Abstract
BACKGROUND: Global dietary guidelines for polyunsaturated fatty acids (PUFAs), especially linoleic acid (LA) and its metabolite arachidonic acid (ARA), remain debated. Almost all research to date has used fatty acid (FA) data expressed as percent of total FA (% total). OBJECTIVE: The objective of this study was to determine whether expressing fatty acid (FA) data as % of total or as absolute concentrations alters associations with clinical biomarkers. METHODS: Serum FA data obtained via electron capture negative-ion mass spectrometry was obtained from NHANES. Each FA was expressed both as % total and absolute concentration (µmol/L). Associations were examined between individual and total FAs and a panel of lipid and non-lipid biomarkers, including total cholesterol, LDL-C, HDL-C, triglycerides, blood pressure, body mass index, waist circumference, glucose, and insulin. RESULTS: Associations between LA and clinical biomarkers including triglycerides, cholesterol, HDL-C, BMI, glucose, and insulin, reversed direction depending on whether LA was expressed as % total or as a concentration. Similar reversals were observed for ARA, DHA, DPA, and stearic acid. Increases in total FA levels were accompanied by decreases in % total of several PUFAs and HUFAs, despite rising absolute concentrations. Total FA was positively associated, often strongly, with nine clinical markers and negatively associated with HDL-C. CONCLUSIONS: Expression format significantly impacts observed FA associations. Reliance on % total FA values alone may misrepresent true associations between individual FAs and clinical endpoints, especially when the total fatty acid pool also changes size. To develop effective dietary guidance or clinical recommendations, it is essential to consider the underlying FA biology and total FA pool size when determining whether % total or absolute FA concentrations are more appropriate.