Optimal cut-off of obesity indices to predict cardiovascular disease risk factors and metabolic syndrome among adults in Northeast China

Background CVD risk factors (hypertension, dyslipidemia and diabetes) and MetS are closely related to obesity. The selection of an optimal cut-off for various obesity indices is particularly important to predict CVD risk factors and MetS. Methods Sixteen thousand seven hundred sixty-six participants aged 18–79 were recruited in Jilin Province in 2012. Five obesity indices, including BMI, WC, WHR, WHtR and BAI were investigated. ROC analyses were used to evaluate the predictive ability and determine the optimal cut-off values of the obesity indices for CVD risk factors and MetS. Results BMI had the highest adjusted ORs, and the adjusted ORs for hypertension, dyslipidemia, diabetes and MetS were 1.19 (95 % CI, 1.17 to 1.20), 1.20 (95 % CI, 1.19 to 1.22), 1.12 (95 % CI, 1.10 to 1.13), and 1.40 (95 % CI, 1.38 to 1.41), respectively. However, BMI did not always have the largest adjusted AUROC. In general, the young age group (18 ~ 44) had higher ORs and AUROCs for CVD risk factors and MetS than those of the other age groups. In addition, the optimal cut-off values for WC and WHR in males were relatively higher than those in females, whereas the BAI in males was comparatively lower than that in females. Conclusions The appropriate obesity index, with the corresponding optimal cut-off values, should be selected in different research studies and populations. Generally, the obesity indices and their optimal cut-off values are: BMI (24 kg/m2), WC (male: 85 cm; female: 80 cm), WHR (male: 0.88; female: 0.85), WHtR (0.50), and BAI (male: 25 cm; female: 30 cm). Moreover, WC is superior to other obesity indices in predicting CVD risk factors and MetS in males, whereas, WHtR is superior to other obesity indices in predicting CVD risk factors and MetS in females. Electronic supplementary material The online version of this article (doi:10.1186/s12889-016-3694-5) contains supplementary material, which is available to authorized users.

ously, other indices may be used to measure obesity, but 48 we do not consider all of them here. 49 Some studies indicated that WC or WHtR might be 50 better predictors for CVD risk factors or MetS in 51 Korean/Chinese populations [9,13], whereas, Mbanya 52 et al. noted that WC was the best predictor in Cameroo-53 nians [14]. Moreover, Bergman et al. found that BAI was 54 a better predictor for African-Americans and Mexican-55 Americans [12], However, Lam et al. proposed that BAI 56 is not likely to be better than BMI and does not apply to 57 Asians [11]. Therefore, selection of the proper obesity 58 index for specific research and study populations was a 59 challenge. 60 In our study, the predictive ability and the optimal 61 cut-off values of five obesity indices (BMI, WC, WHR, 62 WHtR and BAI) for CVD risk factors and MetS are 63 comprehensively investigated. Data from 16,766 partici-64 pants aged 18-79 in Jilin Province were used to evaluate 65 the obesity indices. Jilin is in central northeast China 66 and has an annual average temperature 4.8°C (latitude 67 40°~46°, longitude 121°~131°) [15]. Therefore, the re-68 sults can be instructive and meaningful for studies re-69 lated to obesity in northeast China. WC and WHtR are 70 superior to other obesity indices in predicting CVD risk 71 factors and MetS in our study, with optimal cut-off 72 values of WC and WHtR of 85 (male)/80 (female) and 73 0.5, respectively. 74 Methods 75 Study population 76 A large-scale cross-sectional survey was implemented in 77 Jilin Province in 2012. A total of 16,766 participants who 78 had lived in Jilin Province for more than 6 months and 79 were 18-79 years old were selected through multistage 80 stratified random cluster sampling (see details in Part 1 81 of the Additional file 1). 82 Data measurement 83 Height, weight, WC and HC were measured according 84 to a standardized protocol and techniques, with the par-85 ticipants wearing light clothing but no shoes. Blood 86 pressure was measured by trained professionals using a 87 mercury sphygmomanometer. After an overnight fast, 88 FBG and serum lipids were measured before breakfast 89 using a Bai Ankang fingertip blood glucose monitor 90 (Bayer, Leverkusen, Germany) and a MODULE P800 91 biochemical analysis machine (Roche Co., Ltd., 92 Shanghai, China), respectively (see details in Part 2 of 93 the Additional file 1).

94
The various obesity indices were calculated as follows: The continuous variables were expressed as the means ± 117 standard deviations (SD) and compared using the t test.

118
The categorical variables were expressed as counts or 119 percentages and compared using the Rao-Scott-χ 2 test.

120
ROC analyses were used to compare the predictive abil- Inc., New York, NY, USA) Statistical significance was set 130 at a P value < 0.05.

132
The characteristics of the participants are shown in 133 Table   T1 1. Females had a higher age, TC, LDL-C and 134 HDL-C than males (P < 0.05), but other anthropomet-135 ric indices were significantly higher in males than 136 those in females (P < 0.01). The prevalence of hyper-137 tension, dyslipidemia, diabetes, and MetS differed sig-138 nificantly by gender and were higher in males than in 139 females (P < 0.05).

140
For an overview of each obesity index, Then, the detailed performance of 5 obesity indices 151 associated with CVD risk factors and MetS was inves-152 tigated. For females ( Table   T3 3), the ORs and AUROCs 153 of the obesity indices for CVD risk factors and MetS 154 were the largest in the 18~44 age group, followed by 155 the 45~64 group. Thus, obesity in the younger age 156 groups was at a higher risk for CVD risk factors and 157 MetS (higher ORs), and it had better predictive ability 158 for CVD risk factors and MetS as well (larger 159 AUROC). Further, the AUROC for males had a simi-160 lar tendency and characteristics as that of females 161 (see Additional file 1: Table S3).

162
The detailed optimal operating points (OOPs) for 163 BMI, WC, WHR, WHtR and BAI to predict CVD risk 164 factors and MetS are given in Table   T4 4, in which the OOP 165 is the cut-off value that leads to the maximum Youden  Finally, we investigated the adjusted ORs and AUROC 186 of each obesity index for CVD risk factors and MetS 187 (Table   T5 5) using the optimal cut-off values determined 188 above. In general, the WC and WHtR had higher ad-189 justed ORs and AUROCs for CVD risk factors and 190 MetS, regardless of the small difference between gen-191 ders. WC was superior to other obesity indices in pre-192 dicting CVD risk factors and MetS in males, but WHtR 193 was superior to other obesity indices in predicting CVD 194 risk factors and MetS in females. Abnormal WC or 195 [24,25] to de-205 scribe obesity. Unfortunately, no obesity index was con-206 sistently superior in predicting CVD risk factors and 207 MetS, and the selection of an obesity index depended on 208 the study population and other factors [11]. Thus, in this 209 study, we investigated the proper obesity index and opti-210 mal cut-off values to predict CVD risk factors and MetS 211 for a population in northeast China. 212 In this study, obesity in younger age groups was a 213 higher risk and had better predictive ability for CVD risk 214 factors and MetS than in older groups. It was implied 215 that obesity might have more influence on young people. 216 One possible reason was that the young people took part