Abstract
Early in the COVID-19 pandemic, disruptions to sexual health services and changes to sexual behavior due to the first COVID-19 lockdowns were common among U.S. gay, bisexual, and other men who have sex with men (GBMSM). Less is known about the persistence of these changes after this initial lockdown period. These changes have long-term implications for HIV prevention for current and future pandemic periods. This study collected information on COVID-related impacts on sexual behavior and HIV-related health service disruptions from a cohort of U.S. GBMSM at three time points during the COVID-19 pandemic. We observed that COVID-related disruptions to sexual behavior continued from early lockdown periods through December 2020. Although early interruptions to pre-exposure prophylaxis (PrEP) access resolved in later 2020 and interruptions to antiretroviral therapy (ART) adherence were minimal, extended disruptions were observed in HIV testing, sexually transmitted infection (STI) testing, HIV care clinical visits, and HIV viral load testing. Although sexual behavior did not return to prepandemic levels in late 2020, the reduced access to HIV prevention, testing, and treatment services during this period could result in an overall increased HIV transmission rate, with long-term impacts to the trajectory of the U.S. HIV epidemic. Additional resources and programs are needed to address challenges created by the COVID-19 pandemic, as well as prepare for future potential pandemics and other disruptive events.
Introduction
International restrictions to social contact and mobility (“lockdowns”), spurred by the spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), have caused social and economic disruptions since March 2020. In the United States, reports from early 2020 have identified that COVID-19 has prompted major behavioral changes related to the prevention and control of HIV (Jiang et al., 2020; McKay et al., 2021; Pampati et al., 2021; Sanchez et al., 2020; Santos et al., 2021; Stephenson et al., 2021). These changes include reductions in sexual activity (“sexual distancing”) as well as disruptions to patient access to HIV prevention, screening, and clinical care services (Jiang et al., 2020; McKay et al., 2021; Pampati et al., 2021; Sanchez et al., 2020; Santos et al., 2021; Stephenson et al., 2021).
Present-day HIV prevention efforts for gay, bisexual, and other men who have sex with men (GBMSM) focus on reducing HIV acquisition and transmission through promoting safer sexual behaviors, increasing the availability and use of pre-exposure prophylaxis (PrEP), and promoting the consistent and correct use of antiretroviral therapy (ART) so persons living with diagnosed HIV can maintain a suppressed HIV viral load (Centers for Disease Control and Prevention [CDC], 2023; Eisinger et al., 2019; Foss et al., 2004). The latter two strategies require ongoing access to clinical services (CDC, 2019a); HIV transmission remains high partly due to gaps in access to these tools (CDC, 2019b).
COVID-related disruptions have the potential to affect the trajectory of the U.S. HIV epidemic. For example, clinical interruptions that lead to decreased HIV and sexually transmitted infection (STI) testing, PrEP use, STI treatment, and HIV care may increase HIV incidence—decreasing the proportions of GBMSM who know their status, have access to PrEP, or are virally suppressed. Conversely, reductions in sexual risk behaviors may decrease the spread of HIV. The impact of COVID-related disruptions first depends on the demographic distribution of disruptions. Like HIV burden, disruptions have not been uniform across the United States (CDC, 2021; Millett, 2020). For example, Black individuals experience a higher risk of HIV and may also experience more COVID-related disruptions in HIV prevention and care (due to disproportionate impacts of COVID-19 on Black communities in addition to existent decreased access to HIV prevention and care programs) (CDC, 2021; Fields et al., 2021; Millett, 2020; Poteat et al., 2020). In addition to demographic variations, the impact of pandemic-related disruptions on the HIV epidemic depends on the relative extent and timing of changes in sexual behavior and clinical interruptions. A 2021 modeling study assessing the impact of the COVID pandemic on HIV incidence identified that if sexual behavior rebounded while clinical interruptions persisted, excess HIV infections would be expected because clinical interruptions outweighed transmission-reducing impacts of sexual distancing (Jenness et al., 2021).
