Depression and Anxiety

TL;DR. Depression is not sadness and anxiety is not worry. Both are disorders of regulation: systems that evolved to lower mood after loss and to raise vigilance under threat, stuck on, disconnected from circumstances, and persisting long after any triggering event. Together they affect close to a billion people and rank among the leading causes of disability worldwide. The honest state of the science is that we have treatments that help a lot of people (psychotherapy, several drug classes, exercise, and for severe cases ECT, which is far more effective and far less frightening than its reputation), and we do not have a validated biological explanation of what causes them. The "chemical imbalance" story you were probably told is a simplification that the field abandoned decades ago.

Key takeaways

• About 332 million people have depression, and WHO estimates over a billion people live with a mental health condition of some kind.
• Depression is diagnosed by a pattern, duration, and functional impairment, not by severity of unhappiness. Five or more specific symptoms, most of the day, nearly every day, for at least two weeks, with impairment.
• The serotonin deficiency theory was never established and is not the reason antidepressants work. They do work for moderate to severe depression, and the mechanism is probably downstream effects on neuroplasticity rather than simply raising a chemical.
• Suicide kills over 700,000 people a year and is among the leading causes of death in people aged 15 to 29.
• Exercise, psychotherapy, and medication all have real effect sizes, and combining psychotherapy with medication outperforms either alone in moderate to severe depression.
• Sexual side effects from SSRIs affect a large minority and are systematically under-disclosed; discontinuation symptoms are real and are not addiction.

What they are

In short: Both are defined by a pattern, a duration, and functional impairment rather than by how unhappy or worried someone is.

Major depressive disorder requires at least five of the following for at least two weeks, including one of the first two, causing significant distress or impairment:

• Depressed mood most of the day, nearly every day
• Anhedonia: loss of interest or pleasure in nearly all activities
• Significant weight or appetite change
• Insomnia or hypersomnia
• Psychomotor agitation or retardation, observable by others
• Fatigue or loss of energy
• Feelings of worthlessness or excessive guilt
• Reduced concentration or indecisiveness
• Recurrent thoughts of death or suicide

Note what is not on that list: circumstances. Depression can follow a loss and can arrive without one. Note also how much of it is physical: sleep, appetite, energy, and movement. Depression presents as bodily symptoms at least as often as emotional ones, and in many cultures it is described primarily in physical terms, which is a reason it is missed.

Persistent depressive disorder (dysthymia) is a lower-grade depression lasting two years or more, often for so long that people describe it as their personality.

Anxiety disorders are a family:

Depression and anxiety co-occur so often (roughly half of people with one meet criteria for the other) that some researchers argue they are better understood as varying expressions of a shared underlying vulnerability.

Don't be confused: a panic attack is not a heart attack, and it is not dangerous. It feels dangerous, which is the point: the body has activated a full threat response with no threat present. Racing heart, chest tightness, tingling, and a sense of unreality are the physiology of that response, not signs of damage. First attacks routinely present to emergency departments, and they should be checked once, because the symptoms genuinely overlap. After that, the most useful information a person can have is that the attack will peak within about ten minutes and resolve, and that trying to fight it prolongs it.

The history

In short: Every antidepressant class was discovered by accident, and the chemical imbalance explanation spread through advertising long after researchers had doubted it.

Melancholia appears in the Hippocratic corpus, attributed to an excess of black bile, a humoral theory that gave the condition its name and lasted two thousand years. Robert Burton's Anatomy of Melancholy (1621) is still readable and still recognisable.

Modern treatment began, as so often, by accident.

What actually goes wrong

In short: The serotonin deficiency story was never established, which does not mean the drugs do not work, and the better-supported account involves neuroplasticity, stress signalling, and circumstances.

This section requires unusual honesty, because the popular account is wrong and the scientific account is incomplete.

The serotonin hypothesis, assessed. Depleting tryptophan (serotonin's precursor) does not cause depression in healthy people. Serotonin metabolite levels do not reliably differ between depressed and non-depressed people. A 2022 umbrella review by Moncrieff and colleagues surveyed this literature and concluded there is no consistent evidence that depression is caused by low serotonin. That conclusion was widely reported as "antidepressants don't work," which does not follow: a drug can work through a mechanism other than correcting a deficiency, exactly as aspirin relieves headache without headaches being caused by low aspirin.

What the evidence better supports:

• Neuroplasticity. Chronic stress reduces production of BDNF (brain-derived neurotrophic factor), shrinks dendritic branching in the hippocampus and prefrontal cortex, and reduces synaptic connectivity. Antidepressants of several classes increase BDNF and promote synaptic growth over weeks, which matches the delay before clinical effect far better than monoamine levels do (which change within hours). Ketamine's rapid antidepressant effect appears to work through a fast burst of synaptogenesis.
• The stress system. The hypothalamic-pituitary-adrenal axis is dysregulated in a substantial subset of depressed patients, with elevated cortisol and impaired feedback. Early-life adversity produces lasting changes in this system.
• Inflammation. A subset of depressed patients have elevated inflammatory markers. Interferon treatment causes depression in a large fraction of recipients. Anti-inflammatory drugs have modest antidepressant effects in inflamed subgroups. This looks like one route in, not the route.
• Circuits. Imaging consistently shows altered activity and connectivity between prefrontal regulatory regions, the amygdala, and the default mode network, with rumination associated with default mode overactivity.
• Genetics. Roughly 40 percent heritable, highly polygenic, no individual gene of large effect. The once-famous serotonin transporter gene-by-stress interaction did not replicate in large samples, which was an important corrective for the field.
• Environment. Childhood adversity, poverty, unemployment, discrimination, loneliness, chronic pain, and physical illness all raise risk substantially. These are not softer causes than biology; they are causes with measurable biological signatures.

The honest summary is stress-diathesis: inherited and acquired vulnerability meeting circumstances, with the resulting state maintained by changes in brain circuits, stress signalling, and behaviour (withdrawal and inactivity, which remove the very inputs that would lift mood, which is the target of behavioural activation therapy).

Anxiety has a somewhat clearer circuit story: an overactive amygdala threat detector with insufficient prefrontal regulation, plus learned associations that generalise. Avoidance is the maintaining mechanism, because avoiding a feared situation produces immediate relief that reinforces the avoidance and prevents the learning that would extinguish the fear. That is why exposure-based therapy works and why reassurance does not.

What it does to the body

Depression is not confined to mood.

• Cognition: slowed thinking, poor concentration, impaired memory and decision-making, which is why depression in older adults can be mistaken for dementia.
• Sleep: early morning waking is characteristic, as is unrefreshing hypersomnia in atypical presentations.
• Pain: depression amplifies pain perception, and chronic pain causes depression, in a loop.
• Cardiovascular: depression roughly doubles the risk of coronary heart disease and worsens outcomes after a heart attack, through effects on inflammation, autonomic tone, adherence, and behaviour.
• Immune and metabolic: associations with diabetes, obesity, and inflammation run in both directions.
• Life expectancy: people with severe mental illness die roughly 10 to 20 years earlier than average, and the great majority of that gap is from physical illness, not suicide.

Is it deadly?

Yes, in two ways.

Suicide accounts for over 700,000 deaths a year, roughly one every 40 seconds, and is among the leading causes of death in people aged 15 to 29. Depression is present in a large fraction of suicides, though most people with depression do not die by suicide. Risk is raised by prior attempts, hopelessness, substance use, access to lethal means, and social isolation. Restricting access to means is one of the best-evidenced interventions in public health: barriers on bridges, reduced pack sizes of paracetamol, pesticide bans in agricultural countries, and firearm safety measures all reduce total suicide rates rather than displacing them.

Physical illness. The excess mortality gap in severe mental illness is driven by cardiovascular disease, cancer, respiratory disease, smoking, poor access to physical healthcare, and the metabolic effects of some psychiatric medications.

Is it contagious?

No. Depression and anxiety are not transmissible.

Two nuances that get distorted. Living with someone with depression raises the risk of depression in partners and children, through stress, caregiving burden, disrupted family functioning, and shared genes and environment. That is influence, not infection.

And suicide can cluster. Media reporting that describes methods, sensationalises, or romanticises deaths is associated with increases in suicides (the Werther effect), while reporting that focuses on help-seeking and recovery is associated with decreases (the Papageno effect). This is why media guidelines on reporting suicide exist and why they are worth following.

Who gets it

In short: 332 million with depression, twice as common in women, strongly linked to adversity and poverty, and untreated in most of the world.

• Prevalence: about 332 million people with depression; anxiety disorders are similarly or more common. WHO reported in 2025 that over a billion people live with a mental health condition.
• Sex: depression and most anxiety disorders are roughly twice as common in women, a gap appearing at puberty. Explanations include hormonal, social, and diagnostic factors, and none fully accounts for it. Men have lower diagnosed rates and substantially higher suicide rates in most countries, which suggests under-recognition is part of the picture.
• Age: onset is typically in adolescence and young adulthood. Depression in older adults is common and frequently attributed to ageing rather than treated.
• Adversity: childhood abuse, neglect, and household dysfunction show strong dose-response relationships with adult depression, anxiety, substance use, and physical illness.
• Poverty and inequality: strong associations, running in both directions. Cash transfer studies in low-income settings have shown measurable improvements in mental health, which is some of the best causal evidence available about social determinants.
• Specific periods: perinatal depression affects roughly 10 to 15 percent of mothers and a smaller proportion of fathers, is underdiagnosed, and is treatable. Seasonal patterns are real at higher latitudes.
• Treatment gap: the majority of people with depression worldwide receive no treatment; in many low-income countries the figure exceeds 90 percent.

Treatment, and how it works

In short: Psychotherapy, several drug classes, exercise, and for severe cases ECT, which is the most effective treatment available and the most stigmatised.

Psychotherapy

• Cognitive behavioural therapy (CBT): identifies and tests the thought patterns and behaviours maintaining the disorder. The most studied, effective for depression and for most anxiety disorders.
• Behavioural activation: a simpler, equally effective approach for depression that schedules rewarding activity, directly interrupting the withdrawal-worsens-mood loop.
• Exposure therapy: the core treatment for phobias, panic, social anxiety, OCD (as exposure and response prevention), and PTSD. Graded, repeated contact with the feared stimulus without the avoidance response, which allows the fear response to extinguish.
• Interpersonal therapy, psychodynamic therapy, mindfulness-based cognitive therapy (which reduces relapse in recurrent depression), and EMDR for PTSD.
• Digital and guided self-help CBT, which has reasonable evidence and improves access.

Medication

How well do they work? The largest network meta-analysis (Cipriani et al., 2018, 522 trials, over 116,000 participants) found all 21 antidepressants studied more effective than placebo, with modest average effect sizes. Benefit is clearest in moderate to severe depression and smaller in mild depression, where psychotherapy and exercise are reasonable first choices. Response takes 2 to 6 weeks, and about a third of people do not respond to the first drug, which is why sequential trials and augmentation strategies exist.

Anxiety-specific notes. SSRIs and SNRIs are first-line drug treatment. They can transiently worsen anxiety in the first two weeks, so they are started at low dose. Benzodiazepines work within minutes and are appropriate for short-term crisis use; used regularly, they produce tolerance, dependence, cognitive effects, and falls in older adults, and they interfere with the learning that makes exposure therapy work, so they are not a long-term solution.

Other treatments

• Exercise. Meta-analyses find effect sizes for depression comparable to psychotherapy and medication for mild to moderate disease. It is not a substitute for treatment in severe depression, and it is a real treatment rather than lifestyle advice.
• Electroconvulsive therapy (ECT). The most effective treatment available for severe, psychotic, or life-threatening depression, with response rates around 60 to 80 percent, often within weeks. Given under general anaesthesia with muscle relaxation, it bears no resemblance to its cinematic depiction. Its principal side effect is memory disturbance, usually around the treatment period, occasionally longer, and this is a genuine trade-off rather than a myth. It is dramatically underused relative to its efficacy because of stigma.
• Repetitive transcranial magnetic stimulation (rTMS): non-invasive magnetic stimulation of the prefrontal cortex over several weeks, moderate efficacy, few side effects.
• Light therapy for seasonal patterns.
• Psilocybin and MDMA-assisted therapy: promising early trials, methodological problems (blinding is nearly impossible when the drug is psychoactive), and regulatory decisions so far cautious. Worth watching, not yet established.

What treatment costs

In short: Sexual side effects are common and systematically under-disclosed, and discontinuation symptoms are real, were understated for years, and are not addiction.

