AI Insight
Researchers analyzed 5,043 patients with schizophrenia, schizoaffective disorder, or bipolar disorder type 1 from UK clinical cohorts and identified three distinct classes based on premorbid functioning, illness course, and outcomes rather than traditional diagnostic categories. These classes showed different patterns of genetic liability for psychiatric disorders and behavioral traits, with associations only partially explained by diagnostic labels. The findings suggest that classification systems based on functioning and outcomes may better capture the biological underpinnings of psychotic disorders than current diagnostic categories alone.
Why it matters
This research could inform more personalized treatment approaches in psychiatry by identifying patient subgroups that cut across traditional diagnoses and align more closely with genetic risk factors. The alternative classification system may improve prediction of illness outcomes and treatment response, potentially advancing precision psychiatry beyond conventional diagnostic boundaries.
Understand the Science
⚠️ Preprint – Noch nicht peer-reviewed
Dieser Artikel wurde noch nicht von unabhängigen Experten begutachtet. Die Ergebnisse sind vorläufig und sollten mit Vorsicht interpretiert werden.
Introduction Limitations of current classifications of schizophrenia, schizoaffective disorder, and bipolar disorder are evident from their overlapping symptoms, aetiologies, treatments, and outcomes, and present a barrier to novel treatment discovery. Alternative conceptualisations are needed to address nosological validity, align diagnosis to aetiology, and improve prognostication and treatment choice. We aimed to identify latent classes across the psychosis spectrum based on premorbid functioning and outcomes, and assess these in relation to genetic liability and symptom dimensions. Method Participants with a diagnosis of schizophrenia, schizoaffective disorder, or bipolar disorder type 1, were ascertained from four UK clinical cohorts (total n=5,043). Latent class analysis was conducted using phenotypes not included within the diagnostic criteria, including premorbid functioning, age at illness onset, and measures of severity and course. Polygenic scores (PGS) for psychiatric disorders and behavioural traits were tested for associations with latent classes. We tested if diagnosis explained associations between PGS and classes. Results A three-class model provided the best fit. Class one had poorer premorbid functioning, lower rates of recovery, and higher PGS for schizophrenia and ADHD. Class three had the highest functioning, higher rates of psychosocial stressors before onset, higher intelligence PGS and lower PGS for psychiatric disorders. Class two was intermediate between classes one and three on measures of functioning, but was characterised by high levels of involuntary hospital admissions and high bipolar disorder PGS. Diagnosis only partially explained associations between PGS and class membership. Conclusions We identified classes across the psychosis spectrum characterised by different premorbid functioning and outcomes, that cut across diagnostic categories and captured genetic liability not explained by diagnosis. Our findings suggest alternative conceptualisations of psychotic disorders may complement diagnoses in mapping to the aetiology of these conditions, and could be useful to advance precision psychiatry.
Source: Polygenic associations with phenotypic classes across the psychosis-affective spectrum