Application of the Hegsted Equation to Low Carbohydrate‐Low Fat Diet Comparisons

Abstract

The Keys and Hegsted equations continue to be used to predict serum total cholesterol (TC) or LDL‐C based on dietary saturated fat (SFA), polyunsaturated fat (PUFA) and cholesterol (Chol). The equations frequently give good fit but are limited in that they are primarily descriptive without an underlying mechanism. As predictors of CVD risk, TC and LDL‐C are also not optimal: LDL particle size (and its surrogate TG/HDL‐C) are generally better. Carbohydrate restricted diets (CRD) effectively reduce the more atherogenic small LDL, due to the role of insulin in lipid metabolism. Because the Hegsted equations are still used, we evaluated their predictive value in subjects who consumed a CRD (n = 16: 1504 kcal, %CHO:fat:protein=12:59:28) and a low‐fat diet (n = 18: 1478 kcal,%CHO:fat:protein=56:24:20). The actual change in fasting serum lipid values after 12 weeks of diet was compared to predicted values using recent versions of the Hegsted equation:[ΔTC = 2.10 ΔSFA − 1.16 ΔPUFA + 0.067 ΔChol and LDL‐C = 1.74 ΔSFA − 0.766 ΔPUFA + 0.0439 ΔChol]TC and LDL‐C were weakly predicted for low‐fat diets (R = 0.22, 0.10) but not CRD (R = −0.12, 0.01). In distinction, we found that dietary carbohydrate was a good predictor of LDL particle size or TG/HDL‐C (R = 0.52, 0.51) and may be more generally useful in the assessment of the influence of diet on CVD risk than those indicators, like the Hegsted equations, based primarily on lipid components.