Changes in Substance Use and Sexual Behaviors Following an STI Diagnosis among a Cohort of MSM in Los Angeles, CA

Abstract

Background:

Sexually transmitted infections (STIs), STI reinfection, Human Immunodeficiency Virus (HIV) acquisition and changes in behaviors following an STI were examined in a cohort of men who have sex with men (MSM) in Los Angeles, CA.

Methods:

Data from a longitudinal study of MSM enrolled from 2014 with at least one follow-up visit through March 2020 were analyzed (n=447; 1,854 visits). Study visits every 6 months included self-interviews for sexual behaviors, substance use, and specimen collection for chlamydia, gonorrhea, syphilis, and HIV testing. Changes in behaviors were assessed using McNemar’s test and participants not diagnosed with an STI served as controls for a difference-in-differences (DiD) analysis of changes over time.

Results:

Cumulative incidence of an STI was 55% (248/447). At 24-months post STI diagnosis methamphetamine use declined from 50% to 35% (p<.01), and median number of sex partners declined from 5 (IQR: 2–11) to 2 (IQR: 1–6)(p<.01). Among participants at risk for HIV and diagnosed with an STI (n=102), PrEP use was 35% and HIV-seroconversion was 6%. Based on DiD analyses, participants diagnosed with an STI had higher levels of substance and higher number of sex partners when compared to those with no STIs, however, declines in these behaviors were comparable to participants not diagnosed with an STI (p_DID>0.05).

Conclusions:

Despite behavior modifications following an STI diagnosis, STI/HIV incidence was high, suggesting that MSM with STIs occupy sexual networks where reductions in sexual and substance using behaviors do not protect them from ongoing exposure to STIs and HIV.

Short Summary

Despite declines in substance use and sexual behaviors following an STI, reinfections were high suggesting that MSM with STIs occupy sexual networks with high transmission probabilities and prevention efforts should consider sexual network characteristics.

Introduction

Incidence of acute sexually transmitted infections (STIs) – including chlamydia, gonorrhea and syphilis – have increased in recent years, especially among gay, bisexual and other men who have sex with men (MSM).¹ According to the most recent Centers for Disease Control and Prevention (CDC) report, MSM accounted for over three-quarters of all primary and secondary syphilis cases in the United States (US) and nearly half of all reported gonorrhea cases, with rates that were 42 times higher than those among men who have sex with women only.¹ Moreover, MSM account for more new Human Immunodeficiency Virus (HIV) infections in the US than any other group, representing 69% of new HIV diagnoses.²

Certain biological and behavioral factors for HIV and other STIs are shared and STIs can increase the likelihood of HIV transmission through these factors. Biologically, people with STIs may be susceptible to HIV acquisition and persons living with HIV who are co-infected with an STI may have higher HIV viral loads compared to those who do not have STIs.^3–6 Beyond biological factors, a number of behavioral factors influence STI transmission, including number of sexual partners,⁷ shared sexual networks,⁸ low uptake of preventive measures including condom use,⁹ HIV pre- and post-exposure prophylaxis (PrEP and PEP),¹⁰ HIV and STI testing and treatment,^{11, 12} and substance use.¹³ While substance use has been associated with a number of sexual practices, stimulant use in particular has been associated with increased number of sex partners, condomless sex, transactional sex, and even gut microbiome and inflammatory responses that may increase the potential for STI/HIV acquisition and transmission.¹⁴

Despite an increased risk for reinfections with STIs as well as HIV transmission,^{5, 6} there is a dearth of research examining behavior changes following an STI diagnosis. While previous work indicates possible behavior changes after an HIV diagnosis, including a reduction in number of sexual partners,^{15, 16} it is unknown whether similar behavior changes are observed after an STI diagnosis. Additionally, the few studies that have been conducted were limited by a one-group pretest-posttest design. The lack of a comparison group (i.e., individuals without an STI/HIV) limits interpretation of any changes in behavior, given that this design does not account for factors such as regression to the mean or any maturation effects that would be observed regardless of STI status.¹⁷ Therefore, the objective of this study was to examine changes in sexual behaviors, substance use, and subsequent STI and HIV diagnoses over time, among a cohort of men diagnosed with chlamydia, gonorrhea, or syphilis, comparing changes in these behaviors to men not diagnosed with these infections. Given the availability of STI and HIV prevention methods such as expedited partner therapy (EPT) for STIs and HIV pre-exposure prophylaxis, understanding the behavioral trajectories and reaching MSM with acute STIs stands as a public health priority for prevention of both HIV and STI reinfection.

