Meta-regression analysis of the effect of trans fatty acids on low-density lipoprotein cholesterol

Food Chem Toxicol. 2016 Dec;98(Pt B):295-307. doi: 10.1016/j.fct.2016.10.014. Epub 2016 Oct 14.


We conducted a meta-regression of controlled clinical trial data to investigate quantitatively the relationship between dietary intake of industrial trans fatty acids (iTFA) and increased low-density lipoprotein cholesterol (LDL-C). Previous regression analyses included insufficient data to determine the nature of the dose response in the low-dose region and have nonetheless assumed a linear relationship between iTFA intake and LDL-C levels. This work contributes to the previous work by 1) including additional studies examining low-dose intake (identified using an evidence mapping procedure); 2) investigating a range of curve shapes, including both linear and nonlinear models; and 3) using Bayesian meta-regression to combine results across trials. We found that, contrary to previous assumptions, the linear model does not acceptably fit the data, while the nonlinear, S-shaped Hill model fits the data well. Based on a conservative estimate of the degree of intra-individual variability in LDL-C (0.1 mmoL/L), as an estimate of a change in LDL-C that is not adverse, a change in iTFA intake of 2.2% of energy intake (%en) (corresponding to a total iTFA intake of 2.2-2.9%en) does not cause adverse effects on LDL-C. The iTFA intake associated with this change in LDL-C is substantially higher than the average iTFA intake (0.5%en).

Keywords: CHD; LDL-C; Meta-regression; PHO; TFA.

MeSH terms

  • Bayes Theorem*
  • Cholesterol, LDL / metabolism*
  • Humans
  • Meta-Analysis as Topic
  • Models, Biological*
  • Regression Analysis*
  • Risk Factors
  • Trans Fatty Acids / pharmacology*


  • Cholesterol, LDL
  • Trans Fatty Acids