practised RAI, usual sexual positioning as receptive or versatile, engaging in group sex, and reporting elevated depressive symptoms (PHQ-9 ≥5). Participants aged 30 years and above, reporting food insecurity, reporting ever having insertive anal intercourse, having sex with a cisgender female, using a condom during insertive anal sex, and not having RAI in past 6 months were less likely to test positive for rectal CT/NG. Several factors remained significant in multivariable adjusted analyses (Table 4). Being usually receptive or versatile during sexual intercourse, participating in group sex in the past 6 months, being a transgender/non-binary person, and reporting elevated depressive symptoms (PHQ-9 ≥5) were significantly associated with increased odds of rectal CT/NG, while participants reporting food insecurity, recent antibiotic use, or hazardous or harmful alcohol use (AUDIT ≥8) had lower odds of rectal STI.

Table 3.

Factors Associated with Rectal Chlamydia and/or Gonorrhea in Kisumu and Nairobi

	Negative, N=412 (83.2%) n (%)	Positive, N=83 (16.8%) n (%)	OR (95% CI)	p-values
Study site
Kisumu	236 (96.3)	9 (3.7)
Nairobi	176 (70.4)	74 (29.6)	11.0 (5.37–22.6)	<0.001
Age
18–24 years	180 (75.3)	59 (24.7)	Ref
25–29 years	136 (88.9)	17 (11.1)	0.38 (0.15–1.00)	0.050
30 years and above	96 (93.2)	7 (6.8)	0.22 (0.06–0.86)	0.030
Median Age in years (IQR)	25 (22–29)	22 (20–25)	0.87 (0.78–0.97)	0.014
Participant Gender
Cis-male	370 (85.1)	65 (14.9)	Ref
Transgender, non-binary	42 (70.0)	18 (30.0)	2.44 (0.94–6.33)	0.067
Worried on whether food would run out in the past 30 days
No	130 (79.8)	33 (20.2)	Ref
Yes	282 (84.9)	50 (15.1)	0.70 (0.64–0.77)	<0.001
Antibiotics use past 30 days
No	302 (81.4)	69 (18.6)	Ref
Yes	110 (88.7)	14 (11.3)	0.56 (0.32–0.97)	0.039
Known HIV positive
No	340 (85.4)	58 (14.6)	Ref
Yes	72 (74.2)	25 (25.8)	2.04 (0.27–15.4)	0.491
Receptive Anal Intercourse Ever
No	110 (97.4)	3 (2.6)	Ref
Yes	302 (79.1)	80 (20.9)	9.71 (2.36–39.9)	0.002
Receptive anal intercourse past 6 months
No	49 (90.7)	5 (9.3)
Yes	228 (76.0)	72 (24.4)	3.09 (0.97–9.89)	0.057
No receptive anal intercourse ever/Missing	135 (95.7)	6 (4.3)	0.44 (0.38–0.50)	<0.001
Partner gender at last sex
Male	377 (82.5)	80 (17.5)	Ref
Female	35 (92.1)	3 (7.9)	0.40 (0.30–0.55)	<0.001
Usual positioning with male partner
Usually insertive	197 (95.2)	10 (4.8)	Ref
Usually receptive	94 (71.2)	38 (28.8)	7.96 (2.41–26.3)	0.001
Usually versatile	121 (77.6)	35 (22.4)	5.70 (2.43–13.3)	<0.001
Currently haveg a main male sex partner
No	93 (73.8)	33 (26.2)	Ref
Yes	284 (85.8)	47 (14.2)	0.47 (0.15–1.46)	0.189
Missing	34 (91.9)	3 (8.1)	0.25 (0.07–0.86)	0.027
Partner circumcised
Uncircumcised	14 (87.5)	2 (12.5)	Ref
Circumcised	269 (85.7)	45 (14.3)	1.17 (0.25–5.54)	0.842
Don’t Know	127 (77.9)	36 (22.1)	1.98 (0.83–4.74)	0.112
The last time you had receptive anal intercourse with a male did he use a condom
No	67 (63.2)	39 (36.8)	Ref
Yes	162 (83.1)	33 (16.9)	0.35 (0.31–0.39)	<0.001
No receptive anal intercourse in past 6 months	182 (94.3)	11 (5.7)	0.10 (0.03–0.40)	0.001
Given money or gifts by partner at last sexual encounter
No	230 (81.8)	51 (18.2)	Ref
Yes	181 (85.0)	32 (15.0)	0.83 (0.55–1.24)	0.360
Refuse to answer	1	0
Group sex past 6 months
No	361 (84.5)	66 (15.5)	Ref
Yes	50 (74.6)	17 (25.4)	1.86 (1.72–2.00)	<0.001
Refuse to Answer	1	0
Number of different male sex partners in past 6 months 1
0–1	96 (85.7)	16 (14.3)	Ref
2	94 (86.2)	15 (13.8)	0.96 (0.50–1.85)	0.911
3–4	110 (82.7)	23 (17.3)	1.21 (0.67–2.18)	0.524
5 or more	75 (74.3)	26 (25.7)	1.80 (1.03–3.16)	0.040
Missing	34 (91.9)	3 (8.1)	0.57 (0.17–1.84)	0.346
Refused or Don’t Know	2	0
PrEP use by HIV status
HIV negative, never taken PrEP	85 (79.4)	22 (20.6)	Ref
HIV negative, taken PrEP in past	133 (86.4)	21 (13.6)	0.66 (0.28–1.60)	0.361
HIV negative, currently taking PrEP	122 (89.0)	15 (11.0)	0.53 (0.19–1.50)	0.232
HIV positive	72 (74.2)	25 (25.8)	1.25 (0.42–3.70)	0.683
Mean stigma score (SD)	24.3 (19.7)	26.5 (19.5)	1.01 (0.99–1.03)	0.584
Elevated depressive symptoms
No (PHQ-9 <5)	215 (86.4)	34 (13.6)	Ref
Yes (PHQ-9 ≥5)	197 (80.1)	49 (19.9)	1.57 (1.03–2.40)	0.036
Hazardous or harmful alcohol use
No (AUDIT <8)	265 (81.3)	61 (18.7)	Ref
Yes (AUDIT ≥8)	147 (87.0)	22 (13.0)	0.65 (0.61–0.70)	<0.001
Hazardous or harmful drug use
No (DAST <3)	290 (83.3)	58 (16.7)	Ref
Yes (DAST ≥3)	122 (83.0)	25 (17.0)	1.02 (0.86–1.22)	0.783

