Adolescent social media use and psychiatric outcomes: a longitudinal mediation analysis via interpersonal distrust, sleep, and self-image.

Purpose: The present study investigated the longitudinal associations between social media use (SMU) in early adolescence (age 11) and psychiatric outcomes (age 17) via interpersonal distrust, later bedtime, and negative self-image (age 14) after controlling for prior mental health (age 7) and a range of confounders.

Methods: A structural equation model linked SMU to psychological distress (Kessler-6) via distrust, time-to-sleep, and negative self-perception, using data from a birth cohort in the United Kingdom. From 12,732 eligible adolescents at age 11 (interviewed January 2012 to February 2013), 8,913 participants (52% female, 18% non-White) had complete data on exposure and outcome, thus included in the analytic sample. Sex-stratified analyses were performed, as well as secondary outcome analyses for internalizing/externalizing problems, and a latent variable of 'psychiatric problems' (depression/anxiety diagnosis, self-harm, suicidality).

Results: There were significant indirect paths through distrust (standardized [Formula: see text]), later time-to-sleep ([Formula: see text]), and negative self-perception ([Formula: see text]) after adjustments. The path through distrust was significant for females but not for males, whereas the two other indirect paths were significant for both males and females (this also held true for internalizing, externalizing, and psychiatric problems). Use of self-report measures and a lack of detailed information on the nature of SMU limit these findings.

Conclusion: Social media use is prospectively associated with psychiatric symptoms in adolescence to the extent that it fosters interpersonal distrust, delays bedtime, and degrades self-image, especially for females. Interventions aimed at promoting trust and belonging, good sleep hygiene, and positive self-image, should be considered from a public health perspective.