Methods

Study population and design

Data for this study were based on those collected from participants in the mSTUDY – an NIH/NIDA funded cohort of racial/ethnically diverse MSM living with or at high risk for HIV. Study enrollment for mSTUDY started in August 2014 and both enrollment and follow-up are ongoing. Participants were recruited from two different study sites in Los Angeles, CA including a community-based organization providing services for the lesbian, gay, bisexual, and transgender community, and a community-based university research clinic. For this analysis we include data collected between August 2014 and March 2020. Participants were eligible for mSTUDY if they were: (1) between 18 and 45 years of age at enrollment, (2) assigned male sex at birth, (3) if HIV-negative, reported condomless anal intercourse with a male partner in the past 6-months, (4) capable of providing informed consent, and (5) willing and able to return to the study every six months to complete study-related activities. By design, half of the participants were living with HIV. Furthermore, for inclusion in this analysis participants had to have at least one follow-up visit through March 2020 (n=447; 1,854 study visits; 80% of parent cohort).

Study procedures and data collection

The study was approved by the Institutional Review Board at the University of California Los Angeles and prior to study participation all participants provided written informed consent. At each visit participants completed a self-administered, computer-based questionnaire which was available in both English and Spanish. As part of the questionnaire, participants were asked to report on recent (past six months) sexual behaviors including information on number of sex partners, reports of new sex partners, concurrent partnerships (i.e., sexual partnerships that overlap in time), and transactional sex defined as an exchange of money, drugs, shelter, or other goods for sex. Partnership-specific information was only collected for the last sexual partner (in the past six months) and included information on partnership type, partners substance use status, partner HIV status, and whether their partner had concurrent partnerships. Participants were also asked to report on substances used in the past six months including the use of the following drugs: (1) cocaine/crack; (2) ecstasy; (3) heroin; (4) cannabis; (5) methamphetamine; (6) ‘party drugs’ including GHB and ketamine; (7) poppers; and (8) illicit use of prescription medications. Binge drinking was based on the question “how often did you have 6 or more drinks on one occasion.”

At each study visit samples were collected for STI/HIV testing. Urine samples as well as rectal and pharyngeal swabs were collected for Chlamydia trachomatis and Neisseria gonorrhoeae testing using nucleic acid amplification testing (NAAT) technology (Aptima Combo 2^®, GenProbe, San Diego, CA). Blood samples were collected for HIV testing among those who were HIV-negative and HIV-1 RNA levels for those who were living with HIV. Blood samples were also collected for syphilis testing using the rapid plasma regain test (RPR), with confirmatory testing using the Treponema pallidum particle agglutination test (TPPA). Syphilis disposition (i.e., primary, secondary, early latent, or latent syphilis) was also obtained for each participant and based on standard of care health department investigation of syphilis cases as specified by CDC Treatment guidelines.¹⁸ Primary, secondary, or early latent syphilis was defined as infectious syphilis. Those testing positive for an STI were referred to treatment and the treatment status was verified. All participants were scheduled to return every six months and the study procedures were repeated at each visit. Visits typically lasted 60–90 minutes and participants were compensated between $70 and $90 depending on the visit.

