Homozygosity caused by consanguineous union has long been associated with an increased prevalence of rare Mendelian disorders. In contrast, the role of homozygosity in relation to quantitative traits and complex disease suscep- tibility is less well established. Recent work has shown that an increased bur- den of homozygous DNA segments (defined as the fraction of each genome in any Run of Homozygosity > 1.5 Mb, or FROH) is associated with reduced height and cognitive ability in diverse human populations. Here, we assess the scope of inbreeding depression in humans by analysing the effect of FROH on 41 quantitative traits of public health and evolutionary interest, in >500,000 individuals from >100 cohorts worldwide. Phenotypes studied include anthro- pometry, lipids, liver, lung and kidney function, cognition, glycaemia, fertility, inflammation, electrocardiography and haematology. We confirm the findings of previous work, and also show that increased FROH is associated with reduced values of several components of evolutionary fitness. The constancy of these effects across different levels of homozygosity and across diverse populations in which social confounding more plausibly engenders the opposite effect, strengthens the argument that FROH has a causal effect. In addition, we inves- tigate the magnitude of inbreeding depression between men and women to ex- plore whether sex-specific dominance may provide a mechanism, independent of the sex-chromosomes, by which males and females might adapt to different selective pressures. We investigate the mechanism of inbreeding depression by exploring different measures of homozygosity. Long homozygous segments arise from recent common ancestry and bring all variants, including rare causal variants, into a homozygous state. In contrast, shorter homozygous segments may arise from more ancient common ancestry and thus capture only the dominance effects of more common variants. Using multivariate linear models we show that FROH is more predictive of inbreeding depression than traditional estimates of inbreeding coefficient based on individual SNP homozygosity.
