58 [95% confidence interval (CI) 0.73–3.42] compared with placebo and 2.37 (95% CI 1.38–4.05) compared with mood stabilizers, although this is debated by Sidor and MacQueen, whose meta-analysis suggests no increase in switching for selective serotonin reuptake inhibitors (SSRIs) or bupropion [Sidor and MacQueen, 2010]. However, they nuance this by arguing that cut-off criteria for switching vary: some authors have used the Young Mania Rating Scale (YMRS) to quantify affective change, but the score required to qualify as switching differs between studies, and

indeed some Inhibitors,research,lifescience,medical research reports do not clearly delineate how such change was defined. Agreement on what constitutes switching remains a challenge, as does the nosological FXR agonist validity and long-term significance of so-called subthreshold categories. This is undoubtedly problematic for research and clinical practice, although, given the

demonstrated issue of pathological mood elevation being missed, particularly in milder instances, it would seem reasonable to conclude that many patients fall into this category Inhibitors,research,lifescience,medical with detrimental longer-term outcomes. There is evidence for interclass differences in switching, with tricyclic antidepressants showing an absolute risk difference of 6.8% (95% CI 1.7–11.9%) compared with other classes [Gijsman et al. 2004]. Bupropion has a lower rate of switching than dual-acting venlafaxine and tricyclic antidepressants Inhibitors,research,lifescience,medical but shows no difference compared Inhibitors,research,lifescience,medical with SSRIs [Sidor and Macqueen, 2010].

Thus, there is little valid research to support antidepressant use, and there is evidence that such medications increase the risk of affective switching. Antidepressant augmentation therapy Nemeroff and colleagues failed to show any statistically Inhibitors,research,lifescience,medical significant difference in the addition of either imipramine or paroxetine compared with placebo in subjects on lithium; indeed, subanalysis showed that in the high-lithium subgroup (serum levels >0.8mmol/l) lithium and placebo was superior to lithium and an antidepressant [Nemeroff et al. 2001]. Ghaemi and colleagues undertook a meta-analysis of RCTs in which patients were followed up for more than 6 months: they found that addition of antidepressants to mood stabilizers was not superior to mood stabilizer monotherapy in longer-term management [Ghaemi et al. 2008]. Furthermore, whilst long-term treatment regimens that included antidepressants lowered the risk of recurrence of depression by 27% (pooled relative risk 0.73, 95% CI 0.55–0.97) relative Molecular Cell to mood stabilizer monotherapy, but at the cost of a 72% increase in the risk of new episodes of mania (relative risk 1.72, 95% CI 1.23–2.41). The Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD) is a recent, large, multisite, collaborative study in BPAD that used the more naturalistic outcome measure of remaining euthymic for 8 consecutive weeks. One arm looked at bipolar depression [Sachs et al.