Physical health

Autoimmune disease and schizophrenia: NMDA encephalitis and beyond

April 25, 2026 8 min read

For most of the twentieth century, schizophrenia and autoimmune disease were considered unrelated worlds. That has changed. Large epidemiological studies now suggest that people with schizophrenia have higher rates of certain autoimmune conditions, and people with autoimmune diseases have somewhat higher rates of schizophrenia. More dramatically, a small but important fraction of cases that look exactly like first-episode psychosis turn out to be anti-NMDA receptor encephalitis — a treatable autoimmune brain disease that needs immunotherapy, not just antipsychotics. Knowing where the overlap is helps everyone get the right diagnosis sooner.

In one sentence

Schizophrenia clusters with several autoimmune diseases, and a small but important subset of psychotic presentations are caused by autoimmune encephalitis — which is treatable when identified early.

The epidemiology: schizophrenia and autoimmune disease

Large Danish national register studies have repeatedly shown elevated rates of autoimmune disease in people with schizophrenia. A 2014 study in American Journal of Psychiatry (Benros et al.) found that prior autoimmune disease was associated with a roughly 30% increase in the risk of later schizophrenia, and prior severe infection (which often involves immune activation) added further risk. The relationship is bidirectional: people with schizophrenia also have higher rates of subsequent autoimmune diagnoses.

Conditions with consistent associations include:

The mechanisms are not fully understood. Theories include shared genetic vulnerability (HLA region overlap), inflammation as a contributor to brain development, and gut-immune interactions. See neuroinflammation in schizophrenia.

Anti-NMDA receptor encephalitis

This is the autoimmune disease most relevant to psychiatry. First clearly described by Dalmau and colleagues in 2007, anti-NMDA receptor encephalitis is caused by antibodies against the NMDA receptor — the same receptor implicated in many models of schizophrenia.

Who gets it?

It can affect anyone but is most commonly diagnosed in younger women (often in their 20s and 30s). In a subset, it is associated with an underlying tumour, most commonly an ovarian teratoma; in many cases, no tumour is found.

What does it look like?

The presentation often unfolds in stages over weeks:

  1. Prodromal — flu-like symptoms, headache, low-grade fever for 1–2 weeks
  2. Psychiatric — anxiety, agitation, paranoia, hallucinations, disorganised speech, often misdiagnosed as first-episode psychosis
  3. Neurological — abnormal movements (orofacial dyskinesias, choreoathetoid movements), seizures, autonomic instability (fluctuating blood pressure and heart rate), reduced consciousness
  4. Hypoventilation — sometimes requiring intensive care

The combination of new-onset psychiatric symptoms with seizures, abnormal movements, autonomic instability, or rapid cognitive decline should raise the question. The criteria proposed by Graus et al. in Lancet Neurology (2016) are widely used.

How is it diagnosed?

Treatment

Treatment is immunotherapy: high-dose corticosteroids, IV immunoglobulin (IVIG), plasma exchange, and second-line agents like rituximab or cyclophosphamide. If a tumour is found, removing it is part of treatment. Most patients improve substantially with treatment, though recovery can take months and may be incomplete.

Seek urgent neurological evaluation if

New-onset psychotic symptoms are accompanied by seizures, abnormal involuntary movements, severe autonomic instability (fluctuating blood pressure, heart rate, temperature), rapidly progressive cognitive decline, or reduced level of consciousness. Autoimmune encephalitis is treatable when caught early.

Other autoimmune psychoses

Several other antibody-mediated encephalitides can cause psychiatric symptoms (anti-LGI1, anti-CASPR2, anti-AMPAR). The general principle is the same: psychiatric symptoms with neurological features (seizures, movement disorder, rapid cognitive change, autonomic dysfunction) deserve neurological evaluation alongside psychiatric care.

Who should be tested?

Routine antibody testing is not recommended for typical first-episode psychosis. Most cases of psychosis are not autoimmune. However, the following features should lower the threshold for testing:

Coeliac disease and gluten sensitivity

Some studies have suggested elevated rates of anti-gliadin antibodies and coeliac disease in schizophrenia. The story is not fully settled, but in patients with significant GI symptoms, weight loss, anaemia, or family history of coeliac, screening with tissue transglutaminase antibodies is reasonable. Whether gluten-free diet helps psychiatric symptoms in antibody-positive patients without coeliac is an active research question and not yet established practice.

Implications for general care

For the vast majority of people with schizophrenia, autoimmune diagnoses are the same as for anyone else: get evaluated when symptoms suggest, follow up with appropriate specialists, and recognise that some autoimmune conditions (thyroid disease, type 1 diabetes) are simple to monitor with routine labs. Annual TSH is part of standard psychiatric monitoring for many patients on lithium and is worth doing in any case.

Inflammation in "ordinary" schizophrenia

Even outside the rare autoimmune encephalitides, low-grade inflammation appears to play a role in many cases of schizophrenia. Some trials have explored anti-inflammatory adjuncts (aspirin, NAC, minocycline, omega-3) with modest and mixed results. None of these are yet standard treatments, but the line of research is active. See NAC and omega-3.

The bottom line

The boundary between psychiatry and immunology is more porous than it once seemed. For most people, this matters as background context — useful to know but not requiring action. For a small group, it matters urgently: a treatable autoimmune brain disease can hide behind what looks like a first psychotic episode, and the earlier it is recognised, the better people do. If a presentation does not fit a standard psychiatric picture, it is reasonable to ask whether neurological and immunological evaluation is warranted.


This article is for educational purposes only and is not medical advice, diagnosis, or treatment. Always consult a qualified mental health professional. If you or someone you know is in crisis, call or text 988 in the US, or your local emergency number.

Frequently asked questions

Should everyone with first-episode psychosis be tested for NMDA antibodies?
No — routine antibody testing is not currently recommended for typical first-episode psychosis. The yield is low. Testing is recommended when neurological features, autonomic instability, catatonia, or other atypical features are present.
If I have autoimmune disease, am I at higher risk of schizophrenia?
On average, the population-level risk is modestly elevated, but the absolute risk for any individual remains low. Family history, genetic vulnerability, and other factors weigh much more heavily for any one person.
Will a gluten-free diet help my schizophrenia?
If you have coeliac disease, yes — for the coeliac. Whether it improves psychiatric symptoms in non-coeliac patients is unproven. If you are considering a major dietary change, talk to your prescriber and a dietitian.
Why does NMDA receptor encephalitis cause psychosis?
The NMDA receptor is central to glutamate signalling in the brain. Antibodies against it disrupt normal synaptic function, which can produce a clinical picture that closely resembles schizophrenia — supporting the broader 'glutamate hypothesis' of schizophrenia. See our glutamate hypothesis article for more.

Try Frida — your calm companion

Frida helps people living with schizophrenia track moods, manage medication, and build stability. 7-day free trial.

Get the app →