• Sexual dysfunction from SSRIs and SNRIs: reduced libido, delayed or absent orgasm, and erectile difficulty, affecting a large minority to a majority of users depending on how it is asked about. Patients are frequently not warned, and they frequently do not raise it. Persistence of these symptoms after stopping (PSSD) is reported and is now acknowledged by regulators, though its frequency is unknown.
• Discontinuation symptoms: dizziness, "brain zaps," irritability, flu-like symptoms and anxiety on stopping, especially with short-half-life drugs like paroxetine and venlafaxine. These are not evidence of addiction (there is no craving or dose escalation), and they were understated for years. Current guidance is slow, often hyperbolic tapering over months.
• Other effects: nausea and headache early, weight gain over time with several agents, emotional blunting reported by a significant minority, hyponatraemia in older people, and increased bleeding risk with NSAIDs.
• The suicidality warning: regulators added warnings about increased suicidal thoughts in people under 25 in the first weeks. The absolute risk is small, the risk from untreated depression is larger, and the practical response is close monitoring at the start rather than withholding treatment.
• Benzodiazepines: dependence, cognitive impairment, falls and fractures in older people, and dangerous interaction with opioids and alcohol.

What the person can do

In short: Motivation follows action in depression rather than preceding it, and in anxiety every avoidance strengthens the fear.

• Get assessed if symptoms persist for two weeks or more, and sooner if there are thoughts of self-harm. Most depression is treated in primary care, and effective treatment exists.
• Exercise, especially aerobic, at whatever level is achievable. Start below what feels ambitious, since a failed plan reinforces the problem.
• Protect sleep. Insomnia both precedes and worsens depression, and CBT for insomnia improves both.
• Reduce alcohol, which reliably worsens both depression and anxiety despite providing short-term relief, and which interacts badly with the medications.
• Behavioural activation on your own terms: schedule specific, small, previously enjoyable activities, and do them regardless of motivation. Motivation follows action in depression rather than preceding it, which is the single most useful practical fact in this chapter.
• For anxiety, approach rather than avoid. Every avoidance strengthens the fear. Graded, deliberate exposure is what weakens it.
• Tell someone. Isolation is both a symptom and an accelerant.
• If you are in crisis, contact emergency services or a crisis line, and remove access to means. Suicidal states are typically time-limited, and getting through the acute period matters enormously.

Living with it

Depression is uniquely self-deceiving, in that it distorts the evidence a person uses to judge whether treatment is worth attempting. Hopelessness is not a rational assessment produced by a working mind; it is a symptom, and it lifts when the depression does.

Stigma remains substantial. People conceal diagnoses at work, and the disclosure calculation is genuinely difficult. Mental health conditions are also the single largest cause of long-term work absence in several countries, and workplace interventions have been slower to arrive than the rhetoric suggests.

For families: what helps is practical presence rather than advice, taking suicidal talk seriously and asking directly (asking does not plant the idea, and the evidence on this is clear), and understanding that "just get out more" describes the destination rather than the route.

What's next

• Rapid-acting antidepressants beyond ketamine, targeting the glutamate and synaptogenesis pathway without dissociation.
• Psychedelic-assisted therapy, if the methodological problems of blinding and expectancy can be handled convincingly.
• Precision psychiatry: predicting who responds to which treatment, currently a genuine failure of the field, with imaging, EEG, and inflammatory markers all under investigation and none clinically validated.
• Anti-inflammatory approaches for the inflamed subgroup.
• Scaling delivery: task-shifted psychological therapy delivered by trained lay counsellors (the Friendship Bench in Zimbabwe, and similar programmes) has strong trial evidence and is the most plausible route to closing the global treatment gap.

Sources and notes

Depression prevalence (about 332 million) and the estimate of over a billion people living with mental health conditions are WHO figures (2025). Suicide mortality (over 700,000 deaths a year) is WHO. Diagnostic criteria are DSM-5-TR and ICD-11. Serotonin umbrella review: Moncrieff et al., Molecular Psychiatry, 2022, and the substantial published debate that followed. Antidepressant efficacy: Cipriani et al., The Lancet, 2018. Neuroplasticity and BDNF: Duman and Aghajanian, Science, 2012. Serotonin transporter gene-by-environment non-replication: Culverhouse et al., Molecular Psychiatry, 2018. Exercise for depression: Noetel et al., BMJ, 2024, network meta-analysis. ECT efficacy: UK ECT Review Group, The Lancet, 2003, and subsequent reviews. Means restriction: Hawton et al. and the WHO Preventing Suicide resource. Media reporting effects: Niederkrotenthaler et al., BMJ, 2020. Friendship Bench: Chibanda et al., JAMA, 2016.

Open questions. No biological test exists for any condition in this chapter. Why antidepressants take weeks, why they work for some people and not others, and whether depression is one condition or many are all unresolved. The frequency and mechanism of persistent sexual dysfunction after SSRIs is not established.

Next: the two conditions that shaped how society thinks about madness. 👉