A Cannabis–Education Association Was No Longer Significant After Confounder Adjustment
In a Finnish birth-cohort analysis, the association between adolescent cannabis use and later educational attainment was no longer significant after additional adjustment. That finding neither proves a causal cannabis effect nor establishes that cannabis cannot affect education.
Adolescent reports linked to later registers
Researchers related substance-use reports at ages 15–16 to educational attainment through age 33. The analysis considered family characteristics, childhood behavioral and emotional problems, psychiatric diagnoses and alcohol use. Exposure was self-reported rather than established by blood, urine or hair testing.
The model changed the question
An earlier adjusted model linked cannabis or other illicit-substance use with lower odds of college or university education. After family-related factors and alcohol-intoxication frequency were added, the cannabis comparisons were nonsignificant. The estimate therefore depended on which overlapping risks were accounted for.
Alcohol findings were different
Several adolescent alcohol measures retained associations with poorer educational outcomes. These should not be silently transferred to cannabis, nor should the cannabis null be used to erase the alcohol results. Each exposure and model has its own interpretation.
What the null does not establish
Cannabis use was relatively uncommon in this cohort, limiting precision, and exposure was assessed at one adolescent time point. The study could not fully characterize later use or eliminate residual confounding and reverse pathways. Its historical Finnish setting may also differ from current products and patterns. The useful conclusion concerns the robustness of this particular association, not a prediction of an individual student’s future or a universal statement of safety.
Related analysis: evaluating cannabis tolerance claims.
Primary research
Levola J; Alakokkare AE; Denissoff A; Mustonen A; Miettunen J; Niemelä S. Adolescent alcohol and cannabis use and early adulthood educational attainment in the 1986 Northern Finland birth cohort study. BMC public health (2024). DOI: 10.1186/s12889-024-17693-w.
