Objective: Concerns have been raised that antidepressants may be overused. This study aimed to provide an estimate of antidepressant overuse in a more recent, nationally representative sample of adults and with a more contemporary set of antidepressants than has been covered in prior studies.

Methods: The data set included adult (weighted N=23,026,608) respondents who self-reported antidepressant treatment in the household and prescription drug components of the 2005 Medical Expenditure Panel Survey. Overuse was defined as off-label antidepressant prescribing with limited or no scientific support for use as a treatment for the diagnosis, according to the Physicians' Desk Reference, the United States Pharmacopeia-National Formulary, and the Micromedex DrugDx data system. Stratification and multivariate logistic regression was used to examine clinical and socioeconomic predictors of overuse.

Results: Overuse was estimated at 20%, with the majority concentrated in newer-generation antidepressants (74% of overuse). Another 30%-40% of overuse was associated with documented diagnoses that may represent a reasonable clinical rationale for antidepressant use or suggest underdiagnosis of possible depressive and anxiety syndromes. Older age (odds ratio [OR]=.95, p=.03) and self-report of poor mental health (OR=.80, p=.02) were negatively associated with overuse.

Conclusions: Antidepressant overuse among adults is less common than previously reported. Our results suggest that the actual extent of overuse may be lower than 20%. To improve treatment quality and the efficiency of the U.S. health care system, nationally representative data collection efforts on prescription drug use should aim to include enhanced measures of need in order to further refine future estimates of antidepressant overuse.