¹For number of different male sex partners in the past 6 months: category 0–1 includes n=9 participants who reported 0 sex partners, though in other questions they did indicate having male sex partners; n=38 responses were missing due to a skip pattern error in survey programming

Table 4.

Results of Logistic Regression: Factors Associated with Rectal Chlamydia and Gonorrhea in Kisumu and Nairobi

	Odds Ratio (95% CI)	Adjusted Odds Ratio1 (95% CI) N=490
Age in years (continuous)	0.87 (0.78–0.97)	0.90 (0.77–1.05)
Usual positioning with male partner
Usually insertive	Ref	Ref
Usually receptive	7.96 (2.41–26.3)	5.84 (1.68–20.3)
Usually versatile	5.70 (2.43–13.3)	5.29 (2.03–13.8)
Group sex past 6 months
No	Ref	Ref
Yes	1.86 (1.72–2.01)	2.00 (1.93–2.06)
Participant Gender
Cis-male	Ref	Ref
Transgender, non-binary	2.44 (0.94–6.34)	1.37 (1.22–1.55)
Worried on whether food would run out in the past 30 days
No	Ref	Ref
Yes	0.70 (0.64–0.77)	0.88 (0.83–0.94)
Antibiotics use past 30 days
No	Ref	Ref
Yes	0.56 (0.32–0.97)	0.52 (0.44–0.61)
Elevated depressive symptoms
No (PHQ-9 < 5)	Ref	Ref
Yes (PHQ-9 ≥ 5)	1.57 (1.03–2.40)	1.43 (1.11–1.83)
Hazardous or harmful alcohol use
No (AUDIT <8)	Ref	Ref
Yes (AUDIT ≥ 8)	0.65 (0.61–0.70)	0.71 (0.42–1.19)

¹Model is simultaneously adjusted for all variables presented.

Site Specific Results

In Kisumu, only age was significantly associated with urethral STI (Supplemental Table 1), with older participants having reduced odds of urethral STI. Several factors were associated with rectal infection (Supplemental Table 2): ever practising RAI, being transgender/non-binary, and having known HIV positive status were strongly associated with higher odds of rectal STIs. Increase in age also yielded reduced odds of having a rectal STI. Surprisingly, reported condom use was associated with higher odds of rectal CT/NG.

