Basic Neurosciences · Paper A
Schizophrenia endophenotypes
This concept matters because it lets researchers detect genetic risk for schizophreniaA long-term mental illness involving episodes of psychosis such as hallucinations and delusions.Read more → (a psychotic illness with delusionsA fixed false belief, held with total conviction, that is not shared by the person's culture and is not changed by evidence against it.Read more →, false perceptions and disorganised thinking) in people who show no psychotic symptoms at all. Endophenotypes are heritable, measurable traits that sit between genes and the full clinical disorder, found in relatives even when those relatives never develop the disorder themselves. Testing the unaffected sibling of a patient with schizophrenia, e.g. one who shows abnormal smooth pursuit eye movements despite no psychotic symptoms, works because the endophenotype tracks genetic risk rather than the clinical illness itself. In practice, testing unaffected relatives for these traits can reveal genetic risk that would otherwise stay completely hidden.
Smooth pursuit eye movement abnormalities are a candidate endophenotype, found more often in relatives of patients than in the general population. Prepulse inhibition (the normal dampening of a startle response by a preceding weak stimulus) and P50 sensory gating deficits reflect impaired sensory filtering, another candidate endophenotype distinct from the clinical symptom of delusionsA fixed false belief, held with total conviction, that is not shared by the person's culture and is not changed by evidence against it.Read more →. Working memory deficits represent a cognitive endophenotype, again distinct from overt psychosis. An unaffected sibling showing reduced P50 suppression despite never having psychosis, e.g. one recruited purely for a genetic-risk study, demonstrates that the measurable trait can still track inherited vulnerability. Endophenotypes are studied specifically because they sit closer to the underlying genetics than the full clinical syndrome, which is what makes them useful for genetic research even in unaffected family members. The endophenotype concept was introduced to psychiatry by Gottesman and Shields. Their formal criteria for a valid endophenotype are association with the illness in the population, heritability, and state-independence, meaning it is present whether the illness is currently active or in remission. The remaining criteria are co-segregation with the illness within families and, where possible, association with a candidate gene together with biological plausibility. Stated together as the four core formal criteria: association with a candidate gene, biological plausibility, state-independence, and co-segregation with the disorder within families. Neurological soft signs (subtle motor and sensory-integration abnormalities) are common in relatives of people with schizophreniaA long-term mental illness involving episodes of psychosis such as hallucinations and delusions.Read more → but are NOT accepted as a validated endophenotype, because they lack the specificityThe proportion of people who truly do not have a condition that a test correctly identifies as negative.Read more → and state-independence the formal criteria require.
(a psychotic illness with , meaning fixed false beliefs, false perceptions such as , and disorganised thinking) is strongly heritable, but its genetics have proved difficult to pin down because the diagnosis itself is a broad clinical category built from symptoms, not a single measurable biological trait. Endophenotypes were developed as a way around this problem, and this matters because it means genetic risk can be studied in people who are completely well, expanding the pool of research subjects far beyond just diagnosed patients and getting closer to the underlying biology than symptom checklists ever could.
An endophenotype is a heritable, measurable biological or cognitive trait that sits somewhere between the underlying genetic risk and the full clinical disorder, and, crucially, it can be detected in relatives who never develop themselves.
It is important to be precise about what an abnormal endophenotype result means for an individual: it is a research tool for understanding genetic architecture, not a diagnostic test. An unaffected relative with an abnormal endophenotype finding has not been diagnosed with anything and will not necessarily develop . There is currently no blood test, scan, or single measurable trait used in routine NHS practice to diagnose schizophrenia, which still relies on structured clinical assessment against operationalised criteria such as those in ICD-11 or DSM-5.
The basic idea is that endophenotypes sit on the causal pathway between genes and the clinical phenotype (the observable, diagnosed illness), so they should be simpler, more directly biological, and more common in unaffected relatives than the full disorder itself. A trait present in a small minority of the general population but in a much larger share of unaffected first-degree relatives of patients with is behaving exactly as you would expect if it tracks inherited genetic risk, even without any psychotic symptoms present at all. Four candidate endophenotypes are consistently discussed for schizophrenia: smooth pursuit eye movement abnormalities, prepulse inhibition deficits, P50 sensory gating deficits, and working memory impairment.
Smooth pursuit eye movement abnormalities describe a difficulty smoothly tracking a slowly moving target with the eyes, instead producing jerky, saccadic (rapid, stepwise) movements. Prepulse inhibition (PPI, the normal dampening of a startle response when a weak warning stimulus, the prepulse, precedes a louder startling stimulus) is typically reduced in and in unaffected relatives, reflecting impaired sensory filtering rather than a symptom of psychosis itself. P50 sensory gating deficits are a related electrophysiological finding: normally, a repeated identical sound produces a smaller brain response the second time it is heard, because the brain filters out predictable, repeated information; this suppression is blunted in schizophrenia and its relatives. Working memory deficits, difficulty holding and manipulating information over short periods, follow the same pattern in milder form. Suppose a research team wants to identify genes contributing to schizophrenia risk. Rather than only recruiting diagnosed patients, they recruit unaffected siblings of patients and measure their P50 suppression in the laboratory; one sibling shows clearly reduced suppression despite never having experienced any psychotic symptoms. Because the endophenotype tracks genetic risk rather than the clinical illness itself, this sibling's result still carries useful genetic information, placing them closer to their affected relative's underlying genetic risk than an unrelated healthy control would be.
An endophenotype is easily confused with a clinical biomarker (a measurable indicator used at the bedside to help diagnose or monitor disease), but the two occupy different worlds: a biomarker is validated for individual clinical decision-making, while an endophenotype is a group-level research construct with no individual diagnostic meaning. It should also be distinguished from a symptom (an early warning sign that someone is about to develop the full illness), since an unaffected relative with an abnormal endophenotype is not necessarily prodromal and may never convert to the disorder at all. Finally, it is distinct from a simple risk factor like family history alone, because an endophenotype is a specific, laboratory-measurable trait rather than a general statistical association.
Endophenotypes are identified and measured with specific laboratory and psychophysiological paradigms, not clinical interview. Smooth pursuit is assessed with eye-tracking equipment during a visual tracking task; prepulse inhibition is measured via the startle reflex, typically an eyeblink response recorded electromyographically; P50 suppression is measured with EEG (electroencephalography, a recording of the brain's electrical activity) during a paired-click paradigm; and working memory is assessed with standardised neurocognitive tasks. None of these produces a diagnosis; each produces a quantitative trait score that can be compared statistically between patients, unaffected relatives, and controls.
Endophenotype research sits firmly in the domain of basic and translational neuroscience rather than everyday clinical management, and "management" here means how the finding is used in research design and genetic counselling contexts rather than at the bedside. Endophenotype studies expand in genetic research by capturing risk information from people who would be excluded entirely if a study only compared diagnosed patients against unrelated healthy controls, which is exactly why researchers deliberately recruit unaffected first-degree relatives rather than only patients. In everyday clinical practice, an endophenotype finding in a relative does not currently change that person's management, since it is not a validated predictor of future illness at the individual level; it is used to refine which genes and biological pathways researchers pursue.
If a stem describes a trait found in unaffected relatives at a rate between the general population and affected patients, and asks what kind of finding this is, the answer is an endophenotype, not a diagnostic marker. A distractor claiming this trait means the relative "will develop " or "has a subclinical form of the illness" is testing whether you remember that endophenotypes carry no individual predictive or diagnostic weight.
An endophenotype is a heritable, laboratory-measurable trait sitting between genes and the clinical disorder, present in unaffected relatives and useful for genetic research power, but it is never itself a diagnosis.