Clinical Study Of Metabolically Obese Normal-Weight Related Issues In Changchun

BackgroundBody mass index(BMI)is a universal criterion to define normal weight,overweight and obesity.Metabolism disorders and cardiovascular events have been associated with an increase of BMI.But some normal-weight people are likely to develop type 2 diabetes mellitus(T2DM),high blood pressure(HBP)and hypertriglyceridemia,which is defined as metabolically obese normal-weight(MONW).The risk of cardiovascular disease-related metabolic disorders are markedly overlooked in patients with MONW.At present,there are few studies on MONW in China.Objectives1.To investigate the epidemiology,clinical characteristics and risk factors of MONW in Changchun,Northeastern China.To identify the body composition,including mass and distribution of fat,and correlation between body composition and metabolic disorders.2.To observe annual incidence of T2DM and hypertriglyceridemia in normalweight adults.To assess the predictive value of the indicators,and establish models to predict T2DM and hypertriglyceridemia in normal-weight adults.3.To assess the predictive value of the anthropometric indicators,and find the best predictive indicators for MONW in normal-weight adults.Methods1.In the first part,a cross-sectional study was conducted and normal-weight people aged 40 years or older in Changchun were selected as objects.In the first section,we collected basic information and anthropometrics,performed laboratory tests and questionnaire surveys.Then we calculated the prevalence of T2DM,HBP,hypertriglyceridemia and MONW in normal-weight population.Univariate and multivariate analyses were used to identify the risk factors for MONW.In the second section,dual-energy X-ray absorptiometry(DXA)was used to detect fat mass(FM),and distribution of fat.Then we calculated body fat percentage(BF%),fat mass index(FMI),fat mass percent(FM%)to identify characteristics of fat and muscle distribution for MONW.Partial correlation analysis after adjustment for age,sex,and BMI was used to clarify the associations between body composition and metabolic disorders.2.The second part was a cohort study.Firstly,we selected normal-weight people without T2DM at baseline as objects,and the progression to T2DM was considered as the endpoint to build a predictive model for T2DM.Univariate and multivariate analyses were used to identify prognostic indicators.Receiver operating characteristic(ROC)and Youden index was used to evaluate the cutoff value of indicators.We used multivariable logistic regression analysis to construct the T2DM prediction model and Nomogram.ROC and Calibration curves were performed and area under the curve(AUC)and Concordance index(C-index)were calculated to assess the predictive performance of the model.Internal model validation was carried out by bootstrap validation.Secondly,we build a predictive model for hypertriglyceridemia using the same method.Then we intend to build a novel model based on the two models to predict the risk of T2DM and hypertriglyceridemia for normal-weight people,which can be considered as a new diagnostic criterion for MONW.Net reclassification index(NRI)and AUC were used to compare the accuracy of the old and new criteria.3.In the third part,a cross-sectional study and a cohort study was performed.We collected anthropometric indicators indexes which are novel or commonly used clinically.Then we performed ROC and calculated AUC to find the most accurate indicators for MONW using baseline data.Similarly,we intend to find the best predictor for MONW using follow-up data.Results1.In the first section of part one,a total of 3303 subjects with normal weight were included,including 931 males and 2372 females.There were 726 adults(21.98%)with T2DM,1465 adults(44.35%)with HBP,474 adults(14.35%)with hypertriglyceridemia,and 774 adults(23.43%)with MONW.BMI,waist circumference(WC),systolic blood pressure(SBP),diastolic blood pressure(DBP),heart rate,fasting plasma glucose(FPP),2-hour postprandial plasma glucose(PPG),glycosylated hemoglobin(Hb A1c),fasting serum insulin(FINS),the homeostasis model assessment-insulin resistance(HOMA-IR),triglyceride(TG),low density lipoprotein cholesterol(LDL-C)were significantly higher in MONW group,and high density lipoprotein cholesterol(HDL-C)was lower in MONW group than metabolically healthy normalweight(MHNW)group(p < 0.05).The prevalence of T2DM(45.99%),hypertriglyceridemia(43.67%),and HBP(70.03%)were also higher in MONW group(14.63%,19.07% and 36.50%,respectively,p<0.001).No work and daily