Abstract
Background
Clients seeking sexual health care often need concurrent mental healthcare services, which were disrupted during the COVID-19 pandemic. Using data from two studies conducted before and after the onset of COVID-19, we examined changes in self-reported depression and anxiety scores among clients in a sexual health clinic in New York City (NYC).
Methods
We enrolled 144 participants pre-COVID-19 and 319 post-COVID-19 pandemic onset. Participants completed questionnaires assessing demographics, sexual behaviors, and mental health status. Primary mental outcomes included depression (Patient Health Questionnaire [PHQ-9]) and anxiety (Generalized Anxiety Disorder Scale [GAD-7]). We conducted descriptive analyses and used generalized linear mixed models (GLMM) to estimate predictors of mental health changes.
Results
Cohorts were comparable by age, self-identified gender, race/ethnicity, income, HIV, and sexually transmitted infection assessment scores. Post-COVID-19 participants reported significantly higher mean PHQ-9 and GAD-7 scores compared with pre-COVID-19 participants (3.6 ± 4.2 vs 5.7 ± 5.3, P < .001; 3.9 ± 4.3 vs 5.1 ± 5.0; P = .019). Post-COVID-19 participants were also more likely to be uninsured or on Medicaid (2.7% vs 20%, 18% vs 30%, P < .001), and to report intimate partner violence victimization (24% vs 45%, P = .003). Adjusted GLMM showed post-COVID-19 was associated with a 1.55 (95% CI: .07, 3.03, P < .04) mean increase in PHQ-9 scores, but not GAD-7.
Conclusions
Depression and anxiety scores increased after the onset of the COVID-19 pandemic in this NYC sexual health clinic sample. The sustained impact of the COVID-19 pandemic on depression calls for integrated, accessible mental health services within sexual health care settings.
Keywords: COVID-19, men who have sex with men, mental health, sexual health services
Globally, adverse mental health outcomes—including depression and anxiety—increased during and following the coronavirus disease 2019 (COVID-19) lockdowns. Evidence of this relationship emerged early in the pandemic. A meta-analysis of studies with a pre-pandemic baseline showed that increases in daily SARS-CoV-2 (a virus that causes COVID-19) infection rates and reduction in human mobility were associated with 27.6% and 25.6% rise in the prevalence of major depressive and anxiety disorders, respectively [1].
Some studies suggest that the mental health impact of COVID-19 was more pronounced among minoritized and marginalized groups, including sexual and gender minorities like lesbian, gay, bisexual, transgender, and queer (LGBTQ+) individuals [2]. Research indicates that LGBTQ+ people experienced higher levels of unmet mental health care needs during the pandemic compared with their non-LGBTQ+ counterparts [3, 4]. This excess mental health burden reflects compounding factors, including stigma related to sexual and gender identity, lower income or socioeconomic status, lack of health insurance, experienced or perceived or internalized stigma, and limited access to comprehensive care. An integrative review qualitatively highlighted the contribution of COVID-19 on the mental health of LGBTQ+ individuals, along with access to physical and psychological healthcare, challenges with alcohol and substance use as coping strategies, housing, economic, and other population-specific vulnerabilities [5]. However, there remains a dearth of quantitative data on the effects of COVID-19 on LGBTQ+ persons in the United States.
Even before the COVID-19 pandemic, the United States faced a growing mental health crisis, particularly among youth and LGBTQ+ populations [6]. Minority stress theory posits that members of marginalized or minoritized groups often face elevated levels of mental illness due to intersecting stressors stemming from a hostile social system that operates throughout the course of their lives [7, 8]. A national survey of LGBTQ+ youth found that 54% and 67% of respondents had depression or anxiety, respectively [9]. Other US national data showed LGB adults and transgender persons were at least two and four times as likely to have mental health conditions when compared with their heterosexual or cisgender peers [10–12].
Increased rates of mental health challenges among men who have sex with men (MSM) and transgender women (TW) are linked to individual-level factors such as substance use, low socioeconomic status, experienced and internalized stigma, and exposure to violence. Structural determinants—including social isolation, financial hardship, community stigma, and violence from both intimate partners and society—exacerbate these risks. The co-occurrence of HIV and its risk factors has been well documented. However, the effect of emerging infections and broader societal disruptions like COVID-19 has received less attention. Moreover, these evolving health crises were met with inadequate and delayed healthcare system responses.
Past studies have shown that access to adequate HIV prevention and treatment care can improve an individual's mental health through increased HIV knowledge, development of behavioral skills to protect against HIV and sexually transmitted diseases, and psychological relief from disease-related anxiety. Additionally, for many MSM and TW, sexual health services, including HIV prevention and care, are a gateway into broader healthcare engagement including mental health care for those in need [13–18]. Many health services providing LGBTQ+ care or linking people to mental health care closed temporarily to curb disease transmission and manage resources to respond to the pandemic [19–23]. The combination of lack of access to comprehensive sexual health care, at the same time when care needs were the highest, could have exacerbated the acute and lasting effects of COVID-19 on various outcomes including mental health.
To better understand the impact of COVID-19 on mental health outcomes among MSM and TW engaged in comprehensive HIV prevention services, we used baseline data from two similarly designed cohort studies conducted before and after the onset of the COVID-19 pandemic. We applied a serial cross-sectional design to examine group-level changes in self-reported levels of depression and anxiety pre-COVID-19 onset (using the Stick2PrEP2 cohort) and post-COVID-19 onset (using Stick2PrEP3 cohort). To further inform inference on the attribution of COVID-19 to the mental health of LGBTQ+ individuals, we triangulated our findings from the serial cross-sectional analysis with an assessment of individual-level changes in mental outcomes among a subset of those repeatedly sampled.
METHODS
Study Setting
The Stick2PrEP program comprises a series of studies evaluating the effectiveness of short messaging service interventions to increase healthcare engagement and HIV pre-exposure prophylaxis (PrEP) use among MSM and TW receiving care at the Columbia University Irving Medical Center/New York-Presbyterian Hospital (CUIMC/NYP) in northern Manhattan, New York. Stick2PrEP2 (S2PrEP2) recruited participants attending the sexual health program between November 2018 and February 2020, shortly before the pandemic onset. Stick2PrEP3 (S2PrEP3) recruited the same type of participants in the same clinic settings from May 2021 to December 2023.