Analytic Strategy

Univariate analyses provided descriptive statistics for the sample overall and by STI status. Changes in sexual and substance-using behaviors following an STI diagnosis were examined by comparing each follow-up visit to the index visit, which was defined as the first visit a participant tested positive for chlamydia, gonorrhea, or was diagnosed with infectious syphilis. Participants not diagnosed with an STI at any of their study visits served as a comparator group, with their mSTUDY baseline visit defined as the index visit for this analysis. Differences in behaviors from each of the follow-up visits to the index visit were evaluated using paired t-tests or McNemar’s test, as appropriate. Changes in behaviors of interest over time were also evaluated using the Cochran-Armitage test for trend. In order to account for any potential factors beyond an STI diagnosis which may impact behavior change we conducted a difference-in-differences (DiD) analysis. While DiD methods are most commonly used in public health policy research, the design is a quasi-experimental approach that is useful in any setting where a randomized control trial is not feasible or ethical.^{19, 20} This approach involves comparing changes in the outcomes of interest (i.e., sexual and substance use behaviors) in those diagnosed with an STI (i.e., exposed group) to those not diagnosed with an STI at any study visit (i.e., control group). By including participants who were not diagnosed with an STI, the secular trends common to both groups are subtracted from the association between the exposure, in this case an STI diagnosis, and the outcomes of interest, in this case sexual and substance-using behaviors. Given the unequal censoring rates between the groups (i.e., longer follow-up time for those with no STIs), we limited follow-up to four follow-up visits (24-months) in order to limit the differential censoring rates and resultant loss of power between our comparison groups. All analyses were conducted using SAS version 9.4 (SAS Inc., Cary, NC).

Results

Characteristics of study population

Of the 557 participants enrolled in mSTUDY, 447 (80%) were eligible and included in this analysis. At the index visit, the average age of participants was 32.1 years with 42% identifying as African American/Black, followed by 39% Hispanic/Latinx (Table 1). Unstable housing was reported by 29% of participants with nearly the same amount reporting unemployment. Among the 447 participants, 55% (n=248) tested positive for an STI at least once during the study, with chlamydia being the most common STI (47%; 116/248), followed by gonorrhea (42%; 104/248) and infectious syphilis (26%; 65/248). The most common site of infection for chlamydia was rectal (72%; 83/116), followed by urethral (25%; 29/116) and pharyngeal infections (16%; 19/116). Rectal infections were also highly prevalent among those testing positive for gonorrhea (56%; 58/104) as were pharyngeal infections (61%; 63/104). Furthermore, among the 248 participants with an STI, 26% (64/248) tested positive for chlamydia, gonorrhea, or were diagnosed with infectious syphilis over the follow-up period (i.e., reinfection).

Table 1.

Baseline characteristics of study participants, by STI status (8/2014 – 3/2020)

	Total (n=447)^		STI-positive (n=248)^		STI-negative (n=199)^		P value
	n	%	n	%	n	%
Age, mean(SD)	32.1 (6.9)		31.7 (6.9)		32.6 (6.9)		0.21
Race/ethnicity							0.26
African American	187	41.9	97	39.1	90	45.5
Hispanic/Latinx	173	38.8	101	40.7	72	36.4
Other	62	13.9	33	13.3	29	14.6
White	24	5.4	17	6.9	7	3.5
Unemployed	122	27.9	65	26.5	57	29.5	0.49
Unstable Housing, past 6 months*	130	29.3	79	31.9	51	26.2	0.18
HIV-positive	230	51.5	146	58.9	84	42.2	<.01
STI diagnosis							--
Chlamydia	116	26.0	116	46.8	0	0
Gonorrhea	104	23.2	104	41.9	0	0
Infectious Syphilis	65	14.5	65	26.2	0	0

Abbreviations. SD=Standard deviations

^{^}Sum may not equal total due to missing information

^*Defined as not having a regular place to stay in the past 6 months

At the index visit, no differences in sociodemographic characteristics were noted among those with and without an STI, however, participants positive for an STI were more likely to be living with HIV as compared to those without an STI (59% vs. 42%; p<.01). Additionally, differences were noted in substance-using and sexual behaviors. For instance, at the index visit, those diagnosed with an STI had a higher prevalence of methamphetamine use as compared to those without an STI (50% vs. 35%; p<.01)(Figure 1) and reported higher levels of concurrent partnerships (53% vs. 38%; p<.01) and transactional sex (23% vs. 11%; p<.01)(Figure 2).

