In the 1960s, a working hypothesis emerged from early antidepressant research: depression might involve too little serotonin, one of the brain’s chemical messengers. It was a compelling and communicable idea, and over the following decades it became the prevailing public explanation for why depression happens. Pharmaceutical advertisements made it familiar. Clinicians used it as shorthand. Patients found it offered a coherent frame for an experience that had otherwise been hard to explain.

The problem is that the research never fully confirmed it. Over the past several years, scientists reviewing the accumulated evidence have reached a more complicated conclusion — one that points toward a condition shaped by multiple systems rather than a single shortfall. Understanding what that research shows does not change what someone living with depression experiences. It does change the questions researchers and clinicians are now asking about why it develops, why it affects people so differently, and why treatments that work well for some people show little effect for others.

How the Chemical Imbalance Idea Took Hold

The monoamine hypothesis of depression dates to the early 1960s. Researchers noticed that drugs which depleted monoamines — a category that includes serotonin, norepinephrine, and dopamine — could sometimes produce depressive symptoms. Separately, medications that increased monoamine availability seemed to have antidepressant effects. From these observations, a theory took shape: depression might be caused by insufficient activity in these neurotransmitter systems.

Serotonin became the most prominent target. Selective serotonin reuptake inhibitors, introduced in the 1980s, were effective for a meaningful proportion of people with depression, and their mechanism of action — slowing the reabsorption of serotonin at synapses — appeared to support the deficit model. By the 1990s and 2000s, the idea that depression was caused by a “chemical imbalance” in serotonin had moved from a research hypothesis to something close to established fact in mainstream culture.

What the hypothesis never had, though, was direct and consistent evidence. Studies looking at serotonin levels, serotonin metabolites in cerebrospinal fluid, or serotonin receptor availability in living human brains produced inconsistent results. The mechanism of antidepressants was inferred backward from their effects — but a drug that relieves a symptom does not necessarily reveal the cause. Aspirin relieves a headache; that does not mean headaches are caused by an aspirin deficiency.

When Researchers Systematically Reviewed All the Evidence

In 2022, a team of researchers published what may be the most thorough examination of the serotonin hypothesis to date. Moncrieff and colleagues conducted an umbrella review — a systematic review of systematic reviews and meta-analyses — examining the accumulated evidence across multiple lines of serotonin research: serotonin metabolite levels, serotonin receptor binding, serotonin transporter levels, the effects of tryptophan depletion, and genes related to serotonin function (Moncrieff J et al., Molecular Psychiatry, 2023).

Their conclusion was unambiguous: the evidence did not support the hypothesis that depression is associated with or caused by lower serotonin activity. Levels of the serotonin metabolite 5-HIAA were not consistently lower in people with depression. Serotonin transporter levels — the actual mechanism targeted by SSRIs — were, if anything, sometimes higher in people with depression than in controls, not lower. Tryptophan depletion studies, which reduce available serotonin, did not reliably produce depressive symptoms in people without a history of depression.

The paper attracted significant attention precisely because the deficit model had become so embedded in both clinical communication and public understanding. The authors did not argue that antidepressants are ineffective — the evidence for their effectiveness in some patients is separate from the question of mechanism. What they argued is that the biological rationale that came to define how depression was explained to patients and the public was not well-founded. That gap between explanation and evidence matters, because it shapes what researchers investigate and what patients understand about their own condition.

What Chronic Stress Does to the Brain

If a serotonin shortage does not explain depression, what does current research point toward? One of the most consistently studied biological factors is the hypothalamic-pituitary-adrenal (HPA) axis — the body’s primary stress-response system.

Under acute stress, the HPA axis triggers the release of cortisol, a hormone that helps mobilize energy resources and heightens alertness. In people with depression, a substantial proportion show evidence of HPA axis dysregulation: cortisol output that remains elevated beyond what the situation warrants, and a blunted ability to rein the system back down. Menke, writing in Current Neuropharmacology in 2024, reviews how chronic HPA axis activation can have measurable effects on brain structure and function — particularly in the hippocampus, a region involved in memory and stress regulation. Sustained high cortisol is associated with reduced hippocampal neurogenesis, the process by which the hippocampus generates new neurons. It is also associated with changes in the prefrontal cortex, a region involved in regulating mood and executive function.