napping habits are risk factors for MONW(p<0.05).2.In the second section of part one,a total of 318 subjects with normal weight were included,including 115 with MONW.Visceral adipose tissue(VAT),and subcutaneous adipose tissue(SAT)were higher in the MONW group(p<0.05).VAT,was positively correlated with metabolic disorders(p<0.001).The fat mass(FM),fat mass index(FMI),and fat mass percent(FM%)of the trunk and Android were higher in the MONW group than those of the MHNW group,while the FM,FMI,and FM%of thighs and Gyroid was significantly lower in the MONW(p<0.05).FM,FMI,FM%of Android were positively correlated with metabolic disorders,and FM,FMI and FM%of the thigh and Gyroid were negatively correlated with metabolic disorders(p<0.05).3.In the second part,a total of 1492 adults with normal weight had been followed up,including 397 males(26.61%).At baseline,there were 1133 adults without T2DM and 1285 adults without hypertriglyceridemia.At follow-up,55 adults(4.85%)were diagnosed with new-onset T2DM,and the incidence rate were 1.66/100 person-years;141 adults(10.97%)were diagnosed with new-onset hypertriglyceridemia,and the incidence rate were 3.74/100 person-years.SBP,PPG,Hb A1 C,HOMA-IR and their cut-point values 140 mm Hg,7.15mmol/l,5.95% and 2.3 were included in the T2DM X prediction model and Nomogram.ROC(AUC=0.80),C-index(0.80,95%CI 0.74-0.86),and the internally verified C-index(0.79),indicating that the model has moderate predictive ability.Then we built a hypertriglyceridemia prediction model included FINS,TG,HDL-C,LDL-C and their cut-point values 6.00 u IU/ml,1.45 mmol/l,1.21mmol/ l and 2.98 mmol/l using the same method.The calibration chart,ROC(AUC=0.78),C-index(0.78,95%CI 0.74-0.82)and internally verified C-index(0.78),indicating that the model has moderate predictive ability.Finally,we combined the two models to a scoring table included blood sugar,insulin,blood pressure,and lipids,to predict the risk of T2DM and hypertriglyceridemia for normal-weight adults.The scoring table are bounded by 110,which means a score higher than 110 represents a high risk of T2DM and hypertriglyceridemia,and a score lower than 110 represents a low risk of T2DM and hypertriglyceridemia.The scoring table can be recognized as a new diagnostic criterion for MONW.The new model was proved to have an improved predictive capability compared to the old model(AUC = 0.63,95% CI 0.584-0.680,vs.AUC = 0.58,95% CI 0.525-0.636;NRI = 5.00%>0).4.In the third part,we included WC,BMI,waist-to-hip ratio(WHR),waist-toheight ratio(WHt R),and body adiposity index(BAI),body adiposity index-WC(BAIWC),a body shape index,(ABSI),conicity index(CI),body roundness index(BRI),and TMI(tri-ponderal mass index).The above anthropometric indicators had proved to have certain diagnostic power(p<0.05,AUC>0.5)by ROC using baseline data.Among these,WHt R(AUC=0.674,95%CI 0.647-0.702),BAI-WC(AUC=0.673,95%CI0.646-0.701),and BRI(AUC=0.665,95%CI 0.637-0.692)had the strongest diagnostic ability and the cut-points are 0.48,19.19 and 4.03,respectively.Similarly,we found that WHt R,BAI,BAI-WC,BRI,CI and TMI have certain predictive power(p<0.05,AUC>0.5).Among these,BAI(AUC=0.635,95%CI 0.551-0.719),BAI-WC(AUC=0.617,95%CI 0.531-0.703)and BRI(AUC=0.614,95%CI 0.527-0.701)had the strongest predictive ability and their cut-point were 27.49,19.57,and 4.24,respectively.Conclusions1.The prevalence of T2DM,HBP,hypertriglyceridemia and MONW in the normal-weight population was high in Changchun.The anthropometric indicators and metabolic indicators,and the prevalence of T2DM,HBP,and hypertriglyceridemia were higher in the MONW population than MHNW.No work and daily lunch break were risk factors of MONW.VAT,fat mass of trunk and Android were significantly increased in the MONW population,while the fat of thigh and Gyroid was significantly reduced.2.The T2DM prediction model had a moderate predictive power,which included SBP,PPG,Hb A1 C and HOMA-IR and their cut-points 140 mm Hg,7.15mmol/l,5.95%and 2.3.Similarly,the hypertriglyceridemia prediction model also had a moderate predictive power,which included FINS,TG,HDL-C and LDL-C,and their cut-points6.00 u IU/ml,1.45 mmol/l,1.21 mmol/l and 2.98 mmol/l.The new criteria of MONW,which incorporates of blood glucose,insulin,blood pressure,and lipids,has a better predictive ability than the old standard.3.The best evaluation indicators of MONW are WHt R,BAI-WC and BR,and the cut-points are 0.48,19.19 and 4.03,respectively.The best predictors of MONW are BAI,BAI-WC and BRI,and their cut-point were 27.49,19.57,and 4.24,respectively.