Relationship of substance abuse to dependence in the U.S. general population.

OBJECTIVE: The diagnostic categories of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, for substance abuse and dependence are commonly used in clinical work and research studies, but whether abuse and dependence represent two different syndromes has been debated. The purpose of this article is to investigate the relationship of substance abuse and dependence for cannabis, cocaine, stimulants and sedatives among lifetime users of these substances in the National Epidemiologic Survey on Alcohol and Related Conditions, a nationally representative survey conducted in 2001-2002.

METHOD: The multiple indicators multiple causes (MIMIC) model addresses three sets of relationships: those between (1) diagnostic criteria and latent factors, (2) latent factors and covariates, and (3) criteria and covariates. This approach allows for the detection of and compensation for noninvariance of the measurement of criteria across subgroups.

RESULTS: Compared with one-factor models, two-factor models (factors roughly corresponding to abuse and dependence) fit significantly better across all substances, with abuse and dependence factors highly correlated. The MIMIC model indicated that race/ethnicity, age, income, and marital status showed some differential relationships across substance groups, although most covariates showed similar associations to dependence and abuse factors. Noninvariance of criteria measurement by demographic covariates was most pronounced for cannabis abuse and dependence criteria.

CONCLUSIONS: The general relationship of abuse to dependence was consistent across substances. Results were equivocal on the value of retaining separate factors; therefore, investigating the relationships of specific genetic variants and treatment outcomes to dimensional indicators of abuse, dependence, and measures combining these criteria is warranted. Measurement of cannabis abuse and dependence criteria appears most affected by demographic characteristics.