Addressing the risks of antidepressants among people with alcohol use disorders

We thank Bahji and colleagues1 and Elefante and colleagues2 for their interest in the guideline on the management of alcohol use disorder (AUD).3 We appreciate the opportunity to explain that the guideline’s caution against prescribing certain antidepressants was based not only on the evidence that some antidepressants appear to be largely ineffective among those with AUD and comorbid depression or anxiety,4^,5 but also the literature demonstrating risks of increased alcohol use with certain antidepressants.

Bahji and colleagues cited several individual studies suggesting benefits of selective serotonin reuptake inhibitors (SSRIs), but most of these were not placebo-controlled trials. Conversely, various meta-analyses have found SSRIs to have no effect on depressive symptoms among people with AUD.4^–6 In particular, the Cochrane review by Agabio and colleagues5 found no effect with SSRIs on final depression score, change in score, number of responders, or number of remissions relative to placebo. To Elefante and colleague’s comments, the review also found no benefit of SSRIs versus placebo for either abstinent or heavy drinking days.5 Similarly, the relevant Cochrane review by Ipser and colleagues6 concluded that “the effectiveness of medication in treating anxiety disorders and comorbid alcohol use disorders is currently inconclusive.”

Bahji and colleagues also cited a study showing efficacy for combined SSRI with naltrexone in the treatment of patients with AUD and major depressive disorder, but a subsequent meta-analysis including that study concluded SSRIs “either alone or in combination with relapse prevention medications such as naltrexone, had no significant effect on depressive symptoms.”4

In addition, the guideline summarized an underappreciated literature demonstrating that certain antidepressants appear to increase alcohol use among some patients. Although Elefante and colleagues described a single trial,7 at least 6 double-blind placebo-controlled trials have shown risks of increased alcohol use with SSRIs.7^–12 Elefante and colleagues are correct that, as reviewed in the guideline, some studies have found this concern most relevant among those who had earlier onset and more severe AUD. In addition, a functional polymorphism (5-HTTLPR) in the serotonin transporter gene may predispose those with AUD treated with SSRIs to either positive (i.e., less drinking) or negative outcomes (i.e., increased drinking).6 A randomized controlled trial (RCT) that critically examined this question concluded that, since the genotype implicated in SSRI-induced heavier alcohol use was more common than the genotype affording benefit, an estimated twice as many patients would be “adversely affected” by SSRI prescribing and that the “widespread use of antidepressants suggest that the findings reported here are relevant to a substantial proportion of the US population.”7

To the comments about trazodone, a serotonin receptor antagonist and reuptake inhibitor, this RCT was presented to further demonstrate the risks of increased alcohol use with another serotonergic medication that is commonly used in the treatment of AUD, in comparison to placebo.13 Although increasing a medication’s dose would generally not be expected to reduce adverse effects, additional studies would be required to assess the suggestion that increasing the trazodone dose or combining trazodone with anti-craving medications would limit the possible adverse effect of increased drinking.13 In sum, the current evidence justifies the current recommendations in the guideline.