Abstract
Objectives. To examine linear and nonlinear associations between psychosocial syndemic factors and HIV risk and engagement in HIV prevention care among sexual and gender minority (SGM) youths.
Methods. Between February and October 2022, we recruited 17 578 SGM youths aged 13 to 18 years in the United States for an online survey. We examined the relationship of syndemics (i.e., binge drinking, drug use, sexual victimization, and anti–lesbian, gay, bisexual, and transgender discrimination) with sexual behaviors (i.e., sexual initiation, condomless anal or vaginal sex, and number of sexual partners) and HIV prevention care (i.e., HIV testing, preexposure prophylaxis awareness and utilization) using regression.
Results. Psychosocial syndemic burden (number of syndemic factors reported) was linearly and cubically associated with engagement in sexual behaviors. Psychosocial syndemic burden was linearly associated with higher HIV testing and preexposure prophylaxis awareness and cubically associated with higher preexposure prophylaxis utilization.
Conclusions. Our findings are evidence of synergism across psychosocial syndemic factors regarding HIV risk and engagement in HIV prevention care among SGM youths in the United States.
Public Health Implications. Multicomponent interventions may help reduce HIV risk and promote access to HIV prevention services among SGM individuals aged 13 to 18 years. (Am J Public Health. 2024;114(9):892–902. https://doi.org/10.2105/AJPH.2024.307753)
Sexual and gender minority (SGM) youths aged 13 to 24 years bear a disproportionate burden of the HIV epidemic, accounting for more than 80% of new HIV diagnoses in their age group.1 Behaviors associated with HIV transmission, including condomless sex, multiple sexual partners, and sexualized drug use, are more prevalent among SGM youths than among their heterosexual counterparts, contributing to heightened HIV risk among this population.2,3 Suboptimal engagement in HIV prevention care among SGM youths also contributes to continued HIV transmission.4 In particular, preexposure prophylaxis (PrEP) uptake among youths younger than 18 years is low, with only 1.1% of commercially insured individuals in the United States with PrEP indications having received a prescription.5
Increased HIV risk and lower engagement in care cooccur with psychosocial problems, commonly referred to as syndemics. Initially developed to explain interrelationships between substance use, violence, and HIV risk,6 the concept of syndemics has been expanded to include factors such as depression,7,8 alcohol use,8,9 stigma,10 and poverty.7 There is robust evidence that several syndemic factors are more prevalent among SGM youths than among their heterosexual counterparts.3,11–13 Mounting evidence also shows that syndemic factors are associated with behaviors that may lead to HIV acquisition (e.g., condomless sex),9,14 increased HIV and sexually transmitted infection transmission,8 and poor engagement in HIV treatment and prevention care15,16 among SGM youths and adults.
Much of the literature on syndemics has focused on demonstrating additive dose–response relationships between syndemic factors and negative health and social outcomes.7,8,10,14 These studies show linear, progressive worsening of social and health conditions as the number of syndemic factors experienced increases. However, these studies are not able to adequately examine whether syndemic factors interact with one another to contribute to negative health outcomes, a key assumption of syndemic theory.17,18 The distinction between additive and synergistic relationships is of relevance because the latter would suggest that interventions addressing more than 1 syndemic factor would likely have greater health and social effects.19
Studies examining syndemic interactions have found conflicting results. For example, Tomori et al. found a significant dose–response relationship between the number of syndemic conditions and condomless anal sex and a similar, albeit not statistically significant, pattern for syphilis.20 Synergistic interactions were observed between intimate partner violence and depression for condomless anal sex, and between alcohol dependence and illicit drug use for syphilis.20 Similarly, Bulled identified synergism between substance use and violence with respect to HIV infection among White individuals but not individuals of color.21 Other studies examining interactions between syndemic factors have found inconsistent results, including null findings22 or antagonistic interactions (the combined effect of 2 syndemic factors being smaller than the sum of the individual effects of each).23
Inconsistent findings regarding synergism across studies (i.e., different pairs of syndemic factors showing significant interaction) and within studies (i.e., different interaction results for similar outcomes, such as condomless sex and sexually transmitted infections20) make interpretation of findings and their implications less clear. Additionally, research that focused on 2-way interactions may fail to capture more complex interrelationships across multiple syndemic factors, which is at the core of theoretical conceptualizations of syndemics.18 Moreover, although several studies have examined relationships between psychosocial syndemics and HIV-related outcomes among SGM youths, few studies have focused on individuals younger than 18 years. The transition from adolescence to young adulthood is marked by changes in how SGM individuals experience psychosocial problems and HIV risk,12,24 and engagement in HIV prevention care is especially low among individuals younger than 18 years.5 Therefore, the dearth of studies with SGM individuals younger than 18 years is another important gap in the existing literature.
To address these gaps, we examined linear and synergistic associations between psychosocial syndemic burden (i.e., an index that included substance use, binge drinking, sexual victimization, and anti-LGBTQ [lesbian, gay, bisexual, transgender, queer or questioning] discrimination) and HIV risk and engagement in HIV preventive care in a large online US sample of SGM youths aged 13 to 18 years.
