Optimal anthropometric measures to predict incidence of coronary heart disease in adults in Turkey.

Abstract

We aimed to identify the most suited anthropometric measure for the prediction of risk for incident coronary heart disease (CHD) among the Turkish population. We collected data on body mass index, waist circumference (WC), hip circumference, waist-to-hip ratio (WHR), waist-to-height ratio (WHtR) and a body shape index. We analysed these using both C-statistics and Cox regression models adjusted for age, systolic blood pressure, glucose and high-density lipoprotein (HDL)-cholesterol for assessing risk of incident CHD among 3203 Turkish Adult Risk Factor (TARF) study participants (mean [SD] age 48.5 [11] years). Over a mean follow-up of 9.93 years, new CHD developed in 573 individuals. Multi-adjusted C-statistics were highest for WHtR followed by WC, in both sexes. Except WHR, all measures were significantly associated with incident CHD in combined sexes in the full model. There was a sex difference, however, in the mediation of the three risk factors for adiposity; these attenuated hazard ratios (HRs) in males, whereas in females, significant prediction of incident CHD persisted for each measure. WC (HR 1.36 [95% CI 1.13; 1.64]), followed by WHtR (HR 1.24 [95% CI 1.10; 1.40]), were in combined sex, as in females, the most informative surrogates of adiposity. Hip circumference did not protect, but rather conferred modest CHD risk, especially in females, rendering a low utility of predictive value for WHR. The CHD risk curve did not have a J shape. WC is the most suitable of five adiposity surrogates for CHD risk among Turkish adults, while in males, unmediated adiposity risk was similarly identified by WHtR. Retention of the large part of CHD risk in females perhaps reflects the underlying autoimmune activation.