Vitamin E Serum Levels and the Challenge to Correct for Lipids: Accounting for the Usual Double Correction for Variance Shared by Total Cholesterol and Fasting Triglycerides Reveals New Insights into the Association with the One-Carbon Pathway


Vitamin E has important antioxidant properties, and low vitamin E status is therefore implicated in accelerated development of age-related diseases, including type 2 diabetes and atherosclerosis [7, 13, 23]. Importantly, both these conditions are characterized by high circulating concentrations of fasting triglycerides and total cholesterol, which may complicate evaluation of the association of vitamin E status with these conditions. The reason is that in the circulation, vitamin E is carried by lipoproteins, which usually results in a positive correlation of circulating concentrations of vitamin E with those of fasting cholesterol and fasting triglycerides and the need to one way or another correct for this [32, 33]. This is usually done by dividing circulating vitamin E concentrations by the concentration of total lipids, which is calculated as total cholesterol plus triglycerides [32, 33]. Importantly, because there usually also is a positive correlation between total cholesterol and fasting triglycerides, simple summation of total cholesterol and fasting triglycerides to calculate total lipids and correction for this by dividing vitamin E concentrations may result in an unintended double correction for variance shared by total cholesterol and fasting triglycerides and unwanted weakening, disappearance, or even inversion of actually existing associations. If such an effect would exist, it would predict otherwise positive associations of vitamin E with fasting circulating lipids to become significantly inverse rather than absent after correction for total lipids. In this field HDL cholesterol might be an outlier. Because it contributes to total cholesterol, it is likely positively correlated with total cholesterol, but because of the action of circulating cholesterol ester transfer protein (CETP), it is likely inversely correlated with fasting triglycerides [5, 20], while, because of its antioxidant activity [16–18], one would anticipate a positive correlation with circulating vitamin E concentrations. Here it is quite unpredictable what the effect of correction of total cholesterol and triglycerides will be. Other interesting domains to be investigated from this perspective are components of the metabolic syndrome, including domains of body composition, blood pressure, glucose homeostasis, inflammation, uric acid, and liver enzymes, particularly, alanine amino transferase and gamma-glutamyl transferase, all of which are known to be relatively strongly and positively correlated with fasting triglycerides [1, 2, 12, 27, 34]. Concerning correlations of vitamin E with other circulating vitamins and effects of correction for total lipids with double correction for variance shared by total cholesterol and fasting triglycerides versus adjustment by means of linear regression, circulating concentrations of vitamin A might be interesting to have a closer look at, because in the circulation, vitamin A is also transported by lipoproteins [11, 21]. We hypothesized that the effect of correction for total lipids on the correlation of circulating vitamin E concentrations with circulating vitamin A concentrations will materially differ from the effect of this correction correlations with circulating concentrations of other vitamins. One way to overcome I. Minović et al. 203 double correction for variance shared by total cholesterol and fasting triglycerides may be by adjustment of vitamin E concentrations via linear regression rather than correlating variables to the quotient of vitamin E and total lipid concentrations. We therefore set out to compare correlations of circulating vitamin E concentrations, correlations of the quotient of circulating vitamin E concentrations, and linear regression-derived standardized regression coefficients in which associations are adjusted for total cholesterol and fasting triglycerides with circulating concentrations of fasting lipids, components of the metabolic syndrome, and circulating vitamin concentrations.

year of publication



  • Vitamin E in Human Health


  • Minović, I
  • Sotomayor, CG
  • Eggersdorfer, M
  • Riphagen, IJ
  • de Borst, MH
  • Dekker, LH
  • et al.

full publication

click here to view the full publication