In Nairobi, there was higher prevalence of rectal STI among participants (Supplemental Tables 3–4) who reported ever having RAI, and usual sexual positioning as receptive or versatile. Those who used condoms or did not have RAI in the past 6 months had significantly lower odds of contracting a rectal STI as compared to those who had unprotected receptive sex. Increasing age and having used an antibiotic in the past 30 days also reduced the odds of rectal STI.

Factors Associated with Elevated DAST

To better understand the inverse association between DAST and STI risk, we conducted bivariate analyses to examine factors associated with elevated DAST score (Supplemental Table 5). Elevated DAST score was more prevalent among participants aged 25–29, cis-males, Kisumu participants, those with elevated depressive symptoms (PHQ-9 ≥ 5), potentially hazardous alcohol use (AUDIT ≥8), and those who were worried about food running out. Adding these variables to the multivariable model that were not already included did not attenuate the inverse association between elevated DAST score and STI, nor were any of these variables significant, thus not explaining the independent association between DAST and STI.

Sensitivity and Specificity of Symptoms and Signs for Laboratory Detected STI

Self-reported symptoms and clinician detected signs of infection had poor sensitivity for laboratory-detected infections, and moderate to high specificity (Supplemental Table 6). Participants’ self-reported dysuria or penile discharge detected only 38.6% and 30.6% of urethral infections, respectively. Many participants reported having these symptoms despite testing negative, resulting in 86.3% and 80.5% specificity for dysuria and discharge, respectively. Clinician-detected urethral discharge was rarely observed even when infection was present. Participant-reported anal ulcer/sore in the past 6 months had 38.6% sensitivity and 85.0% specificity, T the most sensitive rectal symptom associated with infection. Anal discharge detected on exam yielded extremely low sensitivity (3.6%) and excellent specificity (97.8%).

Discussion

Urethral and rectal CT/NG infections were prevalent, with substantially higher rectal infection prevalence observed in Nairobi. Urethral infections were more frequent among men who engaged in sex with cisgender females, and those who typically assumed the insertive role during anal sex. Rectal infections predominated in younger individuals, those practicing RAI, and those identifying as transgender or non-binary.

Our findings align with studies demonstrating elevated STI risk among young adults and gender-diverse populations due to structural and behavioural vulnerabilities.²⁶ Increased urethral STIs among those with female sex partners may point to underrecognized transmission pathways in bisexual networks.²⁷ Interestingly, participants who reported hazardous or harmful drug use had lower odds of urethral STIs, a finding that contrasts with literature linking substance use to increased STI risk.²⁸ The observed inverse association could suggest that individuals who use drugs may engage in less frequent sexual activity, possibly due to the physical or psychological effects of substance use.

Participants who reported usual sexual positioning as receptive or versatile, engaging in group sex, identifying as transgender or non-binary, and having depressive symptoms had increased likelihood of rectal CT/NG infections. These associations are consistent with prior studies showing that receptive and versatile positioning increase risk of rectal STIs due to mucosal vulnerability and higher likelihood of exposure to infected secretions.²⁹ Group sex has also been documented as a facilitator of STI transmission, often involving multiple partners and reduced condom use, which intensifies exposure risk.³⁰ Transgender and non-binary individuals often face structural barriers to healthcare and stigma, which can delay diagnosis and treatment.²⁶ The link between depressive symptoms and rectal STIs mirrors research indicating that poor mental health is associated with higher sexual risk behaviours and reduced health-seeking behaviour.³¹

The contrasting patterns observed in Kisumu and Nairobi highlight how local context can influence STI risk. In Kisumu, older age was generally protective against urethral and rectal STIs, aligning with studies that show younger MSM and transgender individuals engage more frequently in high-risk sexual behaviours. The strong association between RAI and rectal STI positivity is well-documented, given the vulnerability of rectal mucosa to bacterial infections.³² The higher odds of rectal STIs among transgender women and those with HIV infection may reflect overlapping risk networks and barriers to comprehensive care.²⁶ In Kisumu, condom use was linked with increased odds of rectal STIs, possibly indicating inconsistent or incorrect use. The inverse risk observed with food insecurity may suggest that economic hardship constrains sexual activity or reduces opportunities for high-risk partnerships.

