The hippocampus in schizophrenia

Of all the brain regions implicated in schizophrenia, the hippocampus has the strongest and most replicated evidence behind it. Decades of structural MRI, functional imaging, and post-mortem work converge on the same picture: the hippocampus is smaller than expected, regionally overactive, and structurally abnormal at the cellular level — and these changes are present at the time of the first psychotic episode, not only after years of illness.

What the hippocampus does

The hippocampus is a curved, seahorse-shaped structure tucked deep in the medial temporal lobe. It is essential for forming new episodic memories, for spatial navigation, and for distinguishing similar contexts from each other (a process called pattern separation). It is divided into anatomical subfields — CA1, CA2, CA3, the dentate gyrus, and the subiculum — each with distinct cell types and connections.

The hippocampus is also one of the brain's most plastic regions. It generates new neurons in adulthood (in the dentate gyrus), and its function is exquisitely sensitive to stress, sleep, and inflammation.

Volume reductions: what imaging shows

Structural MRI studies have consistently found that hippocampal volume is reduced in schizophrenia. The largest single analysis to date comes from the ENIGMA Schizophrenia Working Group, a collaboration that pooled MRI scans from thousands of patients and controls across more than a dozen sites. Their 2016 paper in Molecular Psychiatry reported significantly smaller hippocampi (and amygdalae, thalami, and accumbens) in patients compared with controls, with effect sizes that grew with illness duration but were already present at first episode.

Subfield-specific studies suggest the reductions are not uniform. The CA1 subfield, and to a lesser extent CA4 and the dentate gyrus, tend to show the largest losses. Work from the Schobel and Small labs at Columbia has been particularly influential in localising the abnormality to CA1.

Hyperactivity: a paradox at first glance

If the hippocampus is smaller, you might expect it to be underactive. The opposite is true. Resting-state and arterial spin labelling studies show elevated baseline blood flow and metabolism in the hippocampus, especially in CA1, in people with schizophrenia and in those at clinical high risk for psychosis. This pattern has been replicated in multiple cohorts and is one of the more robust functional findings in the field.

Importantly, the hyperactivity is regional, not global. CA1 looks overactive while other subfields show different patterns. The hyperactivity is thought to reflect a loss of inhibitory control, possibly from reduced parvalbumin-expressing GABA interneurons, which has been documented post-mortem.

The hippocampus–striatum link

Why does this matter? The hippocampus projects, via the subiculum and accumbens, to the ventral tegmental area, which contains the dopamine neurons that send projections back to the striatum. In animal models — most notably the methylazoxymethanol acetate (MAM) rat model developed by Anthony Grace and colleagues — hippocampal hyperactivity drives an increase in the number of spontaneously active dopamine neurons. This in turn elevates striatal dopamine release, the change most closely linked to positive psychotic symptoms.

Translated to humans, the working model is: hippocampal hyperactivity → excess dopamine neuron firing → striatal dopamine dysregulation → psychosis. Imaging studies that combine arterial spin labelling and PET have begun to show correlations consistent with this model, though direct causal proof in humans is still elusive.

Post-mortem findings

Post-mortem brain banks have allowed cellular-level study of the schizophrenia hippocampus. Findings include:

• Reduced density of parvalbumin-positive GABA interneurons, especially in CA2 and CA3/4 regions, reported by groups including Francine Benes and Stephan Heckers.
• Reduced expression of GAD67, an enzyme essential for GABA synthesis.
• Subtle reductions in pyramidal neuron size and dendritic spine density.
• No consistent neurodegeneration or gliosis on the scale seen in dementia — the changes appear developmental rather than degenerative.

Stress, cortisol, and the hippocampus

The hippocampus is dense with glucocorticoid receptors, which makes it sensitive to chronic stress. Childhood adversity — itself a documented risk factor for psychosis — is associated with smaller hippocampal volume across populations. In schizophrenia, the question of how much hippocampal change reflects developmental vulnerability versus the cumulative effects of stress, episodes, and antipsychotic exposure is still being worked out. Longitudinal first-episode studies suggest that some volume loss continues over the first few years of illness, while medication effects on hippocampal volume appear modest.

Implications for treatment

The hippocampal model has begun to inspire treatment ideas:

• Reducing hippocampal hyperactivity with agents that act on GABA or glutamate is an active research direction. A small Columbia trial of low-dose levetiracetam in early psychosis showed a reduction in hippocampal CBF and modest symptom benefit, supporting the model.
• Stress reduction interventions — sleep, exercise, mindfulness, trauma-focused therapy — may protect hippocampal function indirectly. See our pieces on exercise and mindfulness for psychosis.
• Cognitive remediation targets memory functions that depend on the hippocampus.

What the hippocampus story does not explain

Schizophrenia is not a "hippocampus disease" any more than it is a "dopamine disease." Many other regions and circuits — prefrontal cortex, thalamus, default mode network — show abnormalities. The hippocampus is one of the most consistently affected nodes, but it sits within a distributed network problem. Treatments aimed only at the hippocampus would not address negative or cognitive symptoms in any complete way.

The bottom line

The hippocampus is one of the most reliably implicated brain regions in schizophrenia. Smaller volumes (especially CA1), elevated baseline activity, and reduced GABA interneuron density are present early in the illness and may contribute to the dopamine abnormalities that drive psychosis. The hippocampal model is also one of the most actionable — it points toward specific therapeutic targets and gives a biological grounding to interventions like sleep, stress reduction, and exercise. For more, see our companion pieces on the dopamine hypothesis and overall brain volume changes.

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.