Depression: The Neuroscience Behind It

Depression is one of the leading causes of disability worldwide. The World Health Organization estimates that in 2023 about 322 million people had depression, including 5.2% of adults (4.1% of men and 6.2% of women), and that an estimated 727,000 people died by suicide in 2021. Yet for decades, the dominant public narrative has been reductive to the point of being misleading: "depression is a chemical imbalance — low serotonin." That explanation fit neatly on a pharmaceutical brochure. It was also, at best, a gross oversimplification — and at worst, a story that delayed real scientific understanding.

Educational content, not medical advice. Talk to your doctor before changing any medication, treatment, or routine. Information here reflects current research but every person's situation differs.

If you are struggling right now. If you are having thoughts of suicide or self-harm, contact local emergency services or a crisis line. In the US you can call or text 988 (Suicide & Crisis Lifeline); in the UK and Ireland, Samaritans answer on 116 123. Elsewhere, findahelpline.com lists free, confidential services by country. You do not need to be in immediate danger to reach out.

What we now know is considerably more interesting. Depression involves a cascade of biological changes spanning brain circuits, the stress-hormone system, the immune system, and even the gut. Understanding that complexity does not diminish the suffering of people with depression — it takes it more seriously.

What Depression Is, Clinically

Before the biology, a definition. The WHO describes a depressive episode as depressed mood or loss of pleasure or interest lasting most of the day, nearly every day, for at least two weeks, usually alongside other symptoms such as poor concentration, excessive guilt, hopelessness, disrupted sleep, appetite change, fatigue, and thoughts of death or suicide. Episodes are graded mild, moderate or severe by the number and severity of symptoms and their effect on daily functioning.

That definition matters for the neuroscience, because it is built from symptoms, not from a biological test. Two people can both meet the criteria while sharing only one or two symptoms. One may sleep 14 hours and gain weight; another may wake at 4 a.m. and stop eating. Many researchers now suspect "major depressive disorder" is an umbrella over several biologically distinct conditions, which helps explain why no single biomarker has emerged and why treatments help some people dramatically and others barely at all.

The Serotonin Debate — and Why It Matters

In 2022, psychiatrist Joanna Moncrieff and colleagues published a landmark umbrella review in Molecular Psychiatry examining decades of research on serotonin and depression. Their conclusion was striking: there is no consistent evidence that depression is caused by low serotonin activity or low serotonin levels.

Studies comparing serotonin metabolites in depressed versus non-depressed people showed no reliable difference. Tryptophan depletion experiments — which sharply reduce serotonin — did not reliably induce depression in healthy volunteers. The "chemical imbalance" framing, it turns out, was a marketing simplification that migrated into clinical culture.

The review was widely covered and heavily criticized. In June 2023, Molecular Psychiatry published a Comment led by Sameer Jauhar with 36 co-authors from King's College London, titled "A leaky umbrella has little value." According to King's College London's summary, the critics argued that the review presented no new analyses, used its own quality-grading criteria, and misreported tryptophan-depletion studies, which can bring on depressive symptoms in people with a history of depression even if they do not in never-depressed volunteers. They also argued it misread serotonin receptor and transporter imaging. Moncrieff and colleagues replied that many brain systems, serotonin included, probably play a complex role in mood, but that this is different from the claim that depression is caused by low serotonin, and that the burden of proof lies with those making that claim.

The fair summary is narrower than either side's headlines. Almost nobody in the field still defends a simple "low serotonin causes depression" model. Many researchers do still argue that the serotonin system is involved in mood regulation and in how some treatments work. The two positions overlap more than the public argument suggested.

Crucially, this does not mean SSRIs do not work. Selective serotonin reuptake inhibitors remain beneficial for many patients. It means we probably do not understand why they work — and that serotonin is one node in a far more complex system, not the root cause.

What Actually Happens in a Depressed Brain

Modern neuroimaging and neuroendocrinology paint a richer picture.