Recruitment and Enrollment
Stick2PrEP2 and Stick2PrEP3 recruited persons assigned male sex at birth who were receiving HIV comprehensive prevention services at the CUIMC/NYP Comprehensive Health Program Sexual Health Program. Participants were approached by the research team in the waiting room or after their visit. They were invited to participate if they were between 18 and 65 years old and understand English. Those who were screened eligible and provided written informed consent were enrolled.
Study Design and Data Collection
We conducted a serial cross-sectional analysis using baseline data from all participants who enrolled in S2PrEP2 and S2PrEP3. The analytic sample was comprised of all participants in S2PrEP2 and S2PrEP3 who completed the enrollment survey. Additionally, a subset of individuals who participated in both cohorts were included in a triangulation sample, enabling individual-level longitudinal assessment. Participants self-reported demographics, sexual behaviors, HIV vulnerability, depression, and anxiety in an enrollment survey completed on their own device or a clinic-provided tablet.
This article adheres to the reporting standards established within The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) [24].
Measurement Definitions
Outcomes
We measured the primary outcome, depression, with the validated Patient Health Questionnaire (PHQ-9), a 9-item scale (PHQ-9) [25]. The questionnaire asked about the frequency when the participants are bothered by a list of biological or psychological problems, ranging from “0” (not at all) to “3” (nearly every day). The co-primary outcome, anxiety, was assessed using the validated 7-item Generalized Anxiety Disorder Scale (GAD-7) where participants also reported the frequency of experiencing the problems described ranging from “0” (not at all) to “3” (nearly every day) [26]. The scales were summarized both as continuous and categorical variables, where any sum of the score <5 is defined as minimal depression or anxiety, 5–9 as mild, and >9 as moderate or severe.
HIV Vulnerabilities
The following were measured by validated scales to identify potential cofounders due to HIV knowledge, stigma, or lifetime adverse events. We measured HIV vulnerability using the 7-item scale HIV Incidence Risk Index for MSM (HIRI-MSM) ranging from 0 to 47, with a score of 10 or higher as the cutoff point for greater HIV risk and when intensive HIV prevention methods are encouraged [27]. Sexually transmitted infection (STI) vulnerability was measured using the 6-item IWantTheKit tool. The score ranged from 0 to 10, with 0–2 defined as low risk, 3–6 as medium risk, and 7–10 as high risk [28]. HIV knowledge was measured using the 18-item HIV knowledge questionnaire HIVKQ18, ranging from 0 to 18, where the participants responded with “Yes,” “No,” or “I don’t know” to a given statement on HIV [29]. We also measured PrEP self-efficacy using the 15-item PrEP Adherence Self-Efficacy Scale (ASES), where participants rated their confidence in performing a certain action from “0” (cannot do at all) to “10” (completely certain can do) [30]. To measure the relationship and interaction between PrEP providers and patients we incorporated the 11-item modified PrEP Index that ranges from 10 to 50, with a score of 45 or higher suggesting no likelihood of continuous HIV care engagement within 6 months [31]. We assessed social and emotional well-being with the Adverse Childhood Experiences (ACE) questionnaire, a 10-item scale modified from the original ACE scale where participants answered in binary, with ACE score ≥1 as experienced adverse events in childhood [32]. Intimate partner violence (IPV) was evaluated through the perspective of a perpetrator and victim individually, each with a 5-item questionnaire where participants reported the frequency by which they experienced certain behaviors during an intimate relationship [33]. HIV stigma was assessed using a shortened 12-item scale that measured participants' attitude toward disease and its prevention method in five categories, from strongly agree to strongly disagree [34]. Participants' knowledge, awareness, and attitude toward PrEP were assessed using a 10-question survey developed from the RADAR cohort where participants responded in four categories from strongly disagree to strongly agree. The scale is divided into 3 questions surveying positive attitude, with a score ranging from 3 to 12, and 7 questions to PrEP associated stigma, with a score ranging from 7 to 28 [35].
Statistical Analysis
We analyzed data collected from November 2018 to April 2023. We conducted descriptive statistics summarizing categorical variables using frequencies and continuous variables using means and standard deviations, if normally distributed, or medians and range, if non-normal. Differences in proportions of categorical variables pre- and post-COVID-19 onset were tested by Pearson's χ2 test and Fisher's exact test when expected cell frequencies were less than five. Means and medians were compared by paired t-tests and repeated measures ANOVA for repeating individual data. Group-level differences were compared by the Wilcoxon rank sum, or the Kruskal–Wallis test was used for continuous variables. For the serial cross-sectional analysis, unadjusted and multiple linear regression was performed to estimate the association between COVID-19 onset and anxiety or depression. Predictors that were identified as confounders were included in the multiple linear regression. To assess potential selection bias, we compared participants enrolled in both cohorts to the entire sample. For participants who were sampled in both S2PrEP2 and S2PrEP3 with repeated measures, we utilized generalized linear mixed models to estimate the effect of COVID-19 onset on anxiety and depression to account for non-independent measures. Fixed effects included study group, insurance status, and IPV score, and number of study enrollment as repeated factor. To address potential selection bias in our findings, we examined selection bias by comparing the characteristics of entire sample with the subgroup enrolled in both studies, allowing for insights into potential differences between these groups. Although our data collection approach minimized missing data, we further assessed potential bias due to missingness. All analyses were performed using R 4.2.1 and RStudio 2022.02.03.
