Major depression

Major depression can affect mood, interest, energy, thinking, sleep and everyday functioning. People may share a diagnosis while having different symptoms, histories and needs. The most convincing scientific account therefore combines inherited vulnerability, development, experiences, present circumstances and interacting processes in the brain and body. No single chemical shortage, brain abnormality or life event explains major depression as a whole. Review of major depressive disorder

It helps to separate why an episode begins from what produces or maintains its symptoms. A difference found during depression could be a preceding vulnerability, a consequence of being ill, a process keeping illness going, or an effect of treatment. These possibilities require different evidence. Researchers are more confident that several influences contribute than that they can trace the complete chain in an individual.

This essay mainly concerns adults with major depression that is not part of bipolar disorder, often called unipolar depression. Depression within bipolar disorder, childhood depression and depression during pregnancy require their own assessments. Longstanding persistent depression is discussed separately in this series. Treatments can be evaluated by whether they help even while the original causes remain incompletely understood.

Many inherited differences make small contributions

Family and twin studies establish an inherited contribution to susceptibility. A foundational review estimated it at 37%, with an uncertainty range of 31–42%. This describes variation between people within particular populations and environments. It does not mean that 37% of one person's depression is genetic. Nor does the remaining percentage consist entirely of identifiable experiences: measurement error contributes too. A substantial inherited influence does not make an episode inevitable or recovery impossible. Review of inherited susceptibility to major depression

Large genetic studies show that many small differences in DNA each contribute small effects. A 2025 study combined 688,808 people classified as having depression with more than 4.3 million comparison participants. It identified 697 associations across 635 regions of DNA and pointed towards processes involving nerve cells and the connections between them. Its definitions included interviews, health records, questionnaires and self-reported diagnoses, rather than one uniform clinical assessment. The findings establish biological leads, not a complete explanation or a test identifying the cause of someone's episode. Representation across ancestries and the usefulness of genetic prediction also remain uneven. Large international genetic study

This differs from older claims about a few supposedly powerful “depression genes”. Large studies failed to reproduce many proposed effects of individual genes, including claims that particular variants strongly alter responses to adversity. Those failures weaken specific stories, not the broader evidence that many inherited differences and environmental experiences both matter. Large study of historical candidate-gene claims

Adversity and social circumstances can contribute

Childhood abuse and neglect, violence within relationships and food insecurity are consistently associated with depression. A review bringing together 26 earlier reviews found substantial support for these connections. Experiences often cluster, however: someone facing one form of adversity may face several. An ordinary comparison cannot readily separate their effects or exclude every other explanation. Review of social influences on major depression

Stronger studies try to address this problem through methods such as comparisons within families and natural experiments. In a review of 34 studies involving 54,646 participants, the association between childhood maltreatment and later mental-health problems became appreciably smaller after adjusting for shared familial and other influences, but did not disappear. The average statistical difference fell from about 0.56 to 0.31 standard deviations, a unit that allows results measured on different scales to be compared. These figures are not percentages of cases explained. The change supports a causal contribution alongside other risks. Because the outcomes included several mental-health conditions, the figure is not a precise causal estimate for major depression alone. Review of childhood maltreatment using stronger causal designs

Economic studies add another kind of evidence. Natural experiments involving income losses, and randomised programmes intended to reduce poverty, indicate that economic circumstances can affect depression and anxiety. The relationship can work both ways: mental illness may make employment harder and reduce earnings. Many studies measure symptoms or common mental-health problems rather than strictly diagnosed major depression, so their findings should not be treated as exact estimates of how a policy would change diagnosed illness. Decisions about particular policies also require evidence about costs, effectiveness and who benefits. Review of poverty and mental health

These results make social conditions part of the explanation, rather than merely a background detail. They do not mean every episode has a discoverable adverse event at its beginning, or that adversity inevitably causes depression.

Thinking patterns can precede illness and keep it going

Prospective studies follow people before illness develops. They show that some patterns of thinking and a tendency towards negative emotions can precede a first clinically assessed depressive episode. A review of 26 studies found that cognitive vulnerabilities were associated with roughly twice the odds of later onset, but estimates varied widely and publication bias weakened confidence. Some apparent vulnerabilities could also have been early symptoms below the threshold for diagnosis. An association with future illness is useful evidence, but does not establish one universal psychological cause. Review of psychological factors preceding first depression

Repetitive negative thinking, low self-esteem and related difficulties can also accompany persistence or recurrence. A person may withdraw from meaningful activities, spend more time dwelling on problems, and have fewer experiences that challenge a hopeless view. These processes overlap with anxiety and are not unique to depression. Findings also depend on measurement: long-term research has produced stronger results for several self-reported characteristics than for some laboratory tests of attentional bias. Review of psychological factors and the course of depression

Understanding these patterns does not imply that someone is choosing to be depressed or could recover simply by thinking positively. It identifies possible points for treatment and support while leaving room for physical illness, adversity and other contributing processes.