Prefrontal Cortex Hypoactivity

The prefrontal cortex (PFC) — particularly the dorsolateral and ventromedial regions — shows reduced activity and grey matter volume in major depressive disorder (MDD). The PFC is the brain's executive center: it regulates emotional responses, suppresses excessive fear signals from the amygdala, and supports goal-directed thinking. When it is underactive, the capacity to "talk yourself down" from negative states is genuinely reduced. This is not a failure of willpower — it is a measurable deficit in top-down inhibitory control.

Amygdala Hyperactivity

At the same time, the amygdala — the brain's threat-detection hub — is chronically overactive in depression. It responds more intensely to negative stimuli and less to positive ones. This creates an asymmetric threat bias: neutral faces are perceived as sad, minor setbacks feel catastrophic. Functional MRI studies consistently show elevated amygdala reactivity that normalizes (partially) with effective treatment.

Hippocampal Volume Reduction

One of the most replicated findings in depression research is hippocampal shrinkage. Large multi-site analyses, including the ENIGMA consortium's pooled imaging data, report a small but consistent reduction in hippocampal volume, most evident in people with recurrent or early-onset depression. The effect is modest on average and cannot diagnose depression in any individual. The hippocampus is critical for memory consolidation and for providing contextual information that modulates stress responses.

The mechanism involves cortisol. Under chronic stress, the HPA (hypothalamic-pituitary-adrenal) axis releases sustained high levels of cortisol. Cortisol is neurotoxic to hippocampal neurons in excess — it suppresses the production of brain-derived neurotrophic factor (BDNF), a protein essential for neuronal survival and the growth of new neurons (neurogenesis). The hippocampus is one of the few brain regions where neurogenesis occurs in adults, and chronic stress demonstrably reduces it.

HPA Axis Dysregulation

In a substantial subset of people with MDD, especially those with severe or melancholic depression, the HPA axis appears stuck in overdrive. The dexamethasone suppression test — a standard clinical tool — reveals that many depressed patients cannot properly turn off their cortisol response. This means the stress system has lost its feedback brakes. Chronically elevated cortisol then feeds forward to damage the very hippocampal circuits that should be dampening HPA activity, creating a vicious cycle.

The Neuroinflammation Hypothesis

One of the most productive recent frameworks is the inflammatory model of depression. A consistent body of evidence shows elevated pro-inflammatory cytokines — particularly interleukin-6 (IL-6), tumor necrosis factor alpha (TNF-α), and C-reactive protein (CRP) — in people with MDD.

Charles Raison and colleagues (2013, JAMA Psychiatry) demonstrated that infliximab — an anti-TNF-α drug used for inflammatory diseases like Crohn's — reduced depressive symptoms specifically in patients with elevated baseline CRP. People with low CRP showed no benefit. This suggests a biologically distinct inflammatory subtype of depression.

Neuroinflammation is particularly relevant to treatment-resistant depression (TRD), where standard antidepressants fail. Up to one-third of patients do not achieve remission with first-line SSRIs. Elevated inflammatory markers are disproportionately prevalent in this population, suggesting that for them, the disease mechanism is fundamentally different and requires anti-inflammatory approaches, not more serotonergic drugs.

Microglia — the brain's resident immune cells — appear to be key mediators. When peripheral inflammation is high (from obesity, stress, poor sleep, or chronic infection), microglia become activated and shift tryptophan metabolism away from serotonin production toward kynurenine pathway metabolites, some of which are neurotoxic.

The Gut-Brain Axis and Microbiome

In a landmark 2019 study published in Nature Microbiology, Valles-Colomer and colleagues analyzed gut microbiome data from 1,054 individuals and found that two bacterial genera — Coprococcus and Dialister — were consistently depleted in people with depression, even after controlling for antidepressant use.

The gut-brain axis operates through multiple channels: the vagus nerve (a direct bidirectional communication highway), short-chain fatty acids produced by gut bacteria (which influence microglial function), and gut-derived neurotransmitter precursors (roughly 95% of the body's serotonin is produced in the gut, though it does not cross the blood-brain barrier). Gut bacteria also modulate systemic inflammation, providing a plausible indirect mechanism for mood effects.