RESULTS
There were 144 participants in Stick2PrEP2 (pre-COVID-19) and 319 in Stick2PrEP3 (post-COVID-19), among whom 34 participated in both studies and were included in the triangulation analysis. Table 1 shows the distribution in the sample enrolled in each study and those who participated in both. Among participants enrolled in one study, the median age was 29 (interquartile range [IQR]: 25, 34) years, 96% self-identified as male, and 1.5% as TW. Most of the samples identified themselves as non-Hispanic white (42%), followed by Hispanic (35%), non-Hispanic Black (14%) and people identifying as any other racial-ethnic identity (9%). Exactly 50% were working part-time, 56% reported an annual income below $40 000, 35% had private insurance, 27% had Medicaid as public insurance, and 11% had other types of health insurance. Several demographic variables differed between the two studies (ie, S2PrEP2 vs S2PrEP3). Racial and ethnic self-identity differed between the pre- and post-pandemic cohorts (P = .002). Hispanic participants declined from 49% to 29%, while non-Hispanic White participants increased from 31% to 46%, people identified as other race from 5.5% to 10%, and Black or African American was 15% and 14%. Approximately 69% of the post-COVID group had a bachelor's degree or higher, compared with 56% in the pre-COVID group (P = .014). More participants earned an annual income of $40 000 or more in the post-COVID group (48% vs 35%, P = .031). Among the participants enrolled in both studies, the distribution of most sociodemographic variables did not differ by study participation. A higher proportion of participants were on Medicaid or uninsured in the post-COVID sample (8.8% vs 21%, 5.9% vs 24%, P = .003).
Table 1.
Baseline Demographics Characteristics of Participants by Number of Studies Enrolled
| Characteristics | One Study Only | Both Studies | ||||||
|---|---|---|---|---|---|---|---|---|
| Overall (N = 395)a | Stick2PrEP 2.0 (N = 110)a | Stick2PrEP 3.0 (N = 285)a | P Valueb | Overall (N = 68)a | Stick2PrEP 2.0 (N = 34)a | Stick2PrEP 3.0 (N = 34)a | P Valueb | |
| Gender | >.9 | >.9 | ||||||
| Male | 380 (96%) | 107 (97%) | 273 (96%) | 68 (100%) | 34 (100%) | 34 (100%) | ||
| Transwoman | 6 (1.5%) | 1 (0.9%) | 5 (1.8%) | 0 (0%) | 0 (0%) | 0 (0%) | ||
| Other | 9 (2.3%) | 2 (1.8%) | 7 (2.5%) | 0 (0%) | 0 (0%) | 0 (0%) | ||
| Race and ethnicity | .002 | >.9 | ||||||
| Non-Hispanic White | 165 (42%) | 34 (31%) | 131 (46%) | 34 (50%) | 18 (53%) | 16 (47%) | ||
| Hispanic | 137 (35%) | 54 (49%) | 83 (29%) | 20 (29%) | 10 (29%) | 10 (29%) | ||
| Non-Hispanic Black of African American | 57 (14%) | 16 (15%) | 41 (14%) | 9 (13%) | 4 (12%) | 5 (15%) | ||
| Other | 35 (9%) | 6 (6%) | 29 (10%) | 5 (7%) | 2 (6%) | 3 (9%) | ||
| Age (y) | 29 (25, 34) | 29 (24, 33) | 29 (25, 34) | .3 | 32 (27, 35) | 30 (26, 34) | 32 (29, 36) | .065 |
| Income | .031 | .15 | ||||||
| <$40 000 | 219 (56%) | 71 (65%) | 148 (52%) | 34 (50%) | 20 (59%) | 14 (41%) | ||
| ≥$40 000 | 173 (44%) | 39 (35%) | 134 (48%) | 34 (50%) | 14 (41%) | 20 (59%) | ||
| Bachelor's degree or higher | 260 (66%) | 62 (56%) | 198 (69%) | .014 | 46 (68%) | 23 (68%) | 23 (68%) | >.9 |
| Employment status | >.9 | .3 | ||||||
| Working full-time | 91 (23%) | 25 (23%) | 66 (23%) | 12 (18%) | 4 (12%) | 8 (24%) | ||
| Working part-time | 196 (50%) | 55 (50%) | 141 (49%) | 36 (53%) | 18 (53%) | 18 (53%) | ||
| Other | 108 (27%) | 30 (27%) | 78 (27%) | 20 (29%) | 12 (35%) | 8 (24%) | ||
| Insurance | <.001 | .003 | ||||||
| Private health insurance or HMO | 138 (35%) | 39 (35%) | 99 (35%) | 29 (43%) | 14 (41%) | 15 (44%) | ||
| Medicaid | 106 (27%) | 20 (18%) | 86 (30%) | 10 (15%) | 3 (8.8%) | 7 (21%) | ||
| Other | 42 (11%) | 9 (8.2%) | 33 (12%) | 6 (8.8%) | 3 (8.8%) | 3 (8.8%) | ||
| Don't know | 48 (12%) | 39 (35%) | 9 (3.2%) | 13 (19%) | 12 (35%) | 1 (2.9%) | ||
| None | 61 (15%) | 3 (2.7%) | 58 (20%) | 10 (15%) | 2 (5.9%) | 8 (24%) | ||
This table shows the demographic features that the participants reported in the enrollment survey, including gender, race and ethnicity, age, level of education, employment status, and insurance at the time of the survey. The results were reported separately by the two sample groups, those who enrolled in one study and those who enrolled in both. Bold values are significant at α = .05.
a N (%); median (IQR).
bPearson's χ2 test; Fisher's exact test; Wilcoxon rank sum test.
Table 2 shows the distribution of the mental health outcomes and other metrics of HIV vulnerability across the two periods of COVID-19 for participants in either or both studies. Participants had higher PHQ-9 and GAD-7 mean scores in the post-COVID-19 period compared with the pre-COVID-19 period (3.6 ± 4.2 vs 5.7 ± 5.3, P < .001; 3.9 ± 4.3 vs 5.1 ± 5.0, respectively; P = .019). Other metrics related to vulnerability were also higher after COVID-19, including the HIRI (18 vs 21, P = .038), HIVKQ18 (14 vs 15, P < .001), and RADAR Stigma score (14 vs 12, P = .058). Participants reported overall lower self-efficacy (ASES, 8.95 vs 8.29, P = .004) and reported an increase in IPV victimization (24% vs 45% P = .003). Participants reported increased engagement in HIV prevention services in the later time period (index score >45, 64% vs 80%, P = .01). Reported rates of adverse childhood events, HIV stigma, and STI risk scores were similar. Similar trends were observed among the 34 participants who enrolled in both studies (Supplementary Table 1). A higher mean score was observed in both PHQ-9 (2.9 vs 5.2, P = .007) and GAD-7 (3.0 vs 4.8, P = .037), so did HIVKQ18 (14 vs 16, P = .018) and HIV prevention services engagement (76% vs 97%, P = .026). The ASES score, on the other hand, was not significant between the two groups (P = .6).
