| ObjectivesTo obtain specific dietary pattern by reduced rank regression analysis,to explore the relationship between this dietary pattern and dyslipidemia,to provide dietary data for the control of dyslipidemia.MethodsA total of 39259 adult subjects were recruited from the Henan Rural Cohort Study(ChiCTR-OOC-15006699),the data were collected by socio-demographic question-naire,dietary inquiry,physical measurement and biochemical examination.After excluding the subjects with missing dietary data,incomplete blood lipid index and missing physical measurement information,29538 rural residents aged 18-79 were included in this study.As changes in visceral adipose tissue may be associated with dyslipidemia,this study used reduced rank regression analysis to establish the abdominal obesity-dietary pattern with the visceral fat index as the dependent variable,and 12 z-transformed dietary food groups data including staple food,red meat,white meat,fish,eggs,milk and its products,fresh fruits and vegetables,beans and their products,nuts,miscellaneous grains and animal oil as independent variables.The differences among variables were analyzed by t test,variance analysis and chi-square test;binary logistic regression analysis and restricted cubic splines were to analyze the association between dietary pattern and dyslipidemia.Mediation analysis using bootstrap was performed to examine the contribution of visceral fat index to the dietary pattern and dyslipidemia.Quality control was carried out throughout the process.All analysis were conducted by SAS 9.1 and SPSS 25.0.Results1.A total of 29538 people were investigated in this study,with an average age of 55.44±12.35.The majority of the subjects in this study were women,accounting for 59.4%.The crude prevalence of dyslipidemia in this population was 34.2%,the prevalence of dyslipidemia was 35.5%in men and 33.4%in women.2.Abdominal obesity-dietary patterns were obtained by reduced rank regression,characterized by higher loading of refined grains,red meat,beans and miscellaneous grains,and lower factor loading of dairy products,fresh fruits,vegetables and other foods.3.According to current study,there was a positive correlation between abdominal obesity-dietary pattern and dyslipidemia(P<0.05);the highest quantile of dietary pattern scores group was associated with a significantly higher risk of dyslipidemia prevalence compared to the lowest quantile,with odds ratio(OR)(95%CI,confidence interval)of 1.328(1.228-1.437).Meanwhile,this dietary pattern was positively correlated with dyslipidemia components(P<0.05).4.There was a non-linear dose-response relationship between the score of abdominal obesity-dietary pattern and dyslipidemia(P for non-linear=0.021 5),and there was also a non-linear dose-response relationship between the score of this dietary pattern and visceral fat index(P for non-linear<0.001).5.In mediation analysis,the results of mediation analysis showed that visceral fat index partly mediated the relationship between abdominal obesity-dietary pattern and dyslipidemia with an indirect effect OR being 1.068(95%CI,1.058-1.077).The proportion explained of this dietary pattern was 53.26%for visceral fat index.Conclusion1.The abdominal obesity-dietary pattern in this study is characterized by high intakes of carbohydrate;low intakes of fresh fruits,vegetables,milk and dairy products.This current dietary pattern may be positively related to abdominal obesity.2.There is a dose-response relationship between the dietary pattern score and dyslipidemia,as well as increased visceral fat index.The dietary pattern is positively correlated with dyslipidemia and each component of dyslipidemia,after adjusting for the confounding factors.3.Abdominal obesity(visceral fat accumulation)has a partial mediation effect on association between this specific dietary pattern and dyslipidemia. |
PDF Full Text Request