What Is Endogenous Depression and Why Did the Term Change?

Endogenous depression is an older psychiatric term for a form of depression thought to arise from within the body, driven by biological processes rather than triggered by life events. The concept dates back to the earliest modern writings on melancholia and has shaped how clinicians think about depression for well over a century. Though the term itself largely dropped out of official diagnostic manuals in 1980, the clinical reality it tried to capture hasn’t gone away. Today it survives as the “melancholic features” specifier in major depressive disorder, and a growing body of research in neuroimaging, immunology, and genetics suggests that the old distinction between biologically driven and situationally triggered depression may have been onto something real, even if the original framing was too tidy.

Where the Term Came From and Why It Disappeared

The idea that there are two fundamentally different kinds of depression is nearly as old as psychiatry itself. From the earliest modern writings on melancholia, clinicians distinguished a depression that seemed to well up from the body with no clear external cause from one that followed grief, loss, or stress. The former was called “endogenous” (from within), the latter “reactive” or “neurotic.” For the major classification builders in psychiatry, melancholic depression always meant something quite different from non-melancholic depression.¹ The debate around these overlapping distinctions picked up steam in the 1920s and generated a wave of statistical research in the 1960s.²

Then came a dramatic shift. In 1980, the third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III) merged what had previously been two separate categories. Depressive states associated with neurosis and those associated with manic-depressive illness (including melancholia and endogenous depression) were folded into a single umbrella category called “major depression.”³ The rationale was partly practical: earlier research hadn’t cleanly separated the two groups using available tools, and lumping them together made large-scale research easier. But the merger came at a cost. What had been recognized for decades as a biologically distinctive condition was now just a modifier tacked onto a broader diagnosis. The symptom element of the old “endogenous” concept survives today as the melancholia or somatic syndrome specifier.⁴

What Melancholic Depression Looks Like

If you or someone you know has been diagnosed with major depression with melancholic features, the symptom profile looks different from what most people picture when they think of depression. The hallmark is a near-total loss of pleasure in all or almost all activities, or a failure to feel better even temporarily when something good happens. In non-melancholic depression, a funny movie or a visit from a friend can lift the mood for a while. In melancholic depression, that emotional responsiveness shuts down.

Beyond the flattened emotional landscape, the physical symptoms tend to dominate. Early-morning waking (often hours before the alarm), a mood that is distinctly worse in the morning, significant weight loss or appetite suppression, psychomotor changes (either marked physical slowing or agitation), and excessive guilt that feels out of proportion to the situation are all characteristic. This cluster gives the condition a bodily, physiological quality that people often describe as feeling less like sadness and more like being sick.

Research backs up the idea that melancholic depression isn’t just “worse depression.” In a large study comparing melancholic and non-melancholic groups, people with melancholic features (roughly a third of the broader depression sample) showed measurably poorer performance across every domain of thinking tested, as well as slower speed of recognizing emotions in others. Compared to the non-melancholic group specifically, the deficits showed up in information-processing speed, decision speed, and the ability to pick up on happy expressions, even after accounting for overall symptom severity.⁵ In other words, the cognitive profile isn’t simply a consequence of being more depressed. Something about the melancholic subtype affects thinking in distinctive ways.

Sleep Disruption and the Body Clock

Sleep problems are almost universal in depression, but the pattern in melancholic depression has some distinctive quirks. People with depression generally show disrupted rapid eye movement (REM) sleep, including falling into REM faster after falling asleep, spending more time in REM during the first sleep cycle, and having more intense REM periods.⁶ Early researchers hoped these REM abnormalities would serve as a biological marker that could separate endogenous from non-endogenous depression. That hope didn’t pan out cleanly. One study of 45 patients with major depression found that shortened REM latency (falling into REM too quickly) was present to a similar degree whether the depression was classified as endogenous, neurotic, or unclassified.⁷ Later reviews confirmed that the early promise of REM sleep as a specific marker for endogenous or melancholic depression had not been fulfilled.⁸

There is, however, evidence that broader circadian (body clock) rhythms are genuinely disrupted in endogenous depression. A study of melatonin levels throughout the day and night found that people with endogenous depression had lower melatonin across the entire 24-hour cycle, a delayed onset of their nighttime melatonin surge, and an earlier peak in melatonin levels compared to healthy controls.⁹ This fits the lived experience many melancholic patients describe: the body’s internal timing feels off, with the worst mood and energy in the morning and gradual improvement as the day goes on.