Brain systems involved in reward, emotion and thought

Brain-imaging research implicates connected systems concerned with reward and motivation, noticing emotionally important information, directing thought and attention, and thinking about oneself. Researchers use names such as the salience, frontoparietal and default-mode networks for some of these systems. Their activity differs with the task, stage of illness and person being studied. There is no uniform pattern shared by every depressed person. Overview of depression mechanisms

Loss of interest or pleasure—sometimes called anhedonia—makes reward processing especially relevant. Across 48 studies with 3,154 participants, depressed groups showed a small-to-moderate average difficulty on behavioural reward tasks. The average difference was about one-third of a standard deviation. Results varied across aspects such as anticipating, experiencing and learning from rewards, and the literature included small studies and publication bias. This supports reward-processing difficulties in some people, not a single universal reward deficit or a simple dopamine shortage. Review of reward-processing research

Structural differences can also be much smaller than headlines suggest. A large international comparison found that the hippocampus, important for memory and context, was on average about 1.24% smaller in major depression. The statistical group difference was small, with substantial overlap between individuals. It was driven by recurrent depression; people in a first episode did not differ significantly from comparison participants. This cannot establish whether the difference preceded depression or followed prolonged illness and associated experiences. It is not a diagnostic scan result. International study of brain structure in depression

Serotonin matters, but a general shortage is not established

The claim that depression generally results from too little serotonin is not supported by the evidence. A widely discussed review found no consistent support for that general shortage explanation. Its searches ended in December 2020, and it did not test whether antidepressants work. The distinction is important: challenging one explanation of a medicine's action does not invalidate trials showing that it can help. Review of the serotonin-deficiency theory

Researchers continue to debate how particular findings should be interpreted. A subsequent review emphasised that serotonin can influence mood, vulnerability and symptoms while rejecting a one-chemical explanation. Experimentally reducing the body's ability to make serotonin generally does not produce depression in healthy people without evident vulnerability, but can provoke symptoms in some people whose depression has remitted. Brain measurements of serotonin receptors or transporters also cannot simply be read as a gauge showing “too little serotonin”. Review of serotonin and depression

A treatment may change a process downstream from the original cause, compensate for a vulnerability or interrupt a pattern maintaining illness. That reasoning applies to psychotherapy as well as medication. Improvement after changing thinking, sleep or chemical signalling does not establish that a deficit in that same system caused every person's illness.

Connections between brain cells can change

Ketamine and other treatments have increased interest in glutamate, another chemical messenger, and in the brain's capacity to change its connections. A leading explanation proposes that stress and other influences disturb communication between nerve cells, while successful treatment helps restore more adaptable functioning. Some interventions promote structural changes in animal experiments. How those changes relate to recovery in humans remains unresolved, and much of the direct evidence about new connections comes from animals. A complete universal explanation based on brain plasticity has not replaced earlier chemical theories. Review of structural changes associated with rapid-acting antidepressants

One molecule involved in the growth and support of nerve cells is brain-derived neurotrophic factor, or BDNF. A 2024 review found that its levels in blood did not reliably track depression severity or treatment outcome. A plausible biological role is therefore different from a useful clinical test. Review of blood BDNF measurements

Stress, inflammation and metabolism vary between people

The body's stress-response system regulates the release of cortisol. A review of 361 studies with 18,454 participants found higher average cortisol in depressed groups. The difference was substantially smaller in studies meeting stronger methodological standards: about 0.33 rather than 0.60 standard deviations. Results also varied between clinical groups, with larger differences among older hospital patients and people with certain severe presentations, including psychotic depression. An individual diagnosis does not justify assuming persistently high cortisol. Sleep disturbance, current circumstances and the illness itself may all affect the measurement. Review of cortisol research in depression

Inflammation is relevant to a subset. A review estimated that 27% of people studied with depression had C-reactive protein above 3 mg/L, with an uncertainty range of 21–34%. This blood marker reflects inflammation somewhere in the body; it does not directly measure brain inflammation or explain an individual's symptoms. About 73% did not cross that threshold. That prevents treating the finding as universal, without proving immune processes are irrelevant below it. Review of low-grade inflammation in depression

