Background: Fatty acid (FA) composition in different blood compartments is used as biomarkers of dietary fat quality, but it is unclear how FA compositions differ between compartments and which biomarker is most appropriate for epidemiological studies. Furthermore, as most cohorts use single FA measurements to reflect long-term intake, reproducibility data of plasma FAs are clearly warranted.
Objective: To investigate inter-correlations of FA compositions in plasma phospholipids (PP) and cholesterol esters (CE), and to evaluate 10-y reproducibility of FA composition in plasma CE.
Methods: Baseline FA composition of PP and CE was assessed in a population-based cohort of 70-year-old Swedes (n=940), and in a subsample (n=539), FA was additionally measured in CE during follow-up at ages 75 and 80 y. FA in PP and CE was measured by gas chromatography. Spearman’s rank correlation coefficients were calculated between proportions of individual FA measured in plasma PP and CE collected at age 70 y. Plasma CE compositions at ages 70, 75, and 80 y were utilized to assess 10-year reproducibility by calculating intraclass correlation coefficients (ICC) in one-way random-effect models.
Results: Strong correlations between PP and CC were observed for most FA, with the only exceptions palmitic acid and stearic acid (Table). Fair to good reproducibility was observed for most FA, and was greatest for palmitoleic acid (ICC [95% CI] = 0.72 [0.69-0.76]) and all n-6 polyunsaturated FA (ICC=0.64-0.68). Poor reproducibility was only observed for α-linoleic acid, pentadecanoic acid, and stearic acid, with ICC (95% CI) 0.19 (0.14-0.25), 0.22 (0.17-0.28), and 0.23 (0.18-0.29), respectively.
Conclusions: Overall, FA levels measured in plasma PP and CE correlates strongly, suggesting that FA composition in these compartments generally can be used interchangeably as biomarkers in epidemiological studies. Overall good reproducibility of FA levels in CE indicates that single a measurement can reliably estimate average levels over a 10-y period.