Short-term COVID-related changes can alter the U.S. HIV epidemic in the long term because HIV incidence and prevalence are affected by changes in HIV risk behavior and HIV care engagement. It is necessary to understand the demographic distribution, magnitude, and timing of COVID-related changes to sexual behavior and disruptions to HIV-related health services to predict their long-term impact on HIV dynamics. Early data have documented changes in early 2020 (Jiang et al., 2020; McKay et al., 2021; Pampati et al., 2021; Sanchez et al., 2020; Santos et al., 2021), but the persistence of these changes remains unclear. It is possible that with increased social mobility following easing of lockdown restrictions in the later months of 2020 (the “post-lockdown” period) (Pan et al., 2020), sexual behavior and access to clinical services may have returned to prepandemic levels. Data on sexual behavior and clinical service disruptions in the post-lockdown period are needed to inform how HIV transmission may have changed at later stages of the COVID-19 pandemic.
In this study, we present the prevalence and trends of COVID-related sexual distancing and clinical service disruptions among a cohort of U.S. GBMSM through December 2020. Outcomes include information on how the COVID-19 pandemic has affected sexual behavior, HIV testing, PrEP use, HIV clinical care, and ART adherence during the first year of the pandemic. Characterizing the impact of the COVID-19 pandemic on HIV-related behaviors of U.S. GBMSM may help guide HIV prevention programs in the post-lockdown era, for example, through highlighting the need for targeted HIV testing, targeted PrEP programs, and home-based HIV care approaches. Furthermore, understanding the impact of the COVID-19 pandemic on GBMSM HIV-related behavior can provide insight into how behavior may alter in future pandemics, and thus aid in pandemic preparedness.
Method
Participants
This study used data from the American Men’s Internet Survey (AMIS) COVID-19 impact survey, collected from a cohort of U.S. GBMSM at three time points during the COVID-19 pandemic: April 2020, July 2020, and September to December 2020. Participants were recruited via participation in the August to December 2019 annual AMIS study (Sanchez et al., 2018); COVID-19 impact survey responses were linked to 2019 AMIS responses. AMIS study participants were recruited through convenience sampling from websites and through social media applications using banner ads and email messages to members. AMIS targets subpopulations of GBMSM that are underserved (with respect to age, race, and geographic area). Participants were eligible to participate in the AMIS study if they were 15 years or older (participants aged 15–17 years had a waiver of parental permission), male sex at birth, resided in the United States (including U.S. territories), and reported oral or anal sex with a man at least once. For the April 2020 COVID-19 impact survey, individuals from the 2019 AMIS study were sent a link to a special COVID-19 impact survey screener where AMIS eligibility was reassessed. Those who were still eligible and consented to participation (provided online written consent) comprised the COVID-19 impact survey participants. The analyses in this study include only the men who completed the April 2020 survey and at least one of the follow-up surveys (either the July and/or September to December follow-up surveys) (Supplemental Figure 1). The study was conducted in compliance with federal regulations governing protection of human beings and was reviewed and approved by Emory University’s institutional review board.
Measures
The goal of the COVID-19 impact survey was to measure COVID-19-related impacts on several areas: general wellbeing, sexual and substance use behavior, HIV and STI prevention, and HIV treatment. The survey collected data on reported changes in the above categories as well as standard demographic information, self-reported HIV status, and COVID-19 mitigation measures in an individual’s local area. This analysis focuses on outcomes related to sexual behavior, HIV and STI prevention, and HIV treatment to assess the impact that the COVID pandemic had on HIV-related behaviors. At each time point, participants were asked whether the COVID-19 pandemic has affected various behaviors/experiences related to sexual health and substance use. Specifically, participants were asked, compared to the time before COVID-19/Coronavirus, please tell us if COVID-19 and the plans used to manage COVID-19 have impacted these things related to related to sexual health and substance use. Please tell us only if it has changed because of COVID-19.