Experiments offer some causal evidence. In a trial of 115 healthy adults, a substance used to activate the immune system temporarily increased depressed mood, with different people showing different susceptibility. That demonstrates that immune activation can affect mood in those circumstances. It does not establish that inflammation generally causes persistent major depression. Human experiment on immune activation and mood

Researchers are investigating a possible pattern combining inflammation, altered metabolism, fatigue, increased appetite and excessive sleep. Such a grouping could eventually improve treatment decisions. For now, it needs reliable identification, replication and trials showing that matching treatment to the proposed mechanism actually helps. A statistical subgroup alone is not a clinically established subtype. Review of immune and metabolic patterns in depression

Sleep, physical health and emerging explanations

Sleep offers unusually helpful intervention evidence. In a trial of 291 adults aged sixty or older who had insomnia, a structured psychological treatment for insomnia was compared with sleep education. Over three years, new or recurrent major depression occurred in about 12% of the treatment group and 26% of the comparison group. The finding supports a modifiable contribution in this population. Because the treatment changes several thoughts and behaviours, it does not show that improved sleep was the sole reason for the difference. Trial of insomnia treatment and depression prevention

Physical illness can contribute through pain, disability, medication effects, bodily changes and the burden of managing illness. Reproductive transitions can increase vulnerability for some women, without explaining every episode or every difference between sexes through hormones. Assessment may include tests for conditions such as thyroid disease that can produce overlapping symptoms. These tests investigate other or contributing problems; they are not biological tests diagnosing major depression. NHS information on assessment

Research into how genes are regulated found some agreement across brain and blood studies about developmental, inflammatory and energy-related pathways, but substantial variation in individual genes. It cannot yet separate causes from consequences or provide a validated clinical classification. Gut-microbe studies also show mixed findings and no consistent overall difference on several broad measures of microbial variety. Diet, geography, medicines and diagnostic differences complicate the picture. Neither area has established a characteristic biological cause shared by major depression. Review of brain and blood biomarkers, review of the gut microbiome

Effective treatment does not require a complete causal explanation

Structured psychological therapies and antidepressants both improve depression on average. Combining them often improves outcomes, especially when depression is substantial, persistent or incompletely treated. Severe illness may require specialist treatments. The important question is which acceptable treatment offers this person a useful balance of benefit, burden and harm, followed by a review of whether it actually helps. Canadian clinical guideline

Treatment studies distinguish response from remission. Response often means at least a halving of symptom scores. Remission means symptoms fall below a specified threshold. Neither alone guarantees lasting recovery in relationships, work or everyday functioning. Early improvement and staying well are different outcomes, and both deserve attention.

Talking therapies and medication as starting options

CBT helps identify and change thoughts and behaviour that maintain depression. Behavioural activation helps rebuild involvement in meaningful activities, usually in manageable steps. Interpersonal therapy focuses on difficulties such as bereavement, conflict and changes in relationships or roles. These are structured treatments requiring appropriate training, rather than interchangeable labels for any supportive conversation. Clinical guideline on psychological treatment

CBT has the largest research base, but does not consistently outperform every established psychotherapy. A review of 331 trials with 34,285 participants found relatively few reliable differences between major approaches. Many comparisons were too small to establish modest differences, and almost half the trials admitted people on symptom scores without requiring a formal diagnosis. Failure to detect a difference does not prove exact equivalence. Review comparing major psychotherapies

Antidepressants outperform placebo on average. A large 2018 review covered 522 trials with 116,477 participants; all 21 medicines studied improved the likelihood of response. The average symptom-score advantage was about 0.30 standard deviations, a modest difference. This does not mean that 30% of people benefit, or that everyone experiences only a small change. An updated review covering 602 trials and 135,180 participants supported the broad conclusion. 2018 antidepressant review, updated review

No antidepressant is best for everyone. SSRIs are often an initial choice because of their overall safety and tolerability. Previous response, other medicines and possible nausea, sexual difficulties, disturbed sleep or weight changes all matter. Benefits take time, and early review is especially important if agitation or suicidal thoughts emerge. Withdrawal symptoms can occur on stopping, and can sometimes be prolonged, so changes should be agreed with the prescriber. NHS information on antidepressants

For less severe depression, NICE generally favours beginning with a suitable psychological intervention and does not routinely offer medication first unless it is the person's informed preference. For more severe depression, psychological treatment, medication and combined approaches can all be reasonable initial choices. Severity includes the impact on life as well as symptom burden. NICE recommendations for depression in adults