This is an early but rapidly growing field. Probiotic trials show modest but real signals for depression reduction, and dysbiosis (disrupted microbiome composition) is increasingly recognized as a potentially modifiable risk factor. You can read more about gut-brain signaling in our article on how gut bacteria affect health.

Genetics and Gene-Environment Interaction

Twin studies put the heritability of depression at roughly 35–40%, meaning genetic factors account for a meaningful but not dominant portion of risk. The environment — particularly early-life adversity and chronic stress — accounts for the majority.

The 5-HTTLPR serotonin transporter gene polymorphism became famous after a 2003 Science paper by Caspi and colleagues found that individuals with the "short" allele were significantly more likely to become depressed following stressful life events. This gene-by-environment interaction generated enormous excitement.

Subsequent meta-analyses have produced mixed results, and the direct serotonin-depression link has been questioned. However, polygenic risk score approaches — which aggregate hundreds of small genetic effects — do identify meaningful depression risk. Large genome-wide association studies (GWAS) have implicated genes involved in synaptic plasticity, glutamate signaling, and neuronal development, not just monoamine systems. This convergence with the neuroplasticity hypothesis is notable.

The Neuroplasticity Hypothesis and How Antidepressants Actually Work

Ronald Duman at Yale proposed that antidepressants may work not primarily by adjusting neurotransmitter levels, but by promoting neuroplasticity — the brain's capacity to rewire and adapt. This explains one of psychiatry's most puzzling clinical observations: SSRIs raise serotonin within hours, but symptom improvement takes 2–6 weeks.

The delay corresponds to BDNF-driven structural changes: new synaptic connections, dendritic growth, and possibly new neurons in the hippocampus. Chronic antidepressant use increases BDNF expression and hippocampal volume. The drug may not be correcting a serotonin deficit so much as triggering a slow cascade of neural remodeling.

This framework also explains why exercise works. Regular aerobic exercise robustly increases BDNF, reduces inflammation, normalizes HPA axis function, and promotes hippocampal neurogenesis. A 2014 meta-analysis by Josefsson and colleagues found exercise broadly comparable to antidepressants in mild-to-moderate depression — and with fewer side effects. For more on the physiology, see the science of exercise. The broader idea that adult brains keep rewiring is covered in neuroplasticity explained.

Antidepressant Classes: A Comparison

Class	Mechanism	Examples	Typical Use	Key Side Effects
SSRI	Blocks serotonin reuptake	Fluoxetine, sertraline, escitalopram	First-line MDD, anxiety	Sexual dysfunction, initial activation, GI upset
SNRI	Blocks serotonin + norepinephrine reuptake	Venlafaxine, duloxetine	MDD, chronic pain, anxiety	Elevated BP, sweating, discontinuation syndrome
TCA	Blocks multiple monoamine transporters	Amitriptyline, clomipramine	Refractory depression, neuropathic pain	Anticholinergic effects, cardiac risk, overdose danger
MAOI	Inhibits monoamine oxidase enzyme	Phenelzine, tranylcypromine	Atypical or refractory depression	Tyramine dietary restriction, drug interactions
Bupropion	Blocks dopamine + norepinephrine reuptake	Wellbutrin	MDD, smoking cessation	Lowered seizure threshold, no sexual side effects
Mirtazapine	Alpha-2 antagonist, serotonin receptor modulator	Remeron	MDD with insomnia or weight loss	Sedation, weight gain
Ketamine/Esketamine	NMDA glutamate receptor antagonist	IV ketamine, Spravato nasal spray	Treatment-resistant MDD	Dissociation, abuse potential, clinic administration required

How Well Do Standard Treatments Work? The STAR*D Story

The largest real-world antidepressant trial, STAR*D (Sequenced Treatment Alternatives to Relieve Depression), enrolled 4,041 outpatients in the US in the early 2000s and moved people who did not remit through up to four treatment steps. Its widely quoted headline was a cumulative remission rate of about 67% across the four steps.

