Effects of the rate of discontinuing lithium maintenance treatment in bipolar disorders.

BACKGROUND Gradual discontinuation of lithium may reduce high risk of early morbidity in bipolar disorder patients discontinuing successful long-term maintenance on lithium, but previous small samples have limited analyses of subgroups. METHOD DSM-IV bipolar disorder patients (N = 161) were pooled from similar samples maintained on lithium for 4.2 +/- 3.1 years. Effects of discontinuing treatment abruptly (1-14 days) or gradually (15-30 days) were compared by survival analysis in clinically closely similar groups. RESULTS After gradual versus rapid discontinuation, the overall median time to recurrence +/- SE differed by 5.0-fold (20.0 +/- 5.8 vs. 4.0 +/- 0.7 months; p < .0001). After rapid discontinuation, the median time in remission was 2.3 times shorter than the mean cycling interval before lithium (6.3 vs. 14.6 months; p < .0001). The proportion of subjects falling ill/month (recurrence rate) was much higher in the first year after rapid discontinuation (6.5% vs. 2.3%), but similar thereafter (0.4% vs. 0.6%); patients remained stable for 3 years when off lithium treatment 20 times more frequently after gradual than rapid discontinuation (37% vs. 1.8%; p < .0001). Ratios of median survival times after gradual/rapid lithium discontinuation were similar for a first recurrence of mania and depression (4.4 vs. 3.4-fold), insignificantly higher (34%) with rapid or continuous cycling before lithium, and greater in Type II than Type I disorder (9.8- vs. 4.0-fold). The polarity of first off-lithium and first lifetime episodes matched in 70% of cases. CONCLUSION These pooled results strengthen the concept or a pharmacodynamic stress factor in early relapse after stopping lithium maintenance and support the conclusion that early recurrence risk can be minimized by discontinuing maintenance treatment gradually in both Type I and II bipolar disorders.