Depression & the Monoamine Hypothesis

In 1952, the French Navy surgeon Henri Laborit — looking for a pre-operative sedative that wouldn't depress respiration — administered chlorpromazine to surgical patients and noticed they became strangely indifferent to their surgery. Within months, French psychiatrists Delay and Deniker had begun trying it on psychiatric patients. Within a year, the first antipsychotic drug had transformed asylum medicine. Within five years, the same accidental empirical pipeline had produced imipramine (originally an antihistamine; psychiatrist Roland Kuhn tried it on depressed patients in 1957) and iproniazid (originally an antitubercular drug observed to elevate mood). Both raised brain monoamine levels. By the early 1960s, the monoamine hypothesis of depression had crystallized: depression is caused by a deficit of monoamine neurotransmitters, and antidepressants work by correcting it. The hypothesis was simple, mechanistically plausible, clinically actionable, and substantially wrong.

The classical monoamine hypothesis crystallized in the mid-1960s as a clean inference from the accidental pharmacology that came before it: antidepressants raised serotonin, norepinephrine, or dopamine; reserpine, which depleted them, sometimes triggered depressive episodes; post-mortem studies showed altered monoamine metabolites. Depression was a deficit of monoamine neurotransmitters, antidepressants corrected the deficit, and the picture was simple and clinically actionable. SSRIs, beginning with fluoxetine in 1987, became the most-prescribed class of psychiatric drug in the world, with rates now exceeding ten percent of adults in many high-income countries. The trouble is that the hypothesis turned out to be substantially incomplete. The evidentiary case comes from several directions. SSRIs raise synaptic serotonin within hours but clinical effect takes weeks, so the proximate mechanism cannot be the rise itself. Tianeptine enhances serotonin reuptake (the opposite of an SSRI) and has comparable efficacy. Acute tryptophan-depletion studies in remitted depressed patients do not reliably trigger relapse. The 2022 Moncrieff systematic review found no robust association between serotonin-system genes and depression vulnerability. Effect-size meta-analyses by Kirsch and Cipriani show real but modest benefits over placebo, and roughly a third of patients meet criteria for treatment-resistant depression after multiple adequate trials. The replacement is messier and more interesting. The contemporary view is that depression is heterogeneous — several distinct biological subtypes (neuroinflammation, HPA-axis dysregulation, BDNF deficits, glutamatergic dysfunction, default-mode rumination) converging on similar phenomenology — and that antidepressants probably work through downstream effects on neuroplasticity and circuit-level reorganization rather than through direct monoamine elevation. The single strongest piece of evidence for this picture is ketamine: a single sub-anesthetic dose of an NMDA-receptor antagonist produces rapid antidepressant effects within hours, on a timescale incompatible with monoamine accumulation but consistent with rapid synaptic plasticity. Psychotherapy is comparably effective to medication for mild-to-moderate depression and better at preventing relapse, supporting the view that circuit-level reorganization is doing much of the therapeutic work.