Using Time-Varying Effect Models to Explore Social Etiology: Sexual Minority Discrimination and Suicidality Across the Lifespan

Introduction. Sexual-orientation-related health disparities constitute a serious public health concern with strong documentation that lesbian/gay and bisexual populations (i.e., sexual minorities [SMs] demonstrate elevated rates of poor health across a number of mental, behavioral, and physical indicators. Accordingly, the National Institute on Minority Health and Health Disparities recently recognized SMs as a health disparity population, citing stigma and discrimination as key barriers to health equity.

Indeed, an emerging body of research demonstrates the effects of discrimination on the health of SMs. Yet, the mechanism by which discrimination may contribute to health disparities is poorly understood. Given that age-specific estimates can provide a more nuanced view of discrimination and its related outcomes, a promising line of inquiry is the investigation of age-graded differences in (a) the experience of discrimination and (b) its influence on health.

We used time-varying effects models (TVEM) to estimate the age-varying prevalence of past-year sexual-orientation-based (1) healthcare discrimination, (2) general discrimination, and (3) verbal and physical victimization using a nationally-representative sample of SM adults from the National Epidemiological Survey of Alcohol and Related Conditions III. We then estimated the time-varying effects between discrimination and recent suicidality among SMs.

Methods. TVEM is a flexible, semi-parametric model that estimates how regression coefficients vary with respect to a time function (i.e., age or years), thus providing precise age ranges of risk. Using TVEM, we (1) estimated the prevalence of stigma-related experiences from ages 18-60 and (2) whether the association between discrimination (yes = 1; no = 0) and recent suicidality (yes = 1; no = 0) varied by age.

Results. Reports of healthcare and general discrimination, and victimization systematically varied by age, with those who identified as LGB reporting the greatest overall rates of stigma compared to SMs defined by attraction and/or behavior. The association between stigma and suicidality also varied by age. For example, the association between experiences of discrimination and recent suicide attempt was strongest at younger ages for SM males (OR ≈ 15.0 at age 20) and females (OR≈ 7.5 at age 18); the association subsequently decreased until the age of thirty, at which point there was no SM disparity.

Discussion. Among SMs, stigma-related experiences and their association with health vary across life course. The results of these analyses have the potential to identify critical periods for targeted policies, programs, and prevention strategies that combat sexual-orientation-related health disparities.