Lithium treatment of mania: clinical characteristics, specificity of symptom change, and outcome.

The effects of lithium treatment and prediction of response in 18 manic patients were studied as part of the National Institute of Mental Health Collaborative Study of the Psychobiology of Depression. Patients were rated using the Mania Diagnostic and Severity Scale (MADS) and an additional battery of behavioral constructs developed for measurement of state and drug response in depressed patients. About 67% of the patients had good treatment outcome after 26 days of lithium treatment. Responders did not differ from nonresponders before treatment with respect to delusions, hallucinations, or irritable-paranoid symptoms. Nonresponders were rated as more anxious than responders before treatment and had higher scores on the Hamilton Rating Scale for Depression. Improvement on the MADS, however, did not correlate with pretreatment behavioral ratings. Patients with relatively high ratings for aspects of behavior not specific to mania tended to improve in these regardless of change in MADS score. Manic patients who were also depressed (44%) had higher mania ratings than manic patients who were not depressed. Patients with concomitant depression and mania had significantly worse overall treatment outcome, although their depression ratings improved during lithium treatment.