Meta‐Regression Analysis of the Effect of Trans Fatty Acids (TFAs) on LDL‐Cholesterol

Abstract

The Food and Drug Administration (FDA) published a Federal Register notice tentatively determining that partially hydrogenated oils (PHOs, the primary dietary source of industrial TFAs – iTFAs) are no longer generally recognized as safe (GRAS) when used in food. This determination was based on the conclusion that there is no threshold intake for iTFA that would not increase an individual's risk of coronary heart disease, based primarily on research showing iTFAs increase LDL‐cholesterol (LDL‐C) at high levels of intake. A meta‐regression analysis was conducted on the relationship between changes in TFA intake and changes in plasma LDL‐C, building on an evidence map designed to capture all relevant data. The purpose of this analysis was to describe the dose‐response in the range of the human data, particularly the low‐intake region. An advantage of the weighted regression (i.e., meta‐regression) is greater accuracy, since studies with smaller variance are given more weight. A variety of different mathematical expressions were fit to the data to determine which one best fit the data. Variability was very large, particularly at low TFA intake, indicating TFA intake is not the primary determinant of LDL‐C levels at low TFA intake. No significant differences were found in the mathematical fit of models across a wide range of shapes, from very flat initial slope to very steep initial slope. Thus, the mathematical modeling cannot inform our understanding of the shape of the dose‐response curve even in the range of human data, and determine whether a threshold exists. Instead, as is generally the case, one needs to rely on a mode of action evaluation.This work was sponsored by the ILSI North America PHO Task Force.