METHODS
Between February and October 2022, we recruited SGM youths online through social media ads, collaborations with social media influencers aligned with the LGBTQ community, and community- and school-based organizations.25 Individuals were eligible to participate if they were aged 13 to 18 years; identified as a sexual minority (e.g., lesbian, gay, bisexual, pansexual), gender minority (e.g., transgender, nonbinary, agender), or both; and lived in the United States.
Individuals took a screener survey and, if eligible, received a link to the online survey. To identify duplicate and fraudulent responses, research staff examined e-mail and IP (Internet protocol) addresses, attention-check questions, time to complete survey, response patterns, and write-in questions for unusual or suspicious patterns. Additionally, we asked individuals without institutional e-mail accounts to verify their identities with pictures of their IDs or video calls with research staff to receive remuneration. Of 24 570 individuals who we found eligible and who consented or assented to study participation, we excluded 6992 who were deemed fraudulent or duplicate or had less than 9% survey completion (i.e., completed demographic questions). This resulted in 17 578 individuals with demographic data, of which we had complete data for 9480 individuals (∼54%). The most common missing data pattern was abandoning the survey immediately after demographic questions.
Measures
HIV risk behaviors
We assessed the following factors associated with HIV transmission: sexual initiation, defined as previous vaginal, anal, or oral sexual activity (yes/no); condomless anal or vaginal sex (CAVS) in the past 3 months (yes/no); and number of sexual partners in the past 3 months (winsorized at 6 partners).
Engagement in HIV prevention care
We assessed PrEP awareness (“Have you heard of PrEP?”, yes/no) and PrEP utilization (“Have you ever taken PrEP in your life?”, yes/no) among all participants. We also asked individuals who reported sexual initiation about whether they had ever been tested for HIV (yes/no).
Psychosocial syndemic burden
Similar to previous studies,10,14 we created a count score of psychosocial syndemic burden ranging from 0 to 4, with higher scores indicating higher exposure to syndemics. We considered 4 syndemic indicators:
-
1.
Recent binge drinking. We assessed the number of times participants had 5 or more drinks of alcohol within a couple of hours in the past 30 days. We dichotomized responses as any versus no recent binge drinking.
-
2.
Lifetime drug use. We asked participants whether they had ever used drugs not prescribed by a doctor, such as stimulants, opioids, and hallucinogens, but excluding alcohol, nicotine, and cannabis. We dichotomized responses as any versus no drug use.
-
3.
Sexual victimization and rape. With the following question, we assessed whether participants had been sexually victimized or raped: “Have you ever been physically forced to have sexual intercourse when you did not want to?” We dichotomized responses as yes versus no.
-
4.
Anti-LGBTQ discrimination. We assessed how often participants had been teased, threatened, or harassed about being LGBTQ in the past 12 months (ranging from “never” to “very often”). We assessed in-person and online discrimination separately and then combined responses into a single dichotomous anti-LGBTQ discrimination measure (any vs no anti-LGBTQ discrimination).
Depressive symptoms
We assessed symptoms of depressive mood in the past 2 weeks with the Patient Health Questionnaire-2, a 2-item screening tool for major depression.26 Responses to both questions ranged from 0 (“not at all”) to 3 (“nearly every day”). We summed responses and considered them to potentially indicate major depression for scores of 3 or higher, as previously recommended.26
We also assessed age, race/ethnicity, sex assigned at birth, sexual orientation, and gender identity.
Analysis
To examine the cooccurrence of psychosocial problems, we examined pairwise associations between hypothesized syndemic factors using logistic regression. We then examined relationships between psychosocial syndemics and depressive symptoms (independent variables) and HIV risk and engagement in HIV prevention care (dependent variables) using linear and logistic regression for continuous and categorical outcomes, respectively. Because depressive symptoms diverged from other syndemic factors in their associations with our outcomes, we included depressive symptoms separately in our models.17
We explored syndemic synergism with the inclusion of psychosocial syndemic burden as a categorical variable in saturated regression models. We also explored P values for linear, quadratic, and cubic trends and obtained adjusted predicted means (for continuous outcomes) and probabilities (for binary outcomes) considering the highest-order significant (P < .05) polynomial term for the effect of psychosocial syndemic burden. Adjusted analyses controlled for age, race/ethnicity, sexual orientation, gender identity, sex assigned at birth, and highest level of education of parents or caregivers. We also conducted post hoc analyses for PrEP outcomes adjusting for HIV risk behaviors to assess the potential mediating role of sexual behaviors, and we examined potential interaction by race/ethnicity and gender identity.
To account for missing data, we conducted sensitivity analyses using 50 imputed data sets, using chained equations and predictive mean matching for continuous variables and logit regression for categorical ones.
We conducted data analyses in Stata/SE version 17.0 (StataCorp LP, College Station, TX).
RESULTS
Participants in our study had a mean age of 15.79 years (SD = 1.49 years). Most were assigned female sex at birth (61.8%) and identified as White (69.9%), with 18.2% reporting Latinx ethnicity. Our sample was diverse in sexual orientation (29.3% identified as gay or lesbian, 28.5% as bisexual, and 14.6% as pansexual) and gender identity (19.3% identified as a cisgender boy, 16.7% as a transgender boy, 16.2% as a cisgender girl, and 15.2% as nonbinary). There were participants from all 50 US states, with 33% from the US South, 24.2% from the West, 23.3% from the Midwest, and 19.5% from the Northeast. Supplemental Table A (available as a supplement to the online version of this article at http://www.ajph.org) shows the sociodemographic characteristics of our sample. Individuals with incomplete data were more likely to be younger, assigned male at birth, non-White, straight identified, and cisgender (all P < .001, not shown).
Psychosocial Syndemics and Depressive Symptoms
Psychosocial syndemic factors were commonly reported among our sample. About three quarters reported at least 1 psychosocial syndemic factor (n = 7700; 73.8%), such as anti-LGBTQ discrimination (n = 7908; 69.8%), sexual victimization or rape (n = 1509; 14.4%), recent binge drinking (n = 737; 6.7%), and lifetime drug use (n = 695; 6.3%). About one half of the sample screened positive for depression (n = 5882; 55.0%; Supplemental Table A).
All hypothesized syndemic factors, including depressive symptoms, were positively associated with one another (all P < .001), indicating cooccurrence of syndemic indicators. Supplemental Table B (available as a supplement to the online version of this article at http://www.ajph.org) shows bivariate associations between hypothesized syndemic factors.
Psychosocial Syndemics and Sexual Behaviors
About one fourth of the sample reported sexual activity ever (n = 2984; 27.5%), and 84% reported no sexual partners in the past 3 months (average number of partners in the period was 0.23; SD = 0.68). CAVS in the past 3 months was reported by 8.3% of the sample (n = 870; Supplemental Table A).
Psychosocial syndemic burden was additively associated with sexual behaviors (Table 1). Sexual initiation was progressively more common as the number of syndemic factors increased: coefficients ranged from adjusted odds ratio (AOR) = 1.70 (95% confidence interval [CI] = 1.49, 1.94) among individuals with 1 syndemic factor to AOR = 84.28 (95% CI = 35.65, 199.24) for those with all 4 syndemic factors, compared with individuals reporting no syndemic factors (P < .001 for linear trend; P = .49 for quadratic trend; P = .07 for cubic trend). Depressive symptoms were not associated with sexual initiation (AOR = 0.96; 95% CI = 0.90, 1.33).
TABLE 1—
Associations Between Psychosocial Syndemic Burden and Sexual Behaviors: United States, February–October 2022
| Sexual Initiation (n = 9888 in Adjusted Models) | CAVS (3 Mo) (n = 9496 in Adjusted Models) | Number of Sexual Partners (3 Mo) (n = 9887 in Adjusted Models) | |||||||
| n(event) Yes/Noa | AOR (95% CI)b | P for Trend | n(event) Yes/Noa | AOR (95% CI)b | P for trend | Mean ±SDc | AORb (95% CI)b | P for trend | |
| Psychosocial syndemic burden, no. of syndemic factors | Linear: < .001; Quadratic: .49; Cubic: .07 | Linear: < .001; Quadratic: .97; Cubic: .04 | Linear: < .001; Quadratic: < .001; Cubic: .01 | ||||||
| 0 | 431/2296 | 1 (Ref) | 110/2549 | 1 (Ref) | 0.11 ±0.41 | 1 (Ref) | |||
| 1 | 1217/4561 | 1.70 (1.49, 1.94) | 314/5251 | 1.78 (1.40, 2.26) | 0.16 ±0.52 | 0.08 (0.06, 0.11) | |||
| 2 | 873/628 | 9.40 (7.98, 11.08) | 251/1149 | 6.21 (4.79, 8.05) | 0.46 ±0.91 | 0.36 (0.32, 0.41) | |||
| 3 | 256/66 | 25.35 (18.53, 34.68) | 106/193 | 15.07 (10.81, 21.02) | 0.92 ±1.29 | 0.80 (0.73, 0.87) | |||
| 4 | 82/7 | 84.28 (35.65, 199.24) | 48/38 | 31.81 (18.86, 53.66) | 1.79 ±1.90 | 1.69 (1.56, 1.83) | |||
| Depressive symptoms | |||||||||
| Yes | 1620/4103 | 0.96 (0.90, 1.33) | 463/5033 | 0.90 (0.76, 1.06) | 0.22 ±0.65 | −0.02 (−0.04, 0.01) | |||
| No | 1205/3462 | 1 (Ref) | 349/4134 | 1 (Ref) | 0.23 ±0.66 | 1 (Ref) | |||
Note. AOR = adjusted odds ratio; CAVS = condomless anal or vaginal sex; CI = confidence interval. Regression models considered psychosocial syndemic burden as a categorical variable and assessed linear, quadratic, and cubic trends.
Number of individuals reporting the outcome of interest (i.e., sexual initiation, CAVS) and not reporting the outcome of interest by level of the covariate.
Adjusted models controlled for the effects of age, race/ethnicity, gender identity, sex assigned at birth, and sexual orientation.
Mean ±SD of outcome of interest (number of sexual partners) by level of the covariate.
Similarly, CAVS was progressively more common for individuals with a higher syndemic burden (AOR = 1.78; 95% CI = 1.40, 2.26 for 1 syndemic factor and AOR = 31.81; 95% CI = 18.86, 53.66 for 4 syndemic factors; P < .001 for linear trend, P = .96 for quadratic trend, and P = .04 for cubic trend). Depressive symptoms were not associated with CAVS (AOR = 0.90; 95% CI = 0.76, 1.06). Number of sexual partners was also cumulatively associated with syndemics (b = 0.08; 95% CI = 0.06, 0.11 for 1 syndemic factor, and b = 1.69; 95% CI = 1.56, 1.83 for 4 syndemic factors; with P < .001 for linear trend, P < .001 for quadratic trend, and P = .01 for cubic trend). Depressive symptoms were not associated with the number of sexual partners (b = −0.02; 95% CI = −0.04, 0.01).
