Cholesterol Interference in the Assessment of Vitamin D Status: A Canadian Health Measures Survey Biobank Project.

ABSTRACT

BACKGROUND: Matrix effects are a known problem with immunoassays measuring serum 25-hydroxyvitamin D [25(OH)D]. OBJECTIVES: To determine if the inverse association between serum 25(OH)D and serum cholesterol concentrations is a function of assay method: Diasorin Liaison 25(OH) Vitamin D Total Assay (Liaison Total Assay), an immunoassay, compared with liquid chromatography tandem mass spectrometry (LC-MS/MS). METHODS: Canadian Health Measures Survey data and biobank serum (White males aged 20-79 y, n = 392) were evaluated for bias in serum 25(OH)D using Bland-Altman plots. Differences in serum 25(OH)D (Liaison Total Assay - LC-MS/MS) were compared among non-HDL-cholesterol <4.2 (n = 295) compared with ≥4.2 (n = 97) mmol/L and total cholesterol groups <5.2 (n = 256) compared with ≥5.2 (n = 136) mmol/L, and associations tested between 25(OH)D and non-HDL-cholesterol or total cholesterol concentrations, using regression. RESULTS: Serum 25(OH)D measured using Liaison Total Assay ranged from 10.7 to 137.0 nmol/L and 14.4 to 137.9 nmol/L by LC-MS/MS. Liaison Total Assay - LC-MS/MS showed a negative bias of 5.5 (95% limits of agreement -23.8, 12.7) nmol/L. Differences in 25(OH)D were -4.0 ± 9.0 (±SD) nmol/L if non-HDL-cholesterol was <4.2 mmol/L and -10.2 ± 8.7 nmol/L if ≥4.2 mmol/L (P < 0.0001). Differences in 25(OH)D, if total cholesterol was <5.2 mmol/L, were -3.4 ± 8.6 nmol/L and -9.6 ± 9.3 nmol/L if ≥5.2 mmol/L (P < 0.0001). Serum non-HDL-cholesterol (beta -3.17, P = 0.0014) and total cholesterol (beta -2.77, P = 0.0046) were inversely associated with Liaison Total Assay 25(OH)D (adjusted for age, fasting, and body mass index), but not with LC-MS/MS measured 25(OH)D. Interference by these lipoproteins was not eliminated by standardization of the Liaison Total Assay. Similar associations were observed with triglycerides as for the lipoproteins. CONCLUSIONS: Total cholesterol inversely associates with 25(OH)D, which is likely due to elevated non-HDL-cholesterol lipoprotein or triglyceride interference with the Liaison Total Assay. This is important as elevated cholesterol is common, and an underestimation of vitamin D status could be an unnecessary cause for concern.