| Background:With the rapid development of the economy,China has experienced nutrition transition in the past decades.At the same time,the prevalence of metabolic syndrome(MetS)among Chinese adults were rising rapidly in the most recent decade.The traditional diet-disease focused on the relationship between singel nutrient,food or food groups and disease risk,which have certain limitations.However,dietary pattern analysis examines the effects of overall diet quality on health outcome.Conceptually,dietary patterns reflect interaction of foods and nutrients and thus may be more effective to shed light on the impact of diet on health outcomes.At present,exploratory factor analysis or principal component analysis(PCA)is prior method which is used to study the association between dietary patterns and MetS.Additionally,most of these studies were cross-sectional properties.Thus,the relationship between dietary patterns and MetS lack of time-causal evidence and the profiles of dietary pattern have a low reproducibility.The correlation analysis method of dietary pattern and MetS is mainly based on single level multifactor Logistic regression,which cannot effectively distinguish the function of different levels of factors.Moreover,there is no report about the association between dietary patterns and MetS prevalence using reduced rank regression(RRR).Objectives:To examine the relationship between dietary patterns and MetS in Chinese adult from two aspects:horizontal and vertical,based on the mixed effect model of PCA and RRR.Methods:The study was based on the data from the China Health and Nutrition Survey(CHNS),an ongoing longitudinal cohort study.According to the aims of the present study,study subjects were divided into three groups.The first group included those adults aged 18 to 75 years(N=3,299 men and 3,703 women)who participated in the 2009 wave of survey.PCA and RRR were performed to identify sex-specific dietary patterns,their nutrient profiles and potential risk factors for each dietary pattern were investigated further.Meanwhile,the association between dietary patterns and MetS prevelance and its components were analyzed.The second group included the adults aged 18 to 75 years(N=2,035 men and 2,458 women)who had completed dietary and anthropometric data from both the 2006 and 2009 survey years.Two-level mixed effect model was conducted to examine the association of different dietary patterns with the prevalence of MetS and its components in the adults.The third group included the subjects aged 18 to 75 years(N=1,778 men and 2,183 women)both in the cross-sectional and longitudinal studies.The stability and change of sex-specific dietary patterns of adults in a certain period were investigated.The prevalence rate of MetS was calculated by using two criteria defined by Chinese Diabetes Society(CDS2013)and modified National Cholesterol Education Program,Adult Treatment Panel III(NCEP-ATPIII 2005)criteria.For abdominal obesity,we used the International Obesity task Force central obesity criteria for Asia,which defined abdominal obesity as a waist circumference of at least 90 cm for men and at least 80 cm for women.The analyzed variables mainly included individual level variables(food and nutrient intake,dietary pattern factor score,age,level of education,smoking,alcohol consumption,physical activity levels,income level)and community level variables(region,regional,urbanization index).The nutrients were calculated from consecutive 72 hours recall questionnaire.and oil and condiment intake was obtained from consecutive 72 hours weighting.Results:1.There are 3 common PC A dietary patterns in male and female adults in 2009.The traditional southern pattern was rich in rice and its products,dark vegetables,pork,fish,and poultry meat.The western pattern was rich in refined vegetable oil,nuts,fruits,eggs,fast food,dairy,other livestock meat,sweets and snacks,fish and shrimp,high-energy high-sodium vegetarian pattern was rich in refined vegetable oil,salt,salt seasoning and light color vegetables.But a PC A3 pattern was identified specially among the men that was rich in fast food,wine,refining animal oil,nuts and animal innards,which was named animal protein pattern.2.One main dietary pattern was extracted using RRR among the men and women respectively.The main composition of MetS pattern in the men was wheat and its products,alcohol,fast food,eggs and refined vegetable oil,and the main composition of MetS