Suicide is a significant public health concern, and it remains the second leading cause of death among individuals aged 10 to 34 years in the United States. Suicide death rates have climbed more than 35% since 1999 (10.5 per 100,000) till 2018 (14.2 per 100,000), with a greater increase among men, young adolescents, racial/ethnic minorities, and sexual minority youth. The risks of engaging in suicidal behaviors differ across developmental stages; however, few studies documented the nuanced differences of suicidal trajectories over time. More importantly, sociodemographic disparities in suicidal trajectories are largely unknown. Understanding suicidal trajectories can improve the effectiveness of suicide interventions by targeting the most sensitive periods, most important risk factors, and at-risk subpopulations.
Despite decades of suicide research, our ability to predict suicidal behaviors is not strong enough and has not improved over the past 50 years. Previous studies have predominately re-reporting on the same individual risk factors (e.g., psychological and substance use disorders) for suicidal ideation and suicide attempts, while the evidence on the influences of social contexts, especially the role of social networks, is substantially limited. Even less research is available in elucidating the effect of social networks across the life course. Little is known about how different social network characteristics evolve, or how such changes contribute to suicidal trajectories as adolescents transitioning to adults. Guided by the Network Episode Model (NEM), this dissertation fills in these gaps by examining the influence of multidimensional and changing patterns of social networks on suicidal trajectories over time.
Using a three-paper format, this dissertation examined: 1) trajectories of suicidal ideation and suicide attempts (Study 1), 2) how social networks during adolescence influence suicidal trajectories (Study 2), and 3) how parental closeness trajectories influence suicidal trajectories, and whether future orientation may moderate the association between parental closeness trajectories and suicidal trajectories (Study 3). Each study further explores the racial/ethnic, sex, sexual orientation, and socioeconomic disparities in predicting suicidal trajectories.
Study 1: Using latent class growth analyses, Study 1 identified three suicidal ideation trajectories (low-stable, moderate-decreasing-increasing, high-decreasing) and two suicide attempt trajectories (low-stable, moderate-decreasing). Results of multinomial logistic regressions found female and sexual minorities were more likely to have a high risk of suicidal ideation and moderate risk of suicide attempts in early adolescence, following by gradual decreases in risks over time. Sexual minorities had a higher probability of showing moderate suicidal ideation during early adolescence, which decreased through emerging adulthood (ages 18 to 26 years), but increased in later stages (ages 24 to 32 years). Black adolescents were more likely to show a lower risk of suicidal ideation and suicide attempts across life stages. Intersections across race/ethnicity, public assistance receipt, and sexual orientation were found to predict suicidal trajectories characterized by high risks of suicidal behaviors in an early stage.
Study 2: After identifying three suicidal ideation trajectories (low-stable, moderate-decreasing-increasing, high-decreasing) and two suicide attempt trajectories (low-stable, moderate-decreasing), Study 2 found a greater family cohesion was significantly associated with lower probabilities of being in high-decreasing (Trajectory 2) and moderate-decreasing-increasing (Trajectory 3) trajectories of suicidal ideation, as well as the moderate-low trajectory of suicide attempts (Trajectory 2) than in low-stable trajectories. Significant moderation effects of race/ethnicity, sex, and sexual orientation were detected in the associations between a specific type of social networks (household size, peer network density, family cohesion, peer support, neighborhood support) and suicidal trajectories.
Study 3: Using latent growth curve modeling, Study 3 examined parental closeness trajectories over time and found both maternal and paternal closeness decreased as individuals grew older. Individuals with low values of parental closeness during early adolescence and steeper decreasing rates of parental closeness trajectories over time had a higher likelihood of being in the moderate-decreasing-increasing trajectory of suicidal ideation (Trajectory 3) and the moderate-decreasing trajectory of suicide attempts (Trajectory 2). Future orientation was found to have a protective-stabilizing effect on the association between paternal closeness trajectories and the probability of being in the moderate-decreasing-increasing suicidal ideation trajectory (Trajectory 3). The moderating effect of future orientation was only found in Black and Hispanic populations to reduce their likelihood of being in the moderate-decreasing suicide attempts trajectory (Trajectory 2) when there was a steeper decline in maternal closeness.
Discussion and limitations of this research and implications for research, policy, and clinical practice are discussed throughout these papers, as well as the overall dissertation conclusion.