What causes depression?

Quick view

If you may be in immediate danger or are thinking about ending your life, call or text 988 in the United States, or call 911.

The short version

Depression is the end result of several systems shifting in the same direction at the same time. Most people who develop a depressive episode have some combination of: a biological vulnerability, often inherited; one or more life experiences that loaded the system, including childhood adversity, trauma, chronic stress, or major loss; a current trigger that tipped the balance, sometimes obvious, sometimes not; and, in many cases, an unaddressed medical or substance contributor that is keeping the system stuck.

No single factor in that list, on its own, causes depression in most people. The combination is what does it.

Biology: genes, brain circuits, and the systems that are dialed down

Depression runs in families. The heritability estimate from twin studies sits near 37 percent (Sullivan et al., 2000), which means that genetic differences account for a meaningful share of who develops depression but do not predict it on their own. Genome-wide association studies have identified more than 100 genetic variants linked to depression risk, none with a large effect on its own (Howard et al., 2019). The biology of depression is polygenic. There is no depression gene.

What is happening in the brain during a depressive episode is more complicated than the older shortage-of-serotonin framing suggests. A 2022 systematic umbrella review by Moncrieff and colleagues found no consistent evidence that low serotonin causes depression. That does not mean serotonin is unrelated. It means the simple “chemical imbalance” framing was an oversimplification.

What current research does support.

• Reward and motivation circuits. Activity in the ventral striatum and connected regions is altered during a depressive episode, which is part of why anhedonia and low motivation are core symptoms.
• The HPA axis. The body’s stress response system is often dysregulated in depression. Cortisol rhythms shift. Stress feedback loops do not turn off the way they should (Pariante and Lightman, 2008).
• Inflammation. A subset of people with depression have elevated inflammatory markers. Chronic inflammation can produce depression-like symptoms in animals and humans, and depression appears to involve neuroimmune signaling (Slavich and Irwin, 2014).
• Sleep architecture. REM sleep, slow-wave sleep, and the timing of sleep stages all shift during depressive episodes.
• Neuroplasticity. Many of the treatments that work for depression (antidepressants, ECT, ketamine, exercise, psychotherapy) appear to act partly by supporting neuronal connections and the brain’s capacity to reshape its own circuitry.

Most clinicians now describe depression as a network disorder. The systems involved in mood, reward, sleep, appetite, energy, and self-view shift together. Treatment works partly by helping those systems reset.

Life history: early experience and accumulated load

Some of the strongest predictors of adult depression sit in childhood and adolescence.

• Adverse Childhood Experiences (ACEs). The original ACE Study (Felitti et al., 1998) and many subsequent studies have shown that a history of abuse, neglect, household dysfunction, parental mental illness, or parental substance use raises the risk of depression in adulthood. The risk rises with the number of ACEs.
• Loss in childhood. Loss of a parent, prolonged separation, or significant disruption of attachment in early childhood is associated with higher adult depression risk.
• Chronic stress. Long-running exposure to difficult conditions (poverty, bullying, family conflict, caregiving for a sick parent) loads the system.
• Trauma. A history of physical, sexual, or emotional trauma, particularly when it has not been addressed, is a strong predictor. Trauma-focused therapy can be part of a depression treatment plan in this picture.
• Bullying and social rejection during adolescence have lasting effects on depression risk.

None of these are deterministic. Many people with extensive childhood adversity never develop depression. Many people with little adversity do. But the load matters at the population level and often matters at the individual level when a clinician is trying to understand a person’s pattern.

Current circumstances: what is happening right now

The third layer is the current environment.

• Major losses. Death of a partner, parent, or child. End of a long relationship. Loss of a job, a home, a role.
• Chronic stress. Financial stress. Caregiving. A demanding job without recovery. A difficult living situation. The body’s stress system was not designed for prolonged activation, and over time it shifts in ways that look biologically similar to depression.
• Isolation. Loneliness is an independent risk factor for depression and shows up in research as comparable in size to many other major risk factors (Holt-Lunstad et al., 2015).
• Role transitions. Becoming a parent. Retirement. A child leaving home. A career change. Many people develop a first depressive episode at one of these transitions.
• Sleep loss. Sustained sleep deprivation can produce depressive symptoms and worsen an existing episode.
• Substance use. Alcohol, cannabis at high frequency, stimulants, and opioids can all contribute to or worsen depression.
• Light exposure. Reduced light in winter contributes to seasonal patterns of depression for some people.

Stress, loneliness, and disrupted sleep affect the same biological systems that are involved in depression. The line between “environmental cause” and “biological cause” is fuzzier than it sounds.

