Choice of phenotype scale is critical in biobank-based G × E tests

The importance of gene-environment interactions (G × E) for complex human traits is heavily debated. Recently, biobank-based genome-wide association studies (GWASs) have revealed many statistically significant G × E signals, though most lack clear evidence of biological significance. Here, we partly explain this discrepancy by showing that many G × E signals simplify to additive effects on a different phenotype scale, a classical concern that is currently underappreciated. Our results clearly distinguish G × sex effects on height, which vanish on the log scale, from G × sex effects on testosterone, where the log scale uncovers biologically meaningful female-specific effects. Across 32 phenotypes in UK Biobank, we find that scaling by a power transformation can explain 45% of polygenic score (PGS) × sex interactions and that simple log transformation can explain 23%, with similar results for other environments. We also show that the phenotype scale can substantially impact GWAS discovery and the construction and evaluation of polygenic scores. Finally, we provide a set of guidelines to consider and choose phenotype scales in modern genetic studies.