Background: Gradient gel electrophoresis (GGE) and nuclear magnetic resonance (NMR) spectroscopy are both widely accepted methods for measuring LDL and HDL particle size. However, whether or not GGE- or NMR-measured LDL or HDL particle size predicts coronary heart disease (CHD) risk to a similar extent is currently unknown.

Methods: We used GGE and NMR to measure LDL and HDL particle size in a nested case-control study of 1025 incident cases of CHD and 1915 controls from the EPIC (European Prospective Investigation into Cancer and Nutrition)-Norfolk study. The study sample included apparently healthy men and women age 45–79 years followed for an average of 6 years.

Results: Pearson correlation coefficients showed that the overall agreement between NMR and GGE was better for the measurement of HDL size (r = 0.78) than for LDL size (r = 0.47). The odds ratio for future CHD among participants in the bottom tertile of LDL size (smallest LDL particles) was 1.35 (95% CI, 1.12–1.63) for GGE and 1.74 (1.41–2.15) for NMR. For HDL size, these respective odds ratios were 1.41 (1.16–1.72) and 1.85 (1.47–2.32). After adjustment for potential confounders, the relationship between small LDL or HDL particles and CHD was no longer significant, irrespective of the method.

Conclusions: In this prospective population study, we found that the relationships between NMR-measured LDL and HDL sizes and CHD risk were slightly higher than those obtained with GGE.