The Optimal Cut-off Values Of Different Obesity Indicators For Risk Of Hypertension And Dyslipidemia In Adults

Background:Obesity is caused by the increased number or volume of fat cells,which leading to excessive accumulation and abnormal distribution of body fat.Obesity is not only an independent disease,but also a risk factor for chronic noncommunicable diseases such as type 2 diabetes,cardiovascular disease and malignant tumor.It has become a serious problem that threatens people’s health in China.In the past,body mass index(BMI)and waist circumference(WC)were widely used in the field of scientific research for obesity.In recent years,the frequency of human body composition analyzer has become more common in epidemiological investigation.Body fat percentage(BF%),as a physical measure of obesity,has gradually been recognized by the academic community.However,in China,there still lacks the criteria for the optimal cut-off point of BF%.In recent years,the prevalence of hypertension and dyslipidemia has increased significantly all around the world.Using simple obesity indicators to predict the risk of hypertension and dyslipidemia is not only the key work in the field of public health,but also important for the early prevention and control of chronic diseases in our country.Objective:To compare the effects of different obesity indicators on predicting the risk of hypertension and dyslipidemia,and to propose suitable indicators and the cut-off values for predicting hypertension and dyslipidemia.Methods:The data came from the National Natural Science Fundation project "the epidemiological study on the relationship between dietary fat intake and nutrition-related chronic diseases and economic changes".Stratified cluster random sampling method was used.We selected Liaoning,Henan and Hunan in the present study.One city site and one rural site were selected in each province.In each site,200 adults aged 18-69 years old were randomly selected for physical measurement,and 100 subjects were selected from them for hematological examination,including serum total cholesterol(TC),triglyceride cholesterol(TG),high density lipoprotein cholesterol(HDL-C),low density lipoprotein cholesterol(LDL-C)and other blood index.Finally,1022 subjects(462 males and 560 females)were included in the study,and 590(269 males and 321 females)subjects participated in the hematology examination.Receiver operating characteristic(ROC)was used to analyze the effects of physical measurement indicators to predict the risk of hypertension and dyslipidemia,and explore the cut-off values.The consistency of BMI and BF%was determined by kappa coefficient.The Populatiom attributable risk proportion(PARP)was used to determine the attributable risk of hypertension and dyslipidemia by indicators in the population.Results:The results showed that the overweight and obesity rates of BMI,WC,waist to hip ratio(WHR),waist to height ratio(WHtR),were 49.1%,35.3%,50.4%,58.7%and 35.7%respectively,among which the obesity rate of WHtR was the highest.The prevalence rates of hypertension in men and women were 27.1%and 21.6%respectively,and there was a significant difference between different sexes.The prevalence rates of dyslipidemia in men and women was 44.2%and 31.5%respectively,and there was a significant difference between different sexes.The prevalence rates of hypercholesterolemia in men and women were 12.6%and 15.9%respectively,and there was no significant difference between different sexes.The prevalence rates of mixed hyperlipidemia in men and women were 6.7%and 3.4%respectively,and there was no significant difference between different sexes.The average value of male BF%was(22.1±5.3)%,and of female BF%was(33.2 ± 5.9)%.According to the BF%obesity criteria recommended by World Health Organization(WHO),the subjects were divided into two groups:abnormal BF%group and normal BF%group.The results showed that male’s height,weight,WC,BMI,WHR,hip circumference(HC)and waist to thigh ratio(WTR)were significantly higher than that of female’s,while women’s thigh to height ratio(THtR)was higher than that of men,and there was no significant difference between male and female in thigh circumference(ThC).WHtR of normal group was significantly higher than that of the female group,while the difference of WHtR in abnormal group was not statistically significant.The ROC curve shows that WHtR has the largest AUC for predicting hypertension in men and women,and the cut points are 0.53 and 0.56 respectively.WHR is the most meaningful indicator to predicate hypercholesterolemia in men and women,and it is the only indicator in male residents,and the cut-off value is 0.89 and 0.84 in men and women respectively.ThC has a significantly higher AUC in predicting low HLD-C both in male and female adults,and the cut-off value is 52.5;The AUC of WHR and WTR was the largest in predicting LDL-C both in male and female.In the ROC analysis of BF%for predicting the risk of hypertension and dyslipidemia,combined with AUC,Youden index,sensitivity and specificity,this study preliminarily suggested that the cut-off value of BF%for male was 23%,and the appropriate cut-off value for female was 35%.The area under the curve(AUC)was 0.97 for obesity(the cut-off value of BF%proposed in this study)by BMI,which was higher than that obesity determined by BF%cut-off value recommended by WHO.The consistency between two BF%standard and BMI>24kg/m2,BMI≥25kg/m2,BMI≥26kg/m2 is test and the kappa coefficient is calculated.The results show that the kappa coefficient is the largest when BMI is 25.The kappa coefficient of the cut-off value of BF%proposed in this study and BMI is 0.80,which is higher than the kappa coefficient(0.78)of the BF%standard recommened by WHO and BMI consistency test.The BF%cut-off value proposed in this study was in good agreement with BMI.In the analysis of PARP,the male BF%>23%had a higher OR value and PARP for hypertension and other types of Dyslipidemia except that the OR for high TG than BF%≥25%.The OR value and PARP values of WHtR in male population for hypertension,hypercholesterolemia,high LDL-C,and>2 risk factors were higher than that of BMI and BF%,and the PARP of the high TG was higher than BMI and BF%,and the OR value was higher than that of BMI,and the OR and PARP of low HDL-C were all higher than those of BF%.The OR value of high TG is only slightly lower than that of BF%,and the OR value and PARP of low HDL-C are lower than that of BF%.In men,abdominal fat accumulation has a higher risk of hypertension and dyslipidemia than systemic obesity indicators,and this contrast is not obvious in the female population.WHtR has higher PARP for hypertension and dyslipidemia.WHtR controlled below 0.5,which can reduce hypertension and dyslipidemia by about 30%to 50%.The OR value of ThC in male and female for hyponatremia was 6.64 and 3.06,which is significantly higher than 4.45 and 2.38 of BMI,2.68 and 2.53 of BF%,3.61 and 1.65 of WHtR,respectively.The PARP in male and female was also significantly higher than that of BMI,BF%and WHtR,and the value was 57.7%and 42.5%respectively.Controlled ThC under 52.5 could reduce the low HDL-C of 42%～58%.Conclusions:1.The cut-off value of BF%for adult residents in some areas of China was 23%for men and 35%for women by a comprehensive analysis of BF%on predicting the risk of hypertension and dyslipidemia.This value has a higher PARP and OR than the standard recommended by WHO,and also has a stronger consistency with BMI.2.When the cut-off value of BMI is 25kg/m2,effect of BMI to measure obesity(BF%standard)is best,and the consistency with BF%is excellent.3.WHtR was good at predicting the risk of hypertension and dyslipidemia,and the appropriate cut-off value was 0.5.This value has a higher PARP and OR for hypertension and dyslipidemia than indicators of systemic obesity.4.The effection of ThC to predict low HDL-C is better than that of systemic and central obesity indicators,and its cut-off value is 52.5cm.