Marriage was associated with substantial reductions in substance use compared to being divorced/separated or single, and these reductions began prior to marriage. The co-twin comparison and within-person models ruled out the alternative explanation that marriage effects were due to confounding backg …
