Are the predictions of phenotypic cognitive scores by polygenic cognitive scores driven by the cognitive complexity of the phenotypic measures? A meta‐analysis

• Published in: Alzheimer's & dementia Vol. 19; no. S15
• Main Authors: Nijenhuis, Jan te, Balaji, Kannappan, Lee, Kun Ho
• Format: Journal Article
• Language: English
• Published: 01.12.2023
• ISSN: 1552-5260; 1552-5279
• DOI: 10.1002/alz.075917

Background Intelligence is highly heritable and plays an important role in behavioral genetic research. Recent genome‐wide association studies (GWASs) have identified inherited genomic variants that account for a modest portion of the heritability of intelligence. There are many studies where cognitive polygenic scores (PGSs) are used to predict scores on intelligence tests; the resulting rs are all over the place. Various explanations are offered for this high variability in resulting rs, but none of them are convincing. There is another approach, building upon a fundamental finding from intelligence research, namely that the most important difference between intelligence tests is their difference in cognitive complexity. We hypothesize that the predictions of phenotypic cognitive scores by polygenic cognitive scores are a function of the cognitive complexity of the phenotypic cognitive scores. Method We carried out a meta‐analysis of studies reporting 1) cognitive polygenic scores (IQ, Cognitive achievement, Educational attainment) and 2) the correlation between cognitive polygenic scores and at least four cognitive test scores. We computed the cognitive complexity of the cognitive tests. We then computed the correlation between 1) the cognitive complexity of the cognitive tests and 2) the correlation between cognitive polygenic scores and at least four cognitive test scores. A strong positive correlation means that the predictions of phenotypic cognitive scores by polygenic cognitive scores are strongly driven by the cognitive complexity of the phenotypic measures. Result A substantial number of studies were found and many of them were based on a large number of subjects. The studies analyzed so far show large to very large correlations, strongly confirming the hypothesis. Conclusion Based on an analysis of a subset of the studies in the meta‐analysis, cognitive polygenic scores predict most strongly for difficult IQ tests and least strongly for easy IQ tests. So, we conclude that the predictions of phenotypic cognitive scores by polygenic cognitive scores are strongly driven by the cognitive complexity of the phenotypic measures.