The "Schizophrenia is 80% Heritable" Fallacy: Deconstructing a Twin Study Meta-Analysis.

This article critically examines the widely cited 2003 meta-analysis by Sullivan, Kendler, and Neale (SKN) that estimated schizophrenia to be approximately 81% heritable based on twin studies. It argues that this heritability estimate is invalid due to three main issues: the unsupported "equal environments assumption" (EEA) underlying twin studies, unreliable and inconsistent schizophrenia diagnoses—especially in older studies—and the conceptual limitations of heritability estimates, which do not measure the strength or causality of genetic influences. The article also highlights that SKN included older twin studies conducted by researchers with strong eugenic biases and methodological flaws, while excluding or omitting other relevant studies with lower concordance rates. Given the failure of decades of DNA-based gene discovery efforts and these methodological concerns, the article concludes that the 81% heritability figure should be disregarded and calls for a comprehensive re-evaluation of schizophrenia genetics research and the assumptions underlying behavioral twin studies.