That number has been contested. In a 2023 reanalysis in BMJ Open, Pigott and colleagues used the outcome measure specified in the original protocol and counted dropouts as not remitted, and arrived at roughly 35%, about half the published figure (BMJ Open, 2023). The original investigators responded that the reanalysis excluded 561 of 3,671 Level 1 patients and that this exclusion drives most of the difference. Both sides agree on one point: many participants dropped out, so any remission figure depends heavily on assumptions about those who left.

Measure	Figure	Source
People with depression worldwide (2023)	~322 million	WHO fact sheet
Adult prevalence (2023)	5.2% (men 4.1%, women 6.2%)	WHO fact sheet
Suicide deaths (2021)	~727,000	WHO fact sheet
STAR*D cumulative remission, as originally reported	~67%	Rush et al., 2006
STAR*D remission, protocol-based reanalysis	~35%	Pigott et al., 2023
Psilocybin 25 mg remission at week 3 (TRD)	~29%	Goodwin et al., 2022

The practical lesson for understanding the science is not that treatment is hopeless. It is that response is highly variable, that many people need more than one attempt, and that headline percentages depend on how a study handles the people who stop.

Treatment-Resistant Depression and Emerging Therapies

Approximately 30% of patients with MDD do not respond to two or more adequate antidepressant trials — a definition of treatment-resistant depression (TRD). For this group, conventional SSRI-based approaches are insufficient, and biological heterogeneity is likely high.

Ketamine and Esketamine

Ketamine, an NMDA glutamate receptor antagonist developed as an anesthetic, produced rapid (within hours) antidepressant effects in TRD trials, although effects of a single dose often fade within days to weeks. The FDA approved intranasal esketamine (Spravato) in 2019. Unlike SSRIs, ketamine works partly by triggering a rapid surge in BDNF and by promoting synaptogenesis — essentially growing new synaptic connections within hours. This discovery challenged the assumption that only monoamine systems were relevant to depression pharmacotherapy.

Transcranial Magnetic Stimulation (TMS)

TMS uses magnetic pulses to stimulate the dorsolateral prefrontal cortex — the region that is hypoactive in depression. It is non-invasive and FDA-cleared. Standard repetitive TMS (rTMS) courses run over several weeks, and trials report remission in a minority of TRD patients. Accelerated protocols (intensive daily sessions over a week) are showing even stronger results in recent trials.

Electroconvulsive Therapy (ECT)

ECT is consistently among the most effective treatments studied for severe or treatment-resistant depression, with remission rates in many studies well above those seen with medication switches. It is also deeply stigmatized, partly due to its historical misuse and its portrayal in media. Modern ECT is performed under general anesthesia with muscle relaxants; the patient experiences no convulsion. Side effects include transient memory disruption. For suicidal, psychotic, or catatonic depression, ECT is often life-saving.

Psilocybin

A phase 2 trial sponsored by COMPASS Pathways and published in the New England Journal of Medicine in 2022 (Goodwin et al.) randomized 233 adults with treatment-resistant depression to a single 25 mg, 10 mg or 1 mg dose of psilocybin with psychological support. At week 3, depression scores fell 6.6 points more on the 25 mg dose than on 1 mg, and about 29% of the 25 mg group were in remission. The effect was not clearly sustained at 12 weeks, and adverse events, mostly headache, nausea and dizziness, were common; suicidal ideation or self-injury was reported in all dose groups. The mechanism appears to involve 5-HT2A receptor agonism, large-scale neural network desynchronization and re-integration, and — consistent with the neuroplasticity hypothesis — rapid synaptogenesis. Larger and longer phase 3 trials have been run to test whether the benefit holds up; psilocybin is not an approved treatment for depression in most countries.

Learned Helplessness, Stress and the Circuit View

Some of the earliest biological models of depression came from animal work on stress. In the 1960s Martin Seligman and Steven Maier found that dogs exposed to shocks they could not control later failed to escape shocks they could, a phenomenon they called learned helplessness. Decades later, Maier and Seligman revised the interpretation: passivity appears to be the default response to prolonged aversive events, and what animals actually learn, via the prefrontal cortex, is that they do have control. The history is told in the Seligman learned helplessness experiment.

