Discrepancies in the Efficacy of Lithium

In her letter titled "Forty Years of Lithium Treatment," Moncrieff¹ attempts to explain the striking discrepancy between the results of the early placebo-controlled trials of lithium and the low efficacy reported in later observations. She believes the evidence for lithium's efficacy is invalid, as several of the early placebo-controlled studies used discontinuation design, and points to the mounting reports that lithium discontinuation may induce manic relapse. However, this phenomenon did not exist in the early systematic studies of lithium discontinuation and was not observed in the discontinuation trials to which Moncrieff is referring. In fact, in comparison with the recent observations, patients at that time responded to lithium discontinuation in a quantitatively and qualitatively different manner.

The early studies focused on typical, episodic affective disorders diagnosed in the International Classification of Diseases (Kraepelinean) tradition. The recurrences after discontinuation developed after a period expected from their untreated, prelithium course,^{2 - 4} and were predominantly depressive. Similarly, the overactive, manialike rebound was not observed after discontinuation in recent long-term follow-up studies of episodic affective disorders.^{5 - 6}

An explanation alternative to Moncrieff's speculation is that patient groups given lithium in the early and recent trials were dissimilar. Since the 1970s, there has been a significant change in the way mood disorders are diagnosed. Baldessarini and colleagues^{7 - 8} have shown how psychiatrists shift their diagnostic preferences Patients diagnosed with recurrent mood disorders in recent studies differ markedly from those given the diagnosis of recurrent affective disorders in 1970.⁹ Affective disorders are diagnosed much more frequently,¹⁰ age at onset is much lower, premorbid psychopathologic condition is fully acceptable and genetic loading among the relatives has increased markedly. The clinical course of mood disorders is much more chronic¹¹ and recurrences and rapid cycling are more frequent than in the 1970s. With the inclusion of mood incongruent psychotic phenomena among the symptoms of bipolar illness, one has departed far from the typical, classic manic-depressive illness, for which lithium was proven effective.

When patients are diagnosed with bipolar disorder according to one primarily cross-sectional system, Diagnostic and Statistical Manual of Mental Disorders, they cannot be expected to respond to long-term lithium treatment and its discontinuation in the same manner as patients diagnosed with bipolar disorder according to a different—mainly longitudinal (International Classification of Diseases, Kraepelinean)—system. Having this expectation, Moncrieff arrives at the wrong conclusions.