Predicted probabilities and means shown in Figure 1 indicate nonlinear relationships between the psychosocial syndemic burden and sexual behavior outcomes. Specifically, there is multiplicative growth in the probability of CAVS as the number of syndemic factors increase (Figure 1b), with growth leveling off toward higher numbers of syndemic factors. Regarding the number of sexual partners, multiplicative growth does not appear to plateau in the observed range of syndemic exposure (Figure 1c).
FIGURE 1—
Number of Reported Psychosocial Syndemic Factors Based on Highest-Order Significant Polynomial Regression Model and Adjusted Predicted Probabilities and Means of (a) Sexual Initiation, (b) Condomless Anal or Vaginal Sex (CAVS), and (c) Number of Sexual Partners: United States, February–October 2022
Note. Part a shows linear models, and parts b and c show cubic models of the predicted probabilities of sexual behavior outcomes and psychosocial syndemic burden. In all models, psychosocial syndemic burden was considered as a continuous predictor. Adjusted models controlled for the effects of age, race/ethnicity, gender identity, sex assigned at birth, and sexual orientation.
Psychosocial Syndemics and HIV Prevention Care
Utilization of HIV prevention care was low among our sample. Among individuals who reported sexual initiation, 15.8% (n = 471) had been tested for HIV. In the overall sample, 26.2% (n = 2844) were aware of PrEP for HIV prevention and only 1.1% had ever utilized PrEP (n = 114).
Contrary to expectations, syndemic burden was positively associated with engagement in HIV care (Table 2). Among participants reporting sexual initiation, HIV testing was progressively more common among individuals with higher syndemic burden: AOR = 1.42 (95% CI = 0.97, 2.09) among individuals with 1 syndemic factor, and AOR = 7.25 (95% CI = 3.97, 13.25; P < .001 for linear trend, P = .75 for quadratic trend, and P = .41 for cubic trend). Depressive symptoms were not associated with HIV testing (AOR = 0.88; 95% CI = 0.69, 1.11).
TABLE 2—
Associations Between Psychosocial Syndemic Burden and Measures of Engagement in HIV Prevention Care: United States, February–October 2022
| HIV Test (Lifetime) (n = 2700 in Adjusted Analyses)a | PrEP Awareness (n = 885 in Adjusted Analyses) | PrEP Utilization (n = 9890 in Adjusted Analyses) | |||||||||
| n(event) Yes/Nob | AOR (95% CI)c | P for trend | n(event) Yes/Nob | AOR (95% CI)c | P for Trend | Fully Adjusted OR (95% CI)d | n(event) Yes/Nob | AOR (95% CI)c | P for Trend | Fully Adjusted OR (95% CI)d | |
| Psychosocial Syndemic burden (Ref = 0 syndemic factors), no. syndemic factors | 43/388 | 1 (Ref) | Linear: P < .001; Quadratic: P = .75; Cubic: P = .41 | 662/2066 | 1 (Ref) | Linear: P < .001; Quadratic: P = .36; Cubic: P = .30 | 1 (Ref) | 10/2719 | 1 (Ref) | Linear: P < .001; Quadratic: P = .98; Cubic: P = .01 | 1 (Ref) |
| 1 | 145/1072 | 1.42 (0.97, 2.09) | 1456/4316 | 1.12 (1.00, 1.26) | 1.08 (0.96, 1.21) | 28/5745 | 1.42 (0.67, 2.97) | 0.97 (0.45, 2.10) | |||
| 2 | 149/723 | 2.55 (1.72, 3.78) | 437/1063 | 1.26 (1.08, 1.47) | 1.10 (0.93, 1.30) | 30/1474 | 6.06 (2.87, 12.83) | 3.25 (1.45, 7.31) | |||
| 3 | 74/182 | 4.62 (2.94, 7.27) | 134/190 | 2.09 (1.62, 2.69) | 1.68 (1.27, 2.21) | 25/298 | 23.86 (10.92, 52.15) | 10.87 (4.57, 25.90) | |||
| 4 | 32/50 | 7.25 (3.97, 13.25) | 44/45 | 2.43 (1.53, 3.86) | 1.71 (1.05, 2.81) | 13/75 | 35.73 (13.40, 95.28) | 10.15 (3.34, 30.79) | |||
| Depressive symptoms (yes/no) | 247/1370 | 0.88 (0.69, 1.11) | 1385/4328 | 0.82 (0.74, 0.90) | 0.82 (0.74, 0.90) | 59/5647 | 0.85 (0.55, 1.32) | 0.90 (0.56, 1.44) | |||
| 190/1015 | 1330/3332 | 47/4612 | |||||||||
| Sexual initiation (yes/no) | … | … | … | … | 1.25 (1.09, 1.44) | … | … | 1.68 (0.88, 3.20) | |||
| CAVS, 3 mo (yes/no) | … | … | … | … | 0.96 (0.77, 1.19) | … | … | 1.57 (0.84, 2.91) | |||
| No. of sexual partners | … | … | … | … | 1.14 (1.04, 1.25) | … | … | 1.28 (1.07, 1.53) | |||
Note. AOR = adjusted odds ratio; CAVS = condomless anal or vaginal sex; CI = confidence interval; OR = odds ratio; PrEP = preexposure prophylaxis. All regression models considered psychosocial syndemic burden as a categorical variable and assessed linear, quadratic, and cubic trends.
