Possible age-associated bias in reporting of clinical features of drug dependence: epidemiological evidence on adolescent-onset marijuana use.

AIMS: To probe recent evidence on apparent excess occurrence of marijuana dependence when marijuana smoking starts in adolescence. DESIGN AND PARTICIPANTS: A national sample of recent-onset marijuana users was identified within public data files of the National Household Survey on Drug Abuse (NHSDA), 1995-98 (1,866 adolescents and 762 adults). MEASUREMENTS: Marijuana dependence was assessed via seven standardized questions about its clinical features, such as being unable to cut down. Multivariate response models (GLM/GEE and MIMIC) were used to evaluate adolescent excess risk and possible item biases.
FINDINGS: Among people who had just started to use marijuana, clinical features of marijuana dependence occurred twice as often among adolescents compared to adults, even with statistical adjustment for other covariates (P < 0.01 from GLM/GEE). MIMIC analyses suggest that adolescent-onset users have somewhat higher levels of marijuana dependence, and they also provide evidence of age-associated response bias for some but not all clinical features of marijuana dependence. That is, even with level of marijuana dependence held constant, adolescent recent-onset users were more likely than adults to report being unable to cut down (P = 0.01) and tolerance (P = 0.029).
CONCLUSION: Nosologic, methodological and substantive reasons for observed age-related excess in occurrence of marijuana dependence problems among early onset users deserve more attention in future research.