Umbrella review finds no serotonin link to depression

The idea that depression is caused by a chemical imbalance — specifically too little serotonin — has been enormously influential in how the condition is explained to the public and how drug treatment is justified. It shapes what people believe is wrong with them and why a medicine might help. Despite this, the underlying evidence had never been pulled together across all the different research strands that bear on the question. This review set out to synthesise and evaluate that evidence systematically, asking a narrow but foundational question: is depression actually associated with lowered serotonin concentration or activity?

The authors searched PubMed, EMBASE and PsycINFO from inception to December 2020 for systematic reviews, meta-analyses and large data-set analyses in the principal areas of serotonin research: serotonin and its metabolite 5-HIAA in body fluids, 5-HT1A receptor binding, serotonin transporter (SERT) levels from imaging or post-mortem tissue, tryptophan depletion experiments, and SERT gene associations and gene–environment interactions. Depression linked to physical illness and subtypes such as bipolar depression were excluded. Two reviewers independently extracted data and rated quality with AMSTAR-2, an adapted AMSTAR-2 or STREGA, and certainty was graded with a modified GRADE. Seventeen studies were included and the protocol was pre-registered.

Two meta-analyses of the metabolite 5-HIAA showed no association with depression (largest n = 1002), and a meta-analysis of cohort studies of plasma serotonin found no relationship, but did find lowered serotonin associated with antidepressant use (n = 1869). Studies of 5-HT1A receptors (largest n = 561) and SERT binding (largest n = 1845) gave weak, inconsistent signals that, if anything, would be consistent with increased rather than reduced synaptic serotonin — though prior antidepressant exposure was not reliably excluded. Tryptophan depletion produced no effect in most healthy volunteers (n = 566), with only weak evidence in those with a family history (n = 75). The two largest, highest-quality genetic studies (n = 115,257 and n = 43,165) found no association with depression and no gene–stress interaction.

This is a synthesis of existing syntheses rather than new data, and it directly challenges a mechanism that many people still take as established fact. The review does not claim depression has no biological basis, nor does it test whether antidepressants reduce symptoms — that is a separate literature with its own trials. What it does say is that the specific low-serotonin explanation is not supported by the main lines of evidence. The suggestion that long-term antidepressant use may itself lower serotonin concentrations is raised as a possibility consistent with some data, not as a demonstrated effect, and open questions remain about tryptophan depletion, which has not been systematically reviewed since 2007.

This is a reason to treat the familiar chemical-imbalance story as an explanatory shorthand rather than an established fact. It says nothing about whether any individual's treatment is working or should change, and decisions about antidepressant medication belong with a prescriber. For anyone interested in mental health more broadly, it is a useful reminder that a mechanism can be widely repeated for decades without the underlying evidence ever having been assembled and graded. The honest position from this review is that the cause of depression remains unresolved.