Response to Cameron et al.

We welcome Cameron et al.’s letter highlighting the interesting findings of their recent review.¹ We agree that ratio measures of anthropometry such as the waist to hip ratio (WHR) may be less predictive for mortality than their underlying pair of parameters, that is, waist (WC) and hip circumference (HC), although neither our nor their review was designed to specifically address this comparison. As indicated by Cameron et al. HC was associated with mortality in some studies.^{1, 2} However, it is important to note that the significance of this association heavily depends on the degree of adjustment for other anthropometric variables. In contrast to WC, HC was not significantly associated with mortality in a multivariable adjusted model that also included body mass index (BMI) in the European Prospective Investigation into Cancer and Nutrition (EPIC) and other cohort studies.^{3, 4} It was the rationale of our review to analyse and compare cohort studies investigating BMI and at least one abdominal obesity measure, such as WC, WHR or WHtR.⁵ We feel that our findings provide further evidence that mortality prediction may be improved by combining BMI and an abdominal obesity measure such as WHR or WC.