Table 2.
Baseline Screening and Measuring Instruments of Participants Enrolled in One Study
| Characteristics | One Study Only | ||||
|---|---|---|---|---|---|
| N | Overall (N = 395)a | Stick2PrEP 2.0 (N = 110)a | Stick2PrEP 3.0 (N = 285)a | P Valueb | |
| PHQ-9 score | 395 | 5.1(5.1) | 3.6(4.2) | 5.7(5.3) | <.001 |
| PHQ-9 score categorical | 395 | .01 | |||
| <5 | … | 214 (54%) | 70 (64%) | 144 (51%) | |
| 5–9 | … | 110 (28%) | 30 (27%) | 80 (28%) | |
| >9 | … | 71 (18%) | 10 (9.1%) | 61 (21%) | |
| GAD-7 score | 395 | 4.8(4.8) | 3.9(4.3) | 5.1(5.0) | .019 |
| GAD-7 score categorical | 395 | 0.2 | |||
| <5 | … | 224 (57%) | 69 (63%) | 155 (54%) | |
| 5–9 | … | 111 (28%) | 29 (26%) | 82 (29%) | |
| >9 | … | 60 (15%) | 12 (11%) | 48 (17%) | |
| HIV stigma | 395 | 42 (36, 48) | 42 (36, 49) | 42 (36, 48) | .7 |
| HIRI score | 340 | 21 (15, 25) | 18 (15, 22) | 21 (15, 26) | .038 |
| HIRI score ≥ 10 | 340 | 299 (88%) | 48 (87%) | 251 (88%) | .9 |
| IWantTheKit risk score | 340 | 7 (5, 8) | 6.00 (5, 8) | 7.00 (5, 8) | .2 |
| IWantTheKit risk score categorical | 340 | .3 | |||
| 0–2 | … | 17 (5.0%) | 5 (9.1%) | 12 (4.2%) | |
| 3–6 | … | 145 (43%) | 24 (44%) | 121 (42%) | |
| >6 | … | 178 (52%) | 26 (47%) | 152 (53%) | |
| Intimate partner violence—victim | 337 | 141 (42%) | 13 (24%) | 128 (45%) | .003 |
| Intimate partner violence—perpetrator | 337 | 62 (18%) | 10 (18%) | 52 (18%) | >.9 |
| HIVKQ18 | 340 | 15 (13, 16) | 14 (12, 15) | 15 (13, 17) | .001 |
| RADAR—positive attitude | 358 | 11 (9, 12) | 11 (9, 12) | 11 (9, 12) | .8 |
| RADAR—stigma | 358 | 13 (11, 15) | 14 (11, 15) | 12 (11, 15) | .058 |
| ACE score | 395 | 1 (0, 3) | 1 (0, 2) | 1 (0, 3) | .5 |
| ACE score categorical | 395 | .6 | |||
| 0 | … | 139 (35%) | 41 (37%) | 98 (34%) | |
| ≥1 | … | 256 (65%) | 69 (63%) | 187 (66%) | |
| ASES score | 333 | 8.40 (1.70) | 8.95 (1.43) | 8.29 (1.73) | .004 |
| Index score | 339 | 52.0 (46.0, 55.0) | 51.0 (44.5, 55.0) | 52.0 (47.0, 55.0) | .2 |
| Index score categorical | 339 | .01 | |||
| ≤45 | … | 78 (23%) | 20 (36%) | 58 (20%) | |
| >45 | … | 261 (77%) | 35 (64%) | 226 (80%) | |
Abbreviations: ACE, Adverse Childhood Experiences; ASES, Adherence Self-Efficacy Scale; GAD-7, Generalized Anxiety Disorder Scale; PHQ,-9, Patient Health Questionnaire.
This table presents the distribution of the mental health outcomes and other metrics of HIV vulnerability before and after the onset of COVID-19 in the cross-sectional cohort. Bold values are significant at α = .05.
a N (%); median (IQR); mean (SD).
bWilcoxon rank sum test; Fisher's exact test; Pearson's χ2 test.
Differences between the mean scores of PHQ-9 and GAD-7 by time periods and other HIV vulnerabilities are reported in Table 3. The mean scores of PHQ-9 and GAD-7 were higher in participants who experienced IPV (4.48 vs 6.47, P < .001; 3.89 vs 6.16, P < .001) and who inflicted violent behaviors (4.98 vs 7.18, P = .01; 4.49 vs 6.37, P = .02). In Supplementary Table 2, PHQ-9 mean scores were different in mean values among participants covered by various insurance plans. Mean PHQ-9 scores were the highest among participants without health insurance (P = .01), compared with others covered by private insurance, Medicaid, or unlisted insurance plans.
Table 3.
