Rising ADHD and autism diagnoses have long been attributed to shifting awareness and diagnostic criteria — but a new population-scale genomic analysis suggests the story is more complex. If the composition of genetic risk among newly diagnosed individuals is changing over time, it implies that the expanding diagnostic net is capturing meaningfully different populations, with implications for how clinicians, researchers, and health systems interpret prevalence trends.
Drawing on the iPSYCH2015 cohort — a population-based case-cohort from Denmark encompassing incident ASD and ADHD diagnoses spanning 1994 to 2016 — researchers constructed polygenic scores across a broad panel of psychiatric and cognitive-behavioral traits, including depression, bipolar disorder, schizophrenia, IQ, educational attainment, neuroticism, and risk-taking propensity. Regression models assessed year-of-diagnosis as a predictor of mean polygenic score, controlling for age, sex, and ancestry. Crucially, empirical trends were benchmarked against simulated scenarios to distinguish genuine shifts in genetic risk architecture from artifacts of diagnostic expansion.
This work sits at a genuinely underexplored intersection: most prior research either tracks diagnostic trends epidemiologically or examines genetic architecture statically. The iPSYCH dataset, one of the largest national psychiatric biobanks in the world, lends unusual statistical power and population representativeness. However, key limitations apply: Denmark's healthcare system and diagnostic culture may not generalize globally, polygenic scores capture only a fraction of heritable variance, and the observational design cannot establish causal mechanisms for any observed shift. The study also cannot rule out gene-environment correlation — for instance, if broadened diagnostic criteria coincidentally captured individuals with different family genetic backgrounds who were always present in the population. Whether observed polygenic trends reflect true changes in who is developing these conditions versus who is being recognized deserves careful interpretation. Overall, this represents a methodologically sophisticated, potentially paradigm-nudging contribution to the neurodevelopmental epidemiology literature.