This links to the circuit findings above. Chronic stress, which the psychology of stress covers in detail, weakens prefrontal regulation and strengthens amygdala and habit-like responding. Depression and anxiety share much of this circuitry, which is one reason they so often occur together; see anxiety neuroscience.

Expectation also matters. In antidepressant trials, placebo groups often improve substantially, and the gap between drug and placebo is usually larger in more severe depression. That is not evidence that depression is "imaginary"; it reflects real brain responses to expectation and care, explored in how placebos change real symptoms.

Common Misconceptions

• "Depression is just sadness." Clinical depression is a persistent syndrome lasting at least two weeks that affects sleep, energy, concentration and functioning, not a passing mood.
• "It's a chemical imbalance." This slogan oversimplifies; current evidence implicates multiple interacting systems.
• "If serotonin isn't the cause, antidepressants are fake." Whether a drug works and why it works are separate questions. Many people benefit, even while the mechanism is debated.
• "Brain scans can diagnose depression." Group differences in imaging are real but small and overlapping; no scan or blood test diagnoses depression in an individual today.
• "Strong people don't get depressed." Reduced prefrontal control in depression is a measurable change in brain function, not a character flaw.

Lifestyle Factors Studied in Research

Beyond pharmacology, researchers have studied several lifestyle factors. These are population-level findings, not a substitute for individual care; the WHO notes that psychological treatments are first-line and that antidepressants are not needed for mild depression, decisions that belong with a clinician.

• Exercise: Supervised aerobic exercise programmes have shown effects comparable to antidepressants in some trials of mild-to-moderate depression (Blumenthal et al., Archives of Internal Medicine, 1999; Josefsson et al., 2014 meta-analysis), though study quality varies.
• Sleep: Disrupted sleep architecture is both a symptom and a driver of depression. Slow-wave sleep suppression accelerates depressive relapse.
• Light therapy: First-line for seasonal affective disorder (SAD); bright-light treatment has shown efficacy comparable to antidepressants in some trials of seasonal depression, and early evidence explores use in non-seasonal MDD.
• Social connection: Loneliness elevates inflammatory markers and HPA axis activity. Meta-analyses consistently link social isolation to depression risk.
• Diet: Observational studies link Mediterranean-style eating to lower depression risk, and the small SMILES randomized trial (Jacka et al., 2017) reported that dietary support reduced symptoms more than social support, though larger trials are needed.

To understand how sleep connects to broader brain health, see our explainer on how sleep affects the brain. Loneliness, one of the social risk factors above, has its own biology; see the psychology of loneliness.

---

FAQ

Q: Is depression really just low serotonin? A: No. The 2022 Moncrieff umbrella review in Molecular Psychiatry found no consistent evidence for a serotonin deficit in depression. Depression involves multiple systems: HPA axis dysregulation, neuroinflammation, reduced neuroplasticity, and gut-brain signaling. Serotonin is one actor in a complex network, not the single broken "chemical."

Q: If serotonin isn't the cause, why do SSRIs work? A: The honest answer is we are not fully certain. The best current hypothesis (Duman's neuroplasticity model) is that SSRIs promote BDNF-driven synaptic remodeling over weeks. They may also modulate neuroinflammation and HPA axis feedback. The serotonin changes are real but probably not the primary therapeutic mechanism.

Q: Why do antidepressants take 2–6 weeks to work? A: Because their likely mechanism is structural neuroplasticity — growing new synaptic connections, promoting hippocampal neurogenesis, and resetting HPA axis feedback. These biological processes take weeks to manifest, unlike the rapid neurotransmitter effects that occur within hours.

Q: What is treatment-resistant depression? A: TRD is typically defined as failing to respond to at least two adequate antidepressant trials. It affects ~30% of people with MDD. These patients often have elevated inflammatory markers, genetic variants in drug metabolism, or structural brain changes that require different therapeutic approaches (ketamine, TMS, ECT, or psilocybin in trials).