What Biology Reveals

One of the strongest biological threads running through melancholic depression is an overactive stress-hormone system. The dexamethasone suppression test (DST), which checks whether the body can properly dial down its cortisol response, became a landmark tool in melancholia research. A groundbreaking 1981 paper reported that the DST detected melancholic depression with about 67% sensitivity and 95% specificity.¹⁰ That high specificity means the test rarely flagged someone as melancholic who wasn’t, even though it missed a fair number of cases. While the DST never became a routine clinical diagnostic tool (its sensitivity was too low), it pointed clearly toward something going haywire in the stress-hormone axis in this form of depression.

Inflammation also looks different in melancholic depression compared to other depression subtypes. Multiple studies have found that people with melancholic features have elevated levels of the inflammatory marker IL-6 compared to healthy controls, whereas people with atypical depression (the opposite profile, with oversleeping and overeating) do not show the same IL-6 elevation.¹¹ Separate research confirmed this pattern, finding that melancholic patients had higher IL-6 and lower levels of TGF-β, a regulatory molecule that normally helps calm the immune system.¹² The IL-6 finding appears tied to symptom severity: among patients with melancholic features, those with the most pronounced psychomotor slowing had the highest IL-6 levels.¹³

Brain-imaging research adds another layer. People with melancholic depression who have never taken medication show differences in gray matter volume that correlate spatially with the distribution of serotonin transporters, cannabinoid receptors, and opioid receptors in the brain.¹⁴ Functional imaging has revealed disrupted activity in the default mode network, the brain system most active when you’re mind-wandering or reflecting inward. People with melancholic depression showed abnormal patterns in parts of the temporal lobe within this network, suggesting a physiological basis for the rumination and self-focused thinking that characterizes the condition.¹⁵

Heritability and Why Families Matter

If endogenous depression is truly driven by biology, you’d expect it to run in families, and it does. But the genetic picture is less straightforward than you might guess. A study that estimated heritability for depression subtypes found that the melancholic subtype had a heritability of about 0.33, meaning roughly a third of the variation in who develops it can be attributed to genetic factors. The atypical subtype actually had a somewhat higher heritability, around 0.46.¹⁶ The fact that both subtypes show significant heritability, but through partially different genetic channels, supports the idea that they are biologically distinct conditions rather than simply mild and severe versions of the same thing.

A heritability of 0.33 also means that about two-thirds of the risk comes from non-genetic factors. This is an important check against the simplistic reading of “endogenous” as “purely internal.” Even in the subtype most traditionally associated with biological causation, life circumstances, stress exposure, developmental history, and other environmental influences play the larger role. The old endogenous-versus-reactive dichotomy was always an oversimplification, and genetics confirms it.

Treatment Differences That Actually Matter

Whether melancholic depression responds differently to treatment is one of the most practically important questions surrounding the diagnosis, and the answer is genuinely interesting. People with melancholic features do not consistently differ from those with non-melancholic depression in how well they respond to antidepressant medication or electroconvulsive therapy (ECT). But there is one striking and consistent difference: people with melancholia show a much lower rate of placebo response.¹⁷ This has major implications. In a condition where the body’s reward and pleasure systems appear biologically disrupted, it makes sense that the expectation of getting better (which is what drives the placebo effect) doesn’t carry as much weight. It also means that clinical trials for antidepressants are more likely to show a real drug effect in melancholic patients, because the placebo arm improves less.

