← Reference library

REFERENCE LIBRARY

Serotonin SystemSerotonin System

A few tens of thousands of cells in the brainstem, wiring axons across the entire brain — and then more than a dozen different receptors translating one molecule into a dozen different meanings. Which is why "more serotonin = happier" was the wrong question from the start.

Tiny origin, enormous reach

Serotonin (5-HT) is made in a string of small nuclei along the brainstem midline called the raphe nuclei. There are only tens of thousands of these cells in total — negligible against the brain's ~86 billion neurons. But their axons spray out like a sprinkler truck and cover nearly the whole central nervous system: cortex, thalamus, hippocampus, amygdala, hypothalamus, all the way down to the spinal cord.

That "few cells, whole-brain projection" architecture dictates how it works. It usually doesn't order downstream neurons to fire or not fire; it changes how they respond — same input arriving, but the reaction is stronger or weaker, faster or slower. This job description is called neuromodulation, and dopamine, noradrenaline and acetylcholine all play the same kind of role.

One molecule, a dozen different locks

Here's the step that matters: serotonin carries no meaning of its own. The receptor supplies the meaning. Humans have more than a dozen known 5-HT receptors across seven families, and their effects can be flatly opposite — some make a cell easier to fire, others harder. So asking whether serotonin is "high or low" is like asking whether the key went in without asking which door.

5-HT1A

Mostly inhibitory. On the raphe cells themselves it also acts as a thermostat: more serotonin around means the raphe fires less. A frequent target for anti-anxiety drugs.

5-HT2A

Excitatory, densest on layer-5 cortical pyramidal neurons. Every classic psychedelic (LSD, psilocybin, DMT, mescaline) works by activating it; many antipsychotics do the opposite and block it.

5-HT2B

Abundant on heart valves. Chronic repeated stimulation thickens and fibroses them — the reason fenfluramine and pergolide were withdrawn. Any "take a little every day for years" regimen has to clear this hurdle first.

5-HT3

The one exception to slow modulation: it is itself an ion channel, so it acts fast. Best known in the gut and the brainstem vomiting centre — the most widely used class of anti-nausea drugs simply blocks it.

The rest (1B/4/6/7…)

Vascular tone, gut motility, circadian rhythm and more. The triptans used for migraine act mainly at 5-HT1B/1D.

What it actually governs — and it isn't "happiness"

Pooling animal and human data, serotonin looks less like a volume knob on mood and more like a set of parameters about time and cost: patience and waiting (with more serotonin around, animals hold out longer rather than grabbing the small immediate reward), behavioural inhibition (braking when braking is called for), sensitivity to aversive outcomes, plus the baseline rhythms of sleep, appetite and body temperature. If dopamine is roughly "is it worth going for," serotonin is closer to "is it worth waiting a bit longer."

One statistic is routinely misread: roughly 90% of the body's serotonin sits in the gut, not the brain. But serotonin barely crosses the blood–brain barrier, so that 90% does not translate into brain levels — gut-to-brain influence runs through other routes entirely (vagus nerve, immune signalling, metabolites).

SSRIs and the retreat from "chemical imbalance"

SSRIs, the most prescribed antidepressant class, do something very specific: they block the transporter that hauls serotonin back out of the synaptic cleft, so serotonin piles up there. That step is done within hours — yet patients typically take two to three weeks to improve. The lag is itself the strongest clue available: whatever does the treating isn't the restocking, it's the slow downstream adaptation that follows — receptor density shifting, neurotrophic signalling, synaptic remodelling.

Alongside this sits an influential 2022 umbrella review that gathered the evidence for "depression = low serotonin" and found it unsupported. That is not the same as "SSRIs don't work" — that's a separate question, settled by trials — but it does mean the reason they work is probably not the story we were told.