The higher prevalence of hazardous drug use among MSM aged 25–29 may reflect increased exposure to social networks where substance use is normalized, and aligns with existing literature showing that young adults often engage in higher-risk substance use due to transitional life stages and increased autonomy.³³ Transgender and non-binary individuals in our cohort reported lower rates of harmful drug use compared to cisgender males, a contrast to global data suggesting heightened substance use among gender minorities due to stigma and mental health burdens.^{34, 35} This discrepancy may reflect contextual differences in social support or access to harm reduction services, lower socioeconomic status (i.e., less purchasing power), or sampling bias of transgender persons in our study sample. The significantly higher prevalence of harmful drug use in Kisumu compared to Nairobi could point to regional disparities in drug availability or intervention access. Specific drugs particularly bhang, miraa, and cocaine were closely tied to elevated DAST scores, which is consistent with evidence that stimulant and cannabis use are linked with higher addiction risk and co-occurring mental health issues.³⁶ The strong correlation between depressive symptoms and problematic drug use underscores the bidirectional relationship between mental health and substance misuse. We did not find significant confounding of the association between elevated DAST score and reduced prevalence of urethral CT/NG; this may be due to differences in sexual networks, as there was no association between drug use and rectal CT/NG.

The evaluation of symptoms and exam findings against laboratory-confirmed CT/NG infections revealed a consistent pattern of low sensitivity and high specificity, underscoring the inadequacy of syndromic management in identifying asymptomatic infections. Current clinical guidelines have questioned the reliability of symptom-based screening, as studies show that over 70% of rectal and more than 50% of urethral infections may remain asymptomatic.³⁷ Self-reported symptoms occurring in the past 6 months had lower specificity of 80–86%, and may reflect other etiologic causes of urethral discharge and dysuria, such as Mycoplasma genitalium or Trichomonas vaginalis.³⁸ The high specificity of current penile or anal discharge aligns with literature suggesting that when symptoms are present, they are predictive of infection but their rarity limits diagnostic utility. Surprisingly, among participants with self-reported anal ulcer/sore, 38.6% tested positive for rectal CT/NG. Because this was self-reported, the question (“In the past 6 months, have you had any ulcers or sores in your rectum”) may have been interpreted as “soreness” or pain. Anecdotally, clinicians and participants indicate reports of “anal ulcers or sores in rectum” may also be related to rough sex, which could increase risk of rectal STI acquisition and perception of anal ulcers or sores. Conversely, self-reported anal ulcers or sores may represent mechanically-induced mucosal disruption, stemming from infection-related inflammation. This suggests the potential role of this symptom as a red flag in clinical settings, and questioning around rough sex may help understand reporting of this symptom.

As noted, syndromic management has limited utility for these STIs that are primarily asymptomatic. As such, the Kenyan National AIDS and STI Control Program (NASCOP) is currently revising STI Prevention and Treatment guidelines to rely on etiologic management, rather than syndromic management. However, the cost, infrastructure, and assays needed to adopt etiologic testing broadly will likely take several years to roll out. In the interim, DoxyPEP may be a consideration, and a randomized controlled trial of DoxyPEP for MSM is currently underway in Kenya (Clinicaltrials.gov NCT06468462).³⁹ This open label randomized controlled trial of 2,900 participants in Nairobi, Kisumu, and Mombasa will compare standard practice (syndromic management) to DoxyPEP and periodic presumptive treatment, to generate data on acceptability, adherence, effectiveness, and cost-effectiveness to inform guidelines.

Limitations

The cross-sectional design limits temporal inference between risk factors and STI outcomes. Reliance on self-reported sexual behaviour, drug use, and symptom data introduces potential for recall and social desirability biases, which may under- or overestimate true associations. Participants were recruited from urban settings with access to health services and taking part in programs for sexual and gender minority individuals, which may not reflect the experiences of MSM in rural or lower-access regions, limiting generalizability.

Future Study Recommendations

Future research should incorporate qualitative studies to explore the decision-making processes surrounding specific sex acts, partner selection, condom use, and how these intersect with drug and alcohol use among MSM and gender-diverse individuals. Such insights would provide context to the behavioural patterns underlying STI transmission. Longitudinal studies are essential to examine patterns of recurrent infections, changes in sexual behaviour over time, and the impact of interventions such as PrEP and mental health support.

Conclusion

This study demonstrates a substantial burden of urethral and rectal CT and NG infections among MSM in Nairobi and Kisumu, with rectal infections significantly more prevalent in Nairobi. Infections were frequently asymptomatic, underscoring the need for routine STI screening. While some modifiable sexual exposures were identified, results also highlight the need for integrated mental health and substance use services.

Availability of data and materials:

The data used in this analysis may be available upon reasonable request and with appropriate ethical approvals.