Cox and logistic regression models have been previously compared in cohort studies3, 4, 5, 6, 7, 8 but not in case-cohort studies. For cohort studies, it was reported that: (i) the Cox model yields more precise estimates of association; (ii) odds ratios (ORs) and hazard ratios (HRs) diverge as follow-up time, cumulative disease incidence and the strength of the association increase; (iii) in certain situations the Cox model has greater statistical power; and (iv) the Cox model takes a longer time to compute than the logistic regression model. However, most of these studies are at least 20 years old when computational power was limited. Thus, inferences were based on mathematical theory as well as results from small observational studies, and as such do not fully answer the relevant questions for modern GWASs, which are large studies where a small to modest increase in power and precision would probably be sacrificed for an increase in computational efficiency and for practical reasons (such as the availability of suitable software). A more recent study investigated the differences in power between these methods (using
