Correlation Between Atherogenic Index Of Plasma [lg(TG/HDL-C)] And Both Functional And Structural Markers Of Atherosclerosis

Background and Objective:Cardiovascular disease is still the leading cause of death in China.Previous studies have confirmed the important role of lipids in the formation and development of atherosclerosis.Therefore,we have been looking for economic and feasible means to detect early subclinical atherosclerosis.The plasma atherogenic index of plasma(AIP)was calculated from the logarithm conversion value of the ratio of plasma triglyceride(TG)to high-density lipoprotein cholesterol(HDL-C)concentration based on 10,in which TG and HDL-C concentration were expressed in molar concentration.Previous studies have shown that AIP can sensitively reflect the changes of lipids,especially lipid subfractions,and may become an indicator to assess the severity of atherosclerosis and predict cardiovascular events.Hypertension is the most important risk factor of cardiovascular disease,and hypertension is also a risk factor of atherosclerosis.Noninvasive examination of brachial ankle pulse wave velocity(baPWV)and ankle brachial index(ABI)can reflect functional and structural atherosclerotic changes,respectively.This study aims to explore the relationship between AIP and functional and structural changes of atherosclerosis in Chinese hypertensive population by analyzing the correlation between AIP and baPWV and ABI.The stability of this correlation was further explored in subgroups of important covariates.Methods:This study is an integral part of a multicenter,prospective,observational Chinese Hypertension Registry.From March 2018 to August 2018,a total of 14268participants were recruited in Wuyuan County,Jiangxi Province.Among them,34participants were reexamined as non hypertensive patients.Demographic information and fasting venous blood samples of all participants were collected.5232participants were randomly selected.94 participants were excluded because of TG≥500 mg/dl(5.64 mmol/L),176 participants were excluded because of taking lipid-lowering drugs,and 195 participants were excluded because of ECG or history of atrial fibrillation.A total of 4767 participants were included in the final analysis.General information(age,gender,education,etc.),lifestyle(smoking,drinking,labor intensity,etc.),disease history(diabetes,coronary heart disease,stroke,etc.)and drug use(antihypertensive medications,hypoglycemic medicantions,lipid-lowering medicantions,antiplatelet medications,etc.)were collected by questionnaire.Physical examination included height,weight,blood pressure,pulse rate and waist circumference.The blood samples were sent to the automatic clinical analyzer(Beckman Coulter AU680,USA)of the laboratory of Guangzhou National Research Center of clinical kidney disease to detect total triglyceride(TG),total cholesterol(TC),high density lipoprotein cholesterol(HDL-C)and low density lipoprotein cholesterol(LDL-C),LDL-C,fasting blood glucose(FBG),serum uric acid(UA),serum creatinine(Scr),blood urea nitrogen(BUN)and plasma total homocysteine(Hcy)were measured.Bp-203 RPEⅢarteriosclerosis detector(Omron Health care,Kyoto,Japan)was used to detect baPWV and ABI on the day of blood collection.The data of population characteristics were expressed as mean±standard deviation(SD)of continuous variables,and the categorical variables were expressed as frequency(%).Multiple linear regression model(βcoefficient and 95%confidence interval)and multiple logistic regression model(odds ratio and 95%CI)were used to evaluate the relationship between AIP and baPWV,increased baPWV and ABI,low ABI,and the five main covariates were adjusted.Model 1:adjusted according to gender and age;Model 2:adjusted according to gender,age,body mass index(BMI),waist circumference,duration of hypertension,systolic blood pressure,diastolic blood pressure and heart rate;Model 3:adjusted according to gender,age,BMI,waist circumference,duration of hypertension,systolic blood pressure,diastolic blood pressure,heart rate,diabetes history,self-reported history of coronary heart disease and stroke,labor intensity,taking antihypertensive medications and taking antiplatelet medications;Model 4:adjusted according to gender,age,BMI,waist circumference,duration of hypertension,systolic blood pressure,diastolic blood pressure,heart rate,diabetes history,self-reported history of coronary heart disease and stroke,labor intensity,taking antihypertensive medications,taking antiplatelet medications,smoking and drink status;Model 5:adjusted according to model 4+TC,LDL-C,estimated glomerular filtration rate(e GFR),UA and Hcy levels were adjusted.In the regression analysis model,according to the clinical importance of the variables,the statistical significance