Q: How does ketamine work so fast for depression? A: Ketamine blocks NMDA glutamate receptors and triggers a rapid release of BDNF, which drives synaptogenesis — the growth of new synaptic connections — within hours. This is fundamentally different from SSRIs and explains why depressed patients often feel relief within 24 hours of infusion, versus weeks with conventional antidepressants.

Q: Does the microbiome really affect depression? A: Evidence is accumulating. The Valles-Colomer 2019 Nature Microbiology study found consistent microbiome differences in depression. The gut-brain axis operates via the vagus nerve, short-chain fatty acids, inflammatory signaling, and precursors to neurotransmitters. Probiotic trials show modest signals. It is a real biological mechanism, though causal evidence in humans is still developing.

Q: Is depression genetic? A: Partly — around 40% heritable based on twin studies. But no single gene "causes" depression. It is polygenic (many small genetic effects), and gene-environment interaction is critical: genetic risk is often activated by early adversity, chronic stress, or trauma. Genome-wide studies implicate genes in synaptic plasticity and glutamate signaling, not just serotonin.

Q: What is the most effective treatment for severe depression? A: In research on severe and treatment-resistant depression, electroconvulsive therapy (ECT) has consistently shown some of the highest remission rates of any treatment, and ketamine or esketamine can act within hours to days. Choosing between options depends on an individual assessment by a clinician; this article cannot recommend a treatment.

---

Sources

1. Moncrieff, J., Cooper, R.E., Stockmann, T. et al. (2022). The serotonin theory of depression: a systematic umbrella review of the evidence. Molecular Psychiatry, 28, 3243–3256. https://doi.org/10.1038/s41380-022-01661-0
2. Raison, C.L., Rutherford, R.E., Woolwine, B.J. et al. (2013). A randomized controlled trial of the tumor necrosis factor antagonist infliximab for treatment-resistant depression. JAMA Psychiatry, 70(1), 31–41.
3. Valles-Colomer, M., Falony, G., Darzi, Y. et al. (2019). The neuroactive potential of the human gut microbiota in quality of life and depression. Nature Microbiology, 4, 623–632.
4. Duman, R.S., Aghajanian, G.K. (2012). Synaptic dysfunction in depression: potential therapeutic targets. Science, 338(6103), 68–72.
5. Castrén, E., Monteggia, L.M. (2021). Brain-derived neurotrophic factor signaling in depression and antidepressant action. Biological Psychiatry, 90(2), 128–136.
6. Carhart-Harris, R., Giribaldi, B., Watts, R. et al. (2021). Trial of psilocybin versus escitalopram for depression. New England Journal of Medicine, 384, 1402–1411.
7. Goodwin, G.M., Aaronson, S.T., Alvarez, O. et al. (2022). Single-dose psilocybin for a treatment-resistant episode of major depression. New England Journal of Medicine, 387, 1637–1648.
8. Jacka, F.N., O'Neil, A., Opie, R. et al. (2017). A randomised controlled trial of dietary improvement for adults with major depression (the 'SMILES' trial). BMC Medicine, 15, 23.
9. Blumenthal, J.A., Babyak, M.A., Moore, K.A. et al. (1999). Effects of exercise training on older patients with major depression. Archives of Internal Medicine, 159(19), 2349–2356.
10. Malhi, G.S., Mann, J.J. (2018). Depression. The Lancet, 392(10161), 2299–2312.
11. Jauhar, S. et al. (2023). A leaky umbrella has little value: evidence clearly indicates the serotonin system is implicated in depression. Molecular Psychiatry. https://doi.org/10.1038/s41380-023-02095-y
12. Pigott, H.E. et al. (2023). Reanalysis of the STAR*D study's patient-level data. BMJ Open, 13, e063095. https://bmjopen.bmj.com/content/13/7/e063095.abstract
13. World Health Organization. Depressive disorder (depression) fact sheet. https://www.who.int/news-room/fact-sheets/detail/depression