Increased REM sleep density at admission predicts relapse by three months in primary alcoholics with a lifetime diagnosis of secondary depression

Background: Having previously reported that 3-month relapse was associated with increased admission REM pressure in nondepressed primary alcoholics, we hypothesized that baseline polysomnography would predict outcome in primary alcoholics with a lifetime diagnosis of secondary depression. Methods: Twenty-one primary alcoholics with secondary depression received polysomnography and the Hamilton Depression Rating Scale during the first and fourth weeks of a 1-month inpatient alcohol treatment program, Exclusion criteria included serious illness, current major alcohol withdrawal symptoms, other Axis I diagnoses, sleep apnea, nocturnal myoclonus, and psychoactive substances within 14 days of polysomnography. Relapse was defined as drinking any alcohol between hospital discharge and 3-month follow-up. Results: Relapsers' total sleep time was reduced and REM density (reflecting REM sleep ocular activity) was increased significantly throughout admission compared with abstainers. Sleep continuity and Hamilton scores improved by discharge in sober and relapsing alcoholics. Factors derived from admission REM latency, REM percent, and REM density predicted sobriety vs. relapse within 3 months after hospital discharge in 76% of patients. Admission REM density was greater, and total sleep time was less in relapsers than in patients sober at 3 months. Conclusions: Results suggest that increased REM density and decreased total sleep time at about 2-4 weeks of abstinence predict relapse by 3 months in depressed alcoholics. Published 1998 Society of Biological Psychiatry.