Eur Arch Psychiatry Clin Neurosci (2003) 253 : 115–119 DOI 10.1007/s00406-003-0427-4 ■ Abstract Before the rise of atypical antipsychotics, lithium used to be the most frequently investigated sub- stance in the acute treatment of bipolar disorders, al- though studies are not always of the highest method- ological standard. Due to the doubt about a sufficient efficacy of lithium expressed in recent years from vari- ous sides, and the simultaneous availability of newer treatment alternatives, this paper attempts to make a critical stocktaking of our knowledge about lithium in the acute treatment of bipolar disorders. Aspects con- cerning the changed disorder concept through the broadening of the bipolar spectrum, together with the available results from controlled and open studies with lithium, are presented and appraised. This shows that lithium should still be seen as an essential, but not the only corner stone in the differentiated treatment of bipolar patients. Provided that it is taken reliably and well-tolerated, lithium represents a first choice treat- ment, particularly for a classical course of manic-de- pressive illness (Bipolar I disorder), especially for mild to moderate manic syndromes. However, as antidepres- sive treatment, lithium should rather not be applied as a monotherapy, particularly in severe bipolar depression, since with the new generation of antidepressants and anticonvulsants well-tolerated and very effective alter- natives are available. In combination treatment, lithium should be applied particularly when it has already shown good prophylactic efficacy and/or in patients for whom a high suicide risk must be presumed. ■ Key words bipolar disorder · depression · lithium · mania Introduction Over 50 years ago, Cade first described the antimanic ef- ficacy of lithium. Even though his studies must rather be described as not complying to modern methodological standpoints, it was still the first time that psychiatrists were given an effective drug treatment for mania. Ex- tensive further studies, especially by Mogen Schou, fol- lowed in the subsequent decades, so that by the begin- ning of the 1970s lithium treatment of bipolar patients had become established in antimanic acute therapy, and particularly in relapse prophylaxis (Goodwin and Zis 1979; Schou 1997; Bowden 1998). Even before the dis- cussion about the switch risk of antidepressants was ini- tiated, the acute antidepressive efficacy of lithium con- tinued to be followed in smaller studies (Overview in Adli et al. 1998). However, due to the lack of commercial interest from pharmaceutical companies, large con- trolled studies were never performed, which are consid- ered conclusive for acute antidepressive efficacy accord- ing to the methodological standards expected today (EMEA guidelines). As regards antimanic treatment, in the meantime lithium has also satisfactorily delivered placebo-controlled proof of efficacy, so to speak via the back door, as a comparator substance for trials with new potential antimanic compounds (e. g., versus valproate as a test substance and placebo (Bowden et al. 1994)). Besides the question of the efficacy and side effect profile, the most important aspect of lithium treatment is the question which predictors make an individual pa- tient suitable for lithium treatment, i. e. make treatment success with simultaneous good tolerability at least probable. Most studies with lithium were conducted when the diagnostic criteria for a manic-depressive dis- order were still very tight. They corresponded to today’s bipolar I disorder according to ICD 10 with omission of so-called atypical symptoms (e. g., psychotic features, mixed states). Accordingly, most studies with lithium also just included such patients as a homogenous group. The newer Phase III studies also include only bipolar I SPECIAL ISSUE Heinz Grunze Lithium in the acute treatment of bipolar disorders – a stocktaking Accepted: 7 May 2003 EAPCN 427 Dr. H. Grunze () Psychiatrische Klinik der LMU Nussbaumstr. 7 80336 München Tel.: +49-89/5160-5335 Fax: +49-89/5160-5330 E-Mail: grunze@psy.med.uni-muenchen.de