Figure 1.

Changes in binge drinking, methamphetamine use, and popper use after 24-months of follow-up among mSTUDY participants, by STI status (8/2014 – 3/2020)

Abbreviations. FU=Follow-up; STI=Sexually Transmitted Infection; DiD=Difference in Differences

Figure 2.

Changes in sexual behaviors after 24-months of follow-up among mSTUDY participants, by STI status (8/2014 – 3/2020)

Abbreviations. FU=Follow-up; STI=Sexually Transmitted Infection; DiD=Difference in Differences

Changes in sexual and substance using behaviors and STI positivity

Following an STI diagnosis, significant declines were noted in substance use and sexual behaviors (Table 2). At 12-months, binge drinking declined from 50% to 36% (p<.01) as did other substance use including methamphetamine (50% to 41%; p<.01) and poppers (42% to 34%; p=0.02). These changes were maintained at 24-months post-STI diagnosis. For instance, methamphetamine use at 24-months (i.e., follow-up visit 4) was reported among 35% of participants as compared to 50% at the index visit (p<.01). Notable differences in sexual behaviors included a decline in new or concurrent sexual partnerships and transactional sex. The median number of sex partners also dropped from 5 (interquartile range [IQR] 2–11) at the time of STI diagnosis to 3 (IQR 1–10; p=0.02) at 12-months and 2 (IQR 1–6; p<.01) after 24 months of follow-up. STI positivity declined over time when compared to the index visit, however, the prevalence of STIs remained high, with at least 12% of participants testing positive for an STI at each of the follow-up visits. For instance, after 24 months of follow-up, 11% tested positive for chlamydia, 10% for gonorrhea, and 9% were diagnosed with infectious syphilis. Among the 102 participants who tested HIV-negative at baseline and who were diagnosed with an STI, six seroconverted during the study resulting in a cumulative incidence of 5.9% (95% CI 1.3–10.5) as compared to a cumulative incidence of 4.3% (n=5; 95% CI 1.0–8.1) among those who were not diagnosed with an STI. Regardless of STI status, only two participants who seroconverted reported a history of PrEP use with none reporting PrEP use in the six months prior to their visit when HIV seroconversion occurred.

Table 2.

Substance use and sexual behaviors overtime following an STI diagnosis among mSTUDY participants (August 2014 - March 2020)

