Social media use and risky behaviors in adolescents: A meta-analysis

This systematic review and meta-analysis examined the associations between social media use and risky behaviors during adolescence, and evaluated study characteristics (e.g., sample age, type of social media platform assessed) that may moderate these relationships.

A comprehensive search strategy identified relevant studies from PsycInfo, PubMed, Google Scholar, and Proquest Dissertations and Theses Global.

The final sample included 27 independent cross-sectional studies with a total of 67,407 adolescents (M_age = 15.5, range: 12.6-18.0 years; 51.7% girls; 57.2% White). Results from random effects models indicated that there were positive, small-to-medium correlations between social media use and engagement in risky behaviors generally (r = 0.21, 95% CI = 0.16-0.25), substance use (r = 0.19, 95% CI = 0.12-0.26), and risky sexual behaviors (r = 0.21, 95% CI = 0.15-0.28). There were an insufficient number of independent samples available to conduct a random effect models for violence-related behaviors (k = 3). Moderator analyses suggested that studies assessing solely early social media platforms (e.g., Facebook/MySpace only) in relation to substance use had smaller effect sizes than substance use studies assessing a broader range of contemporary social media platforms. In addition, younger samples had larger effect sizes for studies focused on social media use and risky sexual behaviors.

The positive links identified between social media and risky behaviors during adolescence in this meta-analysis suggest that developmental theories of risk taking would benefit from incorporating the social media context. Longitudinal studies are needed to clarify directionality and make more specific practice and policy recommendations so that social media is a safe place in which adolescents can thrive.