Mean Differences Comparison of PHQ-9 and GAD-7 of Participants Enrolled in One Study
| Characteristics | One Study Only | |||
|---|---|---|---|---|
| PHQ-9 | P Valuea | GAD-7 | P Valuea | |
| Study | <.001 | .01 | ||
| Stick2PrEP 2.0 | 3.62 (4.17) | 3.85 (4.27) | ||
| Stick2PrEP 3.0 | 5.69 (5.29) | 5.10 (5.01) | ||
| Gender | .29 | .3 | ||
| Male | 5.07 (5.12) | 4.70 (4.87) | ||
| Transwoman | 4.17 (3.13) | 4.50 (3.39) | ||
| Other | 7.67 (4.64) | 7.22 (4.32) | ||
| Race and ethnicity | .52 | .16 | ||
| White | 5.21 (4.72) | 4.78 (4.80) | ||
| Hispanic | 5.08 (5.79) | 4.93 (5.23) | ||
| Black or African American | 4.44 (4.60) | 3.65 (3.85) | ||
| Other | 6.06 (4.62) | 5.91 (4.83) | ||
| Income | .71 | .16 | ||
| <$40 000 | 5.23 (5.21) | 4.47 (4.58) | ||
| ≥$40 000 | 5.03 (4.97) | 5.17 (5.17) | ||
| Bachelor's degree or higher | .3 | .12 | ||
| No | 5.48 (5.04) | 5.30 (5.02) | ||
| Yes | 4.93 (5.11) | 4.47 (4.74) | ||
| Employment status | .37 | .83 | ||
| Working full-time | 5.71 (5.38) | 4.82 (4.86) | ||
| Working part-time | 4.81 (5.19) | 4.86 (5.18) | ||
| Other | 5.17 (4.64) | 4.51 (4.19) | ||
| Insurance | .48 | .12 | ||
| Private health insurance or HMO | 5.06 (5.15) | 4.94 (4.91) | ||
| Medicaid | 5.20 (5.08) | 4.26 (4.38) | ||
| Other | 6.05 (6.35) | 6.12 (5.96) | ||
| Don’t know | 4.08 (4.03) | 3.69 (4.17) | ||
| None | 5.28 (4.75) | 5.08 (4.97) | ||
| HIRI score ≥10 | .67 | .16 | ||
| No | 5.00 (5.68) | 3.95 (4.38) | ||
| Yes | 5.40 (5.18) | 5.00 (5.01) | ||
| IWantTheKit risk score categorical | .21 | .28 | ||
| 0–2 | 3.29 (4.55) | 3.24 (4.37) | ||
| 3–6 | 5.28 (5.24) | 4.72 (4.99) | ||
| >6 | 5.61 (5.27) | 5.15 (4.96) | ||
| Intimate partner violence—victim | <.001 | <.001 | ||
| No | 4.48 (4.67) | 3.89 (4.21) | ||
| Yes | 6.47 (5.49) | 6.16 (5.43) | ||
| Intimate partner violence—perpetrator | .01 | .02 | ||
| No | 4.89 (4.96) | 4.49 (4.62) | ||
| Yes | 7.18 (6.36) | 6.37 (5.70) | ||
| ACE score categorical | <.001 | <.001 | ||
| 0 | 3.91 (4.36) | 3.37 (4.23) | ||
| ≥1 | 5.77 (5.34) | 5.51 (5.00) | ||
| Index score categorical | .17 | .47 | ||
| ≤45 | 6.03 (4.78) | 5.22 (4.60) | ||
| >45 | 5.15 (5.37) | 4.78 (5.06) | ||
Abbreviations: ACE, Adverse Childhood Experiences; GAD-7, Generalized Anxiety Disorder Scale; PHQ,-9, Patient Health Questionnaire.
a t-Test.
This table shows the differences between the mean values of PHQ-9 and GAD-7 comparing two by time periods and other HIV vulnerabilities in the cross-sectional sample by t-test. Bold values are significant at α = .05.
Table 4 shows unadjusted and adjusted correlates of PHQ-9 and GAD-7 mean scores from linear regression analyses of the cross-sectional samples. Before adjustment, COVID-19 period was associated with a 2.08 (95% CI: .97, 3.18, P < .001) mean increase in PHQ-9 score. Multiple linear regression showed that after adjustment, COVID-19 period was associated with a 1.55 (95% CI: .07, 3.03, P = .04) mean increase in PHQ-9 score. IPV was associated with a 1.78 (P < .001) mean increase, but ASES score was no longer associated with PHQ-9 mean score (P = .19). Before adjustment for potentially confounding variables, COVID-19 period was associated with a 1.25 (95% CI: .18, 2.31, P = .02) mean increase in GAD-7 score. After adjustment, COVID-19 period association reduced to a 0.84 mean increase and was no longer associated with GAD-7 score. IPV experience was associated with a 2.12 (P < .001) mean increase in GAD-7 score, and ASES score was no longer associated (P = .42).
Table 4.
Unadjusted and Adjusted Linear Regression Models of Participants Enrolled in One Study or Two Studies
| Characteristic | PHQ-9 1 Study | GAD-7 1 Study | ||||
|---|---|---|---|---|---|---|
| Beta | 95% CI | P Value | Beta | 95% CI | P Value | |
| Crude | ||||||
| Study | ||||||
| s2p2 | … | … | … | … | ||
| s2p3 | 2.08 | .97, 3.18 | <.001 | 1.25 | .18, 2.31 | .02 |
| Adjusted | ||||||
| Study | ||||||
| s2p2 | … | … | … | … | ||
| s2p3 | 1.55 | .07, 3.03 | .04 | .84 | −.57, 2.25 | .24 |
| Intimate partner violence—victim | ||||||
| No | … | … | … | … | ||
| Yes | 1.78 | .67, 2.89 | <.001 | 2.12 | 1.07, 3.18 | <.001 |
| ASES score | −.21 | −.54, .11 | .19 | −.13 | −.43, .18 | .42 |
| Characteristic | PHQ-9 2 Studies | GAD-7 2 Studies | ||||
|---|---|---|---|---|---|---|
| Beta | 95% CI | P Value | Beta | 95% CI | P Value | |
| Crude | ||||||
| Study | ||||||
| s2p2 | … | … | … | … | ||
| s2p3 | 2.35 | 1.19, 3.52 | <.001 | 1.76 | .37, 3.16 | .02 |
| Adjusted | ||||||
| Study | ||||||
| s2p2 | … | … | … | … | ||
| s2p3 | 2.40 | .42, 1.97 | .01 | 2.11 | −.02, 4.23 | .06 |
| Insurance | ||||||
| Private health insurance or HMO | … | … | … | … | ||
| Medicaid | 2.59 | −4.59, −1.08 | .09 | … | … | |
| Other | 1.63 | −2.13, 1.75 | .37 | … | … | |
| Don’t know | 3.92 | −3.80, 1.46 | .01 | … | … | |
| None | 5.80 | −.86, 3.12 | <.001 | … | … | |
| Intimate partner violence—victim | ||||||
| No | … | … | … | … | ||
| Yes | 4.20 | .88, 3.40 | .003 | 3.06 | .71, 5.40 | .02 |
Abbreviations: CI, confidence interval; GAD-7, Generalized Anxiety Disorder Scale; PHQ,-9, Patient Health Questionnaire.
