I was interested by the retrospective by Schou1 on lithium treatment and the accompanying commentaries. However, some important recent developments and debates concerning lithium were not mentioned. Several commentators speculate on the reasons for the discrepancy between the dramatic results of the early placebo-controlled trials of lithium and subsequent comparative trials and naturalistic data. In an article published in 1995,2 I pointed out that a possible explanation for this situation lies in the designs of these early studies. Many of them were discontinuation studies in which patients who had been receiving lithium were allocated to either continue receiving lithium or to have placebo substitution. This type of design cannot address the question of the efficacy of prophylaxis, particularly because of mounting evidence that lithium withdrawal induces manic relapse. Recent reviews on the subject have concluded that there is substantial evidence of a lithium withdrawal–related phenomenon.3,4 It is most surprising that this issue was not addressed in the series of articles in the ARCHIVES, in view of its implications for the use of lithium and its bearing on the evidence for efficacy.
Moncrieff J. Forty Years of Lithium Treatment. Arch Gen Psychiatry. 1998;55(1):92-93. doi: