Major depressive disorder · Lobeworks/17

Major depressive disorder is a mood disorder defined by episodes of at least two weeks of low mood or lost interest and pleasure, joined by changes in sleep, appetite, energy, concentration and self-worth, and it is one of the leading causes of disability in the world.

Major depressive disorder. Major depressive disorder is a mood disorder defined by episodes of at least two weeks of low mood or lost interest and pleasure, joined by changes in sleep, appetite, energy, concentration and self-worth, and it is one of the leading causes of disability in the world.

The diagnosis is a count of symptoms, not a lesion. DSM-5 asks for five of nine symptoms, one of them low mood or anhedonia, which allows at least 227 combinations: two people with the same diagnosis may share a single symptom. That heterogeneity is why the brain findings are averages. Groups of patients show a amygdala that answers negative faces and words more strongly, a dorsolateral prefrontal cortex less engaged when a thought has to be reappraised, abnormal activity in the subgenual part of the cingulate cortex (the region whose activity changes most consistently with recovery), and a slightly smaller hippocampus: in the largest pooled analysis (ENIGMA, 1,728 patients and 7,199 controls) about 1.2 % smaller, driven by recurrent and early-onset illness, with no difference at a first episode. The negative thoughts that keep returning and take over are the felt side of that circuit: a threat detector turned up and a control system that pulls less.

The monoamine hypothesis came from drugs found by accident in the 1950s that raised serotonin and noradrenaline. Its simple form, a lack of serotonin, has no consistent support (an umbrella review in 2022, itself contested), and antidepressants change monoamines in hours but mood in weeks, which points to slower changes downstream.

Stress enters through the HPA axis: an overactive axis is one of the most replicated findings in depression, thought to come in part from weakened cortisol feedback. Inflammation is another route: on average IL-6 and CRP are raised, a link that weakens once body weight is taken into account (see neuroinflammation).

The treatments act at different levels: medicines on monoamine signalling, psychotherapy on the patterns of thought, TMS on the prefrontal cortex, ECT for severe or resistant cases, and esketamine, approved in 2019, on NMDA receptors.

No single brain marker has been found, and the search itself is changing: normative modelling asks where each patient sits against a reference population instead of comparing group means.

Depression is diagnosed by symptoms and seen in the brain only on average.

The group differences are real and small; any one patient may show none of them.

Questions: What does the amygdala do differently in depression? On average it answers negative faces, words and memories more strongly and for longer, so threat and loss gain weight in what is noticed and remembered. At the same time the prefrontal regions that would damp it engage less, the circuit behind the felt experience of negative thoughts that keep returning and take over. These are group findings: many patients show no measurable difference, and the amygdala is one node of a network, not the seat of the disorder. What is the inflammation hypothesis of depression? It proposes that for some people depression is partly the brain's response to inflammation. The starting observation is that sickness behaviour, the fatigue, withdrawal and low mood of an infection, is produced by cytokines acting on the brain, and that treating patients with the cytokine interferon-alpha brings on depression in part of them. On average people with depression have raised IL-6 and C-reactive protein, but the differences are small, overlap widely with healthy people and shrink after accounting for body weight, so inflammation looks like one route into depression for a subgroup, not its general cause. Why has no single brain marker of depression been found? Because the diagnosis gathers different conditions under one name: five of nine symptoms allow at least 227 combinations, and the routes in (stress, inflammation, genetics, illness) differ between people. Comparing patients with controls finds real but small group differences, such as a hippocampus about 1.2 % smaller in recurrent illness, by which no single patient can be diagnosed. Normative modelling changes the question: instead of looking for one difference shared by the group, it measures how each individual departs from a reference population, and finds that patients scatter. How can chronic stress turn into a loop that keeps the stress axis on? Normally cortisol switches its own release off by acting on receptors in the hypothalamus, the pituitary and the hippocampus. Under prolonged stress that feedback weakens: the receptors respond less, and the hippocampus, which helps hold the axis down, loses dendrites in animal studies. The axis then stays on, cortisol stays high, and high cortisol further weakens the regions that would turn it off. An overactive HPA axis is one of the most replicated biological findings in depression, though it is not present in every patient.