Background: The use of serum fatty acid biomarkers in nutritional epidemiology is increasingly common; however, there is an absence of scientific evidence to substantiate whether the measurement of fatty acids as molecular percentages (which is the conventional approach) or as absolute concentrations is more informative.
Objective: To advance understanding about this fundamental concept, we examined the ability of serum myristic acid and linoleic acid, expressed as molecular percentages or as concentrations, to predict dietary fat and serum cholesterol concentrations.
Design: A cross-sectional analysis of a population-based survey of New Zealand adults (n = 2732) was undertaken. The association of myristic and linoleic acids in serum cholesterol ester and phospholipid with dietary fat or serum cholesterol was assessed.
Results: Intake of saturated fat, dairy fat, and polyunsaturated fat was predicted similarly with the use of both units of measurement. After adjustment for confounders, mean total cholesterol decreased by 0.18 mmol/L from the lowest to the highest quintile of serum cholesteryl-linoleate as a molecular percentage (P = 0.027). In contrast, mean total cholesterol increased by 1.09 mmol/L from the lowest to the highest quintile of serum cholesteryl-linoleate concentration (P < 0.001). The molecular percentage and concentration of serum cholesteryl-myristate were positively associated with total cholesterol (P < 0.001). Results for serum phospholipid fatty acids were similar.
Conclusion: Serum myristic acid and linoleic acid measured as molecular percentages, but not as concentrations, predict serum total cholesterol in a manner that distinguishes between the differential cholesterolemic effects of dietary saturated and polyunsaturated fats.