This table shows unadjusted and adjusted correlates of PHQ-9 and GAD-7 from linear regression analyses of the cross-sectional samples and repeated analyses of the longitudinal samples. Bold values are significant at α = .05.
In the repeated measures analysis of participants in both studies, COVID-19 period was associated with a 2.35 (95% CI: 1.19, 3.52, P < .001) mean increase in PHQ-9 score before adjustment. After adjustment, COVID-19 period was associated with a 2.40 (95% CI: .42, 1.97, P = .01) mean increase in PHQ-9 score. Compared with private insurance, insurance through Medicaid, other forms of insurance, unsure insurance status, or no insurance was associated with mean increases in PHQ-9 scores of 2.59, 1.63, 3.92, and 5.80 (P = .09, P = .37, P = .01, P < .001), respectively. Participants who experienced IPV had a 4.20 (P = .003) mean increase in PHQ-9 score. Mean increases in GAD-7 score changed from 1.76 (95% CI: .37, 3.16, P = .02) before adjustment to 2.11 (95% CI: −.02, 4.23, P = .06) after adjustment. Participants reporting IPV had a 3.06 (P = .02) mean increase in GAD-7 score.
DISCUSSION
This sample of MSM and TW attending HIV prevention services exhibited higher PHQ-9 and GAD-7 mean scores after the onset of COVID-19 compared with the pre-COVID-19 sample. This trend was observed both among participants enrolled in only one study and those with repeated measures across both studies. Even after adjusting for potential confounding variables such as insurance status, IPV experience, and adherence self-efficacy, the association between COVID-19 period and elevated depression or anxiety persisted. These findings are consistent with broader literature reporting increased mental health challenges during the pandemic, particularly among individuals receiving HIV prevention or treatment services [36–38].
Mental health disparities have long been documented among MSM and TW populations, with elevated rates of anxiety, depression, and substance use compared with the general population [39, 40]. These mental health challenges are often attributed to minority stress, stigma, discrimination, and other structural factors. The pandemic appears to have amplified these vulnerabilities through mechanisms such as social isolation, economic uncertainty, and disrupted health services. Other studies have also highlighted the unique burden of COVID-19 on individuals engaged in sexual health services, including increased stress, disruptions in care, and heightened interpersonal violence, all of which exacerbate mental health concerns and potentially increase HIV vulnerability [41, 42]. The observed rise in anxiety and depression rates should therefore be understood in the context of these intersecting factors rather than attributed solely to COVID-19. The combined impact of the mental health crisis and the COVID-19 pandemic, driven by underlying social inequities, disproportionately affects minority groups. These intersecting challenges are best understood through a syndemic framework, which highlights how co-occurring health and social issues compound one another and worsen outcomes [43].
The study's finding of a shifting sociodemographic profile post-COVID, with a higher proportion of white participants, those earning more than $40 000 annually, and individuals with at least a bachelor's degree, raises important questions about access to sexual health services. These shifts could indicate systemic barriers for marginalized groups, such as loss of health insurance or employment during the economic downturn, potentially driving disparities in who remains engaged in care. The association between changes in insurance coverage and increased depression or anxiety further underscores the role of economic instability in shaping mental health outcomes. While the higher HIV engagement service score in the post-COVID cohort might suggest better engagement among participants, it could also reflect the loss of patients most likely to disengage from care, necessitating further investigation.
Despite these insights, the lack of significant differences in STI vulnerability (IWantTheKit), PrEP attitude (RADAR), and adverse childhood events suggests that certain confounding factors were ruled out. However, these findings warrant deeper exploration to fully understand how such variables influence mental health outcomes and HIV vulnerability. For instance, while IPV experience increased during the pandemic, it did not fully explain the observed rise in depression, indicating the multifactorial nature of these trends. Future studies should examine these relationships more comprehensively to identify specific pathways linking structural determinants, mental health, and HIV risk.
This study's limitations include its cross-sectional design, which restricted the ability to infer causality. To address this, participants with repeated measures across pre- and post-COVID periods were included. Though the limited sample size precluded robust statistical analyses, the consistency in the findings between the serial cross-sectional analysis in the directions of results with the scales across the broader and repeated measures samples suggests that our serial cross-sectional analysis reliably approximated a longitudinal study. Additionally, the potential impact of neurological effects from COVID-19 infection on mental health metrics was not examined, representing another limitation. Although all participants were assigned male sex at birth, the majority of the study cohort self-identified as male, with only a small number of TW included. As a result, conclusions related to TW are limited by the small sample size. Finally, we lack detailed data on pre- and post-COVID clinic enrollment to determine whether the study population changes reflect broader trends within the clinic.
These findings underscore the critical need to integrate mental health services into comprehensive sexual health programs, particularly during times of public health emergency. Long-term follow-up on the mental health trajectories of individuals engaged in these programs is essential to understand changes over time and to identify interventions that mitigate adverse outcomes as a consequence of syndemics. Future pandemic planning must prioritize mental health, especially for marginalized and minoritized populations at heightened risk like MSM and TW. A feasible alternative to in-person clinical services is virtual health care, which provides accessible mental health support with the flexibility to engage in care regardless of geographic location. This approach becomes especially valuable when traditional services are disrupted during public health emergencies. Lessons from the COVID-19 pandemic, along with ongoing efforts to strengthen the health care system, can help build greater resilience for future crises.
