Adolescent cannabis and tobacco use and educational outcomes at age 16: birth cohort study (2015)

Aims: To investigate the relationship between cannabis and tobacco use by age 15 and subsequent educational outcomes.
Design: Birth cohort study.
Setting: England.
Participants: The sample was drawn from the Avon Longitudinal Study of Parents and Children; a core sample of 1155 individuals had complete information on all the variables.
Measurements: The main exposures were cannabis and tobacco use at age 15 assessed in clinic by computer-assisted questionnaire and serum cotinine. The main outcomes were performance in standardized assessments at 16 [Key Stage 4, General Certificate of Secondary Education (GCSE)] in English and mathematics (mean scores), completion of five or more assessments at grade C level or higher and leaving school having achieved no qualifications. Analyses were sequentially adjusted for multiple covariates using a hierarchical approach. Covariates considered were: maternal substance use (ever tobacco or cannabis use, alcohol use above recommended limits); life course socio-economic position (family occupational class, maternal education, family income); child sex; month and year of birth; child educational attainment prior to age 11 (Key Stage 2); child substance use (tobacco, alcohol and cannabis) prior to age 15 and child conduct disorder.
Findings: In fully adjusted models both cannabis and tobacco use at age 15 were associated with subsequent adverse educational outcomes. In general, the dose-response effect seen was consistent across all educational outcomes assessed. Weekly cannabis use was associated negatively with English GCSE results [grade point difference (GPD), -5.93, 95% confidence interval (CI) = -8.34, -3.53] and with mathematics GCSE results (GPD, -6.91, 95% CI = -9.92, -3.89). Daily tobacco smoking was associated negatively with English GCSE (GPD, -11.90, 95% CI = -13.47, -10.33) and with mathematics GCSE (GPD, -16.72, 95% CI = -18.57, -14.86). The greatest attenuation of these effects was seen on adjustment for other substance use and conduct disorder. Following adjustment, tobacco appeared to have a consistently stronger effect than cannabis.
Conclusions: Both cannabis and tobacco use in adolescence are associated strongly with subsequent adverse educational outcomes. Given the non-specific patterns of association seen and the attenuation of estimates on adjustment, it is possible that these effects arise through non-causal mechanisms, although a causal explanation cannot be discounted.