DataSheet_1_Analysis of the mediating role of BMI in associations of different folate forms with hepatic steatosis and liver fibrosis in adolescents in the USA: results from the NHANES 2017-2018.docx

BackgroundMost studies have explored the relationship between serum total folate and nonalcoholic fatty liver disease (NAFLD) in adults, but there has been no study on the relationship between different folate forms and hepatic steatosis or liver stiffness in adolescents.ObjectiveTo investigate the association of different folate forms with hepatic steatosis or liver stiffness in adolescents, and further explore the intermediary role of BMI in this relationship.MethodsThe cross-sectional study included 549 participants from the 2017-2018 National Health and Nutrition Inspection Survey (NHANES) survey cycle who had complete data. Four folate data (red blood cell folate, serum total folate, 5-methyl-tetrahydrofolate and folic acid) were included in our study. Controlled attenuation parameters (CAP) and liver stiffness came from the results of liver ultrasound transient elastography. We used linear regression to analyze the relationship between different forms of folate and CAP or liver stiffness, and logistic regression to analyze the relationship between different forms of folate and NAFLD or significant fibrosis. We also used restricted cubic splines to analyze the nonlinear relationship between different forms of folate and NAFLD or significant fibrosis. Finally, we used regression-based intermediary analysis to distinguish the direct and BMI-mediated effects of folate on CAP or liver stiffness. All the analyses adjusted the relevant covariates.ResultsThe means of CAP and liver hardness in this study were 223.02dB/m and 5.03kPa, respectively. We found that in model 2, there was a negative correlation between serum total folate (β: -18.53; 95%CI: -29.32 to -7.73) or 5-methyltetrahydrofolate (β: -14.13; 95%CI: -28.98 to -7.86) and CAP. However, when the BMI was further adjusted in model 3, this negative correlation no longer existed (serum total folate: β: -8.36; 95%CI: -17.69 to 0.97; 5-methyltetrahydrofolate: β: -8.05; 95%CI: -17.19 to 1.09). Similarly, we found a negative correlation between serum total folate or 5-Methyl-tetrahydrofolate and liver stiffness in model 2. There was no significant correlation between red blood cell folate or folic acid and CAP or liver stiffness in either model 2 or model 3. The nonlinear relationship between different folate forms and NAFLD or significant fibrosis was not significant. It is estimated that 76% of the total association between serum total folate and CAP is mediated by BMI. The mediating proportion of BMI in the total correlation between serum total folate and liver stiffness was 50%. Similarly, we found that BMI significantly mediated the relationship between 5-Methyl-tetrahydrofolate and CAP or liver stiffness, with a mediating ratio of 77% and 49%, respectively.ConclusionOur results show that serum total folate or 5-Methyl-tetrahydrofolate are negatively correlated with hepatic steatosis or liver stiffness in adolescents, and BMI plays major mediating role in this relationship. Our findings emphasize the importance of monitoring the concentration of serum folate, not just the serum total folate concentration.