Among individuals who reported sexual initiation.
Number of individuals reporting the outcome of interest (i.e., HIV testing, PrEP awareness, and PrEP use) and not reporting the outcome of interest by covariate level.
Adjusted models controlled for age, race/ethnicity, gender identity, sex assigned at birth, and sexual orientation.
Fully adjusted models controlled for sexual initiation, CAVS in the past 3 months, and number of sexual partners in addition to sociodemographics controlled for in adjusted models.
Syndemic burden was also progressively associated with PrEP awareness (ranging from AOR = 1.12; 95% CI = 1.00, 1.26 for 1 syndemic factor to AOR = 2.43; 95% CI = 1.53, 3.86 for 4 factors; P < .001 for linear trend, P = .36 for quadratic trend, P = .30 for cubic trend). Depressive symptoms were negatively associated with PrEP awareness (AOR = 0.82; 95% CI = 0.74, 0.90). Finally, there was also a cumulative association between syndemic burden and lifetime PrEP utilization, ranging from AOR = 1.42 (95% CI = 0.67, 2.97) for 1 syndemic factor to AOR = 35.73 (95% CI = 13.40, 95.28) for 4 syndemic factors (P < .001 for linear trend, P = .98 for quadratic trend, and P = .01 for cubic trend). Depressive symptoms were not associated with PrEP utilization (AOR = 0.85; 95% CI = 0.55, 1.302).
As shown in Figure 2a and 2b, the probabilities of previous HIV testing and PrEP awareness increased linearly with the psychosocial syndemic burden. Figure 2c shows multiplicative growth in the probability of PrEP utilization by syndemic burden and an indication of a potential plateau in the upper range of observed psychosocial syndemic burden, although with wide CIs.
FIGURE 2—
Number of Reported Psychosocial Syndemic Factors Based on Highest-Order Significant Polynomial Regression and Adjusted Predicted Probabilities of (a) HIV Testing, (b) Preexposure Prophylaxis (PrEP) Awareness, and (c) PrEP Utilization: United States, February–October 2022
Note. Parts a and b show fitted generalized linear models of predicted probabilities of HIV testing (among individuals who have had sex) and PrEP awareness (among all individuals) and psychosocial syndemic burden. Part c shows fitted cubic models of the relationship between predicted probabilities of PrEP utilization and psychosocial syndemic burden. Wide confidence intervals in part c reflect low prevalence of highest syndemic counts and PrEP utilization. In all models, psychosocial syndemic burden was considered as a continuous predictor. Adjusted models control for the effects of age, race/ethnicity, gender identity, sex assigned at birth, and sexual orientation.
Post Hoc Analyses
To probe the unexpected association between psychosocial syndemic burden and greater, rather than lower, engagement in PrEP care, we conducted further analyses controlling for the effect of sexual behaviors that could be proxies for PrEP indication (i.e., sexual initiation, CAVS, and number of sexual partners). These models indicated that the relationship between psychosocial syndemic burden and previous HIV testing, PrEP awareness, and PrEP utilization remained positive, yet with a smaller magnitude (Table 2). Additionally, the number of sexual partners was associated with PrEP awareness (AOR = 1.14; 95% CI = 1.04, 1.25) and PrEP utilization (AOR = 1.28; 95% CI = 1.07, 1.53). Sexual initiation was associated with PrEP awareness only (AOR = 1.25; 95% CI = 1.09, 1.44), and CAVS was not associated with PrEP awareness or utilization (Table 2).
Additional post hoc analyses showed that race/ethnicity and gender identity significantly moderated the relationship between psychosocial syndemics and number of sexual partners (P = .02 and P < .001, respectively), such that this relationship was stronger among non-White and cisgender individuals (Supplemental Tables C–F; Supplemental Figures A and B, available as a supplement to the online version of this article at http://www.ajph.org).
Sensitivity Analysis
Findings from regression models in 50 imputed data sets showed consistent results regarding the directionality and magnitude of relationships between syndemic burden and our outcomes of interest (Supplemental Tables G–J, available as a supplement to the online version of this article at http://www.ajph.org).
DISCUSSION
Our study adds to the growing literature on syndemics and HIV prevention care among SGM youths. By examining linear and nonlinear relationships between syndemic factors and health outcomes, we have contributed to a more detailed understanding of the complex interplay between psychosocial syndemic factors and HIV risk and engagement in preventive care. Similar to previous studies,10–12 we have shown a high prevalence of psychosocial syndemic factors among SGM youths. In our study, almost 20% of our sample reported 2 or more psychosocial syndemic factors, with anti-LGBTQ discrimination being particularly common, indicating that many SGM youths may need interventions to address multilevel psychosocial inequities.
