| BACKGROUNDA marked increase in the prevalence of overweight and obesity has contributed to the worldwide spread of type 2 diabetes mellitus over the past decades.Although it is always warned of the deterious effect of obesity on health by public health organizations and physical exercise is advocated to keep body weight in suitable range,there are still controversies existing on the consistent association between body weight and risk of diabetes.Indeed,many obese persons appear resistant to the development of metabolic disease.On the other hand,many persons appear to develop metabolic disorders in the normal range of body weight,which has been termed as normal weight metabolic unhealthy.The heterogeneous effect of obesity on diabetes also exists.The drastic increase in the number of diabetic patients in China is of great concern.At a given level of body mass index or waist circumference,Chinese are prone to develop diabetes than Europeans.Adipose tissue dysfunction is characterized by ectopic fat deposition in the abdominal viscera and liver,and may be a more important mediator of diabetes development than total fat mass.It is recently reported that,regular aerobic exercise might not lead to weight loss.Of interest,the fat deposited in viscera and liver is decreased with the metabolic disturbance ameliorated,which hinted the important meditor effect of visceral fat and liver fat.However,this issue is rarely studied.OBEJECTIVEIn the present study,we aimed to investigate the following issues in community-based cohort:1)the association of total body fat,fat distribution with glucose metabolism,and the various effect of fat distribution on diabetes risk;2)the association of overall obesity,abdominal obesity with risk of diabetes incidence,and the effect of body mass index,waist circumference at baseline,and the change of body mass index and change of waist circumference at follow up on the development of diabetes;3)develop a simple and reliable algorithm to predict visceral adiposity and explore its relationship with risk of incident diabetes;4)determine whether a change in fatty liver status(either development of new fatty liver,or resolution of existing fatty liver)modified risk of incidence diabetes at 4~5 years follow-up.STUDY DESIGN AND METHODSThe community-based cohorts in Jiading,Shanghai,and Lianqian,Xiamen were investigated.At baseline survey,standardized questionnaires were used to obtain the information on demographic characteristics,lifestyles,history of diseases and medication usage,and height,weight and waist circumference.All the participants underwent standardized oral glucose tolerance tests.Blood samples were collected to measure fasting plasma glucose,2h post load glucose,and lipid profiles.512participants in Lianqian,Xiamen cohort received abdominal computed tomography for measurement of subcutaneous fat area and visceral fat area and received magnetic resonance spectroscopy for measurement of intrahepatic triglyceride content.All participants in Jiading,Shanghai cohort received bioelectrical impedance for measurement of body fat mass and lean mass,and ultrasonography for determination of fatty liver.The follow-up survey of Jiading cohort was conducted after mean 4.4 years.Standardized questionnaires were used to obtain the information on change of lifestyles,medication usage,and height,weight,waist circumference.All the participants underwent abdominal ultrasonic examination and standardized oral glucose tolerance tests at follow-up.SAS version 9.3(SAS Institute,Cary,NC)was used for database management and statiscal analysis.RESULTS1.The association of total body fat,adipose distribution and diabetes:In Xiamen cohort,participants were divided into quartiles by body fat percent,an association across quartiles of body fat percent was observed for 2h post-load glucose levels,but not for fasting glucose levels.When participants were divided into quartiles by subcutaneous fat area,an association across quartiles was neither observed for fasting glucose levels nor for 2h post-load glucose levels.However,fasting glucose levels and 2h post-load glucose levels increased significantly according to quartiles of visceral fat area and intrahepatic triglyceride content.In Jiading,Shanghai cohort at baseline,overall obesity(reflectd by body mass index),abdominal obesity(reflected by waist circumference),and fatty liver(reflected by fatty liver detected by ultrasound)were positively associated with prevalence of diabetes.On the basis of overall obesity,abdominal obesity and fatty liver categories,the participants were classified into 9 groups.Taken the group without overall obesity,abdominal obesity and fatty liver as the reference group,the odds ratio of diabetes for overall obesity group was 1.37(95%CI:0.75-2.51),abdominal obesity group was 1.54(95%CI:1.23-1.93),combined overall obesity and abdominal obesity group was 1.68(95%CI:1.27-2.22),fatty liver group was 3.97(95%CI:3.34-4.73).Compared with the group without overall obesity,abdominal obesity and fatty liver,the odds ratio of diabetes in group with fatty liver,regardless of combined with either overall obesity or abdominal obesity or both,was similar to fatty liver alone,increased approximately by 300%.In Jiading,Shanghai follow up cohort,waist circumference and fatty liver were independent risk factors of diabetes incidence.Total body fat did not remain in the final model for predicting diabetes incident risk.With per SD increasement of waist circumference,the odds ratio of incident diabetes increased