# 20 — Autoimmune Encephalitis: When Psychosis Is a Neurological Emergency

Series: Neuroimmune Psychiatry
Evidence class: Mechanistic review and hypothesis development; see cited evidence

Editorial status: Third substantive revision, 17 October 2026. Not independently peer reviewed.

## Abstract
Autoimmune encephalitis demonstrates that a clinical presentation resembling primary psychiatric illness can arise from an identifiable immune attack on neural function. Its existence is a decisive objection to treating symptom categories as complete causal explanations. The diagnostic challenge is to recognize neurological danger early without turning every unusual psychiatric presentation into presumed encephalitis.

## A symptom can be psychiatric in appearance and neurological in cause
Rapid deterioration accompanied by seizures, altered consciousness, abnormal movements, autonomic instability or marked cognitive change changes the diagnostic problem. In that setting the time course and associated findings are more informative than whether the patient initially reported hallucinations or agitation. Antibody tests, MRI, EEG and cerebrospinal fluid studies have different sensitivities and limitations; a negative or incidental result cannot be interpreted in isolation. A multidisciplinary clinical assessment is necessary because both missed encephalitis and unnecessary immunotherapy can cause serious harm.

The lesson for oxalate research is methodological rather than a direct analogy between mechanisms. Autoimmune encephalitis became clinically actionable through convergent syndrome patterns, laboratory and physiological evidence, and treatment response—not through the assertion that all psychosis is immune-mediated. An oxalate-associated syndrome would need an equivalent chain of distinctive evidence before it should alter urgent diagnostic priorities.

## From antibodies to circuitry
In anti-NMDA receptor encephalitis, pathogenic immunoglobulins interact with neuronal NMDA receptors and disrupt receptor-mediated synaptic function. Early features may include behavioral disorganization, hallucinations, insomnia or agitation; evolving neurological manifestations may include seizures, dyskinesias, catatonia, autonomic disturbance and decreased consciousness. Other antibody-mediated encephalitides have different clinical signatures, demographics and associated conditions. Autoimmune encephalitis is therefore not one disease with one blood test.

## Why routine psychiatric categories can mislead
Diagnostic classification helps describe symptoms, but a first episode unfolding over days with profound sleep disruption and rapid neurological change deserves a broader approach. EEG, CSF studies, MRI, antibody testing and assessment for alternative causes may be appropriate according to the clinical picture. MRI can be unrevealing in some encephalitis cases, and serum antibody results alone may be misleading; interpretation belongs in a neurological and clinical context. The goal is neither indiscriminate panels nor reflex attribution to psychiatric illness.

## The postpartum example
In a study involving 96 women with postpartum psychosis, Bergink and colleagues found two patients with anti-NMDA receptor antibodies, alongside other neuronal antibody findings. This identifies a rare but important differential; it does not establish that all postpartum psychosis is autoimmune. The ordinary clinical emergency of postpartum psychosis remains urgent regardless of antibody status. Likewise, patients with primary bipolar disorder may become very unwell without having encephalitis.

## Treatment and safety
Immune-directed treatment for confirmed or strongly suspected encephalitis is a specialty decision, often involving intensive neurological assessment. Psychiatric symptom management, seizure control and supportive care may also be necessary. High-potency dopamine blockade may create particular challenges in some encephalitic syndromes, but drug selection must be individualized. The appropriate lesson is coordinated care, not substitution of a single unproven immune intervention for established emergency treatment.

## Diagnostic error in both directions
The spectacular clinical stories of autoimmune encephalitis can distort base-rate reasoning. Because a small fraction of psychosis may reflect immune disease, it does not follow that every person with first-episode psychosis needs a broad commercial antibody panel. Conversely, because most psychiatric episodes are not encephalitis, rapid neurological deterioration must not be dismissed as ordinary psychotic behavior. The correct threshold for testing depends on red flags, examination, timing and specialized interpretation.

This problem is particularly important with serum-only testing. Some antibody results may be false positive or clinically nonspecific; some conditions require cerebrospinal fluid or other contextual evidence. A negative imaging study cannot reliably exclude all forms of encephalitis. The best care occurs when psychiatry and neurology treat the patient as one person, share the evolving timeline and revise the differential as new findings appear.

## Why the time course matters
Autoimmune encephalitis can begin with changes in behavior, memory, sleep or emotional regulation that appear psychiatric, then evolve to seizures, movement abnormalities, autonomic instability or fluctuating consciousness. The sequence varies, and not all patients display every feature. Anti-NMDA receptor encephalitis is a familiar illustration, particularly because psychiatric symptoms can appear early. A rapidly progressive presentation should therefore prompt attention to associated neurological findings and medical context rather than immediate certainty that the illness is a primary psychotic disorder. Conversely, an isolated chronic psychiatric syndrome without supportive findings has a different prior probability and should not be labeled autoimmune merely because the symptoms are dramatic.

