Highlights
- Recurrent copy number variants (rCNVs) and polygenic scores (PGSs) provide complementary rather than redundant risk information for ADHD, ASD, and schizophrenia.
- PGSs can effectively stratify absolute risk among carriers of medium-to-high impact rCNVs, potentially identifying individuals where common genetic liability attenuates or amplifies rare variant effects.
- Genetic heterogeneity in disorders like schizophrenia and depression is increasingly being mapped to specific neurocognitive pathways and brain morphological changes during adolescence.
- The integration of lifecourse perspectives and multi-omic data (single-cell/spatial) is essential to transition from genotype association to mechanistic understanding.
Background
The genetic architecture of psychiatric disorders is characterized by a spectrum of variation, ranging from rare, highly penetrant mutations like recurrent copy number variants (rCNVs) to thousands of common, small-effect single-nucleotide polymorphisms (SNPs) summarized in polygenic scores (PGSs). While both contribute significantly to the heritability of neurodevelopmental and psychotic disorders, clinical application has been hindered by a lack of clarity regarding their joint contribution. For clinicians, the challenge lies in understanding whether a specific rCNV carrier’s risk is modified by their background polygenic liability—a concept central to the “two-hit” or multi-hit models of psychiatric pathogenesis. As the field moves toward precision psychiatry, synthesizing how these genetic markers interact is critical for early intervention and personalized risk management.
Key Content
The Joint Influence of rCNVs and PGSs on Absolute Risk
Recent large-scale genetic association studies, notably the 2026 iPSYCH cohort analysis (Vaez et al., JAMA Psychiatry), have provided a survival analysis framework to estimate the absolute risk associated with these variants. In a sample of 94,276 European-ancestry individuals, rCNV carriage was robustly associated with autism spectrum disorder (ASD), attention-deficit/hyperactivity disorder (ADHD), and schizophrenia spectrum disorder (SSD), but notably not with major depressive disorder (MDD). Conversely, PGSs were positively associated with risk across all corresponding disorders.
Critically, the study found that PGSs could stratify risk among rCNV carriers. For example, a negative interaction was identified between the 16p13.11 duplication and ADHD-PGS on ADHD risk, suggesting a “saturation effect” where the impact of the rCNV might be less pronounced in individuals already at high polygenic risk, or conversely, that rCNV-associated risk is attenuated in those with very low polygenic liability. This suggests that the clinical penetrance of a rare variant is partially dependent on the common genetic background.
Dissecting Heterogeneity via Genetic Subtyping
Evidence is mounting that psychiatric diagnoses often mask distinct biological pathways. In schizophrenia, genomic structural equation modeling has “split” the genetic risk into two divergent pathways: one specific to schizophrenia (SZ-specific) associated with lower IQ and educational attainment, and another shared with bipolar disorder (PSY-shared) associated with higher educational attainment (Watson et al., Mol Psychiatry, 2026).
Similarly, depression research has identified atypical depression (characterized by weight gain and hypersomnia) as a distinct subtype with a unique genetic profile. Individuals with atypical depression exhibit higher PGSs for metabolic, inflammatory, and circadian traits compared to those with non-atypical presentations. This subtyping has direct clinical implications, as those with atypical genetic/phenotypic profiles show poorer responses to standard SSRIs and SNRIs but higher side-effect burdens.
Developmental Trajectories and Brain Morphology
Genetic risks for psychiatric disorders do not remain static but manifest dynamically across the lifecourse. Longitudinal neuroimaging studies, such as those leveraging the ABCD and Generation R cohorts, show that high genetic susceptibility to schizophrenia (SCZ-PGS) is associated with divergent cortical surface area trajectories in the frontal regions beginning in early adolescence. While children with low genetic risk show typical cortical expansion, those with high risk exhibit early decreases, serving as a potential biomarker for disease onset.
In ADHD, polygenic risk has been linked to neural signatures of cognitive control, specifically midfrontal theta dynamics. Higher ADHD-PGS predicts increased variability in theta phases, which correlates with response time variability in clinical tasks. This provides a measurable bridge between molecular risk and neurophysiological dysfunction.
Environmental Interplay and Genetic Confounding
The interaction between genetic liability and environmental exposures (GxE) remains complex. While a systematic review of 56 studies confirmed main effects for both PRSD (polygenic risk score for depression) and stress, evidence for GxE interactions often requires massive samples (>40,000) to reach statistical significance. Furthermore, recent studies have cautioned against over-interpreting observational associations, such as maternal diet and offspring ADHD, finding that these relationships may be driven largely by genetic confounding rather than direct causal pathways.
Expert Commentary
The shift from identifying genetic risk to applying it in clinical settings requires a nuanced understanding of effect sizes and absolute versus relative risk. The finding that PGSs can identify more individuals at a comparable level of absolute risk than rare rCNVs (except in ASD) underscores the potential of PGSs as a population-level screening tool. However, the expert consensus emphasizes that genetic scores should not yet be used as standalone diagnostic tests. Instead, they provide a “probabilistic roadmap.”
One significant controversy involves the “saturation” or negative interaction observed between rare and common variants. Biologically, this may suggest that there is a threshold for neurodevelopmental disruption; once a certain degree of biological insult is reached (either via a large CNV or high polygenic burden), additional genetic risk contributes marginally less to the diagnostic outcome. This has profound implications for genetic counseling, as the clinical prognosis for an rCNV carrier may be significantly more optimistic if they reside in the lowest decile of the corresponding PGS.
Conclusion
Psychiatric genetics has moved beyond the discovery phase into a period of integration. The evidence suggests that rCNVs and PGSs provide a multi-layered view of an individual’s susceptibility. Future research must prioritize diverse ancestral populations to ensure the portability of these risk models and integrate single-cell and spatial omics to map exactly how these structural variants disrupt specific neuronal circuits. For the clinician, these findings herald a future where genetic profiling informs not just the risk of diagnosis, but the likely developmental trajectory and treatment response of the patient.
References
- Vaez M, et al. Recurrent Copy Number Variants and Psychiatric Outcomes in the Context of Polygenic Scores. JAMA Psychiatry. 2026. PMID: 42201728.
- Watson HJ, et al. Splitting schizophrenia: divergent cognitive and educational outcomes revealed by genomic structural equation modelling. Mol Psychiatry. 2026. PMID: 41620580.
- Milaneschi Y, et al. Atypical Depression Is Associated With a Distinct Clinical, Neurobiological, Treatment Response, and Polygenic Risk Profile. Biol Psychiatry. 2026. PMID: 41534638.
- Vidal-Pineiro D, et al. Genetic Susceptibility to Schizophrenia and the Onset of Brain Developmental Change During Adolescence. Biol Psychiatry. 2026. PMID: 41833745.
- Ghadirivasfi M, et al. ADHD polygenic risk predicts neural signatures of cognitive control: Evidence from midfrontal theta dynamics. Transl Psychiatry. 2026. PMID: 41916949.
