Authors: Russell Turner, Kristian Daneback, Anette Skårner


Introduction: Longitudinal reciprocal associations between substance use and delinquency are understudied in general adolescent populations, with previous research showing differing findings. This study aims to assess reciprocal associations between drunkenness, drug use and delinquency in a prospective, age-homogenous cohort study, using an analytical strategy that separates within- from between-person variance.

Material and methods: Data comes from the Longitudinal Research on Development in Adolescence (LoRDIA) study in Sweden. Adolescents were surveyed at baseline (age 13, grade 7, N=1409) and followed-up at grades 8 and 9. Random-intercept cross-lagged panel analysis was undertaken.

Results: Between-person variance in the development of drunkenness, drug use, and delinquency were between 26% and 47%. At the within-person level, the links between these behaviors were only weak: delinquency was associated with later drug use across grades 7–9, and with drunkenness between grades 8–9 only; drunkenness was associated with later drug use between grades 7–8. Drug use was not associated with later delinquency or drunkenness at any time point.

Discussion: Stable individual factors play a larger than a hitherto known role in within- and over-time relationships between drunkenness, drug use, and delinquency. Longitudinal reciprocal associations between these behaviors are at best weak and such associations may not be of primary importance in practice with general adolescent populations. Moreover, the behaviors appear to follow three distinct developmental pathways, to which intervention design may need to pay attention.

Conclusion: Reciprocal associations between drunkenness, drug use, and delinquency were assessed, highlighting the differential developmental pathways of these behaviors in early-mid adolescence.

This research is financed by:

Vetenskapsrådet, Vinnova, Formas, Forte.

Contact: Russell Turner, Kristian Daneback, Anette Skårner