	Index visit (n=248)		Follow-up visit 1 post STI diagnosis (n=248)			Follow-up visit 2 post STI diagnosis (n=218)			Follow-up visit 3 post STI diagnosis (n=180)			Follow-up visit 4 post STI diagnosis (n=151)
	n	%	n	%	P value~	n	%	P value~	n	%	P value~	n	%	P value~
Months since index visit, median (IQR)	--		6.2 (6.1–6.9)		--	12.8 (12.2–14.3)			19.3 (18.4–21.1)			25.7 (24.6–27.7)
Substance use behaviors, past 6 months
Binge drinking	123	50.2	111	45.1	0.11	78	36.1	<.01	73	40.6	0.13	60	40.0	0.01
Cocaine	47	19.2	44	17.9	0.77	30	13.9	0.09	22	12.2	0.10	21	14.0	0.18
Ecstasy	35	14.3	41	16.7	0.32	22	10.2	0.02	16	8.9	0.05	17	11.3	0.09
Heroin	9	3.7	4	1.6	0.13	8	3.7	0.76	8	4.4	0.26	3	2.0	0.56
Marijuana	127	51.8	121	49.2	0.41	113	52.3	0.67	87	48.3	0.88	73	48.7	0.88
Methamphetamine	122	49.8	110	44.7	0.05	88	40.7	<.01	73	40.6	0.04	53	35.3	<.01
Poppers	102	41.6	89	36.2	0.06	73	33.8	0.02	56	31.1	0.07	44	29.3	<.01
Prescription drugs	36	14.7	28	11.4	0.13	32	14.8	0.87	19	10.6	0.14	13	8.8	0.03
Sexual behaviors, past 6 months
Number of sex partners (median, IQR)	5 (2–11)		3 (1–8)		0.03	3 (1–10)		0.02	3 (1–9)		<.01	2 (1–6)		<.01
New Sex Partner	185	83.7	173	81.6	0.46	142	77.6	0.30	115	80.4	0.40	79	69.3	0.02
Concurrent Partnership	121	52.6	104	45.4	0.30	83	41.9	<.01	63	38.0	0.01	43	30.3	<.01
Transactional sex*	54	23.0	40	17.4	0.05	34	17.2	0.05	25	15.0	0.04	15	10.3	<.01
Last sexual partner’s characteristics
HIV-positive/unknown	161	65.7	154	62.6	0.14	144	66.4	0.41	116	64.4	0.08	95	63.3	0.12
One-time Partner	66	30.4	48	23.6	0.20	25	16.2	0.01	20	17.5	0.01	25	24.5	0.60
Transactional Sex*	46	21.2	32	15.6	0.06	30	19.5	0.37	15	13.2	0.02	14	13.7	0.07
Partner Concurrent	168	78.5	138	75.0	0.27	96	76.9	0.41	71	78.0	0.56	41	71.9	0.46
Partner Methamphetamine use	75	30.6	70	28.5	0.68	60	27.6	0.28	44	24.4	0.14	39	26.0	0.38
Clinical and Laboratory Factors
HIV-positive	146	58.9	148	59.7	0.16	136	62.4	0.09	114	63.3	0.08	96	63.6	0.16
PrEP use, past 6-months ^	45	34.1	43	42.2	0.14	35	41.2	0.30	23	33.3	0.72	17	29.8	0.99
HIV-1 RNA ≤ 20 copies/mL^^	72	49.7	82	55.4	0.16	72	53.7	0.21	63	55.3	0.32	57	59.4	0.38
STI diagnosis
Chlamydia	116	46.8	31	12.5	<.01	17	7.8	<.01	15	8.4	<.01	16	10.6	<.01
Gonorrhea	104	41.9	25	10.9	<.01	27	12.4	<.01	24	13.4	<.01	15	9.9	<.01
Infectious Syphilis	65	26.2	19	7.7	<.01	26	11.9	<.01	16	8.9	<.01	13	8.6	<.01

Abbreviations. STI=Sexually transmitted infection; SD=Standard deviation; IQR=interquartile range; PrEP=pre-exposure prophylaxis

^{^}Among HIV-negative participants;

^{^^}Among HIV-positive participants

^~p value based on comparisons to baseline using McNemar’s test for paired data

^*Defined as receiving money, drugs, or shelter in exchange for sex

Among participants who were not diagnosed with an STI there were modest reductions in substance use including a decrease in the prevalence of binge drinking and methamphetamine use (Table 3). Overall, sexual behaviors that increase the potential for STI/HIV transmission were lower among those without an STI, with changes noted in the median number of sex partners and new sexual partnerships. For instance, at the index visit 84% of participants with an STI reported having at least one new sex partner in the past 6 months as compared to 70% among those without an STI (p<.01). After 24-months of follow-up new partnerships were reported by 69% of participants with an STI as compared to 52% of participants without an STI (p<.01).

Table 3.

