TY - JOUR T1 - Age-related late-onset disease heritability patterns and implications for genome-wide association studies JF - bioRxiv DO - 10.1101/349019 SP - 349019 AU - Roman Teo Oliynyk Y1 - 2018/01/01 UR - http://biorxiv.org/content/early/2018/07/11/349019.1.abstract N2 - Genome-wide association studies (GWASs) and other computational biology techniques are gradually discovering the causal SNPs and gene variants that contribute to late-onset human diseases (LODs). After more than a decade of GWAS efforts, these can account for only a fraction of the heritability implied by familial studies, the so-called “missing heritability” problem.Computer simulations of an aging population have shown that the risk allele frequency decreases at older ages because the individuals with higher polygenic risk are first to become ill. This effect is most prominent for diseases characterized by high cumulative incidence and high heritability, examples of which include Alzheimer’s disease, coronary artery disease, cerebral stroke, and type 2 diabetes. The LOD incidence rate grows exponentially, doubling in incidence between 5 and 8.5 years, guaranteeing that the cohorts for GWAS studies overrepresent older individuals with lower polygenic risk scores, whose disease cases are disproportionately due to environmental causes such as old age itself. This simultaneously leads to diminished observed heritability and lower GWAS statistical power at older ages; thus, the missing heritability in GWAS is smaller than currently estimated for these LODs.This mechanism also explains the relatively constant heritability with age reported by familial studies for the four most prevalent cancers—breast, prostate, colorectal and lung—due to a combination of their lower heritability and lower cumulative incidence; this should also be true for other LODs with similar characteristics. In addition, this mechanism explains the heritability patterns found by familial studies and in clinical practice as it relates to predicting LOD familial risks for all of these LODs.In conclusion, for LODs showing high cumulative incidence together with high initial heritability, rather than using relatively old age-matched cohorts, study cohorts combining the youngest possible cases with the oldest possible controls may significantly improve the discovery power of GWASs.Author summary We investigated the change in distribution of risk alleles with age progression under a model assuming that relative disease liability remains proportionate to individual polygenic risk. We found that individuals with higher polygenic risk factors become ill proportionately earlier, and the fraction of higher risk alleles diminishes for the remaining unaffected population. This corresponds to diminishing heritability with age and lower GWAS statistical discovery power for LODs with high incidence. Even though incidence for all LODs increases exponentially with age, the effect is minimal for diseases with low prevalence and low heritability, as exemplified by cancer.High cumulative incidence and high initial heritability are the primary determinants of risk allele relative frequency decline and consequent diminishing older age heritability and GWAS statistical power. The effect is very pronounced for a number of prevalent and highly heritable LODs, including Alzheimer’s disease, coronary artery disease, cerebral stroke, and type 2 diabetes, in which the cohorts for GWAS studies over-represent older individuals whose genotype would be considered low risk earlier on and whose disease is due to old age rather than heightened genetic liability. The predictive power of GWASs is closer to the real heritability of this population than it is currently credited with. ER -