Major Depressive Disorder in the Brain

Major depressive disorder is an episode lasting at least two weeks in which mood stays low or interest and pleasure are lost, along with changes to sleep, appetite, energy, concentration and self-worth.

Also called: Depression, Clinical depression, Major depression, Unipolar depression.

ICD-10 F32, F33 · ICD-11 6A70

Nobody can point at a single broken part.

There is no single broken part to point at. It would be easier to explain if there were. The evidence points at systems instead of parts. Three of them matter here: how the body handles stress, how the brain rebuilds itself, and how its networks share control. Several different routes can end in the same illness.

The five stages below run in order. They start with the story you have probably heard. They end with the one researchers currently trust most. Each one is tagged with how settled the evidence is.

The serotonin story

Evidence: contested

This is the explanation almost everyone has been given. Researchers have moved past it.

The idea did not come from studying depression at all. It came from a tuberculosis ward.

In the 1950s doctors treated TB patients with a new drug called iproniazid. The patients started cheering up. They were not recovering from tuberculosis any faster. Their mood was simply lifting. Around the same time another drug, imipramine, was tested for psychosis. It failed at that, but it lifted mood too. Neither drug had been designed to touch mood.

Chemists then asked what the two drugs had in common. Both raised a family of brain chemicals called monoamines. Serotonin and noradrenaline are two of them. In 1965 a psychiatrist named Joseph Schildkraut wrote the conclusion down formally. For the next forty years that was the story. Depression was a shortage, and the drugs topped it back up.

It is easy to see why the story stuck. It was simple. It matched how the drugs worked. It also gave people one short sentence that said "this is an illness, not a character flaw".

The problem is the reasoning. In 2022 a large review gathered decades of studies that looked for the shortage directly. They checked blood, spinal fluid, brain scans and genetics. They did not find a consistent shortage. That review was then criticised heavily over its methods, and the argument is still going on.

The honest answer is narrower than either headline. Serotonin is clearly involved in how antidepressants work. But "a chemical imbalance of serotonin" was never a good explanation of what causes the illness. Both of those are true at the same time. Most people were only ever told the first one.

Think of it this way

Paracetamol relieves headaches. Nobody says headaches are caused by a shortage of paracetamol. When a drug works on a system, that tells you the system is involved. It does not tell you that a shortage in that system started the problem. For forty years, that was roughly the mistake being made.

Why this matters more than a footnote

Imagine being told you have a chemical imbalance, and that the tablets will fix it. Then the tablets do not fix it. That story leaves you with only one explanation left, which is that something is wrong with you in particular. The next four stages are messier. They are also kinder, because none of them describes a single broken part that you happen to have.

A stress system that will not stand down

Evidence: proposed

The body's stress system loses part of its own brake.

Your body has a stress system, and it works as a loop rather than a switch. Doctors call it the HPA axis, short for hypothalamic-pituitary-adrenal, after the three organs in the loop. When you are threatened, the hypothalamus in your brain signals the pituitary gland. The pituitary signals the adrenal glands above your kidneys. They release a hormone called cortisol into your blood. Your heart rate rises, your blood sugar rises, and your attention narrows. This is a good system. It is why you can react before you finish thinking.

The last step of the loop is the one nobody mentions, and it is the one that matters here. Cortisol travels back up to the brain and tells the loop to shut down. The alarm is wired to its own off switch. That is the reason a stress response ends instead of carrying on forever.

In many people with depression, that off switch works poorly. The stop signal is sent but not fully heard. So the loop runs longer and harder than the situation needed. Sometimes it runs when nothing has happened at all.

Cortisol over a long period does damage. It wears down the exact brain structures that manage mood and memory, which is what the next stage is about. A stress response that will not stop is no longer a response. It becomes damage.

This link is strong and has been found many times. It is still not present in everyone with the diagnosis. That is part of why depression behaves like several different illnesses sharing one name.

Think of it this way

Think of a water tank with a broken float valve. The pump is fine. The pipes are fine. But the part that was meant to say "the tank is full, stop now" cannot read the level. So the water keeps running long after anyone wanted it. The fault is not in the pump. It is in the part that switches the pump off.

How anyone can tell

There is a test for this. You give someone a cortisol-like drug overnight. A healthy stress loop reads it as "there is plenty already" and turns its own output down by morning. In some people with severe depression, it does not turn down properly. The test is far too rough to diagnose anyone, so your doctor will not run it. It is accurate enough to show that the brake is real, and that in some people it slips.

The hippocampus gets smaller

Evidence: established

This is a difference you can measure on a scan. When it appears is the interesting part.

The hippocampus is a small structure shaped like a seahorse. There is one deep inside each side of the brain. It does two jobs, and they turn out to be related. It stores new memories, and it helps shut down the stress loop from the last stage.