Substance use and sexual behaviors overtime among participants without an STI diagnosis among mSTUDY participants (August 2014 - March 2020)

	Index visit (n=199)		Follow-up visit 1, no STI diagnosis (n=199)			Follow-up visit 2, no STI diagnosis (n=162)			Follow-up visit 3, no STI diagnosis (n=134)			Follow-up visit 4, no STI diagnosis (n=115)
	n	%	n	%	P value~	n	%	P value~	n	%	P value~	n	%	P value~
Months since index visit, median (IQR)	--		6.2 (6.0–6.6)		--	12.4 (11.7–13.5)		--	19.1 (18.2–21.0)		--	25.7 (24.5–28.6)		--
Substance use behaviors, past 6 months
Binge drinking	119	60.4	98	49.5	<.01	87	54.0	0.10	67	50.4	<.01	56	48.7	<.01
Cocaine	47	23.9	34	17.2	0.03	37	23.0	0.68	25	18.8	0.10	24	20.9	0.68
Ecstasy	33	16.8	21	10.6	0.01	22	13.7	0.32	18	13.5	0.18	10	8.7	0.02
Heroin	10	5.1	6	3.0	0.16	3	1.9	0.18	3	2.3	0.18	5	4.4	0.18
Marijuana	118	59.9	102	51.5	0.01	81	50.3	0.21	60	45.1	<.01	61	53.0	0.99
Methamphetamine	68	34.5	52	26.3	<.01	45	27.9	0.13	40	30.1	0.16	30	26.1	0.05
Poppers	65	33.0	49	24.8	<.01	38	23.6	0.14	33	24.8	0.12	27	23.5	0.16
Prescription drugs	35	17.8	23	11.6	0.02	22	13.7	0.16	14	10.5	0.04	11	9.6	0.22
Sexual behaviors, past 6 months
Number of sex partners (median, IQR)	3 (1–6)		2 (1–6)		<.01	2 (1–5)		0.04	1 (1–4)		0.05	1 (1–3)		0.06
New Sex Partner	138	69.7	116	58.6	<.01	93	57.4	0.01	69	51.9	<.01	52	45.2	<.01
Concurrent Partnership	74	38.1	69	35.6	0.58	46	29.3	0.17	38	29.6	0.02	34	31.2	0.32
Transactional sex*	22	11.3	24	12.4	0.99	18	11.5	0.83	9	8.2	0.82	5	5.2	0.13
Last sexual partner’s characteristics
HIV-positive/unknown	114	57.9	107	54.0	0.82	92	57.1	0.62	72	54.2	0.77	63	54.8	0.48
One-time Partner	55	28.7	47	25.0	0.35	30	21.3	0.24	20	20.8	0.03	19	24.4	0.09
Transactional Sex*	19	9.9	26	13.8	0.06	13	9.2	0.83	9	9.4	0.32	12	15.4	0.99
Partner Concurrent	143	72.0	122	66.7	0.42	92	68.7	0.98	63	69.2	0.98	34	55.7	0.06
Partner Methamphetamine use	38	19.3	40	20.2	0.22	31	19.3	0.72	27	20.3	0.71	20	17.4	0.48
Clinical and Laboratory Factors
HIV-positive	84	42.2	85	42.7	0.32	69	42.6	0.32	56	41.8	0.16	50	43.5	0.42
PrEP use, past 6-months ^	27	23.5	40	34.8	<.01	35	36.1	<.01	28	35.0	0.03	29	43.3	<.01
HIV-1 RNA ≤ 20 copies/mL^^	46	54.8	47	56.6	0.54	39	58.2	0.37	38	67.9	0.07	35	70.0	0.02
STI diagnosis
Chlamydia	0	0.0	0	0.0	--	0	0.0	--	0	0.0	--	0	0.0	--
Gonorrhea	0	0.0	0	0.0	--	0	0.0	--	0	0.0	--	0	0.0	--
Infectious Syphilis	0	0.0	0	0.0	--	0	0.0	--	0	0.0	--	0	0.0	--

Abbreviations. STI=Sexually transmitted infection; SD=Standard deviation; IQR=interquartile range; PrEP=pre-exposure prophylaxis

^{^}Among HIV-negative participants;

