Abstract
BackgroundThis study aimed to assess the predictive performance of diabetes (DM) by using adiposity indices compared to body mass index and waist circumference. MethodsAmong 8249 consecutive subjects who attended the Nationwide Health Check Up System for Senior Citizens between 2008 and 2018 (≥ 65 years), we examined the associations of several adiposity indices with DM risk, and further explored gender differences. ResultsAmong all adiposity indicators, Chinese visceral adiposity index (CVAI) alone demonstrated the highest discriminatory ability for diabetes mellitus by area under receiver operating characteristic curves (AUC) (0.65, 0.68, and 0.66 for men, women, and all participants, respectively), regardless of gender, with optimal cut-offs set as 126.09 in men and 117.77 in women, respectively. Compared with body shape index (ABSI), CVAI was strongly associated with baseline DM (adjusted OR: 4.16 [3.35–5.17] for 4th vs 1st quartile groups, P < 0.001), which was more pronounced in elderly women (P interaction < 0.05). Over a median of 5.25 years (IQR: 3.07–6.44 years) follow-up, Cox regression models showed higher predictive ability of CVAI rather than the ABSI. Further, this independently predicted new-onset DM (adjusted HR: 1.26, 95% CI: 1.18–1.34) and composite endpoint of new DM and death (adjusted HR: 1.17, 95% CI: 1.10–1.25, both P < 0.001) among those without baseline DM. ConclusionsOur population-based data demonstrated that Chinese visceral adiposity index may serve as a superior clinical indicator of diabetes when compared with conventional anthropometric indices among elderly Chinese, especially in women.