Supplementary Material
Notes
Author Contributions. J. Z., D. C., C. C., E. L., D. T., P. G., A. C., K. M., and M. E. S. conceptualized the project. J. Z., E. L., J. K., and T. Y. K. collected the data. S. H., D. Q., D. C., and J. Z. performed data analysis. S. H., D. C., and J. Z. drafted the manuscript. All authors contributed to reviewing the manuscript and have read and agreed to the published version of the manuscript.
Patient consent. All participants provided written consent to the study. The study was approved by the Columbia University Institutional Review Board (IRB).
Data availability. Data were not publicly available.
Financial support. The research was supported by the National Institutes of Health [K23AI150378 (J. Z.), UM1AI069470 (J. Z.)].
Contributor Information
Simian Huang, Division of Infectious Diseases, Department of Medicine, Columbia University Irving Medical Center, New York, New York, USA.
Jason Zucker, Division of Infectious Diseases, Department of Medicine, Columbia University Irving Medical Center, New York, New York, USA.
Delivette Castor, Division of Infectious Diseases, Department of Medicine, Columbia University Irving Medical Center, New York, New York, USA.
Caroline Carnevale, HIV Prevention Program, New York Presbyterian Hospital, New York, New York, USA.
Elijah LaSota, Division of Infectious Diseases, Department of Medicine, Columbia University Irving Medical Center, New York, New York, USA.
Joshua Klein, New Presbyterian Hospital, New York, New York, USA.
Tae Yoon Kim, Division of Infectious Diseases, Department of Medicine, Columbia University Irving Medical Center, New York, New York, USA.
Daniela Quigee, Division of Infectious Diseases, Department of Medicine, Columbia University Irving Medical Center, New York, New York, USA.
Deborah Theodore, Division of Infectious Diseases, Department of Medicine, Columbia University Irving Medical Center, New York, New York, USA.
Peter Gordon, Division of Infectious Diseases, Department of Medicine, Columbia University Irving Medical Center, New York, New York, USA.
Alwyn Cohall, Department of Pediatrics, Columbia University Irving Medical Center, New York, New York, USA.
Kathrine Meyers, Division of Infectious Diseases, Department of Medicine, Columbia University Irving Medical Center, New York, New York, USA.
Magdalena E Sobieszczyk, Division of Infectious Diseases, Department of Medicine, Columbia University Irving Medical Center, New York, New York, USA.
Supplementary Data
Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
References
- 1. Santomauro DF, Mantilla Herrera AM, Shadid J, et al. Global prevalence and burden of depressive and anxiety disorders in 204 countries and territories in 2020 due to the COVID-19 pandemic. Lancet 2021; 398:1700–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Tordoff DM, Wanta JW, Collin A, Stepney C, Inwards-Breland DJ, Ahrens K. Mental health outcomes in transgender and nonbinary youths receiving gender-affirming care. JAMA Netw Open 2022; 5:e220978. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Tabler J, Schmitz RM, Charak R, Propst A. Forgone care among LGBTQ and non-LGBTQ Americans during the COVID-19 pandemic: the role of health, social support, and pandemic-related stress. South Med J 2022; 115:752–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Chen S, Wang Y, She R, Qin P, Ming WK. Disparities in the unmet mental health needs between LGBTQ+ and non-LGBTQ+ populations during COVID-19 in the United States from 21 July 2021 to 9 May 2022. Front Med 2022; 9:995466. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Lima V, Lucena B. (221) Impacts of COVID 19 on the mental health of LGBTQIA+ people: an integrative review. J Sex Med 2023; 20(Suppl 1):qdad060.211. [Google Scholar]
- 6. Office of the Surgeon General (OSG) . Protecting youth mental health: the US Surgeon General's Advisory. Washington, DC: US Department of Health and Human Services, 2021. [PubMed] [Google Scholar]
- 7. Feinstein BA, Goldfried MR, Davila J. The relationship between experiences of discrimination and mental health among lesbians and gay men: an examination of internalized homonegativity and rejection sensitivity as potential mechanisms. J Consult Clin Psychol 2012; 80:917–27. [DOI] [PubMed] [Google Scholar]
- 8. Frost DM, Meyer IH. Minority stress theory: application, critique, and continued relevance. Curr Opin Psychol 2023; 51:101579. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. The Trevor Project . 2023 U.S. National Survey on the mental health of LGBTQ young people. 2023. https://www.thetrevorproject.org/survey-2023/. Accessed 3 December 2024.
- 10. NAMI . LGBTQ+. 2024. https://www.nami.org/your-journey/identity-and-cultural-dimensions/lgbtq/. Accessed 3 December 2024.
- 11. ABES. CDC . Sexual identity. 2024. https://www.cdc.gov/abes/results/sexual-identity.html. Accessed 3 December 2024.
- 12. Price-Feeney M, Green AE, Dorison S. Understanding the mental health of transgender and nonbinary youth. J Adolesc Health 2020; 66:684–90. [DOI] [PubMed] [Google Scholar]
- 13. Santos GM, Ackerman B, Rao A, et al. Economic, mental health, HIV prevention and HIV treatment impacts of COVID-19 and the COVID-19 response on a global sample of cisgender gay men and other men who have sex with men. AIDS Behav 2021; 25:311–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Kumar N, Janmohamed K, Nyhan K, et al. Sexual health (excluding reproductive health, intimate partner violence and gender-based violence) and COVID-19: a scoping review. Sex Transm Infect 2021; 97:402–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Sanchez TH, Zlotorzynska M, Rai M, Baral SD. Characterizing the impact of COVID-19 on men who have sex with men across the United States in April, 2020. AIDS Behav 2020; 24:2024–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Ballivian J, Alcaide ML, Cecchini D, Jones DL, Abbamonte JM, Cassetti I. Impact of COVID-19-related stress and lockdown on mental health among people living with HIV in Argentina. J Acquir Immune Defic Syndr 2020; 85:475–82. [DOI] [PubMed] [Google Scholar]
- 17. Pizzirusso M, Carrion-Park C, Clark US, Gonzalez J, Byrd D, Morgello S. Physical and mental health screening in a New York City HIV cohort during the COVID-19 pandemic: a preliminary report. J Acquir Immune Defic Syndr 2021; 86:e54–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Jones DL, Ballivian J, Rodriguez VJ, et al. Mental health, coping, and social support among people living with HIV in the Americas: a comparative study between Argentina and the USA during the SARS-CoV-2 pandemic. AIDS Behav 2021; 25:2391–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Gonzales G, Przedworski J, Henning-Smith C. Comparison of health and health risk factors between lesbian, gay, and bisexual adults and heterosexual adults in the United States: results from the National Health Interview Survey. JAMA Intern Med 2016; 176:1344–51. [DOI] [PubMed] [Google Scholar]
- 20. HIV.gov . Seeing your health care provider. 2024. https://www.hiv.gov/hiv-basics/staying-in-hiv-care/provider-visits-and-lab-test/seeing-your-health-care-provider. Accessed 17 July 2024.
