The three-class model was the preferred solution based on the adjusted Bayesian information criterion, which favored the three-class structure over models with two, four or five classes, and the Bayesian information criterion. Its entropy value was 0.79, indicating reasonably clear but imperfect separation among the inferred classes. The first class, labeled “Syndemic Adversity Experiences,” contained 160 participants, or 9.2% of the sample. The second, “Psychological Distress Primary,” was the largest class, with 1,208 participants, or 69.7%. The third, “Lower Adversity Experiences,” included 364 participants, or 21.0%. Because latent classes are inferred from response patterns rather than assigned through clinical diagnoses, the labels describe shared statistical profiles within this survey.

Each class had a distinct indicator profile. Within the syndemic class, 91.9% reported sex work, 82.5% met the study’s definition of psychological distress, 78.8% reported hazardous alcohol or substance use, 75.6% met the housing-insecurity definition and 64.4% reported violence victimization. Self-reported HIV-positive serostatus was reported by 21.9% of this class, compared with 9.1% of the lower-adversity class and 0% of the psychological-distress-primary class. The latter group was distinguished by a very high probability of psychological distress—reported for 98.7%—along with substantially lower probabilities of most other measured adversities. The lower-adversity class had low-to-moderate probabilities across the indicators overall, although it was not uniformly free of adversity.

All six adversity indicators were self-reported and covered different, sometimes broad time windows. Hazardous substance use combined hazardous alcohol consumption with misuse of prescription medications or use of cocaine, crack, heroin, methamphetamine, hallucinogens or ecstasy. Psychological distress was defined as a score of at least 13 on the six-item Kessler Psychological Distress Scale or endorsement of suicidal ideation, a suicide attempt or self-harm. Violence victimization referred to unwanted physical violence since March 2020. Housing insecurity included reporting a temporary or unsafe living arrangement, not remaining in the same place or feeling unsafe where currently living; it also included sleeping somewhere not intended for sleeping. Sex work was measured by reporting intercourse in exchange for money, drugs, alcohol, food, a place to sleep or another material good.

The participants were predominantly young, with a median age of 23 years and an interquartile range of 20 to 29 years; 62.8% were between 18 and 25. The sample was largely White: 73.0% identified as White, 11.3% as Black and 15.7% within a combined category for other people of color. A majority reported incomes below $35,000, including 30.1% with household income under $15,000 and 33.2% with income from $15,000 to $34,999. Among respondents with evaluable insurance data, 12.9% lacked health insurance and 35.6% had public coverage, while 51.6% had private insurance. Of 1,544 participants with gender-affirming care data, 42.4% reported barriers after attempting to obtain care, 24.3% reported attempting access without barriers and 33.4% had not attempted to access it.

Discrimination was also common in the sample. Overall, 78.9% of participants with discrimination data reported high or extreme anticipated discrimination, while 38.5% of those with transphobia data reported moderate or extreme enacted transphobia. In multivariable multinomial regression models, participants in the syndemic class had 3.51 times the odds of reporting household income below $15,000, 1.95 times the odds of reporting income from $15,000 to $34,999 and 2.46 times the odds of having no or public rather than private insurance, compared with the lower-adversity class. They also had 2.58 times the odds of reporting barriers to gender-affirming care. Extreme anticipated discrimination was associated with 7.61 times the odds of syndemic-class membership relative to low or moderate anticipated discrimination, while moderate or extreme enacted transphobia was associated with 8.28 times the odds. These are adjusted associations, not causal estimates for individual exposures.

Racial and ethnic marginalization showed a strong association with class membership. Compared with the lower-adversity class, Black participants had 2.16 times the odds of being in the syndemic class, participants grouped as other people of color had 2.79 times the odds and Hispanic or Latine participants had 2.64 times the odds. When the syndemic class was compared directly with the psychological-distress-primary class, Black participants had 9.93 times the odds of syndemic membership, while participants grouped as other people of color had 2.41 times the odds and Hispanic or Latine participants had 2.06 times the odds. Trans feminine participants had 3.85 times the odds of belonging to the syndemic class rather than the psychological-distress-primary class. By contrast, bisexual or queer participants had 2.35 times the odds of belonging to the psychological-distress-primary class than the lower-adversity class and 0.52 times the odds of syndemic membership relative to the psychological-distress-primary class.

The study applied a combined syndemic and intersectional framework. Syndemic theory concerns conditions that cluster within a population because social and contextual factors create the conditions for their co-occurrence and can increase collective burden. Intersectionality examines how overlapping systems of power and identity shape unequal exposure to harm. In this analysis, the statistical clustering of adversities coincided with differences in social capital, economic resources and social processes. The authors emphasize that race, ethnicity, gender modality and sexual orientation are not themselves direct measures of discrimination and should not be interpreted as biological causes. Rather, the findings are consistent with how identities linked to unequal access to resources, protection and opportunity can correspond to distinct adversity profiles in a society structured by White supremacy, cisgenderism and heteronormativity.

Several limitations restrict how the results should be interpreted. The cross-sectional design cannot establish causality, changes over time or the sequence in which adversities occurred. The analysis used an online convenience sample recruited through community-based organizations, so it cannot be generalized to all transgender and nonbinary adults in the United States. The investigators could not determine participants’ motivations for sex work or whether they regarded it as an adversity; the indicator instead reflected the study’s concern with workforce precarity, criminalization and associated risks of violence and incarceration. Collapsing diverse racial and ethnic communities into broad analytical categories may also obscure important differences. The survey was conducted during the COVID-19 pandemic, with data collected from June 14 to December 16, 2021, and the authors noted that pandemic conditions may have contributed to the high level of psychological distress. In addition, all adversity and social-process measures depended on self-report.

The central result is that adversity was not distributed in a single pattern across the sample. Although the syndemic class represented a minority of participants, it contained the most frequent combination of several reported harms and was most strongly associated with racial and ethnic marginalization, low income, limited insurance coverage, barriers to gender-affirming care, anticipated discrimination and enacted transphobia. The psychological-distress-primary class showed a different pattern: severe psychological distress was nearly universal within that group, while most other measured adversities were less common. The authors said the results support public-health and policy responses that address several co-occurring conditions together as well as the surrounding conditions linked to them, including healthcare access, socioeconomic stability and protections against discrimination. They also called for longitudinal research to examine trajectories over time and qualitative research centered on the lived experiences of transgender and nonbinary people.

Subject of Research: Cross-sectional latent class analysis of clustered structural and psychosocial adversities among transgender and nonbinary adults in the United States

Article Title: Characterizing patterns of syndemic structural and psychosocial adversities in transgender and nonbinary, adults in a U.S. nationwide sample: A latent class analysis

Article References: Dobson-Seo, B. F., Silva, S. G., Mulawa, M. I., Wirtz, A. L., Reisner, S. L., Cole, S. W., Adams, D., Brown, C., Miller, M., & Poteat, T. (2026). Characterizing patterns of syndemic structural and psychosocial adversities in transgender and nonbinary, adults in a U.S. nationwide sample: A latent class analysis. SSM – Mental Health, 10, Article 100715. https://doi.org/10.1016/j.ssmmh.2026.100715

Image Credits: AI Generated

DOI: 10.1016/j.ssmmh.2026.100715

Keywords: transgender health, nonbinary health, syndemic adversity, latent class analysis, psychological distress, housing insecurity, transphobia, structural discrimination, gender-affirming care, intersectionality, social determinants of health, Characterizing