β-Blockers and Depression

Objective. —To evaluate the relationship between β-blockers and depression. Design. —Case-control study. Setting. —New Jersey Medicaid recipients during July 1980 to December 1983. Participants. —New depression case patients (N = 4302) were identified from Medicaid claims for depression markers (antidepressant drugs, in-hospital depression diagnosis, or electroconvulsive therapy). Control patients were randomly selected and matched on the basis of Medicaid enrollment on the case patients' date for first depression marker (index date), birth year, sex, race, and nursing home residency status. Main Exposure Measure. —β-Blocker use as evidenced by prescription claims in the year before the index date. Results. —Case patients overall were more likely to have taken β-blockers (simple, matched odds ratio [OR] of 1.45; 95% confidence interval [CI], 1.29 to 1.62). Controlling for confounders (benzodiazepine use, frequent outpatient visits, and frequent use of medications other than β-blockers) resulted in a null effect (OR=0.98; 95% CI, 0.87 to 1.12). The ORs were consistently lower for case patients with a depression diagnosis or electroconvulsive therapy than for cases with only antidepressant use as a marker. These results did not vary by age, sex, race, nursing home status, or use of other selected specific medications. Conclusions. —Ongoing β-blocker use was not causally related to markers of depression. The difference between this study and those it contradicts is that this one identified certain confounding variables that accounted for the apparent relationship. (JAMA. 1992;267:1783-1787)