Erratum: Association between erythrocyte membrane fatty acids and biomarkers of dyslipidemia in the EPIC-Potsdam study

Correction to: European Journal of Clinical Nutrition (2014) 68, 517–525; doi:10.1038/ejcn.2014.18; published online 26 February 2014

Since publication, the authors have found an SAS coding error in the LDL-cholesterol calculation. The other biomarker outcomes of the paper (HDL-cholesterol, non-HDL-cholesterol and triglycerides) are not affected by the error. The authors have corrected the error in the LDL-cholesterol calculation and have recalculated Tables including corrected LDL-cholesterol values.

Regarding Tables 2 and 3 and Online Supplementary Table 1 and 2, the statements in the results part of the paper are still correct for the corrected LDL-cholesterol.

Table 2 Characteristics by tertiles of HDL-cholesterol, LDL-cholesterol, non-HDL-cholesterol and triglycerides in men, EPIC-Potsdam study (n=655)^a

Table 3 Characteristics by tertiles of HDL cholesterol, LDL cholesterol, non-HDL cholesterol and triglycerides in women, EPIC-Potsdam study (n=1104)^a

Regarding the multivariable-adjusted models including LDL-cholesterol in Table 4, we previously reported a lack of associations between individual fatty acids and LDL-cholesterol concentrations and an inverse association between estimated D5D activity and LDL-cholesterol and positive associations between estimated SCD and D6D activity and LDL-cholesterol concentrations.

Table 4 Adjusted arithmetic means (95% CI) of plasma HDL-cholesterol, LDL-cholesterol, non-HDL-cholesterol and adjusted geometric means of plasma triglycerides (mg/dl) by tertiles of erythrocyte fatty acid proportions for men (n=655) and women (n=1104), EPIC-Potsdam study

There is still no significant association between individual fatty acids and the corrected LDL-cholesterol; however, no associations between estimated desaturase activities and LDL-cholesterol despite of a very weak positive trend between estimated D5D activity and the corrected LDL-cholesterol concentrations in women only and a very weak inverse trend of estimated SCD activity with the corrected LDL-cholesterol concentrations in both sexes. Taken together the updated findings, our main conclusion is that triglyceride concentrations may be mediators that link intake and metabolism of fatty acids to metabolic risk whereas HDL-cholesterol, LDL-cholesterol and non-HDL-cholesterol concentrations do not seem to play an important role in linking intake and metabolism of fatty acids to metabolic risk.

Incorrect parts in the Abstract, results and discussion of the manuscript are indicated in the following text.

Abstract section results: ‘Most notably, FA-ratios reflecting activity of Δ6-desaturase (D6D) and stearoyl-coenzyme A-desaturase (SCD) were positively associated with triglyceride and LDL-cholesterol concentrations....’

‘and LDL-cholesterol’ is incorrect.

Section conclusion: ‘Our findings suggest that triglyceride and LDL-cholesterol concentrations may be mediators that link intake and metabolism of FAs to metabolic risk.’

‘and LDL-cholesterol’ is incorrect.

Results section ‘Associations between erythrocyte FAs and biomarkers of dyslipidemia’:

‘In contrast, higher FA-ratios reflecting higher SCD and D6D activities were associated with higher triglyceride and LDL-cholesterol concentrations in both sexes.’

‘and LDL-cholesterol’ is incorrect.

‘These following results satisfy P-values for trend according to correction for multiple testing: associations of D6D and SCD with LDL-cholesterol and triglycerides and associations between D5D and triglycerides in both sexes.’

‘LDL-cholesterol and’ is incorrect.

In the first sentence of the discussion: ‘In this cross-sectional study of middle-aged men and women, D6D and SCD activities calculated from erythrocyte FA-ratios were positively associated with triglyceride and LDL-cholesterol concentrations.’

‘and LDL-cholesterol’ is incorrect.

In the first paragraph of the discussion: ‘Our results of strong associations of estimated desaturase activity with triglycerides and LDL-cholesterol do suggest an involvement of an unfavorable lipid profile in this context.’

‘and LDL-cholesterol’ is incorrect.

In the last sentence of the discussion: ‘Our findings suggest that triglyceride and LDL-cholesterol concentrations may be mediators linking the intake and metabolism of FAs to metabolic risk.’

‘and LDL-cholesterol’ is incorrect.