This matters for understanding why depression can develop after prolonged stress, why it can be difficult to recover from even after the stressor resolves, and why early adverse experiences can raise the risk of depressive symptoms later in life. It also suggests that stress biology — not just neurotransmitter levels — is part of what any complete account of depression needs to explain.

No single person’s biology is identical, and HPA dysregulation is not universal in depression. Population-level associations cannot predict how this pathway will behave in any specific individual.

The Emerging Role of Neuroinflammation

A different line of evidence has pointed toward inflammation as another factor implicated in at least some presentations of depression. Studies have consistently found elevated levels of pro-inflammatory cytokines — signaling proteins of the immune system — in a subset of people experiencing symptoms of depression. The hypothesis is not that depression is an inflammatory condition in the way that arthritis is, but that inflammatory processes in the nervous system can disrupt the function of cells involved in mood regulation, reward processing, and neuroplasticity.

Puentes-Orozco and colleagues, writing in Frontiers in Cellular Neuroscience in 2024, focus specifically on astrocytes — glial cells that are the most abundant cell type in the brain, and that play a central role in regulating synaptic transmission, maintaining the blood-brain barrier, and supporting neuronal function. Their review documents how astrocyte dysfunction and neuroinflammatory signaling interact in major depressive disorder, with consequences for the circuits that govern motivation, energy, and emotional response.

This line of research is notable for what it says about heterogeneity. Depression as currently defined is not a single biological entity — it is a category of symptoms that likely reflects several different underlying processes. The subset of people with depression who also show elevated inflammatory markers may have a meaningfully different biological picture from those who do not. This could help explain why no single treatment works reliably for all people, and why identifying biological subgroups may be a more useful framework than a unified deficit model.

How Imaging Research Has Changed the Picture

Functional neuroimaging has added another dimension to the current understanding. Rather than measuring the levels of a single molecule, brain imaging studies examine how different regions communicate with each other — and whether those patterns of communication differ in people experiencing depression compared with those who are not.

The findings have been consistent enough to shift how many researchers think about the condition. Depression is increasingly understood as a disruption of network-level function — particularly the default mode network, which is active during self-referential thinking and rumination, and the reward circuitry involving the prefrontal cortex and striatum. Alterations in these networks correlate with the core symptoms of anhedonia, low energy, and perseverative negative thinking.

Wang and colleagues, writing in BMC Psychiatry in 2024, examined dynamic functional connectivity in first-episode, drug-naïve patients — people who had not yet received any medication, allowing the researchers to observe the brain’s state before any pharmaceutical intervention. They found altered connectivity patterns in what they describe as antagonistic brain states, adding to the body of evidence that network-level disruption characterizes the depressed brain.

The practical implication of this framework is that depression is not a deficiency of one thing in one place. It is a change in how the brain’s distributed systems communicate with each other — a systems-level problem that may require understanding multiple interacting components: genetic predisposition, stress history, inflammatory burden, sleep quality, and the social and environmental context a person lives in.

What the Evidence Does Not Yet Tell Us

The shift away from the chemical imbalance model does not mean researchers have arrived at a complete and settled understanding of what depression is. The studies described here are informative about patterns observed in groups; they cannot predict how any individual person’s neurobiology is organized, or which factors are most relevant to any one person’s experience.

Several important questions remain open. The neuroimaging research showing network disruption is associational: whether those patterns are a cause of depression, a consequence of it, or a correlated feature of some other underlying process is not yet fully resolved. The inflammation findings are robust enough to support a subgroup hypothesis but not yet robust enough to guide individual clinical decisions — there is no standard test that reliably identifies the subset of people for whom inflammatory pathways are central. The HPA axis findings are similarly population-level: elevated cortisol is more common in people with depression, but many people with depression do not show this pattern, and many people with elevated cortisol do not develop depression.

Genetics adds further complexity. Large-scale studies have identified hundreds of genetic variants that each contribute a small amount to depression risk — a polygenic picture that contrasts sharply with the idea of a single causal mechanism. Environment interacts with genetic predisposition in ways that researchers are still working to characterize. The honest summary of where the science stands is that it has dismantled an oversimple explanation and replaced it with a more accurate but still incomplete one.