These behaviors/items included number of sexual partners, opportunities to have sex, access to STI testing or treatment, use of condoms, getting HIV tested, access to HIV medications, taking HIV medications every day as prescribed, getting HIV care clinical visits, and getting viral loads or other labs done. Participants were asked to select if the behavior/item “has decreased/less because of COVID-19, has not changed or changed for reasons other than COVID-19, or has increased/more because of COVID-19.” Participants also were asked a series of questions related to service interruptions: “Have you had trouble getting [a given service] because of COVID-19 or the public health efforts to manage it?” Clinical services included getting an HIV test, getting PrEP prescription from your doctor, and getting your PrEP prescription filled at the pharmacy.
Analyses
The prevalence of COVID-19-related impacts was calculated overall and stratified by age category and race/ethnicity category. To represent the full U.S. GBMSM population, demographic standardization using 2019 U.S. Census age and race/ethnicity distribution weights (US Census Bureau, 2019) was used to obtain standardized estimates with 95% confidence intervals of sexual distancing and HIV clinical care interruptions for all U.S. GBMSM. Chi-square tests or Fisher’s exact tests, where applicable, were used to determine whether differences by race/ethnicity were statistically significant (with a p value of .05). To examine the impacts of attrition on the study results, a sensitivity analysis that examined the prevalence of COVID-19-related impacts only on the men who completed each of the three study cycles was completed.
Results
Total enrollment in the first COVID-19 impact survey was 1,051 men, but enrollment decreased over the three survey cycles (Table 1). Participants ranged in age from 15 to 82 years, with a median age of 35 years (SD = 15.7 years). Participants were from the United States, with the most represented regions being the South (n = 427, 40.6%). Approximately 70% of participants (n = 740) were non-Hispanic White in the first cycle, but this increased to 75% (n = 278) by the third cycle. Approximately 10% (n = 122) of participants self-reported as HIV-positive.
Characteristics of GBMSM Who Participated in All Three Cycles of the 2020 AMIS COVID-19 Impact Survey, United States, April to December 2020
Note. GBMSM = gay, bisexual, and other men who have sex with men; AMIS = American Men’s Internet Survey.
More than half of the participants (n = 542, 51.5%) reported a decrease in the number of sexual partners in April 2020, relative to sexual partners at any time before the pandemic (Table 2). This continues through 2020. Approximately 5% of participants (n = 57, 5.5%) reported a decrease in condom use because of COVID-19 through December 2020. Reported decreases in the number of sexual partners did not vary by race/ethnicity at any study cycle (Figure 1, Supplemental Table 1), although change in use of condoms did vary (p = .02, .01, and .02 for April, July, and September to December, respectively): at each study cycle, non-Hispanic Black participants reported both more increases in condom use (n = 7, 7.8%; n = 3, 9.1%; and n = 3, 14.3%, respectively) and decreases in condom use (n = 3, 3.3%; n = 2, 6.1%; and n = 3, 14.3%, respectively) (e.g., reported the least amount of no change in condom use) relative to other race/ethnicity groups.
Frequency of Selected AMIS COVID-19 Impact Survey Outcomes, United States, April to December 2020
Note. STI = sexually transmitted infection; PrEP = pre-exposure prophylaxis; AMIS = American Men’s Internet Survey.
For men self-reporting as HIV-negative or with unknown HIV status. bFor men self-reporting as living with HIV. cFor men self-reporting as HIV-negative or with unknown HIV status and currently using PrEP.
Prevalence of Selected AMIS COVID-19 Impact Survey Outcomes Stratified by Race/Ethnicity During April to December 2020
Among men self-reporting as HIV-negative or with unknown HIV status, about 15% of participants reported a decrease in HIV testing in both early and late 2020 (n = 142, n = 47, respectively) (Table 2). In April 2020, approximately 9% of men (n = 18) currently on PrEP reported trouble getting PrEP prescription from their doctor because of the COVID-19 pandemic; by the end of the year, 7% of participants (n = 6) reported trouble getting a PrEP prescription. Although point estimates of the proportion of Hispanic men and non-Hispanic Black men reporting a decrease in HIV testing were higher relative to non-Hispanic White men in the first two study cycles (n = 25, 19.8%; n = 8, 16.3%; and n = 94, 14.7% in April 2020; n = 10, 16.4%; n = 3, 15.8%; and n = 44, 11.9% in July 2020, respectively), differences were not statistically significantly different by race/ethnicity at any study cycle (p = .12, .67, and .10, respectively). Differences in trouble getting an HIV test or getting a PrEP prescription also were not statistically significantly different by race/ethnicity at any study cycle.
