Association With Deviant Peers Across Adolescence: Subtypes, Developmental Patterns, and Long-Term Outcomes

Abstract

Indices of deviant peer group involvement are inconsistent and confound type, frequency, and severity of deviant peer behaviors. These measurement approaches thus obfuscate potential meaningful differences in deviant peer involvement in terms of subtypes, developmental patterns, and long-term outcomes. The current study employed latent class analysis to derive subtypes of deviant peer involvement and examined relations to substance use disorder in adulthood, a common outcome of deviant peer involvement. Youth (76% Caucasian) completed assessments across four time points: ages 10–12 years (Time 1; N = 775, 71% male), 12–14 years (Time 2; n = 649, 72% male), 16 years (Time 3; n = 613, 73% male), and 22 years (Time 4; n = 425, 71% male). At Times 1 to 3, participants completed an interview assessing deviant peer involvement. At Time 4, participants completed a structured interview assessing substance use disorder. Classes of youth with different profiles of deviant peer associations were derived at Times 1, 2, and 3. Classes varied by type (conduct problems vs. substance use) and severity of deviant peer behavior. Youth reported higher levels of involvement with deviant peers across adolescence, suggesting that some of these deviant peer behaviors may be normative. Earlier involvement with deviant peers and involvement with groups defined by severe conduct problems and substance use were related to the greatest risk for substance use disorder at Time 4. Type and severity of peer deviant behavior differentially relate to long-term risk for substance use disorder and should be included in screening and assessment for risk across adolescence.

Additional information

Funding

This work was supported by the National Institute on Drug Abuse and conducted at the Center for Education and Drug Abuse Research at the University of Pittsburgh. The current article resulted from Julia Price’s dissertation, which was supported in part by the Civic Award Foundation. Thank you to Dr. Ralph E. Tarter for his contributions to this work.