pattern in the women was wheat and its products,potato and other grains.3.For men,the correlation coefficients between traditional southern pattern pattern,western pattern,animal protein pattern and MetS pattern in 2006 were-0.45(p<0.001),0.60(p<0.001)and 0.34(p<0.001),respectively.For women,the correlation coefficients between traditional southern pattern,western pattern,high-energy high-sodium vegetarian pattern and MetS pattern in 2006 were-0.71(p<0.001),0.06(p<0.01)and 0.40(p<0.001),respectively.For men,the correlation coefficients between traditional southern pattern,western pattern,high-energy high-sodium vegetarian pattern and MetS pattern in 2009 were-0.56(p<0.001),0.56(p<0.001)and 0.02(p>0.05),respectively.For women,the correlation coefficients between traditional southern pattern,western pattern,high-energy high-sodium vegetarian pattern and MetS pattern in 2009 were-0.72(p<0.001),0.11(p<0.01)and 0.44(p<0.001),respectively.4.For men,the correlation coefficients for traditional southern pattern,western pattern,PCA3 pattern and MetS pattern between 2006 and 2009 were 0.69(p<0.001),0.43(p<0.001),-0.15(p<0.001)and 0.53(p<0.001),respectively.For women,the correlation coefficients for traditional southern pattern,western pattern,high-energy high-sodium vegetarian pattern and MetS pattern between 2006 and 2009 were 0.69(p<0.001),0.48(p<0.001),0.28(p<0.001)and 0.51(p<0.001),respectively.5.In 2006,two-level mixed-effect linear regression model indicated that community-level variation accounts for 58.2%,36.4%,21.8%and 32.8%of total score variation of traditional southern,western,animal protein and MetS dietary patterns among the men respectively,whereas community-level variation explained 61.8%,29.9%,29.7%and 42.2%of each dietary pattern among the women,respectively.In 2009,community-level variation accounts for 51.8%,24.6%,20.8%and 25.4%of total score variation of traditional southern,western,high-energy high-sodium vegetarian and MetS dietary patterns among men respectively,whereas community-level variation explained 52.3%,27.4%,21.0%and 25.7%of each dietary pattern among the women,respectively.Therefore,community-level factor was the major part for score variation of the traditional southern dietary pattern.Individual-level factors in relation to dietary patterns included age,income level,physical activity,smoking,and alcohol consumption.6.For men the traditional southern pattern was negatively associated with waist circumference(p<0.001),western pattern was positively associated with waist circumference(p<0.05)and negatively associated with HDL-C(p<0.05).high-energy high-sodium vegetarian pattern was negatively associated with diastolic pressure(p<0.05),and MetS pattern was positively associated waist circumference(p<0.001).For women western pattern was negatively associated with diastolic pressure(p<0.05),high-energy high-sodium vegetarian pattern was positively associated with systolic blood pressure(p<0.05)and diastolic blood pressure(p<0.05),and MetS pattern was positively associated with waist circumference(p<0.001),systolic blood pressure(p p<0.01)and diastolic blood pressure(p<0.05).7.In 2009,men in the top quintile of the traditional southern pattern had 1.47 times higher risk of high TG compared to those in the lowest quintile.Men in the third quintile of the high-energy high-sodium vegetarian pattern had 1.40 times higher risk of MetS compared to the lowest quintile.Men in the fourth quintile of the high-energy high-sodium vegetarian pattern was 71 percent higher risk of elevated fasting glucose compared to the lowest quintile.For women in the fourth quintile of the western pattern had 1.37 times higher risk of MetS compared to the lowest quintile.Women in the top quintile of the high-energy high-sodium vegetarian pattern had 1.33 times,1.33 times,1.52 times and 1.33 times higher risk of MetS,central obesity,elevated blood pressure and low HDL-C compared to the lowest quintile,respectively.Men in the top quintile of the MetS pattern had 1.28 times,1.45 times,1.46 times and 57 percent higher risk of MetS,central obesity,elevated blood pressure and low HDL-C compared to the lowest quintile,respectively.Women in the top quintile of the MetS pattern had 1.50 times,1.48 times,1.51 times and 1.86 times higher risk of MetS,central obesity,elevated blood pressure and low HDL-C compared to those in the lowest quintile,respectively.The women in the fourth quintile of the MetS pattern had 