As hypothesized, we identified a linear dose–response effect of psychosocial syndemic burden on sexual behaviors that may lead to HIV transmission. Several studies have shown an additive association between psychosocial syndemic factors and HIV risk behaviors in US adult8,14 and youth7,10,27 SGM populations. Our study adds to this body of work and demonstrates similar relationships in a large US sample of SGM individuals younger than 18 years, thus addressing a gap in the existing literature. We encourage further studies to explore mechanisms that could explain our post hoc findings of a stronger relationship between psychosocial syndemics among non-White and cisgender youths (e.g., intersectional discrimination and racism).
Unlike previous studies,9,10,14 we found that the effects of depressive symptoms on sexual behaviors were different from those of the other syndemic factors; namely, depressive symptoms did not have a significant effect on sexual behaviors and were negatively associated with PrEP awareness. Because composite psychosocial syndemic burden variables assume that the individual effects of each syndemic factor are comparable,17 we did not include depression as part of our composite index. Some of the studies that did include depression in counts of syndemic factors did not show depression to be individually associated with sexual behaviors.7,8 Therefore, future syndemic studies should consider the effects of each syndemic factor on each other and on health outcomes individually before creating composite syndemic indicators.
There are several possible explanations for the different effects of depression symptoms and other syndemic factors. Depression symptoms may include decreased sexual drive28 and are linked to social isolation and withdrawal,29 which could limit opportunities for engaging in sexual behaviors and learning about PrEP from social networks. Further studies are needed to identify mediators between depression and sexual behaviors among SGM youths and to study effect modifiers of this relationship, which would help identify groups of youths for whom addressing depression symptoms may decrease HIV risk as well.
Contrary to our expectations, the psychosocial syndemic burden was associated with greater, rather than lower, engagement in HIV prevention care, a relationship that was only partially mediated by sexual risk behaviors. Although some studies have shown syndemic burden to be associated with lower HIV testing and decreased progression along the PrEP cascade,15,30 others have, like our study, shown syndemic burden positively associated with PrEP engagement.31–33 These results may reflect targeted efforts to engage the most at-risk individuals in PrEP care, resulting in higher PrEP awareness and utilization among individuals experiencing higher HIV risk. Importantly, predicted probabilities of PrEP utilization were low (< 15%) even for the highest levels of syndemic burden, indicating that additional efforts are still needed to reach youths experiencing multiple psychosocial problems and potentially at higher risk for HIV acquisition.
An important limitation of existing syndemics research is the dearth of studies examining synergism across syndemic factors.19,34 Our study addressed this gap by examining nonlinear relationships between psychosocial syndemic burden and sexual behavior and HIV prevention outcomes. Specifically, we have shown synergistic polynomial relationships between syndemic burden and sexual behaviors and PrEP utilization. That is, as the burden of psychosocial syndemics increases, the number of sexual partners and the probability of reporting recent CAVS and PrEP utilization grow multiplicatively. These findings have implications for practice, as evidence of syndemic synergism suggests that multicomponent interventions addressing multiple syndemic factors at once may be more effective and efficient than intervening in individual factors. Indeed, existing HIV prevention interventions already incorporate components that address concomitant stigma and substance use, with promising results.35,36 We believe that our study provides support for a combined approach in HIV prevention interventions, which should be further evaluated in future intervention studies.
Limitations
Our findings must be interpreted in light of the limitations of this study. First, as a cross-sectional study, our ability to infer causality and directionality is limited. Moreover, our study had a considerable amount of missing data, and although multiple imputations indicated that missingness did not affect findings significantly, this method relies on unverifiable assumptions. Additionally, our measure of binge drinking (5 drinks within a couple of hours) may have underestimated this behavior among individuals assigned female at birth, and sexual victimization considered only intercourse, which may have underestimated other forms of sexual victimization. Finally, despite our large sample, recruitment efforts relied on community-based organizations and online venues that may be frequented by SGM youths with greater access to affirming social groups and coping resources. Consequently, our findings may not be generalizable to all SGM youths in the United States.
Although Black and Latinx gay and bisexual men and transgender women are disproportionately vulnerable to acquiring HIV, most youths in our study were assigned female sex at birth and were White. Future studies should consider oversampling youths of color to allow more detailed descriptions of how psychosocial factors may affect HIV risk and engagement in care across racial and ethnic groups. These studies should also further examine the role of race, ethnicity, and gender identity as moderators of relationships between syndemics and health outcomes.
Conclusions
These limitations notwithstanding, our study highlights the need to address cooccurring syndemics of HIV risk and psychosocial problems among SGM youths. Multicomponent interventions addressing multilevel syndemic factors have been advocated in HIV research, yet empirical evidence for their improved performance, as opposed to separate interventions, has been relatively scarce. Further research should continue to examine synergism across syndemic factors to identify combinations of intervention components that result in the greatest impact. To that end, novel approaches such as the multiphase optimization strategy37 may be particularly insightful for identifying specific intervention components that should be included in multicomponent interventions for improved effectiveness and efficiency.
ACKNOWLEDGMENTS
This work was supported by the National Institute on Drug Abuse, National Institutes of Health (NIH; grant K01DA047918, principal investigator: R. J. W.) and the National Institute on Minority Health and Health Disparities, NIH (grant R01MD015722-01S1).