in univariate analysis(P<0.1)and the possible influence on the estimated value,the change was more than 10%.The dose-response relationship of AIP with baPWV and ABI was studied by using generalized additive model(GAM)and fitting smooth curve(penalty spline method).In addition,the possible correlation between AIP and baPWV and ABI was evaluated by hierarchical analysis and interaction test.All data statistical analysis and table making use of software statistical package R and empowerstats statistical analysis.Results:1.A total of 4767 patients with hypertension were included,including 2388males 2379(50.1%)and female patients(49.9%),aged 29-93 years(mean 64.60±9.50 years),832(17.5%)patients with diabetes history,316(6.6%)patients with self-reported history of stroke,287(6.0%)patients with self-reported history of coronary heart disease,2877(60.3%)patients taking antihypertensive medications,190(3.98%)patients taking hypoglycemic medications and 99 patients taking antiplatelet medications(2.1%);among them,AIP was-0.00±0.3,baPWV was 18.5±4.2 m/s,ABI was 1.1±0.1,BMI was 23.2±3.5 kg/m~2,waist circumference was82.1±9.5 cm,systolic blood pressure was 147.2±17.6 mm Hg,diastolic blood pressure was 88.6±10.9 mm Hg,heart rate was 75.4±14.3 bpm,Hcy level was 18.6±11.7μmol/L,FBG level was 6.1±1.5 mmol/L,TC level was 5.2±1.1 mmol,TG level was 1.7±0.9 4 mmol/L,LDL-C level was 3.0±0.8 mmol/L,UA level was 431.0±120.7 umol/L,e GFR level was 86.1±19.7 ml/min/1.73 m~2.The basic characteristics of the study population were described according to baPWV quartile(quartile 1:<15.68 m/s;quartile 2:≥15.68,<17.88 m/s;quartile3:≥17.88,<20.60 m/s;quartile 4:≥20.60 m/s).Compared with quartile 1,quartile 2 and quartile 3,the participants in Q4 group(≥20.60 m/s)were more female(54.5%),older(70.2±8.4 years),longer duration of hypertension[6.0(2.0-11.0)years],systolic blood pressure(157.1±18.5 mm Hg),diastolic blood pressure(90.4±11.8 mm Hg)and heart rate(79.8±11.5 bpm),previous smoke(20.2%),mild labor intensity(61.7%),diabetes history(22.1%),FBG(6.3±1.7mmol/L),Hcy[16.3±1.7(13.2-21.1)μmol/l]and the proportion of taking hypoglycemic medications(5.4%)were higher,while the proportion of male participants(45.5%),current smoke(25.7%),current drink(22.4%),moderate labor intensity(19.1%),severe labor intensity(19.2%),BMI(22.4±3.4 kg/m~2),waist circumference(80.9±9.5 cm),LDL-C level was(2.8±0.8 mmol/L),e GFR level was(80.0±20.1 ml/min/1.73 m~2)were lower.The differences were statistically significant(all P<0.05).According to ABI quartile group(quartile 1:<1.08;quartile 2:≥1.08,<1.13;quartile 3:≥1.13,<1.18;quartile 4:≥1.18),the basic characteristics of the study population were described.Compared with quartile 2,quartile 3 and quartile 4groups,quartile 1 group had more patients with mild labor intensity(61.3%),faster heart rate(77.4±15.5 beats/min),waist circumference(82.8±9.9cm),Hcy[15.7(12.8-20.2)μmol/l]and self-reported stroke prevalence(8.1%),while SBP(145.3±18.5 mm Hg),e GFR(84.4±21.7 ml/min/1.73 m~2),moderate labor intensity(19.6%)and severe labor intensity(19.1%)were lower(all P<0.05).2.Univariate analysis showed that age,gender,systolic blood pressure,diastolic blood pressure,heart rate,diabetes history,self-reported history of coronary heart disease,TC,HDL-C,FBG,Hcy were positively correlated with baPWV.BMI,waist circumference,moderate labor intensity,severe labor intensity,current drinking,current smoking and e GFR were negatively correlated with baPWV.3.Univariate analysis showed that systolic blood pressure,diastolic pressure,previous smoking,previous drinking,current drinking,moderate labor intensity,severe labor intensity,HDL-C,e GFR were positively correlated with ABI.Age,sex,heart rate,self reported history of stroke,self-reported history of coronary heart disease,TC,TG,LDL-C,AIP,Hcy were negatively correlated with ABI.4.The dose dependence of AIP and baPWV was evaluated by using the generalized additive model(GAM)and fitting smooth curve(penalty spline method)by fully adjusting the model.It was observed that AIP and baPWV had a linear positive correlation.According to the potential covariates,the multivariate linear regression analysis models of AIP and baPWV were constructed,After adjustment for gender,age,BMI,waist circumference,duration of hypertension,systolic blood pressure,diastolic blood pressure,heart rate,diabetes history,self-reported coronary heart disease and stroke,labor intensity,taking antihypertensive medications,taking antiplatelet medications,smoke status,drink status,TC,LDL-C,e GFR,UA and Hcy,the results showed that there was a significant independent positive correlation between AIP and baPWVβ=1.44(95%CI:1.03,1.86).For the purpose of sensitivity analysis,we also treated the AIP quartile as a categorical variable