1.38 times higher risk of elevated fasting glucose compared to those in the lowest quintile.8.In 2006,men in the top quintile of the traditional southern pattern was 59 percent risk of central obesity compared to those in the lowest quintile.Men in the top quintile of the western pattern had 1.79 times higher risk of central obesity compared to those in the lowest quintile.The men in the fourth quintile of the animal protein pattern was 73 percent lower risk of central obesity compared to those in the lowest quintile.Women in the third quintile of the traditional southern pattern had 1.46 times higher risk of high TG compared to those in the lowest quintile.Women in the top quintile of the western pattern had 1.50 times higher risk of high TG compared to those in the lowest quintile.Meanwhile,the fourth quintile of the western pattern had 1.48 times higher risk of MetS compared to those in the lowest quintile.Women in the second quintile of the high-energy high-sodium vegetarian pattern had 1.43 times higher risk of central obesity compared to those in the lowest quintile.Women in the top quintile of the high-energy high-sodium vegetarian pattern had 1.63 times higher risk of central obesity compared to those in the lowest quintile.The men in the top quintile of the MetS pattern had 1.74 times,2.31 times and 1.71 times higher risk of MetS,central obesity and elevated fasting glucose compared to those in the lowest quintile,respectively.Women in the top quintile of the MetS pattern had 2.00 times,and 2.26 times higher risk of MetS and central obesity compared to those in the lowest quintile,respectively.9.For men,MetS pattern(high intake of pasta and its products,wine,fast food,eggs and vegetable oil),on the one hand,was contrary to the traditional southern pattern(low intake of pasta and its products,dark vegetables,pork),on the other hand,combined the characteristics of the western pattern(high intake of alcohol,fast food,eggs and vegetable oil).For women,MetS pattern(high intake of flour and its products,other cereals and potatoes),on the one hand,was contrary to the traditional southern pattern(low intake of pasta and its products and potatoes),on the other hand,combined the characteristics of the traditional southern pattern.Conclusion:1.From 1991 to 2009,three main PC A dietary patterns were found among China adults.And those dietary patterns kept changing over time.traditional southern dietary pattern remained relatively stable compared to other patterns.Dietary patterns were influenced by community-level factor and individual-level factor.Due to the differences in energy,nutrients and components,the association of dietary patterns with MetS prevalence in the adults were varied,traditional southern pattern was beneficial to MetS prevalence and negatively related to WC and central obesity only for men,but it increased the risk of high TG in both genders.The western pattern was positively correlated with waist circumference and central obesity in the men and high TG in women,and was negatively associated with HDL-C in men and DBP in women,which may be a risk factor for MetS.western pattern was a strong risk factor for MetS in men.The high-energy high-sodium vegetarian pattern was positively correlated with DBP and SBP in women,which may be a risk factor for central obesity and elevated blood pressure in women and serve as a protective factor for elevated fasting glucose in men.The animal protein pattern was a strong risk factor for MetS both in men and women.2.Using the RRR method,one main dietary pattern was extracted from men and women respectively.The main composition of MetS pattern in men was wheat and its products,alcohol and fast food,eggs and refined vegetable oil,which was positively associated with waist circumference and serve as a risk factor for central obesity and elevated fasting glucose.The main composition of MetS pattern in women was wheat and its products,potato and other grains.which was positively associated with waist circumference,SBP and DBP and may serve as a risk factor for the prevalence of MetS,central obesity,elevated fasting glucose and low HDL-C.3.The RRR method can better assume what kind of food and dietary patterns associated with MetS,while the PC A method can explain the existence of groups defined by these dietary pattern,the combination of PCA and RRR method is conducive to examine the relationship between dietary patterns and MetS. |
PDF Full Text Request