Background Adverse childhood experiences (ACEs) are classically defined as physical abuse, sexual abuse, emotional abuse, emotional neglect, bullying, parental substance use or abuse, violence between parents, parental mental health problems or suicide, parental separation, or a parent convicted of criminal offence. Exposure to ACEs can be associated with cannabis use, but no comparisons across all adversities have been made while also considering timing and frequency of cannabis use. We aimed to explore the association between ACEs and cannabis use timing and frequency in adolescence, considering the cumulative number of ACEs and individual ACEs.

Methods We used data from the Avon Longitudinal Study of Parents and Children, a longitudinal UK birth cohort study. Longitudinal latent classes of cannabis use frequency were derived from self-reported data at multiple timepoints in participants aged 13–24 years. ACEs between ages 0 years and 12 years were derived from prospective and retrospective reports at multiple timepoints by parents and the participant. Multinomial regression was used to analyse the effect of both cumulative exposure to all ACEs and the ten individual ACEs on cannabis use outcomes.

Findings 5212 participants (3132 [60·0%] were female and 2080 [40·0%] were male; 5044 [96·0%] were White and 168 [4·0%] were Black, Asian, or minority ethnic) were included in this study. After adjustment for polygenic risk and environmental risk factors, participants who had 4 or more ACEs at age 0–12 years were at increased risk of early persisting regular cannabis use (relative risk ratio [RRR] 3·15 [95% CI 1·81–5·50]), later onset regular use (1·99 [1·14–3·74]), and early persisting occasional use (2·55 [1·74–3·73]) compared with low or no cannabis use. After adjustment, early persisting regular use was associated with parental substance use or abuse (RRR 3·90 [95% CI 2·10–7·24]), parental mental health problems (2·02 [1·26–3·24]), physical abuse (2·27 [1·31–3·98]), emotional abuse (2·44 [1·49–3·99]), and parental separation (1·88 [1·08–3·27]) compared with low or no cannabis use.

Interpretation Risks for problematic adolescent cannabis use are highest for individuals reporting 4 or more ACEs, and were particularly raised for those with parental substance use or abuse. Public health measures to address ACEs might reduce adolescent cannabis use.