by 14%(95%CI:1.01-1.29,P=0.03).Compared with participants without fatty liver,the odds ratio of incident diabetes for participants with fatty liver at baseline was 2.02(95%CI:1.57-2.58,P<0.0001).Further similar analyse was performed for Chinese visceral adiposity index as biomarker of visceral fat mass.The result of Chinese visceral adiposity index was similar with waist circumference.Chinese visceral adiposity index remained in the final model for predicting diabetes incident risk.With per SD increase of Chinese visceral adiposity index,the odds ratio of incident diabetes increased by 50%(95%CI:1.28-1.77,P<0.0001).2.The association of overall obesity,abdominal obesity and diabetes:Participants were divided into three groups(normal weight group,overweight group,overall obesity group)accoding to cutoff points of body mass index of 24 kg/m~2 and 28 kg/m~2.The incidence of diabetes increased in dose-reponse pattern,with 7.8%in normal weight group,with 13.4%in overweight group,and with18.8%in overall obesity group.Participants were divided into abdominal obesity group and non-abdominal obesity group according to the cutoff points of waist circumference at 90 cm for men and 85 cm for women.The incidence of diabetes was 17.1%in abdominal obesity participants,higher than 10.1%in non-abdominal obesity participants.After adjustment for gender,age,smoking and drinking statuses,educational level,phycial activity,and family history of diabetes,with per SD increasment of baseline body mass index,the risk of developing diabetes increased by 43%(95%CI:1.32-1.55);with per SD increase of baseline waist circumference,the risk of developing diabetes increased by 48%(95%CI:1.35-1.62).After adjustment for baseline body mass index and baseline waist circumference,gender,age,smoking and drinking statuses,educational level,phycial activity,and family history of diabetes,with per SD increase of change of body mass index,the risk of developing diabetes increased by 11%(95%CI:1.02-1.20);with per SD increase of change of waist circumference,the risk of developing diabetes increased by 15%(95%CI:1.05-1.25).The risk of body mass index of developing diabetes was weakened by ageing(P for interaction<0.05).However,the risk of waist circumference of developing diabetes did not alter significantly by ageing(P for interaction>0.05).3.Visceral fat and diabetes:We developed an algorithm of Chinese visceral adiposity index in Xiamen cohort.The correlation coefficients of visceral fat mass with CVAI,BMI,WC,VAI(Italy)were 0.68,0.49,0.56,0.19,separately(P all<0.0001).CVAI was more strongly correlated with visceral fat area,than BMI,WC,VAI(Italy).Regarding visceral obesity defined as visceral fat area more than 100 cm~2,the area under the receiver operating characteristic curve of CVAI was higher than BMI,WC,VAI(Italy)(all P≤0.0001).Further analyses were performed to explore the association of CVAI and incident diabetes in prospective Jiading,Shanghai cohort.When participants were divided into sex-specific quartiles of CVAI,a graded association across quartiles of CVAI was observed for the baseline fasting glucose level,baseline 2h post-load glucose level,change of fasting glucose level,change of 2h post-load glucose level,and incidence of diabetes(P for trend<0.0001).After adjustment for age,smoking status,alcohol consumption,physical activity,education level,family history of diabetes,and postmenopausal status only for women,the harzard ratio for incident diabetes increased by CVAI quartile,in comparison with the first quartile,separately in men and women.Concerning diabetes incident risk,ROC curve analysis showed significant differences in C statistics between CVAI and BMI(P<0.0001),and WC(P=0.0001).4.Fatty liver and diabetes:The incidence rate of diabetes was 22.5%in fatty liver participants,higher than 9.0%in non-fatty liver participants at baseline(P<0.0001).New incident fatty liver was associated with risk of incident diabetes after adjustment for age,gender,smoking status,alcohol consumption,physical activity,education level,family history of diabetes(OR:2.41,95%CI:1.87-3.10).Compared with fatty liver sustained group,resolution of fatty liver was associated with decreased risk of incident diabetes(OR:0.45,95%CI:0.31-0.65).Further adjusting for body mass index or waist circumference,the association above did not alter appreciably.CONCLUSIONSFat distribution,rather than total body fat mass,was associated with metabolic disorders and glucose metabolism.Abdominal visceral fat and liver fat were more critical risk factors of metabolic unhealthy obesity and diabetes development than total body fat.Baseline body mass index,baseline waist circumference,change of body mass index,and change of waist circumference were associated with risk of incident diabetes.The effect of body mass index on diabetes development was weakened by ageing.However,the risk of waist circumference of developing diabetes did not alter significantly by ageing.The CVAI is able to predict visceral obesity and to be employed in clinical practice.CVAI was associated with metabolic disorders and incident diabetes,which might therefore be a useful tool for the assessment of visceral obesity and visceral dysfunction.Development of new fatty liver was associated with increased risk of diabetes.And resolution of existing fatty liver was associated with decreased risk of incident diabetes.The associations were independent of baseline body mass index or waist circumference,change of body mass index or waist circumference. |
PDF Full Text Request