If common variants are the small streams that feed into schizophrenia risk, copy number variants are the rarer, deeper rivers. They are large structural changes in DNA — entire chunks of chromosome that are duplicated or deleted — and a small but important set of them have been shown to substantially increase the risk of schizophrenia.
Copy number variants (CNVs) are large duplications or deletions of DNA, and at least eight specific CNVs have been firmly linked to schizophrenia, each rare in the population but each carrying a substantially increased risk in carriers.
What CNVs are
The genome is organised into 23 pairs of chromosomes, with two copies of nearly every gene. A copy number variant is a region where the standard two-copy state is changed — typically by losing a stretch of DNA (a deletion) or gaining one (a duplication). CNVs can be inherited from a parent or arise de novo at conception. Some are common and harmless. Others are rare, span hundreds of thousands or millions of base pairs, and disrupt many genes at once.
CNVs are detected by chromosomal microarray analysis (CMA), MLPA, or sequencing-based methods. The widespread use of CMA in clinical genetics has been one of the main reasons CNVs have become tractable to study.
The Psychiatric Genomics Consortium CNV work
Just as the PGC has driven common-variant GWAS, the PGC's CNV Workgroup has driven CNV studies in schizophrenia. The 2017 paper from Marshall and colleagues in Nature Genetics, using more than 21,000 schizophrenia cases and 20,000 controls, confirmed several specific CNVs as significantly enriched in schizophrenia. Subsequent work has refined and extended that list.
The established schizophrenia CNVs
The CNVs with the most robust evidence for schizophrenia association include:
- 22q11.2 deletion — the strongest single CNV, associated with roughly a 25-fold increase in schizophrenia risk. See our 22q11.2 article.
- 16p11.2 duplication — associated with schizophrenia and also linked to intellectual disability and autism.
- 15q11.2 deletion (BP1–BP2) — modestly increased risk.
- 15q13.3 deletion — increased risk for schizophrenia, epilepsy, and intellectual disability.
- 1q21.1 deletion and duplication — both ends of the dosage spectrum implicated.
- 3q29 deletion — one of the strongest single risk CNVs for schizophrenia after 22q11.
- NRXN1 deletion — disrupts a synaptic gene critical for neuronal connections.
- 17q12 deletion — associated with schizophrenia and renal cysts.
All of these CNVs are rare in the general population — typically affecting fewer than 1 in 1,000 people — but each is significantly enriched in schizophrenia samples.
Pleiotropy: CNVs are rarely specific
One striking feature of schizophrenia-associated CNVs is that none of them are specific to schizophrenia. Almost all of the established CNVs increase the risk of multiple neurodevelopmental conditions — autism, intellectual disability, ADHD, epilepsy — as well as physical features like congenital heart disease or distinctive facial appearance. This pleiotropy supports a model in which CNVs broadly disrupt brain development, with the specific clinical outcome influenced by other factors including the rest of the genome and environment.
Penetrance
Penetrance — the probability that a CNV carrier develops a particular outcome — varies. For 22q11.2 deletion, lifetime schizophrenia risk is roughly 25%. For 3q29 deletion, schizophrenia risk is significantly elevated, though precise estimates depend on the cohort. For most other established CNVs, penetrance for schizophrenia specifically sits in the single digits to low double digits — meaningful, but well short of certainty. The Decipher database (deciphergenomics.org) collects clinical findings on rare CNVs to refine these estimates.
De novo CNVs
De novo CNVs — new mutations not inherited from either parent — appear to be enriched in people with schizophrenia compared with healthy siblings. This suggests that some schizophrenia cases arise from new structural events at conception. The Brain Somatic Mosaicism Network and other groups are now investigating whether even more localised changes within neurons themselves (somatic mutations) contribute.
Clinical implications
For most adults with schizophrenia, CNV testing is not routine. Several scenarios in which it may be considered:
- Childhood-onset or very early-onset schizophrenia
- Schizophrenia with intellectual disability or developmental delay
- Schizophrenia with congenital anomalies (heart, palate, kidney, characteristic facial features)
- Strong family history with multiple affected relatives
- Cases where reproductive counselling would benefit from a clearer recurrence picture
Clinical microarray testing should be ordered through, or in consultation with, a board-certified medical geneticist or genetic counsellor. The yield in unselected adult schizophrenia samples is modest; the yield in selected populations is meaningfully higher.
Discuss potential implications with a genetic counsellor — including for siblings, future children, insurance, and emotional preparation. Some findings will not be specific or actionable.
What CNV findings change in care
Knowing that a person with schizophrenia carries one of these CNVs may change care in several ways:
- Targeted somatic surveillance — for example, cardiac screening in 22q11.2 carriers.
- Family genetic testing and counselling.
- Earlier psychiatric monitoring of children and adolescent siblings who may also carry the CNV.
- Connection to syndrome-specific patient communities and specialised clinics.
A CNV result rarely changes the choice of antipsychotic itself, but it can substantially change the surrounding care plan.
Where the field is going
Whole-genome sequencing, increasingly affordable and increasingly used in research cohorts, is uncovering smaller and rarer structural variants and short rare protein-coding mutations that may further explain schizophrenia genetics. Programs like the SCHEMA consortium have already shown that ultra-rare loss-of-function variants in a small set of genes — including SETD1A and GRIN2A — substantially raise schizophrenia risk in carriers. The line between CNV studies and rare-variant sequencing studies is blurring as technologies converge.
For more, see our pieces on 22q11.2 deletion syndrome, GWAS in schizophrenia, and genetic counselling for schizophrenia.
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.