Abstract
BACKGROUND: Visceral adiposity index (VAI) and a body shape index (ABSI) were newly developed indices for visceral fat mass. Whether they are superior to conventional obesity indices in predicting colorectal cancer (CRC) remains unclear. We examined the associations of VAI and ABSI with CRC risk, and investigated their performance in discriminating CRC risk compared with conventional obesity indices in the Guangzhou Biobank Cohort Study. METHODS: A total of 28,359 participants aged 50 + years without cancer history at baseline (2003-8) were included. CRC were identified from the Guangzhou Cancer Registry. Cox proportional hazards regression was used to assess the association of obesity indices with the CRC risk. Discriminative abilities of obesity indices were assessed using Harrell's C-statistic. RESULTS: During an average follow-up of 13.9 (standard deviation = 3.6) years, 630 incident CRC cases were recorded. After adjusting for potential confounders, the hazard ratio (95% confidence interval) of incident CRC for per standard deviation increment in VAI, ABSI, body mass index (BMI), waist circumference (WC), waist-to-hip ratio (WHR) and waist-to-height ratio (WHtR) was 1.04 (0.96, 1.12), 1.13 (1.04, 1.22), 1.08 (1.00, 1.17), 1.15 (1.06, 1.24), 1.16 (1.08, 1.25)and 1.13 (1.04, 1.22), respectively. Similar results for colon cancer were found. However, the associations of obesity indices with risk of rectal cancer were non-significant. All obesity indices showed similar discriminative abilities (C-statistics from 0.640 to 0.645), with WHR showing the highest whilst VAI and BMI the lowest. CONCLUSIONS: ABSI, but not VAI, was positively associated with a higher risk of CRC. However, ABSI was not superior to the conventional abdominal obesity indices in predicting CRC.