- 21. Nagendra G, Carnevale C, Neu N, Cohall A, Zucker J. The potential impact and availability of sexual health services during the COVID-19 pandemic. Sex Transm Dis 2020; 47:434–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Lee JJ, Katz DA, Kerani RP, Lerner JE, Baral SD, Sanchez TH. Physical violence and psychological distress among Asian and Pacific Islander sexual minority men in the United States before and during the COVID-19 pandemic. LGBT Health 2022; 9:418–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Hong C, Queiroz A, Hoskin J. The impact of the COVID-19 pandemic on mental health, associated factors and coping strategies in people living with HIV: a scoping review. J Int AIDS Soc 2023; 26:e26060. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. STROBE . Checklists. 2024. https://www.strobe-statement.org/checklists/. Accessed 14 November 2024.
- 25. Kroenke K, Spitzer RL, Williams JBW. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med 2001; 16:606–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Spitzer RL, Kroenke K, Williams JBW, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med 2006; 166:1092–7. [DOI] [PubMed] [Google Scholar]
- 27. Wilton J, Mishra S, Tan DHS. Considerations for using the HIRI-MSM screening tool to identify MSM who would benefit most from PrEP. J Acquir Immune Defic Syndr 2017; 76:e58–61. [DOI] [PubMed] [Google Scholar]
- 28. Gaydos CA, Barnes M, Aumakhan B, et al. Can e-technology through the Internet be used as a new tool to address the Chlamydia trachomatis epidemic by home sampling and vaginal swabs? Sex Transm Dis 2009; 36:577–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Carey MP, Schroder KEE. Development and psychometric evaluation of the brief HIV knowledge questionnaire. AIDS Educ Prev 2002; 14:172–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Golub SA, Starbuck L, Fikslin R, Gamarel KE. Psychometric evaluation and predictive validity of an adapted adherence self-efficacy scale for PrEP. AIDS Behav 2023; 27:218–30. [DOI] [PubMed] [Google Scholar]
- 31. Meyers K, Quigee D, Zucker J, et al. The index of engagement in PrEP care: evaluation of psychometric properties and predictive potential. J Acquir Immune Defic Syndr 2023; 94:214–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Felitti VJ, Anda RF, Nordenberg D, et al. Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults: the adverse childhood experiences (ACE) study. Am J Prev Med 1998; 14:245–58. [DOI] [PubMed] [Google Scholar]
- 33. Miltz AR, Lampe FC, Bacchus LJ, et al. Intimate partner violence, depression, and sexual behaviour among gay, bisexual and other men who have sex with men in the PROUD trial. BMC Public Health 2019; 19:431. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Sayles JN, Hays RD, Sarkisian CA, Mahajan AP, Spritzer KL, Cunningham WE. Development and psychometric assessment of a multidimensional measure of internalized HIV stigma in a sample of HIV-positive adults. AIDS Behav 2008; 12:748–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Mustanski B, Ryan DT, Hayford C, Phillips G, Newcomb ME, Smith JD. Geographic and individual associations with PrEP stigma: results from the RADAR cohort of diverse young men who have sex with men and transgender women. AIDS Behav 2018; 22:3044–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Javanbakht M, Rosen A, Ragsdale A, Richter EI, Shoptaw S, Gorbach PM. Interruptions in mental health care, cannabis use, depression, and anxiety during the COVID-19 pandemic: findings from a cohort of HIV-positive and HIV-negative MSM in Los Angeles, California. J Urban Health 2022; 99:305–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Kuman Tunçel Ö, Pullukçu H, Erdem HA, Kurtaran B, Taşbakan SE, Işikgöz Taşbakan M. COVID-19-related anxiety in people living with HIV: an online cross-sectional study. Turk J Med Sci 2020; 50:1792–800. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Marbaniang I, Sangle S, Nimkar S, et al. The burden of anxiety among people living with HIV during the COVID-19 pandemic in Pune, India. BMC Public Health 2020; 20:1598. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Mimiaga MJ, O’Cleirigh C, Biello KB, et al. The effect of psychosocial syndemic production on 4-year HIV incidence and risk behavior in a large cohort of sexually active men who have sex with men. J Acquir Immune Defic Syndr 2015; 68:329–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Mimiaga MJ, Hughto JMW, Klasko-Foster L, et al. Substance use, mental health problems, and physical and sexual violence additively increase HIV risk between male sex workers and their male clients in Northeastern United States. J Acquir Immune Defic Syndr 2021; 86:305–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Auty SG, Aswani MS, Wahbi RN, Griffith KN. Changes in health care access by race, income, and Medicaid expansion during the COVID-19 pandemic. Med Care 2023; 61:45–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Bundorf MK, Gupta S, Kim C. Trends in US health insurance coverage during the COVID-19 pandemic. JAMA Health Forum 2021; 2:E212487. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Singer M, Bulled N, Ostrach B, Lerman Ginzburg S. Syndemics: a cross-disciplinary approach to complex epidemic events like COVID-19. Annu Rev Anthropol 2021; 50:41–58. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.