Genetic evidence suggests cannabis use disorder may lower educational attainment, while education reduces disorder risk

Bidirectional Mendelian randomization found genetic liability to cannabis use disorder was associated with 1.2 fewer months of education (P=0.0008). Lifetime cannabis use showed no significant effect on education. In the reverse direction, higher educational attainment was associated with 61% lower CUD risk (OR 0.39, P=1.69x10^-10) but 35% higher lifetime cannabis use (OR 1.35, P=0.003). The CUD-education relationship may be partially explained by shared ADHD risk factors.

Bidirectional two-sample Mendelian randomization using genome-wide association study summary statistics from European ancestry samples (55,374 to 632,802 participants). Primary method was inverse-variance weighted MR with sensitivity analyses. Multivariable MR adjusted for intelligence, smoking initiation, and ADHD.

This study uses genetic instruments to approximate causal inference, suggesting the relationship between cannabis and education is more nuanced than observational studies indicate. Problematic use may impair education, while education may increase experimental use but protect against problematic use.

The finding that education increases lifetime cannabis use but decreases cannabis use disorder challenges the assumption that any cannabis use leads to problematic outcomes. The distinction between use and disorder matters.

Limited to European ancestry populations. Mendelian randomization assumptions (no pleiotropy, no direct instrument effects) may not fully hold. Shared ADHD risk factors complicate interpretation. Cannot specify mechanisms by which CUD affects education.