Soc Psychiatry Psychiatr Epidemiol (2002) 37: 199 – 206 10.1007/s00127-002-0541-z ■ Abstract Background This is an epidemiological study of a possible causal role of marijuana use in the development of Major Depressive Episode (MDE).Male- female differences in the suspected causal association have also been studied. Method Data are from 6,792 Na- tional Comorbidity Survey participants aged 15–45 years, assessed via the University of Michigan modified version of the Composite International Diagnostic In- terview (UM-CIDI). Survival analysis methods were used to estimate cumulative risk of MDE by levels of marijuana use and to estimate suspected causal associ- ations after adjustment for other influences. Results The risk of first MDE was moderately associated with the number of occasions of marijuana use and with more advanced stages of marijuana use. Relative to never users, non-dependent marijuana users had 1.6 times greater risk of MDE (95 % Confidence Interval: 1.1, 2.2), even with statistical adjustment for sex, birth cohorts, alcohol dependence, and history of daily tobacco smok- ing. Conclusions There was male-female variation in the degree of association between stage of marijuana in- volvement and MDE, but the strength of the association is modest at best. ■ Key words comorbidity – depressive disorder – marijuana abuse – sex – substance-related disorders – survival analysis Introduction The co-occurrence of depression and psychoactive drug use has been investigated in both clinical and commu- nity samples, with most studies focusing on a non-spe- cific category of drug use disorder (Anthony and Helzer 1991; Lin et al. 1996; Kessler et al. 1997; Kandel et al. 1999), on alcohol (Boyd et al. 1984; Blazer et al. 1994), or on tobacco (Breslau et al. 1993; Breslau et al. 1998; Wu and Anthony 1999). By contrast, very few studies have probed for a specific relationship between Major De- pression and marijuana, which is the most widely used illegal drug in North America and Western Europe. Nonetheless, the marijuana – depression association has some biological plausibility and there is some support- ive evidence of causal relationship (e. g., Grant and Pick- ering 1998; Hall and Solowij 1998; Troisi et al. 1998; Green and Ritter 2000). In this research, we have turned to data from public use files of the United States (US) National Comorbidity Survey (NCS) in order to study the co-occurrence of marijuana use disorders and Major Depressive Episode. Rather than estimate cross-sectional associations with- out reference to temporal sequences or age of onset in- formation, we have used survival analysis methods to evaluate marijuana use and other suspected risk factors as time-dependent variables, seeking to estimate the strength of association between frequency of marijuana use, the stage of marijuana involvement, and subsequent Major Depressive Episode. We have attempted to guide this work in relation to major insights gained in prior studies of psychiatric co- morbidities based upon field survey diagnostic inter- view data.For example,we have applied a procedure that makes no assumptions about which came first, the mar- ijuana use or the Major Depression. Instead, the se- quencing is determined by what the respondent says about the age of onset of the first marijuana use (a quite discrete and reliably ascertained event in the life history of users), and the age of onset of the first episode of Ma- ORIGINAL PAPER Chuan-Yu Chen · Fernando A.Wagner · James C.Anthony Marijuana use and the risk of Major Depressive Episode Epidemiological evidence from the United States National Comorbidity Survey Accepted: 15 January 2002 SPPE 541 C.-Y. Chen · F. A. Wagner · Professor James C. Anthony, Ph. D. () Johns Hopkins University Bloomberg School of Public Health Department of Mental Hygiene 624 N. Broadway, 8th fl. ELCID @ Rm. 893 Baltimore, MD 21205, USA E-Mail: janthony@jhu.edu F. A. Wagner Instituto Nacional de Psiquiatría Dirección de Investigaciones Epidemiológicas y Psicosociales Departamento de Investigación en Servicios de Salud México, DF, México