A Cohort Study On The Association Between Chinese Visceral Adiposity Index And Type 2 Diabetes Mellitus Incidence

Objectives1.To explore the association between Chinese visceral adiposity index（CVAI）and risk of type 2 diabetes mellitus（T2DM）.2.To assess the predictive value of CVAI for T2DM incidence,compare predictive power of CVAI with other obesity indices,and estimate its population attributable risk percentage（PAR%）to provide epidemiological evidence for preventing and controlling incident T2DM.MethodsCluster random sampling was used to recruit 20194 participants（age≥18 years）in rural areas of Xin’an County,Henan Province between July to August 2007 and July to August 2008.Baseline data were collected by questionnaire interview,physical examination,and laboratory measurement.Follow-up surveys were conducted during July to August 2013 and July to October 2014,and 17265participants were successfully followed up（response rate 85.5%）.After excluding participants with T2DM,type 1 diabetes mellitus at baseline and gestational diabetes at baseline or follow-up（n=1717）,diabetes status at baseline or follow-up unknown（n=6185）,and missing data for CVAI（n=61）,a total of 12231 individuals were included to explore the association between CVAI level and risk of incident T2DM.Participants were divided into the first,second,third,and fourth quartile groups based on baseline CVAI level.Cox proportional hazards regression models were used to estimate hazard ratios（HRs）and 95%confidence intervals（CIs）for incident T2DM by quartiles of CVAI and per standard deviation（SD）increase,and restricted cubic splines were used to model the dose-response association of CVAI level with incident T2DM.Stratified analyses were conducted by age,sex,smoking status,drinking status,physical activity and blood pressure levels,and multiplicative interaction models were used to test the interaction among these subgroups.Sensitivity analyses were also performed to assess the robustness of the associations through excluding participants with impaired fasting plasma glucose at baseline,excluding participants who developed T2DM within the first two years of follow-up,and further excluding participants with cardiovascular diseases or kidney diseases.Besides,receiver operator characteristic（ROC）curves were used to calculate the area under the curve（AUC）and 95%CI of CVAI for predicting incident T2DM,and compare the predictive power among CVAI,visceral adiposity index（VAI）,conicity index（CI）,waist-to-height ratio（WHt R）,WC,and body mass index（BMI）in total population,men,and women,respectively.Finally,we calculated abnormal rate of CVAI based on the optimal cut-offs,assessed the adjusted HRs and 95%CIs of T2DM for abnormal group versus normal CVAI group,and estimated its PAR%and 95%CI.Results1.During a median follow-up period of 6.01 years,834 T2DM cases were identified among 12231 participants（cumulative incidence:6.82%;incidence density:11.51/1000 person-years）.The cumulative incidence and incidence density of the second,third,and fourth quartile groups of CVAI were all higher than the first quartile group,with an increasing trend(all P_trendvalue<0.05).After adjusting potential confounding factors,the risk of T2DM for the second,third,and fourth quartile groups of CVAI all increased versus the first quartile group,with an increasing trend(all P_trendvalue<0.05),and the corresponding HRs and 95%CIs were2.04（1.45-2.88）,2.82（2.03-3.93）,and 4.72（3.41-6.53）,respectively.Per-SD increase in CVAI was associated with a 68%increased risk of T2DM（HR 1.68,95%CI1.54-1.83）.The results of sensitivity analyses were almost consistent.2.Restricted cubic splines all showed non-linearity dose-response associations between CVAI and incident T2DM in the three models(all P_nonlinear value<0.05).3.In stratified analyses,after adjusting potential confounding factors,compared with the first quartile group of CVAI,increased risks of T2DM for the fourth quartile group were all statistically significant in all subgroups（all P value<0.05）;risks of T2DM were all increased with the increase of CVAI（all P value<0.05）;The results of sensitivity analyses were almost stable.The results of multiplicative interaction models showed that CVAI level had multiple interaction with age,sex,smoking status,physical activity and blood pressure levels(all P_interaction value<0.05).4.In ROC curve analyses,the AUCs and 95%CIs of T2DM incidence for CVAI were 0.703（0.694-0.711）,0.695（0.682-0.708）,and 0.707（0.697-0.717）,respectively,for total population,men,and women.The predictive power of incident T2DM for CVAI were better than VAI,CI,WHt R,WC,and BMI（all P value<0.05）for total population and women;for men,CVAI had the largest AUC among these obesity indices,while no statistical differences were observed between CVAI and CI,WHt R,and WC（all P value>0.05）.5.The optimal cut-offs of CVAI were 100.73,98.21,and 100.73,respectively,with the corresponding sensitivity and specificity all over 60%,for total population,men,and women.The abnormal rate of CVAI were 37.42%,37.75%,and 38.31%according to the optimal cut-offs in our study,respectively;after adjusting potential confounding factors,risks of T2DM for abnormal group of CVAI all increased significantly versus the normal group（total population:HR 2.35,95%CI 1.97-2.81;men:2.74,2.05-3.65;women:2.13,1.68-2.69）,and the corresponding PAR%were40.7%（95%CI,30.3%-50.1%）,45.0%（29.7%-58.0%）,and 38.2%（24.9%-50.1%）,respectively.Conclusions1.Our study found that higher CVAI was associated with increased risk of T2DM,which suggests that increased CVAI may be a risk factor of T2DM incidence.2.Restricted cubic splines showed non-linearity positive dose-response associations of CVAI with incident T2DM,indicating that decrease of CVAI may be great significance in reducing risk of T2DM.3.In total population,especially in women,the predictive power of incident T2DM for CVAI was better than VAI,CI,WHt R,WC,and BMI,suggesting that CVAI might be a reliable and applicable obesity indicator for identifying people at high risk of T2DM.4.Risk of T2DM in abnormal group of CVAI were higher than normal group in total population,men,and women,and maintain normal CVAI could prevent 40.7%T2DM cases in general population,which has important public health significance for preventing incident T2DM.