Understanding why some people are more vulnerable to psychiatric illness while others excel cognitively has long been one of medicine's hardest problems. New findings from one of science's most prestigious journals suggest the answer is partly written into the evolutionary history of our genomes — and that brain-related genes operate under fundamentally different rules than genes governing physical traits.
Analyzing genome-wide association study data across dozens of complex traits, researchers found that genetic variants influencing brain-related characteristics — including cognitive function, psychiatric conditions, and neurological susceptibility — are subject to substantially stronger purifying selection than variants shaping non-brain traits such as height or body composition. This means that harmful mutations in brain-relevant genes are more aggressively weeded out across generations, producing a distinctive genetic architecture: fewer common variants of large effect, more reliance on rare variants, and compressed heritability signals that are harder for GWAS to detect. The implication is structural — current GWAS methodology, optimized around common-variant discovery, may be systematically underpowered for psychiatric and cognitive phenotypes precisely because evolution has constrained the variant landscape.
This finding reframes a long-standing frustration in psychiatric genetics: why schizophrenia, bipolar disorder, and major depression show relatively modest GWAS hits despite high heritability estimates. The evolutionary constraint hypothesis has been discussed theoretically for years, but rigorous empirical quantification across many traits simultaneously strengthens the case considerably. For researchers, this argues for prioritizing whole-genome or whole-exome sequencing approaches that capture rare variants. For clinicians and the public, it sets realistic expectations — polygenic scores for brain traits may inherently plateau below those achievable for metabolic or anthropometric traits. The study is observational and cross-sectional in design, limiting causal inference, and replication in non-European ancestries remains essential.