Among men self-reporting as living with HIV, 28% of men (n = 33) reported a decrease in getting HIV care clinical visits because of the COVID-19 pandemic in April 2020, decreasing to 19% (n = 6) by the end of the year (Table 2). Few participants reported disruptions in their access to ART: Only 5% (n = 6) of participants living with HIV reported a decrease in taking HIV medication every day as prescribed in April 2020, although this increased to 6.7% (n = 2) in late 2020. Although participants reported decreases in access to HIV medication, taking HIV medication every day as prescribed, getting HIV care clinical visits, and getting viral loads or other labs are higher among minority race/ethnic groups relative to non-Hispanic White men, differences were not statistically significantly different by race/ethnicity at any study cycle.
Standardization by Census-derived age and race/ethnicity weights did not greatly affect our results (Table 3). For example, the percent of men who reported a decrease in sexual partners in April 2020 changed from a crude percent of 52% to a standardized percent of 54%. However, the 95% confidence intervals for some standardized estimates are wide due to limited sample size within strata.
Age and Race/Ethnicity Standardization of Selected AMIS COVID-19 Impact Survey Outcomes, United States, April to December 2020
Note. Standardized by age (15–24, 25–29, 30–39, ≥40 years) and race (non-Hispanic Black, non-Hispanic White, Hispanic, Other). CI = confidence interval; STI = sexually transmitted infection; PrEP = pre-exposure prophylaxis; AMIS = American Men’s Internet Survey.
For men self-reporting as HIV-negative or with unknown HIV status. bFor men self-reporting as living with HIV. cFor men self-reporting as HIV-negative or with unknown HIV status and currently using PrEP.
We observed similar results when restricting the study participants to only those who participated in all three study cycles (n = 265) (Table 4). Overall, approximately 55% (n = 143) of these participants reported a decrease in the number of sexual partners through the study period. Among these HIV-negative participants, approximately 14% reported a decrease in HIV testing in both early and late 2020 (n = 34 and n = 39, respectively), and 9% (n = 5) of these participants on PrEP reported trouble getting a PrEP prescription in late 2020. Among these HIV-positive participants, 32% of men (n = 7) reported a decrease in getting HIV care clinical visits because of the COVID-19 pandemic in April 2020, decreasing to 10% (n = 2) by the end of the year. Only one participant (5%) who participated in all three study cycles reported a disruption in ART adherence, which occurred in April 2020 only.
Frequency of Selected AMIS COVID Impact Survey Outcomes Where Participant Participated in All Three Survey Cycles (n = 265)
Note. STI = sexually transmitted infection; PrEP = pre-exposure prophylaxis; AMIS = American Men’s Internet Survey.
For men self-reporting as HIV-negative or with unknown HIV status. bFor men self-reporting as living with HIV. cFor men self-reporting as HIV-negative or with unknown HIV status and currently using PrEP.
Discussion
In this study, we observed that COVID-related disruptions to HIV prevention and treatment services and changes in sexual behavior continued from early lockdown periods through December 2020. Extended disruptions were observed in HIV testing, STI testing, HIV care clinical visits, and HIV viral load testing, with only small improvements over time. Although sexual behaviors including number of sexual partners and opportunities to have sex remained below prepandemic levels in later 2020 for many GBMSM, reduced access to HIV prevention, testing, and treatment services that lasted through the year created additional challenges for the control of HIV, which could result in an overall increased HIV transmission rate.
