Adolescent loneliness in the United States: Prevalence, sociodemographic correlates, and psychological health over time.

Loneliness is a public health concern, but its epidemiology in U.S. adolescence remains poorly described. Using data from the Adolescent Brain Cognitive Development (ABCD) Study (analytic N = 6067), we estimated the prevalence of loneliness at approximately age 14 and tested whether sociodemographic characteristics and psychological health were associated with later loneliness. Adolescents self-reported how often they felt lonely ("not lonely," "sometimes," "often"). Caregivers reported adolescents' psychological symptoms (depression, anxiety, attention-deficit/hyperactivity, conduct, oppositional defiant, somatic) annually from baseline through Year 3. We estimated linear mixed-effects growth models to obtain person-specific intercepts (midpoint symptom level) and slopes (rate of change) and then used these trajectories as predictors of loneliness in multilevel ordinal models. Prevalence estimates indicated that 8.9% of adolescents reported being "often lonely," 31.6% "sometimes lonely," and 59.5% "not lonely." In the fully adjusted multilevel model, females had higher odds of loneliness than males (OR = 1.88), and sexual-minority adolescents had higher odds than non-sexual-minority adolescents (OR = 2.81). Higher depressive symptom levels across late childhood and early adolescence were associated with greater odds of loneliness (OR = 1.29), and modest increases in conduct problems were also related to higher loneliness (OR = 1.12). The remaining symptom trajectories showed small or null associations after adjustment. The results from the present study suggest that, despite growing concern about an "epidemic" of loneliness, the experience is unevenly distributed by mid-adolescence, with elevated risk among female adolescents, sexual-minority adolescents, and those with heightened depressive symptoms.