Abstract
Serum low-density lipoprotein cholesterol (LDLC) value is a recognized target for atherosclerotic risk management, and is generally calculated using the "Friedewald formula". Alternative risk markers include directly measured LDLC, non-high-density lipoprotein cholesterol (non-HDLC) and apolipoprotein B (ApoB). The relationships among such various measured or calculated quantities in medium-sized sets of patient results were investigated. Results from two sets of patients were retrieved from our laboratory information systems. One group (n=8436) included results of cholesterol, HDLC, triglyceride (TG) and glucose measurements. A second group (n=902) included, in addition, results of ApoB measurement. The results confirmed the unreliability of the Friedewald formula at TG>350 mg/dL (3.96 mmol/L), but also indicated TG-linked underestimation of LDLC below such a TG level. By contrast, non-HDLC values were shown to be independent of TG, and better correlated to ApoB than LDLC values. Mathematically, the difference between non-HDLC and LDLC is TG×0.458 (values in mmol/L): therefore, the latter cannot be compared to (or converted into) the former by simply adding a constant amount. The ratio LDLC/ApoB was shown continuously to decrease with increasing TG concentrations, while the ratio non-HDLC/ApoB did not. The TG-dependent underestimation of LDLC may be the reason for the reported better cardiovascular risk predictivity of non-HDLC in diseases associated with TG increase, such as in diabetes. Non-HDLC values are not influenced by TG levels, and are better correlated with ApoB.