A research group called ENIGMA pooled MRI brain scans from 1,728 people with major depression and 7,199 people without it. The size of that study matters. The effect they were looking for is small, so a study of forty people could easily have found it by luck, or missed it. They found that people with depression had a smaller hippocampus.

What they did not find is the interesting part. There was no difference in people having their first episode. The whole difference came from people who had episodes again and again. It also tracked how many episodes they had had, and how long those lasted.

Read in that order, this says the opposite of what people usually assume. It does not look like a weakness you were born with. It looks like the result of the illness running, and running repeatedly. The scan does not show you why it happened. It shows you what it cost.

One warning is needed here. These are averages across large groups, and the difference in any one person is small. No brain scan can diagnose depression. Anyone offering you one is selling something.

Think of it this way

Think of a grass pitch used for football. Play one match on it and the grass recovers within a week. Photograph it a month later and you would see nothing. Play on it every day for two years and bare patches appear that never fill in. The grass was not weaker than any other grass. The wear simply tracks the number of matches.

And it is not one-way

The size of the hippocampus is not a permanent record of everything you have been through. It responds to the conditions it is kept in. The next stage picks that up. The same structure is one of the places where the brain builds new connections, and that building can start again.

The brain stops rebuilding itself

Evidence: proposed

This explanation currently accounts for the most.

Your brain is never finished. Brain cells constantly rebuild their connections. They grow new branches, cut back unused ones, and strengthen the ones you use. Scientists call this plasticity. It is not a comparison for learning. It is the physical thing that learning is made of, and it is happening as you read this.

Long-term stress slows that rebuilding down. The branches that receive signals, called dendrites, shrink back. Connections between cells, called synapses, are lost. The chemicals that support the whole process drop. This happens most in the hippocampus and the prefrontal cortex, the two areas the last two stages were about.

On this explanation, low mood is a knock-on effect. The main problem is not that you feel bad. It is that the system which would normally let you learn your way out of feeling bad has slowed down. A brain that cannot update cannot climb out of a state it has fallen into. It can only sit there.

This also explains something the serotonin story never could. If depression were simply low serotonin, a selective serotonin reuptake inhibitor, or SSRI, should work within a day. Raising serotonin in the synapse only takes hours. But it does not work in a day. It takes weeks, in almost everyone.

Rebuilding connections takes weeks. The drug is not slow to arrive. The drug arrives at once, and the repair takes as long as repair takes.

Think of it this way

Think of a city that has stopped repairing its roads. Nothing collapses. Every road that existed yesterday is still there and traffic still moves. But nothing new gets built. So the first time a road closes, there is no other way round, and you take the same wrong turning every morning. You are not lost because the map is wrong. You are stuck because nobody is drawing new roads.

The most hopeful thing on this page

Suppose the main problem really is slowed rebuilding. Then it is not a missing chemical, and it is not a permanently damaged structure. It is a process, and a process can start again. This also explains why treatments that look unrelated all work. Drugs, therapy, exercise and sleep all push on the same thing from different directions.

The rumination loop

Evidence: proposed

This is why the thinking changes too, and not only the mood.

One set of brain regions switches on whenever you stop paying attention to the outside world. It runs when you remember last Tuesday, replay a conversation, wonder what someone meant, or imagine next year. Scientists call it the default mode network. The brain defaults to it when nothing outside is asking for attention.

This network is not a fault. It holds your memory of your own life, your planning, and most of what people mean by an inner life.

In depression this network runs too much. More importantly, it becomes hard to switch away from. The outward-facing networks should take over when someone speaks to you or something needs doing. In depression they struggle to take over.

From the inside this feels like rumination. Rumination is not the same as thinking about a problem, because thinking ends when the problem is solved. Rumination circles instead. The same thoughts, from the same angle, reaching the same place, then starting again. It often happens at three in the morning.

This is why people say depression changed how they think, not only how they feel. It is also why "just stop dwelling on it" lands so badly. The advice is not wrong about where you want to end up. It is addressed to the exact system that is not working.

Think of it this way

Think of a tab on your phone that will not close. You have read it. There is nothing new in it. It is simply still open and still in front of you. Every time you put something else on the screen, it comes straight back the moment that thing ends.

Which is why therapy is not the soft option

The problem here is being unable to pull attention away from an inner loop. Therapy trains exactly that skill. So it is not a gentler substitute for medication working on the biology. It acts directly on the same mechanism. Brain imaging supports this. Cognitive behavioural therapy (CBT) and antidepressants both change activity in this network, arriving from opposite ends.

The same thing, without the vocabulary

Long stress leaves the body's alarm system stuck on. Months of stress hormones damage the parts of the brain that handle memory and mood. They also slow down the brain's ability to rewire itself.

A brain that cannot rewire properly struggles to get out of a rut. At the same time, the inward-looking part of the brain is running too hard. So the same painful thoughts keep coming round.

Put those three together: an alarm that will not switch off, a reduced ability to change, and thinking that goes in circles. That is much closer to what depression is than "not enough serotonin" ever was.