Discordance between non-HDL-cholesterol and LDL-particle measurements: results from the Multi-Ethnic Study of Atherosclerosis

Abstract

Background: Cardiovascular risk assessment incorporates measurement of atherogenic lipids such as non-HDL cholesterol (non-HDL-C). It remains uncertain under which circumstances atherogenic lipoprotein enumeration such as LDL particle number (LDL-P) differs from simultaneously acquired non-HDL-C.

Methods: Participants of the Multi-Ethnic Study of Atherosclerosis (MESA) were deemed LDL-P > non-HDL-C discordant if they exhibited higher LDL-P than expected for simultaneously measured non-HDL-C, given the observed distribution of both in MESA. Conversely, a lower LDL-P than would be suggested from non-HDL-C characterized LDL-P < non-HDL-C discordance. Regression models were used to estimate associations of demographics and comorbidities with discordance and of LDL-P and non-HDL-C with carotid intima-media thickness (CIMT) and detectable coronary artery calcium (CAC) among discordance groups.

Results: Discordance was observed among 44% of subjects. LDL-P > non-HDL-C compared to LDL-P < non-HDL-C discordance was more common among Hispanics and smokers; among subjects with lower HDL-C, lower triglycerides, or greater insulin resistance by homeostatic model assessment of insulin resistance (HOMA-IR); and among subjects on lipid-lowering therapy, anti-hypertensive therapy, or hormone replacement therapy. In the setting of discordance, LDL-P exhibited a modestly greater association with CIMT than did non-HDL-C (+0.024-0.025 mm vs +0.018-0.021 mm per SD increase). In the presence of LDL-P < non-HDL-C discordance, LDL-P demonstrated a modestly greater association with detectable CAC than did non-HDL-C (OR 1.51 vs 1.46 per SD increase).

Conclusions: Our results demonstrated that disagreement between LDL-P and non-HDL-C was common and significantly associated with several clinical characteristics. In the setting of discordance, LDL-P was more closely associated with CIMT and CAC than non-HDL-C, though observed differences were small.

Keywords: Apolipoprotein B; Cholesterol; LDL-particle number; Lipoproteins; Risk assessment.

Figure 1

Distribution of LDL-particle number and non-HDL-cholesterol in the overall cohort and by ethnicity. LDL-P>non-HDL-C discordant, LDL-P one risk category higher than expected for simultaneously measured non-HDL-C; LDL-P>>non-HDL-C discordant, LDL-P two or more risk categories higher than expected for simultaneously measured non-HDL-C; LDL-P

Figure 2

Subgroup analysis of the prevalence of concordance and discordance between LDL-particle number and non-HDL-cholesterol. “Any clinical predictor” includes study participants with any of the following: HDL-C<50 mg/dL, HOMA-IR≥0.7, lipid-lowering therapy, or hormone replacement therapy (HRT). “No clinical predictor” excludes study participants with any of the four aforementioned characteristics.