The Impact of Historical, Neighborhood, and Individual Factors on Suicidal Thoughts and Behaviors Among Adolescent Females: A Multilevel Structural Equation Modeling Approach - Suicidal thoughts and behaviors (STB) among adolescent girls have reached crisis levels: the proportion seriously considering suicide rose from 17% to 27% between 2009 and 2023,1 with the steepest increase among Black girls, whose suicide rate rose 122% from 2004 to 2019.2–5,15 Despite this, research on suicide outcomes has predominantly focused on person-level factors,6,7 typically examined in sex-aggregated samples that obscure processes specific to girls and rarely follow them prospectively across adolescence; evidence for Black girls is scarcer still, as studies have largely overlooked distinct factors for Black individuals.8,9 Because Black girls exist at the intersection of race and sex, research examining these influences separately by race and by sex fails to account for the unique vulnerabilities and protective factors Black girls experience.10 Emerging evidence suggests that social determinants of health (e.g., long-standing policies, community practices, and institutional norms that shape differing outcomes across racial groups) also contribute to mental health outcomes.11–13 Grounded in an ecological framework to assess community health, this disparities project investigates individual and neighborhood influences, as well as historical housing policies (i.e., redlining), on suicide risk over time. The project employs a multilevel structural equation modeling (MSEM) approach to assess the interplay between individual-level factors (e.g., affective dysregulation, bullying, loneliness) and neighborhood-level factors (e.g., school funding, green space, access to mental health services) using data from the Pittsburgh Girls Study, an NIH-funded 20-wave cohort following girls across racial and ethnic groups from childhood into young adulthood. This project also incorporates sociohistorical context using Home Owners' Loan Corporation (HOLC) categories, which classified neighborhoods based on investment risk assessments in the 1930s. All three aims are estimated in the full sample of girls, with subgroup analyses testing whether pathways differ for Black girls. The specific aims are to: (1) investigate how individual-level factors contribute to age-20 STB among girls clustered in neighborhoods, evaluating both direct and mediated effects; (2) assess the influence of neighborhood-level factors on age‑20 STB, considering the role of HOLC categories, including testing of cross-level interactions between neighborhood- and individual-level mediators; and (3) evaluate developmental timing by testing whether pathway coefficients differ across early, middle, and late adolescence in their associations with age‑20 STB. By identifying which processes confer risk (Aim 1; e.g., bullying, loneliness, affective dysregulation) and when they emerge (Aim 3), findings will inform the timing and targets of screening and preventive intervention delivered by the healthcare system. Because subgroups have unique needs, testing whether pathways differ across groups provides evidence to adapt, not simply extend, existing mental health services. This work will also provide methodological training to support my development as an independent researcher in adolescent mental health, community health, and suicide prevention.