The authors acknowledge the Human Rights Campaign Foundation for their partnership in data collection.
Note. The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
CONFLICTS OF INTEREST
The authors have no conflicts of interests to report.
HUMAN PARTICIPANT PROTECTION
The University of Connecticut institutional review board reviewed and approved all study materials and procedures. Participants received US $5 for participation. All individuals provided informed electronic consent or assent, and we obtained a waiver of parental consent.
REFERENCES
- 1. Centers for Disease Control and Prevention. HIV surveillance re port. May 2021. . Available at: https://www.cdc.gov/hiv/pdf/library/reports/surveillance/cdc-hiv-surveillance-report-2018-updated-vol-32.pdf . Accessed June 26, 2024.
- 2.Centers for Disease Control and Prevention. Sexual risk behavior differences among sexual minority high school students—United States, 2015 and 2017. MMWR Morb Mortal Wkly Rep. 2018; 67(36):1007–1011. 10.15585/mmwr.mm6736a3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Centers for Disease Control and Prevention. Prescription opioid misuse and use of alcohol and other substances among high school students—Youth Risk Behavior Survey, United States, 2019. MMWR Suppl. 2020;69(1):38–46. 10.15585/mmwr.su6901a5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Centers for Disease Control and Prevention. HIV testing, linkage to HIV medical care, and interviews for partner services among youths—61 health department jurisdictions, United States, Puerto Rico, and the US Virgin Islands, 2015. MMWR Morb Mortal Wkly Rep. 2017;66(24): 629–635. 10.15585/mmwr.mm6624a2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Barocas JA, Gai MJ, Nurani A, Bagley SM, Hadland SE. Initiation of HIV pre-exposure prophylaxis among youth in the United States, 2015–2018. AIDS Care. 2023;35(3):431–436. 10.1080/09540121.2022.2067318 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Singer M. A dose of drugs, a touch of violence, a case of AIDS: conceptualizing the SAVA syndemic. Free Inq Creat Sociol. 1996;24(2):13–24. [Google Scholar]
- 7.Teixeira da Silva D, Bouris A, Voisin D, Hotton A, Brewer R, Schneider J. Social networks moderate the syndemic effect of psychosocial and structural factors on HIV risk among young black transgender women and men who have sex with men. AIDS Behav. 2020;24(1):192–205. 10.1007/s10461-019-02575-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.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(3): 329–336. 10.1097/QAI.0000000000000475 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Mustanski B, Phillips G, Ryan DT, Swann G, Kuhns L, Garofalo R. Prospective effects of a syndemic on HIV and STI incidence and risk behaviors in a cohort of young men who have sex with men. AIDS Behav. 2017;21(3):845–857. 10.1007/s10461-016-1607-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Mimiaga MJ, Hughto JMW, Biello KB, et al. Longitudinal analysis of syndemic psychosocial problems predicting HIV risk behavior among a multicity prospective cohort of sexually active young transgender women in the United States. J Acquir Immune Defic Syndr. 2019;81(2):184–192. 10.1097/QAI.0000000000002009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Pellicane MJ, Ciesla JA. Temporal trends in rates of depression, anxiety, and suicidality among cisgender sexual minority and heterosexual college students. Psychol Sex Orientat Gend Divers. 2023;10(4):560–574. 10.1037/sgd0000563s [DOI] [Google Scholar]
- 12.Fish JN, Bishop MD, Russell ST. Developmental differences in sexual orientation and gender identity–related substance use disparities: findings from population-based data. J Adolesc Health. 2021;68(6):1162–1169. 10.1016/j.jadohealth.2020.10.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Meyer IH. Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: conceptual issues and research evidence. Psychol Bull. 2003;129(5):674–697. 10.1037/0033-2909.129.5.674 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Harkness A, Bainter SA, O’Cleirigh C, Albright C, Mayer KH, Safren SA. Longitudinal effects of syndemics on HIV-positive sexual minority men’s sexual health behaviors. Arch Sex Behav. 2019; 48(4):1159–1170. 10.1007/s10508-018-1329-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Blashill AJ, Brady JP, Rooney BM, et al. Syndemics and the PrEP cascade: results from a sample of young Latino men who have sex with men. Arch Sex Behav. 2020;49(1):125–135. 10.1007/s10508-019-01470-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Myers K, Li T, Baum M, Ibanez G, Fennie K. The individual, interactive, and syndemic effect of substance use, depression, education, and ethnicity on retention in HIV care. Int J STD AIDS. 2021;32(2):184–193. 10.1177/0956462419890727 [DOI] [PubMed] [Google Scholar]
- 17.Tsai AC, Venkataramani AS. Syndemics and health disparities: a methodological note. AIDS Behav. 2016;20(2):423–430. 10.1007/s10461-015-1260-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Singer M, Clair S. Syndemics and public health: reconceptualizing disease in bio‐social context. Med Anthropol Q. 2003;17(4):423–441. 10.1525/maq.2003.17.4.423 [DOI] [PubMed] [Google Scholar]
