A-021 Lipoprotein particle measurement by Nuclear Magnetic Resonance (NMR) spectroscopy correlates with diabetes independently from kidney function
Abstract
Nuclear magnetic resonance (NMR)–based lipoprotein analysis enables simultaneous quantification of lipoprotein particle concentration and size. Although primarily applied to cardiovascular disease (CVD) management, characteristic lipoprotein changes are also reported in patients with diabetes and chronic kidney disease (CKD). Both diseases are associated with increases in small low-density lipoprotein (sLDL), large very-low density lipoprotein (LVLDL), and decreased large high-density lipoprotein (LHDL). To elucidate this issue further, we aimed to evaluate NMR-measured lipoprotein size and particle concentrations in patients undergoing clinically ordered measured Iothalamate glomerular filtration rate (GFR) assessment, stratified by kidney function and diabetes status. Lipoproteins were measured using Axinon software (numares AG, Regensburg, Germany) on a 600MHz Bruker NMR in residual clinical sera from patients undergoing iothalamate renal clearance (n=330). Reduced kidney function was defined as a measured GFR <60 mL/min/1.73m2. A lipoprotein insulin resistance (LPIR) score was calculated using LVLDL, sLDL, and LHDL particle concentrations and size using published methods (Otvos, et. al). Associations between lipoproteins and disease status were assessed using Wilcoxon rank-sum or Pearson’s chi-squared tests, as appropriate. Of 330 participants, 159 (48%) had reduced eGFR. No statistically significant differences in lipoprotein parameters were observed by kidney function. Participants with reduced GFR showed non-significant trends toward increased LVLDL, [median (IQR): 1.6(1.0-3.8) vs. 1.5(0.9-2.7) nmol/L; p =0.3], and decreased LHDL [3.4(1.1-7.2) vs. 4.3(1.7-6.4) mcmol/L; p=0.4]. No differences between patients with reduced GFR were observed for sLDL [436(280-604) vs. 455(286-631) nmol/L; p=0.5] or LPIR score [46(33-64) vs. 46(34-60; p=0.5]. Diabetes was diagnosed in 86 (26%) participants. Compared to non-diabetic participants, those with diabetes had significantly increased LVLDL [2.15(1.00-4.50) vs. 1.50(0.90-2.80) nmol/L; p=0.01], decreased LHDL [3.0(1.1-5.3) vs. 4.3(1.9-7.0) mcmol/L; p=0.04], and smaller LDL size [20.9(20.6-21.1) vs. 21.1(20.8-21.3) nm; p<0.001]. Despite increased sLDL in diabetes [462(324-655) vs. 443(273-612) nmol/L], the difference was not significant (p=0.3). The LPIR score was significantly higher among participants with diabetes [53(37-68) vs. 45(32-60); p<0.01]. Our data support prior reports that diabetes is associated with adverse changes in lipoprotein particle size and concentration. In contrast, measured kidney function showed no significant independent association with NMR lipoprotein parameters in this cohort, nor does it skew diabetic evaluations. Our data suggests that the association between kidney function and lipoprotein patterns is not robust and may in part be caused due to diabetic related changes. Irrespective of eGFR, LPIR was a predictor of incident diabetes.