The diagnostic workup cannot be reduced to one antibody result. Clinicians consider neurological examination, EEG, neuroimaging, cerebrospinal fluid findings and appropriate antibody testing in the context of the presentation. Some antibody assays can yield findings that are difficult to interpret, especially if testing is indiscriminate in low-probability populations. A negative MRI does not rule out every encephalitic process, while an unconvincing positive antibody finding does not prove an immune cause. Diagnostic reasoning is necessarily cumulative and evidence-weighted.

## What is actually happening biologically
Different autoimmune encephalitis syndromes have different mechanisms. Some antibodies target extracellular neuronal proteins or receptors and may alter receptor function or internalization. Other immune processes involve different targets and cellular responses. A single phrase such as “brain inflammation” masks this heterogeneity. The phenotype can reflect changes in synaptic signaling, neuronal network stability, seizure propensity and broader immune injury. These processes may manifest as hallucinations, agitation, memory impairment or catatonic features, but symptom overlap with primary psychiatric disorders is not proof that the underlying pathologies are identical.

This distinction has therapeutic consequences. In confirmed autoimmune encephalitis, treatments addressing the immune mechanism and underlying triggers may be essential alongside supportive psychiatric management. Delaying recognition can have substantial consequences. At the same time, unproven immune-directed treatment in patients without compelling evidence may produce serious adverse effects. The objective is neither to overmedicalize every psychiatric symptom nor to ignore a rare but treatable neurological disease.

## A research lesson for the entire volume
Autoimmune encephalitis supplies a methodological template for investigating other biological psychiatric hypotheses, including the proposed oxalate pathways. A credible candidate cause begins with objective biological abnormalities, a plausible temporal sequence and a neurological phenotype that can be assessed independently. Stronger support comes from mechanistic specificity and reproducible effects of appropriate treatment. Weak support consists of a broad symptom list attached to an unvalidated test. The better-defined autoimmune disorders show what successful biological causal inference can look like; they cannot automatically validate a separate, less-tested metabolic theory.

The clinical safety boundary is particularly important. New-onset psychosis with seizures, altered consciousness, fever, severe autonomic symptoms or rapidly progressive cognitive deterioration requires prompt medical assessment. In those circumstances, speculation about diet or a long-term terrain protocol should not delay evaluation. The research value of the paper is its insistence that mind and nervous system belong to the same organism, but that diagnostic claims remain subject to evidence.

## A diagnostic dilemma that illustrates why context outranks isolated tests
Imagine an adult with sudden agitation, insomnia and unusual speech. A psychiatric disorder may be plausible, but a new seizure, fever or rapid memory decline would shift the evaluation toward neurological and medical causes. No single symptom makes autoimmune encephalitis certain. Likewise, finding an antibody in serum without a fitting clinical syndrome can create diagnostic confusion. Depending on the presentation, examination, EEG, imaging and cerebrospinal fluid testing may provide independent clues. The correct combination is determined by clinical judgment and established diagnostic approaches rather than broad screening for every possible immune antibody.

The therapeutic stakes are substantial: early recognition of an immune disorder can change outcomes, while immunosuppression without a compelling diagnosis can cause serious harm. Psychiatric medications may be needed for symptom control, yet they do not substitute for treating a confirmed immune mechanism. This distinction between symptom management and causal treatment is a theme throughout the Institute's work. It also helps evaluate proposed oxalate-associated illness. A metabolite can be biologically consequential without being responsible for every symptom in a complex presentation; a cause must be identified with evidence strong enough to justify the intervention directed at it.

The most valuable lesson for speculative metabolic psychiatry is that a biological cause can be real even when its first symptoms resemble ordinary psychiatric disease. That observation argues against diagnostic complacency. It does not mean that testing for every exotic mechanism is proportionate in every patient. Good medicine begins with syndrome, time course and risk, then uses targeted investigation to identify the biologically plausible causes.

## Conclusion
Autoimmune encephalitis proves the general principle that psychiatric behavior can be produced by a specific biological lesion or process. It also shows the evidentiary and clinical discipline required before announcing a newly discovered biological psychiatric subtype.

## Selected readings
- Dalmau et al., Anti-NMDA receptor encephalitis case series: https://pubmed.ncbi.nlm.nih.gov/18851928/

- Pollak et al., Autoimmune psychosis consensus: https://pubmed.ncbi.nlm.nih.gov/31669058/

- Bergink et al., Postpartum psychosis autoantibodies: https://pubmed.ncbi.nlm.nih.gov/26183699/