The trouble with antidepressants: why the evidence overplays benefits and underplays risks—an essay by John B Warren

Dear Editor,

John Warren's essay rightly notes some of the drawbacks and limitations of antidepressant medication, but gives short shrift to the more nuanced aspects of antidepressant treatment, including its benefits.

Warren begins by asserting that, "A common justification for using antidepressants is that they correct a chemical deficiency in the brain." Few psychiatrists in the U.S. would invoke such an outdated hypothesis, commonly and misleadingly called, "the chemical imbalance theory" of depression. More recent work points to more subtle mechanisms of antidepressant action, such as effects on brain-derived neurotrophic factor (BDNF), which has downstream effects at the level of the gene. [1]

But more to the point: the main justification for antidepressant use is that these agents help patients with moderate-to-severe unipolar depression. The data cited by Warren to the contrary omits discussion of the limitations of, for example, the Hamilton Depression Scale (HAM-D). Specifically, Warren refers to "the 0.3 standardised mean difference in effect size seen with antidepressants" vs. placebo controls. However, Bech demonstrated that a reallocation of HAM-D items focusing solely on those measuring severity of clinical depression—the HAM-D-6—has yielded effect sizes of 0.40 or greater for second-generation antidepressants, in placebo-controlled trials.[2] This compares favorably with the effect size of approximately 0.30 in meta-analyses of studies using some version of the full HAM-D scale.

Moreover, as I have argued in a recent editorial [3], it is easy to become fixated on the HAM-D as the alpha and omega of antidepressant effectiveness. This, in my view, is a serious mistake. When we look, for example, at the construct of "quality of life" (QOL), we find substantial (though not conclusive) evidence that modern-day antidepressants improve QOL for many depressed patients. For example, a study by Demyttenaere et al [4] examined the effect of escitalopram treatment on QOL in patients with major depressive disorder (MDD) (n = 1140) or generalized anxiety disorder (n = 1045), using the Quality of Life Enjoyment and Satisfaction Questionnaire. Data were obtained from 8 randomized, 8-week, double-blind, placebo-controlled clinical trials with escitalopram. Treatment resulted in statistically and clinically significant improvement in QOL enjoyment and satisfaction, and the improvement was greater in patients treated with escitalopram than those treated with placebo.

Finally, in my view, Warren overstates the problem of antidepressant withdrawal syndromes, based on inadequate study methods. For example, he cites the much-publicized study by Davies and Read [5], which--based largely on internet survey data--found that over half of people taking antidepressants experience withdrawal symptoms; and that these symptoms were severe in over half of such cases. First of all, such online surveys can, at most, provide information on what patients retrospectively report to survey takers. Such data are not able to establish the actual incidence (or prevalence) of the phenomenon in question, which would require (1) a validated set of uniform diagnostic criteria for “withdrawal”; and (2) careful--ideally, contemporaneous--clinical observation to confirm that subjects actually meet the specified criteria. Moreover, even if we accept the Davies and Read report at face value, it actually shows that the vast majority of persons reporting withdrawal effects--about 74%--do not experience severe symptoms upon discontinuation of antidepressants. Indeed, when second-generation antidepressants are tapered very gradually--over, say, a period of 2 months--serious withdrawal symptoms are rare, particularly when paroxetine and venlafaxine are avoided. [6]

To be sure, the benefits of antidepressants were probably oversold in the 1990s--largely as a result of "Big Pharma" advertising and selective publication of favorable data--and some of their adverse effects were understated. Nevertheless, there is compelling evidence that these medications can greatly improve the quality of life of many patients with clinical depression. And, when carefully managed, antidepressant treatment is safe, effective, and well-tolerated by most patients.

Ronald W. Pies, MD

1. Cattaneo, A., Cattane, N., Begni, V. et al. The human BDNF gene: peripheral gene expression and protein levels as biomarkers for psychiatric disorders. Transl Psychiatry 6, e958 (2016). https://doi.org/10.1038/tp.2016.214

2 Bech P. Is the antidepressive effect of second-generation antidepressants a myth? Psychol Med. 2010;40:181–186.

3 Pies RW. Antidepressants, the Hamilton Depression Rating Scale Conundrum, and Quality of Life. J Clin Psychopharmacol. 2020;40(4):339-341. doi:10.1097/JCP.0000000000001221

4 Demyttenaere K, Andersen HF, Reines EH. Impact of escitalopram treatment on Quality of Life Enjoyment and Satisfaction Questionnaire scores in major depressive disorder and generalized anxiety disorder. Int Clin Psychopharmacol. 2008;23:276–286.

5 Davies J, Read J. A systematic review into the incidence, severity and duration of antidepressant withdrawal effects: are guidelines evidence-based?Addict Behav 2019;97:111-21. doi: 10.1016/j.addbeh.2018.08.027 pmid: 30292574

6 Pies RW. Antidepressant Discontinuation: A Tale of Two Narratives. J Clin Psychopharmacol. 2019;39(3):185-188. doi:10.1097/JCP.0000000000001021