Combining treatments often helps, but the comparison matters

A review of 101 trials with 11,910 participants found that therapy plus medication increased the probability of response by about 25% relative to medication alone and 27% relative to psychotherapy alone. These are relative increases, not improvements of 25 or 27 percentage points. A 2026 analysis of individual participant records also favoured adding psychotherapy to medication, both after treatment and during follow-up. However, the researchers obtained individual data from only 31 of 75 eligible studies and found no reliable patient characteristics identifying who specifically needed the combination. Review of combined treatment, analysis of individual participant data

A different 2026 review asked a narrower question: does adding an antidepressant improve the results of psychotherapy when the comparison group receives the same therapy plus a placebo pill? Across eleven trials, the additional benefit was uncertain. The estimate ranged from little or no benefit to a potentially meaningful benefit, and the evidence was judged very low certainty. This is not proof that adding medication never helps. It shows why adding therapy to medicine and adding medicine to therapy should not be treated as identical decisions. A placebo pill also controls some expectations and clinical contact more closely than therapy alone. Review of adding antidepressants to psychotherapy

Combined care therefore has a strong place among routine options, while the evidence is less secure for starting everyone on both treatments. A preferred single treatment may offer a better practical balance for some people. Guideline ordering also considers costs and delivery, not effectiveness alone.

ECT and magnetic stimulation meet different needs

Electroconvulsive therapy, or ECT, has an established role in severe depression, particularly when rapid improvement is needed, physical health is deteriorating, other treatment has failed or ECT helped previously. It is given under a general anaesthetic. Short-term memory difficulties are common, and some people experience lasting gaps in personal memories; the risks depend partly on the treatment technique. Relapse is possible, so an acute course needs a plan for continuing care. Royal College of Psychiatrists information on ECT

Depression with delusions or hallucinations requires specialist management. Combining an antidepressant with an antipsychotic is an established option, alongside ECT where appropriate. Benefits and adverse effects require monitoring. If someone cannot stay safe or their physical condition is seriously deteriorating, urgent local medical help is needed. NICE evidence and recommendations on psychotic depression

Repetitive transcranial magnetic stimulation, or rTMS, applies magnetic stimulation while the person is awake. It avoids general anaesthesia and has a favourable profile for effects on memory and thinking. It can be relevant after insufficient benefit or poor tolerability from initial treatment, although repeated clinic visits can be demanding. NICE finds adequate evidence of short-term effectiveness and no major safety concerns, while recognising variable response and questions about maintaining benefit. NIMH explanation of brain stimulation, NICE guidance on rTMS

One review comparing ECT, rTMS and intravenous ketamine in treatment-resistant depression found no statistically clear differences in effectiveness, but rated those comparisons very low certainty. That does not establish that the treatments are equivalent. Different patients, techniques and treatment durations can materially change the result. Review of specialist treatment comparisons

When initial treatment has not been enough

“Treatment-resistant depression” commonly means that at least two adequate antidepressant trials have not helped sufficiently, although definitions vary. Before applying the label, clinicians reconsider the diagnosis, the treatment's dose and duration, whether it could be taken consistently, and other problems affecting recovery. Missed bipolar disorder changes the treatment question. Depending on the situation, options include adding psychotherapy, changing medicine, adding another medicine or considering a specialist physical treatment. Clinical guideline on further treatment

Adding a second medicine to improve a partial response is called augmentation. A 2026 review of 56 placebo-controlled trials supported several newer antipsychotics, including aripiprazole, quetiapine, brexpiprazole and cariprazine. Such medicines can cause restlessness, weight gain and metabolic problems. The same review found older evidence for lithium less robust after adjustments for possible bias, reducing confidence in the size of its short-term additional benefit rather than proving ineffectiveness. Lithium also requires blood-level, kidney and thyroid monitoring. Review of medication augmentation

Intravenous ketamine and esketamine nasal spray can help selected people with treatment-resistant depression. In the ELEKT-D trial, ketamine met a prespecified criterion for being no worse than ECT over three weeks. Response occurred in about 55% receiving ketamine and 41% receiving ECT. The trial enrolled people without psychosis, was not blinded, and some participants withdrew before treatment. It cannot establish a ranking for every severe depressive episode. Trial comparing ketamine with ECT

In ESCAPE-TRD, about 27% receiving esketamine reached remission at eight weeks, compared with about 18% receiving quetiapine augmentation. Both groups also took an oral antidepressant. The trial was manufacturer funded and participants knew their treatment, although outcome assessors were masked. Esketamine requires supervised administration and monitoring because of risks including sedation, a sense of detachment from oneself or surroundings, and raised blood pressure. Rapid improvement in symptoms does not establish prevention of suicide deaths; the regulatory information states that effectiveness in preventing suicide has not been demonstrated. Trial of esketamine and quetiapine, esketamine prescribing information

