Chapter 2 30 and calcium and urinary albumin excretion, and fasting glucose levels, CRP and eGFR. In addition, we performed sensitivity analyses in which we replaced adjustment for antihypertensive treatment in Cox regression analyses by adjustment for presence of cardiovascular disease and in which we replaced adjustment for eGFR in Cox regression analyses by adjustment for presence of chronic kidney disease. Hazard ratios (HRs) are reported with 95% confidence intervals (CIs). Restricted cubic splines with three knots were performed to show the association between ionized Mg and risk of T2D using Cox regression analyses. We evaluated potential effect modification in the analyses of plasma ionized Mg and risk of T2D by fitting models containing both main effects and their cross-product terms. Interaction terms were considered statistically significant at p<0.10. Missing data (present in 0.0-13.4%) in covariables were handled by multiple imputation 30. Results are reported for imputed data, except for the baseline characteristics and the analytic validation data. We considered a two-sided p value <0.05 as statistically significant. Data were analyzed using SPSS Statistics version 23.0 (SPSS Inc, Chicago, IL). Results Analytical performance of the NMR-measured ionized Mg assay The coefficient of variation for the NMR ionized Mg assay ranged from 4.6-7.1% for within-lab imprecision (Table 1). We compared NMR-measured Mg in EDTA plasma specimens with lithium heparin plasma total Mg measured by a Roche Modular colorimetric assay in 799 samples from the PREVEND study. We found a strong linear relationship between NMR-measured ionized Mg and colorimetrically measured total Mg (r=0.90). Bland-Altman analysis showed a systematic bias of 0.07 mmol/L with chemically measured total Mg concentrations being slightly higher than the ionized Mg quantified by NMR (Figure 2). Passing-Bablok regression analysis revealed an intercept of 0.02 (95% CI: -0.02-0.08) and a slope of 1.08 (95% CI: 1.00-1.13) (Figure 3). 2 pools of EDTA plasma with low and high Mg concentrations were tested twice a day in duplicate for 20 days on one instrument.
RkJQdWJsaXNoZXIy MTk4NDMw