Psychotherapy, on the other hand, appears to be at a disadvantage in melancholic depression. A 12-week trial comparing antidepressant medication to cognitive behavioral therapy (CBT) in people with melancholic depression found that the medication group improved substantially more. Scores on a standard depression scale improved by about 61% in the medication group versus roughly 34% in the CBT group, with the difference becoming visible as early as four weeks.¹⁸ This doesn’t mean therapy is useless for people with melancholic features. But it does suggest that treating melancholia with talk therapy alone, without medication, may leave the biological drivers of the condition unaddressed.

ECT remains one of the most effective interventions for severe melancholic depression, though the evidence is more nuanced than often assumed. While ECT generally works well for severe, treatment-resistant depression, one study using a specific melancholia rating scale found that the melancholic designation did not independently predict who would respond to ECT once other factors were accounted for.¹⁹ The picture that emerges is that ECT works for severe depression broadly, and melancholia doesn’t get a special bonus on top of that.

Newer treatments are also reshaping the landscape. Ketamine, which works through entirely different brain pathways than traditional antidepressants, can produce rapid effects. Research shows that within minutes, ketamine promotes new connections between nerve cells, normalizes disrupted brain-network activity, and can relieve symptoms within hours.²⁰ The speed of ketamine’s action is particularly relevant for melancholic depression, where the biological disruption seems to go beyond simple serotonin or norepinephrine deficits. Whether melancholic patients respond specifically better to ketamine than other depressed patients remains an active question, but the drug’s mechanism fits well with the neurobiological picture of the condition.

Somatization and Cultural Context

One commonly overlooked aspect of melancholic depression is its physical symptom burden. People with this subtype don’t just feel emotionally flat; they often experience genuine physical discomfort, heaviness, pain, or a visceral sense that something is physically wrong. Research comparing melancholic patients in Turkey and Germany found an interesting split. Both groups reported similar levels of actual somatic symptoms, like pain or digestive disturbance, suggesting a shared neurobiological core. But the Turkish patients were more likely to show excessive worry about those symptoms and hypochondriacal fears.²¹ The researchers concluded that the raw perception of bodily distress in melancholia may be a built-in neurobiological feature, while the degree of alarm and preoccupation with those symptoms is shaped by culture. This distinction matters for diagnosis, because in cultures where depression tends to be expressed through physical complaints rather than emotional language, the somatic core of melancholia can either be recognized for what it is or misattributed to medical illness.

The Future of Depression Subtyping

The central tension in the story of endogenous depression is that clinicians have always sensed that depression comes in meaningfully different forms, but the tools to prove it have lagged behind the intuition. The DSM’s decision to lump everything into “major depressive disorder” created a diagnosis that many researchers now see as unhelpfully broad. As one recent review put it, the current system groups patients with wildly different presentations together, with no biomarkers to guide treatment, comparable to diagnosing heart disease solely by chest pain without imaging to reveal the underlying problem.²²

Functional brain imaging is now opening a different approach entirely: defining depression subtypes not by symptoms alone, but by patterns of brain-circuit dysfunction. Researchers have developed methods that score dysfunction across multiple large-scale brain circuits relative to healthy norms. These personalized circuit profiles can predict which treatment is most likely to work for a given individual. Early results suggest that matching treatment to a patient’s brain-circuit profile could potentially double remission rates compared with unmatched treatment.²³ This kind of biotype-based approach doesn’t map neatly onto the old endogenous-versus-reactive distinction, but it does vindicate the core insight behind it: that different depressions have different biological fingerprints and may need different treatments.

Whether the field eventually resurrects “endogenous depression” as a formal category, refines “melancholia” into something more biologically precise, or replaces both with brain-circuit biotypes, the practical takeaway is already clear. If your depression is characterized by a total loss of pleasure, morning worsening, significant physical symptoms, and poor response to placebo, it likely reflects a different biological process than depression that tracks closely with life stressors and responds well to therapy alone. Knowing that can change how you and your clinician approach treatment, and how quickly you move toward medication or other biological interventions when the first approach isn’t working.