- 19.Tsai AC, Burns BFO. Syndemics of psychosocial problems and HIV risk: a systematic review of empirical tests of the disease interaction concept. Soc Sci Med. 2015;139:26–35. 10.1016/j.socscimed.2015.06.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Tomori C, McFall AM, Solomon SS, et al. Is there synergy in syndemics? Psychosocial conditions and sexual risk among men who have sex with men in India. Soc Sci Med. 2018;206:110–116. 10.1016/j.socscimed.2018.03.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Bulled N. A new approach to measuring the synergy in a syndemic: revisiting the SAVA syndemic among urban MSM in the United States. Glob Public Health. 2022;17(9):2070–2080. 10.1080/17441692.2021.1974513 [DOI] [PubMed] [Google Scholar]
- 22.Wirtz AL, Burns PA, Poteat T, et al. Abuse in the continua: HIV prevention and care outcomes and syndemic conditions associated with intimate partner violence among Black gay and bisexual men in the Southern United States. AIDS Behav. 2022;26(11):3761–3774. 10.1007/s10461-022-03705-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Bhardwaj A, Comins CA, Guddera V, et al. Prevalence of depression, syndemic factors and their impact on viral suppression among female sex workers living with HIV in eThekwini, South Africa. BMC Womens Health. 2023;23(1):232. 10.1186/s12905-023-02392-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Halkitis PN, Kapadia F, Bub KL, Barton S, Moreira AD, Stults CB. A longitudinal investigation of syndemic conditions among young gay, bisexual, and other MSM: the P18 Cohort Study. AIDS Behav. 2015;19(6):970–980. 10.1007/s10461-014-0892-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Watson RJ, Caba AE, Lawrence SE, et al. Examining mental health and bullying concerns at the intersection of sexuality, gender, race, and ethnicity among a national sample of sexual and gender diverse youth. LGBT Health. 2024;11(1): 20–27. 10.1089/lgbt.2023.0072 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kroenke K, Spitzer RL, Williams JB. The Patient Health Questionnaire–2: validity of a two-item depression screener. Med Care. 2003;41(11): 1284–1292. 10.1097/01.MLR.0000093487.78664.3C [DOI] [PubMed] [Google Scholar]
- 27.Brennan J, Kuhns LM, Johnson AK, et al. Syndemic theory and HIV-related risk among young transgender women: the role of multiple, co-occurring health problems and social marginalization. Am J Public Health. 2012;102(9): 1751–1757. 10.2105/AJPH.2011.300433 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Laurent SM, Simons AD. Sexual dysfunction in depression and anxiety: conceptualizing sexual dysfunction as part of an internalizing dimension. Clin Psychol Rev. 2009;29(7):573–585. 10.1016/j.cpr.2009.06.007 [DOI] [PubMed] [Google Scholar]
- 29.Rueger SY, Malecki CK, Pyun Y, Aycock C, Coyle S. A meta-analytic review of the association between perceived social support and depression in childhood and adolescence. Psychol Bull. 2016;142(10):1017–1067. 10.1037/bul0000058 [DOI] [PubMed] [Google Scholar]
- 30.Turpin R, Slopen N, Boekeloo B, Dallal C, Chen S, Dyer T. Testing a syndemic index of psychosocial and structural factors associated with HIV testing among Black men. J Health Care Poor Underserved. 2020;31(1):455–470. 10.1353/hpu.2020.0033 [DOI] [PubMed] [Google Scholar]
- 31.Chandler CJ, Bukowski LA, Matthews DD, et al. Understanding the impact of a syndemic on the use of pre-exposure prophylaxis in a community-based sample of behaviorally PrEP-eligible BMSM in the United States. AIDS Care. 2020;32(5):551–556. 10.1080/09540121.2019.1659921 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Willie TC, Kershaw TS, Blackstock O, et al. Racial and ethnic differences in women’s HIV risk and attitudes towards pre-exposure prophylaxis (PrEP) in the context of the substance use, violence, and depression syndemic. AIDS Care. 2021;33(2):219–228. 10.1080/09540121.2020.1762067 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Kuhns LM, Perloff J, Johnson AK, et al. A cross-sectional analysis of psychosocial and structural barriers and facilitators associated with PrEP use among a sample of transgender women in Chicago, IL. AIDS Res Ther. 2023;20(1):24. 10.1186/s12981-023-00516-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Singer M, Bulled N, Ostrach B. Whither syndemics? Trends in syndemics research, a review 2015–2019. Glob Public Health. 2020;15(7):943–955. 10.1080/17441692.2020.1724317 [DOI] [PubMed] [Google Scholar]
- 35.Garofalo R, Kuhns LM, Reisner SL, Biello K, Mimiaga MJ. Efficacy of an empowerment-based, group-delivered HIV prevention intervention for young transgender women: the Project LifeSkills randomized clinical trial. JAMA Pediatr. 2018;172(10): 916–923. 10.1001/jamapediatrics.2018.1799 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Bogart LM, Barreras JL, Gonzalez A, et al. Pilot randomized controlled trial of an intervention to improve coping with intersectional stigma and medication adherence among HIV-positive Latinx sexual minority men. AIDS Behav. 2021;25(6): 1647–1660. 10.1007/s10461-020-03081-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Collins LM, Kugler KC, Gwadz MV. Optimization of multicomponent behavioral and biobehavioral interventions for the prevention and treatment of HIV/AIDS. AIDS Behav. 2016;20(suppl 1): S197–S214. 10.1007/s10461-015-1145-4 [DOI] [PMC free article] [PubMed] [Google Scholar]