Consistent with other studies and as previously reported (McKay et al., 2021; Pampati et al., 2021), we observed that measures of sexual behavior decreased in early 2020. In our study, GBMSM reported a decrease in both sexual partners and opportunities to have sex in April to May 2020. This aligns with the findings of Pampati et al, who observed that among a cohort of PrEP-using MSM in the southern United States, MSM had a decrease in the number of sexual partners during February to April 2020 (Pampati et al., 2021). A study by McKay et al of U.S. gay and bisexual men also noted a decrease in sexual partners during April to May 2020 (McKay et al., 2021). Our results expand upon these early findings in finding that changes in sexual behavior persisted through the end of the year: Most participants reported a decrease in both the number of sexual partners and opportunities to have sex in both July 2020 and September to December 2020.
Our results additionally complement the early reports that document decreased utilization of/access to HIV prevention and treatment services in the initial stages of the COVID-19 pandemic (Santos et al., 2021; Stephenson et al., 2021); our study observed that U.S. GBMSM experienced HIV prevention and service disruptions because of the pandemic. For GBMSM not living with HIV, initial disruptions to HIV testing and PrEP prescriptions continued in late 2020 for 15% and 7% of participants, respectively. For GBMSM living with HIV, care access was reduced throughout 2020; in late 2020, approximately 19% of participants reported a decrease in HIV medical care visits, down from 28% in April 2020. As others have reported, we observed that few participants reported disruptions in their access to ART in early 2020, and this continued through the year. Although interruptions to HIV clinical care were not widespread and decreased by the end of 2020, these findings highlight the opportunity for new and targeted HIV clinical care interventions, such as home-based HIV care initiation and retention approaches, including telehealth services and multimonth ART prescriptions.
We observed that most sexual behavior and clinical service disruption measures did not vary significantly by race/ethnicity. Due to the vast racial/ethnic inequities in HIV infection and HIV prevention in the United States that predate the COVID pandemic, we would expect HIV transmission to increase most dramatically as a result of the COVID pandemic in a scenario in which clinical service disruptions are more experienced by Black and Hispanic/Latino GBMSM. Historically and in present day, Black and Hispanic/Latino GBMSM have been the most disproportionately affected populations in the United States (CDC, 2021). This is a result of social and structural factors, including but not limited to structural racism, lack of access to quality health care, provider bias, discrimination, and poverty, which exist in the environments in which sexual risk behaviors occur (Maulsby et al., 2014; Sullivan et al., 2014). In our study, we observed that non-Hispanic White men reported less trouble accessing HIV testing, PrEP, and HIV clinical care services, but these differences were not statistically significant for any measure at any study cycle. However, small population-level changes in health care access and/or behavior might still affect the HIV epidemic because HIV transmission in a community can be driven by a small number of individuals (Carnegie & Morris, 2012), so even nonsignificant differences are of note. Targeted HIV prevention efforts among marginalized communities remain essential due to the historically higher burden of HIV experienced by Black and Hispanic GBMSM populations.
Data that assess the temporal changes of sexual risk behaviors and HIV prevention and treatment service utilization are necessary to determine the impact of the COVID-19 pandemic on HIV transmission. The impact of decreased HIV screening, for example, could be offset by concurrent reductions in sexual risk behavior, but the timing and demographic distribution of changes are important. If service interruptions occur in populations with the highest burden of HIV, for example, there may be greater effects on HIV transmission. Modeling studies that use demographically stratified empirical reports of sexual distancing and HIV clinical service disruptions, such as the data presented in our study, can help examine how pandemic disruptions will affect the trajectory of the U.S. HIV epidemic.
This analysis has several limitations. First, study data were obtained from convenience sampling and may not generalizable to all U.S. GBMSM even after demographic standardization. Study participants were more likely to be of non-Hispanic White race/ethnicity, of higher socioeconomic status (SES), and more likely to be insured than the general U.S. GBMSM population. This was particularly true in the later cycles of this study because there was significant loss to follow-up. Although our sensitivity analysis findings demonstrate that attrition did not affect the overall results (the prevalence of sexual behavior and clinical service disruptions experienced by the subset of men who participated in all three study cycles were similar to those experienced by the full study population), participants in the third study cycle were more likely to be non-Hispanic White, older, and not known to be living with HIV. Lack of generalizability may be particularly important for our race/ethnicity findings, given that minority race/ethnicity participants may be more likely to be insured and of higher SES than minority race/ethnicity GBMSM populations, skewing our results to appear to have less racial/ethnic disparities.