Medical contributors: what is often missed

A meaningful share of depressive episodes are partly or fully driven by an underlying medical condition or a medication. These are the contributors a clinician usually screens for at a first visit.

• Thyroid disease. Hypothyroidism in particular can present as depression. A simple TSH blood test catches it.
• Anemia and vitamin deficiencies. Iron, vitamin D, B12, and folate deficiencies can contribute to fatigue, low mood, and concentration problems.
• Sleep apnea. Untreated obstructive sleep apnea produces fatigue, low mood, and irritability that often looks like depression.
• Chronic pain. Persistent pain alters mood circuits over time.
• Neurologic conditions. Parkinson’s disease, stroke, traumatic brain injury, and multiple sclerosis all raise depression risk.
• Cardiovascular disease and diabetes. Depression is more common in people with these conditions, in both directions.
• Cancer. Especially during active treatment.
• Hormonal shifts. Pregnancy, postpartum period, perimenopause, menopause, and certain hormonal contraceptives can all contribute in vulnerable people.
• Medications. Some medications can lower mood. Examples include certain corticosteroids, interferon, some hormonal contraceptives, certain blood pressure medications, and some treatments for acne and chronic pain. This is not a reason to stop any medication on your own. It is a reason to bring it up with a clinician.
• Substance use. Alcohol is a depressant. Cannabis at high frequency lowers mood for many people. Withdrawal from stimulants or opioids commonly produces depressive symptoms.

A first evaluation usually screens for these. When one is missed, a treatment plan can stall in ways that look like the depression is not responding when in fact the underlying contributor is keeping it in place.

The stress-diathesis model: the useful frame

A frame that brings these layers together is the stress-diathesis model. The idea is simple. Most people have a biological vulnerability to depression that is partly inherited and partly shaped by life history. That vulnerability sits quietly until something tips it. The “something” can be a major loss, a long stretch of stress, an illness, a substance, or a transition. People with a high vulnerability tip easily and with smaller triggers. People with a low vulnerability take more loading to tip.

The model is not perfect. It is useful because it explains both the heritability and the role of life events, why the same event causes depression in one person and not another, and why treatment focuses on multiple layers at once.

What about the “chemical imbalance” idea

For about thirty years, the public conversation about depression centered on a serotonin deficit. Antidepressants were often explained as fixing that deficit. Both halves of that explanation turned out to be too simple.

The Moncrieff review and other recent work make clear that depression cannot be reduced to a shortage of any single neurotransmitter. The biology is a network shift, not a single chemical missing.

What is also true. Antidepressants help many people. The fact that the original mechanism story was oversimplified does not change that. Current understanding is that antidepressants work through several mechanisms, including downstream effects on neuronal connections, the stress response, and the brain’s capacity to reshape itself over weeks.

The practical implication for a patient: choosing an antidepressant is not a matter of fixing a known deficit. It is a matter of finding a medication that produces a meaningful response with tolerable side effects. The first one tried is not always the right one.

Why “what caused mine” is often unanswerable

A common question in a first appointment is “what caused this in me.” The honest clinical answer is often a list. Heritable risk. A few significant life events. Some chronic stress. A sleep pattern that has not been quite right for years. One medical contributor on the edge of normal. A current trigger. The relative weight of each, in any one person, is rarely knowable.

What is more answerable is what is keeping the depression going now, and what is most likely to help. That is what an evaluation produces. The Just Diagnosed article on this site covers what to expect from that conversation, and Depression vs. Sadness covers when a low period crosses into a clinical pattern.

If looking for a single cause is taking up a lot of mental energy, that is worth saying out loud to a clinician. People often arrive at peace with the question after they have started feeling better.

Why understanding causes matters for treatment

Knowing the contributors changes the plan.

• If there is a strong family history of bipolar disorder, treatment is approached differently.
• If there is a clear medical contributor, addressing it is part of the plan.
• If there is significant trauma history, trauma-focused therapy may be part of the plan.
• If chronic stress and isolation are dominant, the plan looks different from a plan where the stress has lifted but the symptoms have not.
• If a substance is in the picture, treatment for the substance use is part of the depression plan, not separate from it.
• If sleep apnea is undiagnosed, a sleep study can be more useful than a second antidepressant.

Understanding causes is also useful because it removes some of the guilt people carry into a first appointment. Depression is not a sign of poor character. It is the result of measurable biological and environmental processes interacting over time. People do not cause their own depression any more than they cause their own pneumonia.

Related: anxiety resources

For the parallel question of what causes anxiety, see our sister publication AnxietyResource.org, edited by the same physician reviewer