^{^^}Among HIV-positive participants

^~p value based on comparisons to baseline using McNemar’s test for paired data

^*Defined as receiving money, drugs, or shelter in exchange for sex

Difference-in-differences analysis

Based on the DiD analyses, participants diagnosed with an STI consistently reported higher levels of substance use and sexual behaviors when compared to those who were not diagnosed with an STI (Figure 1 and Figure 2). Furthermore, we found that the decline in both sexual and substance-using behaviors were not significantly different when comparing those with an STI to those without an STI. For instance, over the course of the four follow-up visits, the absolute decline in binge drinking among participants with an STI was 11%, declining from a high of 60% at baseline to 49% at the fourth follow-up visit (p _trend<.01)(Figure 1). A similar decline in binge drinking was also noted among those who were not diagnosed with an STI, declining from 50% at baseline to 34% at the fourth follow-up visit (p _trend<.01). Based on unadjusted as well as adjusted analyses, controlling for HIV status, the decline in binge drinking was not significantly different when comparing those with an STI to those without (p _DID=0.09) suggesting that being diagnosed with an STI is not associated with this decline.

Among participants who tested HIV-negative, PrEP use across study visits ranged from 30% to 35% among those who were diagnosed with an STI (Figure 3). While PrEP use remained unchanged among those who were diagnosed with an STI, participants who were not diagnosed with an STI reported an increase in PrEP use from 24% at baseline to 43% after 24-months of follow-up (p _trend<.01). Among participants living with HIV, viral suppression was low at baseline and there were no meaningful changes noted over the follow-up period (Figure 3). Fewer than half of participants living with HIV and diagnosed with an STI had HIV-1 RNA levels ≤ 20 copies/mL at baseline, which increased to 54% at 12-months follow-up and 59% at 24-months (p _trend=0.50).

Figure 3.

Changes in PrEP use and HIV-1 RNA levels after 24-months of follow-up among mSTUDY participants, by STI status (8/2014 – 3/2020)

Abbreviations. FU=Follow-up; STI=Sexually Transmitted Infection; DiD=Difference in Differences

Discussion

Our findings demonstrate that while there were significant declines in reported substance use and sexual behaviors following an STI diagnosis, comparisons to participants without an STI suggest that these changes may be linked to factors other than any behavior modification resulting from the STI diagnosis. While few studies have explored behavior changes following non-HIV STIs, HIV-specific studies note significant reductions in risk behaviors following an HIV diagnosis.^{16, 21} Most of these studies have used a one-group design and as our study demonstrates, the lack of a comparison group may limit the interpretation of these findings. Additionally, we note that PrEP use was low at baseline and remained low over the course of the study, despite continued substance use and sexual behaviors that increase the likelihood of HIV acquisition. Taken together with the rates of HIV seroconversion, these findings suggest that MSM with STIs occupy a sexual network with high STI/HIV transmission probabilities where even reductions in some behaviors are not protective of ongoing exposures to STIs and HIV.

Sexual network factors potentiate the onward spread of STIs/HIV based on the dynamics of sexual partnerships and the interconnections within the network.²² This is particularly relevant to STIs such as chlamydia and gonorrhea that have a short window of infectiousness and require either high levels of partner turn-over (i.e., individual level factor) or partnership overlap (i.e., network level factor) to spread. This may partially help to explain why despite reductions in factors such as number of sexual partners or concurrent partnerships, reinfections with STIs continued. In fact, we found that participants diagnosed with an STI were more likely to report that their partner likely had concurrent partnerships when compared to those with no STIs. This relationship held even in instances when the participant themselves did not report having any concurrent sexual partnerships. This suggests that targeted efforts that move beyond individual risk behaviors and give consideration to sexual network characteristics are needed to reduce the burden of STIs among MSM, including approaches such as strategies that expedite and improve partner notification and treatment to avert reinfections with STIs. However, the use of EPT for MSM has been limited, in part due to uncertainties in the CDC recommendations.¹⁸ The CDC first recommended EPT for chlamydia and gonorrhea in 2006 and both the original recommendation as well as the updated guidelines from 2021 express reservations around EPT use among MSM. In particular, the CDC notes that a risk of this approach is potential missed opportunities to provide HIV counseling and testing services to non-index cases. However, to date no published studies have demonstrated reductions in HIV partner services resulting from the provision of EPT among MSM, while an increasing number of studies highlight the potential benefits, including increases in partner notification.^{23, 24} This approach offers a way not only to avert primary infections among persons who may not seek care, but also to reduce risk of STI reinfection within sexual networks with high rates of STIs.