Further, the surveys only involved self-report of COVID-related impacts. Participants might have misreported the impacts that the COVID-19 pandemic has had on their sexual behaviors or service utilization/access or misreported the timing of changes. For example, although all participants in this study participated at each study cycle, participants may have been referring to any time during the COVID-19 pandemic when they complete the impact questions (e.g., referring to a decrease in partners during August when they complete the survey in September). However, this concern is somewhat mitigated in seeing that clear temporal decreases in some outcomes are observed (e.g., trouble getting an HIV test, trouble in getting a PrEP prescription). Furthermore, COVID-19 impact survey measures are primarily categorical (e.g., behavior increased, decreased, no change); continuous measures such as the exact number of sexual partners would be useful to identify more specific changes in sexual behavior. For example, a fraction of participants may have reduced their partners by only one partner, whereas others may have reduced their partners by several partners. The impact of these reductions on population-level transmission dynamics is difficult to predict without data on these nuances. Finally, our findings have limited temporal generalizability, given the ongoing changing nature of the COVID-19 pandemic and local restrictions and social behavior patterns. A major strength of this study is its longitudinal nature, but even within one study cycle, there could be short-term temporal fluctuations.
This study is the first to examine the impact of the COVID-19 pandemic on both sexual behavior and clinical service disruptions among U.S. GBMSM through December 2020. Although our findings demonstrated that GBMSM had continued reductions in sexual behavior in late 2020, that access to PrEP was returned to normal in late 2020, interruptions to ART adherence were minimal, and interruptions did not significantly vary by race/ethnicity, our findings highlight the gaps in HIV prevention and treatment that have worsened in the pandemic era. In addition to elucidating behavioral patterns that may occur during future pandemics (and thus aiding in pandemic preparedness), our findings highlight that additional resources and programs will be needed to address existing disparities in HIV prevention and treatment (such as those increasing uptake of PrEP among indicated GBMSM), in addition to solving the new challenges created by the COVID-19 pandemic (such as decreases in HIV testing).
Supplemental Material
sj-docx-1-jmh-10.1177_15579883231168602 – Supplemental material for The Impact of the COVID-19 Pandemic on Sexual Behavior and HIV Prevention and Treatment Services Among U.S. Men Who Have Sex With Men in the Post-Lockdown Era
Supplemental material, sj-docx-1-jmh-10.1177_15579883231168602 for The Impact of the COVID-19 Pandemic on Sexual Behavior and HIV Prevention and Treatment Services Among U.S. Men Who Have Sex With Men in the Post-Lockdown Era by Laura M. Mann, Travis Sanchez, Rob Stephenson, Patrick S. Sullivan and Samuel M. Jenness in American Journal of Men's Health
Supplemental Material
sj-docx-2-jmh-10.1177_15579883231168602 – Supplemental material for The Impact of the COVID-19 Pandemic on Sexual Behavior and HIV Prevention and Treatment Services Among U.S. Men Who Have Sex With Men in the Post-Lockdown Era
Supplemental material, sj-docx-2-jmh-10.1177_15579883231168602 for The Impact of the COVID-19 Pandemic on Sexual Behavior and HIV Prevention and Treatment Services Among U.S. Men Who Have Sex With Men in the Post-Lockdown Era by Laura M. Mann, Travis Sanchez, Rob Stephenson, Patrick S. Sullivan and Samuel M. Jenness in American Journal of Men's Health
Footnotes
Declaration of Conflicting Interests
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 work was supported by grants from the National Institutes of Health (grant numbers: F31MH130274, P30AI050409, and R01MH110358).
Ethical Statement
The study was conducted in compliance with federal regulations governing protection of human beings and was reviewed and approved by Emory University’s institutional review board.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
