Antidepressants: What They Actually Do, What the Evidence Shows, and What the Criticism Gets Right

Antidepressants are among the most prescribed medications in the world and among the most contested in popular discourse. The contest has two false poles: the pharmaceutical-promotional position ("antidepressants are effective, well-tolerated, and underused") and the antipsychiatry position ("antidepressants are no better than placebo, the serotonin hypothesis is false, and the drugs harm people"). Both are wrong in specific ways.

The Evidence on Efficacy

The largest and most methodologically rigorous meta-analysis of antidepressant efficacy was published by Cipriani et al. in The Lancet in 2018 — pooling 522 RCTs and 116,000+ participants across 21 antidepressants compared to each other and placebo.

Findings: All 21 antidepressants were more effective than placebo for acute-phase treatment of major depressive disorder. The odds ratios for response rates (typically defined as ≥50% symptom reduction) ranged from approximately OR 1.4 to OR 2.1 against placebo. The effect size in response rate terms: approximately 40–60% responding to antidepressant vs. 30–40% responding to placebo.

The absolute difference: approximately 15–20% more patients improve on antidepressant than on placebo. Real. Not dramatic.

Irving Kirsch's critique: Kirsch's analysis of FDA trial data found that the drug-placebo difference in symptom score reduction was below clinical significance thresholds for mild-to-moderate depression, with the difference reaching clinical significance primarily in severe depression. This is a legitimate finding — it does not mean antidepressants don't work; it means their effect is most meaningful for severe depression and less clear for mild-to-moderate.

> 📌 Cipriani et al. (2018) Lancet meta-analysis of 522 RCTs: all 21 antidepressants significantly more effective than placebo; sertraline (Zoloft) and escitalopram (Lexapro) showed favorable efficacy-tolerability profiles; agomelatine, mirtazapine, and amitriptyline showed highest efficacy; placebo response was substantial (30–40%) in all trials. [1]

The Serotonin Hypothesis

The "chemical imbalance" explanation for depression — that depression is caused by low serotonin and antidepressants work by correcting this deficiency — is a significant oversimplification that has been walked back by most of the psychiatric research community since the early 2000s.

A 2022 umbrella review by Moncrieff et al. published in Molecular Psychiatry concluded that the evidence for a simple relationship between serotonin function and depression is weak — a finding that was widely misrepresented as "proving antidepressants don't work." The serotonin hypothesis is not the mechanism of antidepressants; it was a flawed marketing simplification. The mechanisms of antidepressant action are genuinely complex and include neuroplasticity (BDNF upregulation), neurogenesis in the hippocampus, synaptic remodeling, and anti-inflammatory effects — none of which is serotonin balance.

Antidepressants work. The serotonin deficiency explanation for why they work was never accurate.

Withdrawl and Discontinuation

Discontinuation syndrome is underappreciated — abrupt cessation of many SSRIs and SNRIs produces withdrawal symptoms: dizziness, electric shock sensations ("brain zaps"), flu-like symptoms, anxiety, irritability. This is not addiction (cravings, dose escalation patterns are absent) — it is physiological dependence with withdrawal.

The practical implication: antidepressants should be tapered, not stopped abruptly, and patients should be informed of discontinuation syndrome risk before starting medication.

---

Key Terms

• SSRI (Selective Serotonin Reuptake Inhibitor) — the most commonly prescribed antidepressant class; inhibits the reuptake transporter SERT, increasing synaptic serotonin concentration; the mechanism's relationship to antidepressant efficacy is more complex than the simple serotonin level increase the name implies
• Placebo response — the symptom improvement observed in patients randomized to placebo in RCTs; in antidepressant trials, approximately 30–40%; the variable that makes the drug-placebo difference smaller than absolute drug response rates suggest
• BDNF (Brain-Derived Neurotrophic Factor) — the neuroplasticity signaling factor upregulated by antidepressants; promotes hippocampal neurogenesis and synaptic remodeling; the mechanistic candidate for the neurobiological effects of antidepressants that is more consistent with the evidence than serotonin balance
• Discontinuation syndrome — the physiological withdrawal-like symptoms that occur with abrupt cessation of SSRIs/SNRIs; distinguished from addiction by the absence of craving and dose escalation; managed by gradual tapering; should be disclosed before initiating antidepressant treatment

---

Scientific Sources

• 1. Cipriani, A., et al. (2018). Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: A systematic review and network meta-analysis. The Lancet, 391(10128), 1357–1366. PubMed
• 2. Moncrieff, J., et al. (2022). The serotonin theory of depression: A systematic umbrella review of the evidence. Molecular Psychiatry, 28, 3243–3256. PubMed