Fig. 3

Weighted logistic regression analysis between dietary fiber density and CVD risk in age and gender subgroups. Model 1, weighted univariate multinomial logistic regression model; Model 2, weighted multinomial multivariate logistic regression model that adjusted for age/gender, family income, education levels, and marital status; Model 3, weighted multinomial multivariate logistic regression model that adjusted for age/gender, family income, education levels, marital status, total bilirubin, creatinine, and metabolic syndrome