Dietary characteristics associated with plasma concentrations of per- and polyfluoroalkyl substances among adults with pre-diabetes: Cross-sectional results from the Diabetes Prevention Program Trial

Abstract

Diet is assumed to be the main source of exposure to per- and polyfluoroalkyl substances (PFAS) in non-occupationally exposed populations, but studies on the diet-PFAS relationship in the United States are scarce. We extracted multiple dietary variables, including daily intakes of food group, diet scores, and dietary patterns, from self-reported dietary data collected at baseline (1996-1999) from adults with pre-diabetes enrolled in the Diabetes Prevention Program, and used linear regression models to evaluate relationships of each dietary variable with plasma concentrations of six PFAS (perfluorooctane sulfonic acid (PFOS), perfluorooctanoic acid (PFOA), perfluorohexane sulfonic acid (PFHxS), 2-(N-ethyl-perfluorooctane sulfonamido) acetic acid (EtFOSAA), 2-(N-methyl-perfluorooctane sulfonamido) acetic acid (MeFOSAA), perfluorononanoic acid (PFNA) adjusting for covariates. Participants (N = 941, 65% female, 58% Caucasian, 68% married, 75% with higher education, 95% nonsmoker) had similar PFAS concentrations compared to the general U.S. population during 1999-2000. Using a single food group approach, fried fish, other fish/shellfish, meat and poultry had positive associations with most PFAS plasma concentrations. The strongest effect estimate detected was between fried fish and PFNA [13.6% (95% CI: 7.7, 19.9) increase in median concentration per SD increase]. Low-carbohydrate and high protein diet score had positive association with plasma PFHxS. Some food groups, mostly vegetables and fruits, and the Dietary Approaches to Stop Hypertension diet score had inverse associations with PFOS and MeFOSAA. A vegetable diet pattern was associated with lower plasma concentrations of MeFOSAA, while high-fat meat and low-fiber and high-fat grains diet patterns were associated with higher plasma concentrations of PFOS, PFHxS, MeFOSAA and PFNA. We summarized four major dietary characteristics associated with variations in PFAS plasma concentrations in this population. Specifically, consuming more meat/fish/shellfish (especially fried fish, and excluding Omega3-rich fish), low-fiber and high-fat bread/cereal/rice/pasta, and coffee/tea was associated with higher plasma concentrations while dietary patterns of vegetables, fruits and Omega-3 rich fish were associated with lower plasma concentrations of some PFAS.

Keywords: Diet; Diet score; Dietary pattern; Food intake; Per- and polyfluoroalkyl substances; Prediabetic adults.

Figure 1.

Feature expression heatmap on associations of 27 food groups and 6 pyramid food groups (G1-G6) with PFAS plasma concentrations (N=941). Note: PFOS: Perfluorooctane sulfonic acid [sum of linear and branched isomers]; PFOA: perfluorooctanoic acid [sum of linear and branched isomers]; PFHxS: perfluorohexane sulfonic acid; EtFOSAA: N-ethyl-perfluorooctane sulfonamido acetic acid; MeFOSAA: N-methyl-perfluorooctane sulfonamido acetic acid; PFNA: perfluorononanoic acid. M1 represents univariate linear regression model with food group as the independent variable and PFAS plasma concentration as the dependent variable. M2 represents multivariate linear regression model adjusting for sex, age, marital status, education, income, waist circumference, and smoking status. M3 represents M2 with additional adjustment for daily caloric intake. Effect size represents change in PFAS plasma concentration (in standardized z-score) per standard deviation increase in daily intake and are depicted by color intensity (blue for negative effect and red for positive effect). Size of the circles indicates statistical significance (p-value) and circles with a dot in the middle represent significance below the 0.2 false discovery rate (FDR).

Figure 2.

Associations between dietary patterns and PFAS plasma concentrations among prediabetic adults in the Diabetes Prevention Program (N=941). (A) Loading values for factor score of dietary patterns derived using Principal Component Analysis; (B) Relative percent change in median PFAS plasma concentrations per standard deviation increase in factor score. Note: PFOS: Perfluorooctane sulfonic acid [sum of linear and branched isomers]; PFOA: perfluorooctanoic acid [sum of linear and branched isomers]; PFHxS: perfluorohexane sulfonic acid; EtFOSAA: N-ethyl-perfluorooctane sulfonamido acetic acid; MeFOSAA: N-methyl-perfluorooctane sulfonamido acetic acid; PFNA: perfluorononanoic acid. The top 10 PCs explained 60.8% in the variances of baseline self-reported intakes. Loading value were multiplied by 100 and rounded to the nearest integer. Values greater than 40 are colored in dark blue. Effect estimates statically differed from 0 were marked with * (p < 0.05).