and observed a significant trend in Model 1-5(P<0.001).5.The dose-dependent relationship between AIP and ABI was explored by using the fully adjusted generalized additive model(GAM)and the fitting smooth curve(penalty spline method).It was observed that AIP and ABI had a linear negative correlation.According to the potential covariates,different linear regression analysis models of AIP and ABI were constructed.After adjustment for gender,age,BMI,waist circumference,duration of hypertension,systolic blood pressure,diastolic blood pressure,heart rate,diabetes,self-reported coronary heart disease and stroke,labor intensity,taking antihypertensive medications,taking antiplatelet medications,smoke status,drink status,TC,HDL-C,e GFR,UA and Hcy,AIP was negatively correlated with ABI[β=-0.014(95%CI:-0.027,-0.001),P=0.032].For the purpose of sensitivity analysis,we also treated the AIP quartile as a categorical variable,and found no correlation in the fully adjusted model(P>0.05).6.The dose-dependent relationship between AIP and increased baPWV was evaluated by using the generalized additive model(GAM)and the smoothing curve fitting(penalty spline method).AIP was associated with increased baPWV(Q4:≥20.60),Multivariate logistic regression analysis showed that:after adjustment for gender,age,BMI,waist circumference,duration of hypertension,systolic blood pressure,diastolic blood pressure,heart rate,diabetes,self-reported coronary heart disease and stroke,labor intensity,taking antihypertensive medications,taking antiplatelet medications,smoke status,drink status,TC,LDL-C,e GFR,UA,Hcy levels,the OR with increased baPWV was 3.68(95%CI:2.54,5.35).Further trend test showed that significant linear trend was observed in model 1-5 after AIP quartile grouping(P<0.001).In the fully adjusted model,the OR values of Q2-Q4 group of AIP were 1.35(95%CI:1.06,1.72),1.95(95%CI:1.50,2.53),2.27(95%CI:1.71,3.02).The trend test was statistically significant(P<0.001).7.The dose-dependent relationship between AIP and low ABI was explored by using fully adjusted generalized additive model(GAM)and fitting smooth curve(penalty spline method).It was observed that AIP was positively correlated with low ABI.Multivariate logistic regression analysis of AIP and low ABI(Q1:<1.08 group was defined as low ABI population),After adjustment for gender,age,BMI,waist circumference,duration of hypertension,systolic blood pressure,diastolic blood pressure,heart rate,diabetes,self-reported coronary heart disease and stroke,labor intensity,taking antihypertensive medications,taking antiplatelet medications,smoke status,drink status,TC,LDL-C,e GFR,UA,Hcy,the OR of low ABI was1.30(95%CI:0.96,1.78),P=0.093.Further trend test,after grouping AIP quartile,no correlation was observed in the complete model(P>0.05).8.The results of stratified analysis and interaction test between AIP and ABI subgroup showed there was no significant interaction among the following subgroups,including age,gender,BMI,systolic blood pressure,diastolic blood pressure,e GFR,Hcy,self-reported coronary heart disease,self-reported stroke,smoke status,drink status and labor intensity,taking antihypertensive medications and antiplatelet medications.In addition,the correlation between AIP and baPWV was more significant in the participants with diabetes[β=2.16(95%CI:1.14,3.18),P<0.001](P for interaction=0.015).9.The results of stratified analysis and interaction test between AIP and ABI subgroup showed that the interaction test of BMI(≥24kg/m~2 vs.<24kg/m~2;P for interaction=0.013)was significant,and the correlation between AIP and ABI was more significant in hypertensive patients with BMI<24kg/m~2[β=-0.03(95%CI:-0.05,-0.01).There was no significant interaction among the following subgroups,including age,gender,systolic blood pressure,diastolic blood pressure,e GFR,homocysteine,diabetes,self-reported coronary heart disease,self-reported stroke,smoke status,drink status,labor intensity,taking antihypertensive medications,taking antiplatelet medications,Conclusion:1.In Chinese hypertensive population,AIP has independent and significant correlation with baPWV and increased baPWV.baPWV increases with the increase of AIP,that is,arterial stiffness increases with the increase of AIP.The correlation between AIP and baPWV was more significant in the participants with diabetes.2.In the fully adjustment model with further adjustment of blood biochemical parameters,there was no significant correlation between AIP and ABI or low ABI.there is a stronger correlation between AIP and ABI in hypertensive patients with BMI<24kg/m~2.3.AIP is different in reflecting functional atherosclerosis and structural atherosclerosis,and it can better reflect functional atherosclerosis in Chinese hypertensive population.