Associations of dietary macronutrient composition with cardiometabolic health: data from the National Health and Nutrition Examination Survey (NHANES) 1999–2014

Abstract Background Human nutrition is a leading modifiable risk factor for the prevention of cardiometabolic diseases. However, most nutritional research focuses on the role of specific nutrients rather than compositional analysis of the diet. Purpose We aimed to assess the association of dietary macronutrient composition with various markers of cardiometabolic health. Methods This study was undertaken using 24-hour food recall data from the National Health and Nutrition Examination Survey from years 1999–2014 (n=36,637). Participants included both males and females ≥20 years of age (49.8±18.3 years; 53% Female). Caloric outliers for males (<800 kcal/day or >4200 kcal/day) and females (<600 kcal/day or >3500 kcal/day) were removed from the analysis (n=1844). Cardiometabolic health markers were collected via mobile examination center and included triglycerides (n=16,444), total cholesterol (n=31,155), LDL cholesterol (n=15,798), HDL cholesterol (n=33,435), blood pressure (n=34,121), body fat percentage (n=4269), body mass index (n=35,187), glucose (n=17,208), insulin (n=16,273), oral glucose tolerance test (n=8280), and HbA1c (n=27,288). Associations of dietary macronutrients with cardiometabolic health markers were assessed using 3-dimensional general additive models (Protein; Carbohydrate; Fat) with adjustment for age, sex, misreporting, ethnicity, education level, household income, physical activity, alcohol intake, and smoking. Response surfaces on right-angled mixture triangles were generated as predictions from absolute macronutrient intake at the 50th percentile of total energy intake. Macronutrient intakes were transformed to a percentage of energy to visualize and interpret associations compositionally. Results There was a complex 3-way interactive non-linear association of macronutrient composition with lipid profile (Figure A-D; all p<0.04), blood pressure (Figure E-F; all p<0.001), body fat percentage (Figure G; p<0.001), body mass index (Figure H; p<0.001), and all markers of glycemic control (Figure I-L; all p<0.001). Diets comprised of high protein coupled with low fat were associated with higher triglycerides but generally more favorable total cholesterol, blood pressure, and body fatness. Conversely, diets composed of high protein coupled with high fat were associated with higher HDL cholesterol but worse total cholesterol, body fatness, body mass index, and glycemic control. Conclusions These findings display a complex relationship between dietary macronutrient composition and cardiometabolic health. Future research is needed to evaluate how dietary source and quality may modify the observed associations. Funding Acknowledgement Type of funding sources: Private company.