Access is a separate question from whether a treatment can have an effect. NICE does not recommend routine NHS use of esketamine for treatment-resistant depression under its current appraisal. That judgement includes uncertainty about clinical and cost effectiveness. It should not be translated into a claim that the medicine has no antidepressant action. NICE appraisal of esketamine

Exercise can be part of treatment

Exercise can improve depressive symptoms and may be a chosen initial intervention in suitable circumstances or one part of a wider plan. A 2026 Cochrane review included 73 trials with 4,985 participants. Exercise helped compared with control conditions, but confidence was generally low and the estimated benefit became smaller in the better-designed trials. Direct comparisons with psychotherapy and medication involved much less evidence, and longer-term benefits remained uncertain. Cochrane review of exercise for depression

The findings support exercise without establishing that it is superior to, or interchangeable with, other treatment at every level of severity. Comparing exercise against an inactive control with medication against a placebo pill cannot identify a winner. An activity plan also needs to fit the person's physical health, energy, access and preferences.

Recovery includes a plan for staying well

An effective treatment may need to continue after symptoms improve. Antidepressants are commonly continued for at least six months after improvement, with duration reviewed according to the illness course and the person's circumstances. NHS information on continuing treatment

The UK ANTLER trial studied people taking long-term antidepressants who felt well enough to consider stopping. Within a year, relapse occurred in about 39% assigned to continue and 56% assigned to discontinue. Under those conditions, roughly six people continuing for a year would prevent one additional relapse. Continued medication still left a substantial risk, and withdrawal symptoms and subsequent treatment changes complicate interpretation. These participants were not representative of everyone recovering from a first episode. ANTLER trial

Psychological treatment can also reduce recurrence. Mindfulness-based cognitive therapy combines structured cognitive work with mindfulness practice and has trial evidence for relapse prevention after recurrent depression has remitted. This does not establish the same benefit for generic meditation apps. Review of mindfulness-based cognitive therapy

A 2026 review of 76 trials with 17,379 participants adds an important qualification to continuation evidence. In selected people whose depression had remitted, gradual reduction in medication combined with psychological support could provide relapse prevention comparable to continuing medication. The evidence included indirect comparisons and varied forms of support, with some anxiety studies included. It does not mean everyone can stop without increased risk. It supports a planned, supported process that distinguishes withdrawal symptoms from returning depression, rather than abrupt stopping. Review of antidepressant-stopping strategies

Why research cannot yet choose one best treatment for everyone

Comparisons depend on who takes part and what is measured. A waiting list, usual care, a placebo pill, sham stimulation and a credible active treatment are different tests. Psychotherapy participants generally know what they are receiving, and the distinctive effects of ketamine can also weaken blinding. Researcher preferences, industry funding, dropout and incomplete reporting of harms require scrutiny without automatically invalidating a study.

The major antidepressant trials largely excluded psychotic depression and established treatment resistance, while ketamine-versus-ECT research specifically studied treatment resistance without psychosis. Combining their response rates into a league table would mislead. Response at three weeks, remission at eight weeks and relapse over a year are also different outcomes. The evidence is stronger that established treatments help than that a fixed ranking identifies the best one for an individual.

Similar limitations apply to mechanisms. Studies comparing people during an episode with healthy participants cannot readily separate vulnerability from illness effects. Symptoms, self-reported diagnoses and interview-confirmed episodes are not interchangeable. Sleep, diet, smoking, medication, physical health and social conditions can influence both depression and biological measurements, and depression can change those factors in return. A result that predicts illness need not cause it; a causal contributor may still be too nonspecific to diagnose it.

Useful future research would follow diverse people before, during and after first episodes; test whether proposed mechanisms actually explain treatment effects; and establish whether biological subgroups improve care in new populations. Treatment studies need longer, more comparable specialist-treatment trials, reliable tests of individual treatment selection, and stronger comparisons of combination and withdrawal strategies. Memory, functioning, relationships, harms and sustained recovery should receive the same attention as symptom scales.

For current care, the evidence supports choosing an acceptable treatment, checking symptoms and functioning over time, reviewing adverse effects and changing an inadequate plan. It also supports attending to sleep, physical health and practical circumstances. An uncertain universal explanation is compatible with substantial evidence that people can improve through several different routes.