While PrEP use among our study population (who primarily identified as Black/African American or Hispanic/Latinx) was higher than national estimates of 23%, it was significantly lower than the 2025 Ending the HIV Epidemic initiative target of ≥50% of individuals who have an indication for PrEP.^{2, 25} PrEP acceptability and uptake among MSM has increased considerably since it first became available in 2014, but overall coverage remains low, with even lower coverage observed among Black and Latino MSM, despite a higher HIV burden in these groups.²⁶ Consideration of racial/ethnic disparities in uptake of PrEP represents an important challenge in addressing the HIV epidemic among MSM as this population experiences additional individual, network, and structural-level barriers to PrEP use.²⁷ Effective evidence-based approaches to engage minority MSM are needed to address this gap in coverage. Peer-based intervention approaches, such as peer navigation, which bridges the gap between patient and provider by providing peer support in navigating HIV services, are a promising approach to improving HIV service uptake among key populations.²⁸ An acceptability study from Washington State exploring attitudes towards peer navigation for PrEP among MSM supports the potential for this approach.²⁹ Furthermore, ensuring PrEP service delivery models are culturally and ethnically sensitive and free of provider stigma is also critical to reducing racial disparities and improving uptake.³⁰

This study had several limitations of note. Our data—especially data on sexual behaviors and substance use—were based on self-report, which can result in an underestimation of these behaviors. However, the use of computer-assisted self-interviews may help minimize the potential for social desirability bias. There may be some misclassification of STI status, given that participants could receive STI testing and treatment outside of the study and would be considered STI negative for the purposes of this study. In particular, participants receiving PrEP who were previously diagnosed with a bacterial STI may have received quarterly STI testing (as per standard of care) outside of the study. We anticipate that this potential misclassification would bias our effect estimates towards the null given that higher levels of substance use and sexual behaviors associated with an STI diagnosis would be attributed to a participant who tested negative as part of the study, but who was otherwise positive for an STI at some point between follow-up visits. However, given that follow-up visits occur relatively frequently (6-month intervals), the proportion of participants receiving additional STI testing outside of the study period likely remains low. While STI treatment status for all participants was verified, information related to partner STI status, treatment, and resumption of sexual activity following an STI was not collected and thus limits our interpretation related to reinfections. Finally, this study was based on participants recruited from a community-based sexual health clinic and a university-based research clinic in Los Angeles and may not be generalizable to other populations. However, this work also has several strengths. The use of longitudinal data allowed us to assess temporality of measured associations and measure incident STI/HIV infections. Our ability to include a comparator group free of STIs and the use of DiD analyses allowed us to control for secular trends during the study and compare changes in behaviors by STI status. Finally, the cohort is HIV serostatus neutral, allowing for exploration of the association between STIs and HIV among those at risk of HIV as well as those living with HIV.

Conclusion

These findings advance the understanding of sexual behaviors and substance use links with STI/HIV transmission in a group of MSM at increased risk for STI/HIV acquisition by demonstrating that decreases in sexual risk and substance use behaviors were not linked to any behavior modification resulting from the STI diagnosis. Further, these changes in behavior did not correspond with reductions in STI incidence. Given that STI infections are associated with increased risk of HIV acquisition, and coinfection of HIV with other STIs can have synergistic effects, reducing STI burden among MSM is an essential component of comprehensive HIV prevention and treatment in this population. Strategies to improve the reach of partner notification and treatment for STIs in these networks and increase PrEP uptake are critical to efforts to reduce the burden of these infections among MSM.