Psychopharmacologic strategies for the prevention of suicidal behavior in bipolar patients Maria A. Oquendo * , Andres Barrera, J. John Mann New York State Psychiatric Institute, Columbia University, 1051 Riverside Drive, Unit 42, New York, NY 10032, USA Received 5 June 2001; received in revised form 22 August 2001; accepted 27 August 2001 Abstract Current pharmacological strategies for the prevention of the suicide behavior in bipolar patients are reviewed. Additionally, these studies are discussed in the context of a stress–diathesis model, to explore whether this model explains the empirical fact that some drugs appear to have antisuicidal properties while others do not. A review of the relevant literature suggests that lithium and serotonin enhancing anti- depressants reduce suicidal behavior in bipolar patients. A stress–diathesis model explains the differential effect of such medications compared to other antidepressants or mood stabilizers by proposing additional effects of these medications on the diathesis for suicidal behavior. This effect may be mediated by augmentation of serotonergic function, which is linked to suicidal behavior. Serotonergic enhancing drugs therefore can potentially reduce suicidal behavior. q 2001 Elsevier Science B.V. All rights reserved. Keywords: Bipolar disorder; Pharmacotherapy; Lithium; Anticonvulsants; Suicidal behavior; Antipsychotics 1. Introduction Patients with bipolar disorders (BD) are at high risk for suicide completion [1] compared to other major psychiatric disorders. They also have a high rate of suicide attempts [2] the best predictor of future completed suicide. A recent comprehensive review and meta-analysis [3] found a weighted mean of 18.9% mortality from suicide in subjects with BD and that 25–50% of all bipolar patients report at least one suicide attempt. The suicide rate in BD is amongst the highest in major psychiatric disorders. When patients with psychotic disor- ders were followed for up to 8 years, bipolar patients had an estimated annual rate of suicide of 2.5% and a lifetime rate of suicide of 14% [4]. Although BD is less common than unipolar major depression (lifetime prevalence of BD, 1.1%; 4.9% for unipolar depression) [5,6], in a series of 100 suicides of subjects with primary major depression, 47 subjects had bipolar disorder [7], indicating a high-risk of suicide in BD. Many [4,7,8] but not all studies [9,10] have found bipolar subjects to be at higher risk for completed suicide than unipolars. Because a lifetime suicide rate as high as 15 to 20% has been estimated for bipolar patients [3,11], evaluating the antisuicidal efficacy of phar- macologic interventions is crucial. A high-risk of suicide attempts in BD has been documen- ted in community and clinical populations, as well. Twenty nine percent of bipolar subjects interviewed in the commu- nity acknowledged suicide attempts, compared to 15.9% of unipolar subjects and 4.2% of subjects with any Axis I diagnosis [2]. In clinical samples, bipolar subjects are over-represented among depressed suicide attempters [12]. Also, bipolar subjects appear more likely to have a history of suicide attempts with at least a moderate intent to die than unipolar subjects, although the difference appears to be statistically significant only for female subjects [13]. Simi- larly, bipolar I patients who present with a major depression are significantly more likely to attempt suicide during a two year follow-up period than unipolar depressed patients [14]. Thus, patients with BD, regardless of the setting, are at high- risk for both suicide attempt and completion. This review will discuss current pharmaco-therapeutic approaches to bipolar disorder and what is known regarding their effect on suicidal behavior. These therapeutic approaches will be discussed in the context of a stress– diathesis model. 2. A stress–diathesis model of suicidal behavior We have previously described a model in which suicidal Clinical Neuroscience Research 1 (2001) 387–393 1566-2772/01/$ - see front matter q 2001 Elsevier Science B.V. All rights reserved. PII: S1566-2772(01)00042-1 www.elsevier.com/locate/clires * Corresponding author. Tel.: 11-212-543-5835; fax: 11-212-543-6017. E-mail address: moquendo@neuron.cpmc.columbia.edu (M.A. Oquendo).