Are studies comparing only mathematical formulas for determining LDL-cholesterol superfluous?

Overview

Objective: Accurate estimation of low-density lipoprotein cholesterol (LDL-C) is important for guiding cholesterol-lowering therapy. Different methods currently exist to estimate LDL-C. Among them are the Friedewald, Martin-Hopkins and Sampson/NIH equations. In the last three years, many studies have been published that compare only LDL-cholesterol calculated by these equations, without comparison with measured values. The aim of this study is to confirm that, using artificial data or simulated data set, such a comparison can be made with absolute reproducibility.

Design: Methodical study.

Material and Methods: This study sought to assess the discrepancy in estimated LDL-C using the Friedewald, Martin-Hopkins, and Sampson/NIH equations using a synthetic data set with such constructed parameters that meet both the reality criterion and the ideal mathematical-statistical criteria. At the same time, real sets were used for comparison: the set of hypothyroidism, n=40, and the set of hyperthyroidism, n=30.

Results: We documented the differences between the LDL-C values obtained by comparing the three equations against each other with various graphical and numerical statistical comparisons. These differences measured on an artificial set are absolutely reproducible. The differences calculated on the hypothyroid and hyperthyroid sets are the same as on the artificial set.

Conclusion: Repeated comparisons of the three equations for the determination of LDL-C, and in general any other equation, on variously defined real-world patient populations are superfluous. The relationships between the equations are uniquely described by calculations on a synthetic data set.

Keywords:

LDL-cholesterol – Friedewald formula – Martin-Hopkins formula – Sampson/NIH formula – synthetic data set