Heritability and missing heritability
: can twin studies be trusted?

Student thesis: Doctoral ThesisDoctor of Philosophy


'Missing heritability' is the discrepancy between the amount of variance explained by specific single nucleotide polymorphisms (SNPs) identified in genome-wide association (GWA) and twin-estimated heritability. Four categories of explanations have been proposed for missing heritability: (1) additive effect sizes of common single nucleotide polymorphisms (SNPs) used in GWAS are too small to detect with current sample sizes; (2) rare variants are not captured by commercial arrays; (3) nonadditive effects (allelic, gene-gene or gene-environment interactions); (4) twin estimates of heritability are inflated. A recently developed quantitative method that uses GWA data- Genome-wide Complex Trait Analysis (GCTA)- has made it possible to explore these issues as it allows to compare quantitative twin-based estimates with quantitative DNA-based estimates. I use data from an on-going longitudinal study of 14,000 twins (7000 pairs) born in the UK between 1994 and 1996 called the Twins Early Development Study (TEDS) to investigate the following: the proportion of twin heritability that can be explained by additive effects of common SNPs (Chapters 2, 3 and 4); increasing heritability across development in the presence of strong genetic stability (Chapters 5 and 6); and genetic pleiotropy (Chapter 7). In Chapters 2, 3 and 4, Iapply univariate twin, GWA and GCTA methods to demonstrate that although we are still far from closing the gap between heritability and the actual genetic variants, there still is scope for discovery of common additive genetic effects. In Chapters 5, 6 and 7, I employ bivariate GCTA, polygenic predictor scores (PGS) and twin estimates from the same sample to confirm that twin estimates and DNA estimates of genetic pleiotropy and stability concur. In conclusion, in this thesis I provide evidence that much of the so-called 'missing heritability' can be explained by common additive genetic effects and that phenomena from twin research can be replicated using DNA alone.
Date of Award2013
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorRobert Plomin (Supervisor) & Thalia Eley (Supervisor)

Cite this