Marijuana use and depression among adults: testing for causal associations

Aim To determine whether marijuana use predicts later development of depression after accounting for differences between users and non-users of marijuana. Design An ongoing longitudinal survey of 12 686 men and women beginning in 1979. Setting The National Longitudinal Survey of Youth of 1979, a nationally representative sample from the United States. Participants A total of 8759 adults (age range 29-37 years) interviewed in 1994 had complete data on past-year marijuana use and current depression. Measurements Self-reported past-year marijuana use was tested as an independent predictor of later adult depression using the Center for Epidemiologic Studies-Depression questionnaire. Individual's propensity to use marijuana was calculated using over 50 baseline covariates. Findings Before adjusting for group differences, the odds of current depression among past-year marijuana users is 1.4 times higher (95% CI: 1.1, 1.9) than the odds of depression among the non-using comparison group. After adjustment, the odds of current depression among past-year marijuana users is only 1.1 times higher than the comparison group (95% CI: 0.8, 1.7). Similarly, adjustment eliminates significant associations between marijuana use and depression in four additional analyses: heavy marijuana use as the risk factor, stratifying by either gender or age, and using a 4-year lag-time between marijuana use and depression. Conclusions After adjusting for differences in baseline risk factors of marijuana use and depression, past-year marijuana use does not significantly predict later development of depression. These findings are discussed in terms of their relevance for understanding possible causal effects of marijuana use on depression.