ELECTRONIC SUPPLEMENTARY MATERIAL
The Effect of High-Dose Cholecalciferol on Bioavailable Vitamin D Levels in Critically Ill Patients: A Post-Hoc Analysis of the VITdAL-ICU Trial
VITdAL-ICU Trial overview
The VITdAL-ICU trial was a randomized double-blind, placebo-controlled, single-center trial conducted in 5 ICUs of the Medical University of Graz (medical, neurological, cardiothoracic surgery and 2 mixed-surgery units). The trial was approved by the Ethics Committee of the Medical University of Graz and the Austrian Agency for Health and Food Safety, and written informed consent was obtained directly from the patients or from a legal surrogate. For patients in whom this was not possible, the Ethics Committee approved the use of a “surrogate consent.”
Inclusion criteria were 18 years or older, expected ICU stay of >48 hours, and vitamin D deficiency defined as a 25(OH)D level of 20 ng/ml or lower.
Patients were considered not eligible in cases of severely impaired gastrointestinal function, other trial participation, pregnant or lactating women, hypercalcemia, sarcoidosis, nephrolithiasis in the prior year.
Four-hundred and seventy-five patients constituted the intention-to-treat population.
Patients were assigned to either a placebo group (238) or vitamin D3 group (237) in a 1:1 ratio using the Randomizer for Clinical Trials tool developed at the Medical University of Graz. Patients randomized in the treatment group received a loading dose of 540.000 IU of vitamin D3 and, starting 28 days after ingestion of the study medication, 5 monthly maintenance doses of 90.000 IU of oral vitamin D3.
Patients were followed up for 6 months, and outcome end points included length of stay and mortality in the ICU at 28 days, in hospital mortality, and mortality at 6 months.
Patient Selection for Subgroup Analyses on Bioavailable fractions of Vitamin D
To explore the effect of vitamin D3 supplementation on the bioavailable fractions of vitamin D levels, and the impact of such treatment on the predefined mortality outcomes, we calculated the bioavailable 25(OH)D and bioavailable 1,25(OH)2D using previously-used formulas on day 0 (baseline) and day 7. [1]
Such formulas rely on the measurement of the vitamin D binding protein (VDBP) on plasma stored at -80°C.
Among the 475 patients of the intention-to-treat population, in n=87 patients the baseline value was not available (storing samples was only possible on weekdays, or the analysis was not possible for technical reasons, e.g. not enough material, hemolytic or lipemic sample) and consequently they were excluded a priori from the analysis because no imputation would have been possible.
Moreover some patients (n=23) without data on day 7 were also excluded, considering that the data were incomplete, and that this was a negligible number of cases.
The final subgroup of the analyses was composed of 365 patients, 181 in the placebo group and 184 in the vitamin D3 group.
Figure 1. Flow Diagram for the bioavailable vitamin D fractions subgroup analyses.
Reliability of the sample
Considering that a number of patients in the intention-to-treat population were excluded in the analyses, we verified whether the randomization process was still effective, and tested the persistence of the homogeneity of the two groups for several variables: age, gender, body mass index, Charlson comorbidity index, Simplified Acute Physiology Score, therapeutic intervention scoring system (TISS), 25(OH)D, 1,25(OH)D, bioavailable 25(OH)D, bioavailable 1,25(OH)D, severe liver disease, chronic kidney disease, and mechanical ventilation. As in the original intention-to-treat analysis all the p values were > 0.05. (Table 1)
Table 1. Baseline characteristics of the two groups.
No. (%)Placebo
n=181 / vitamin D3
n=184 / P value
Age, mean (SD), y / 64 (13.3) / 63 (15.9) / 0.23
Women / 61 (34) / 68 (37) / 0.58
BMI, mean (SD) / 27.2 (5.8) / 27 (5) / 0.82
Charlson Comorbidity Index,
mean (SD) / 3.1 (2.2) / 2.8 (2.2) / 0.17
SAPS II at ICU admission,
mean (SD) / 34.3 (15.4) / 31.8 (15.5) / 0.12
TISS-28 at study inclusion,
mean (SD) / 38 (8.2) / 37.2 (7.7) / 0.31
25(OH)D day 0
median (IQR), ng/ml / 13.2 (10.5-16.6) / 12.2 (9.2-15.9) / 0.26
1,25(OH)D day 0
median (IQR), pg/ml / 32 (12-63) / 31 (14-57) / 0.67
Vitamin D binding protein day 0
median (IQR), μg/ml / 273 (198-360) / 267 (188-352) / 0.60
Bioavailable 25(OH)D day 0
median (IQR), ng/ml / 0.88 (0.64-1.38) / 0.9 (0.6-1.35) / 0.41
Bioavailable 1,25(OH)D day 0
median (IQR), pg/ml / 3.71 (1.64-8.08) / 3.95 (1.63-7.94) / 0.40
Chronic kidney disease
moderate/severe / 48 (26.5) / 44 (23.9) / 0.62
Severe liver disease / 7 (3.9) / 14 (7.6) / 0.17
Mechanical ventilation / 116 (64.1) / 111 (60.3) / 0.51
Statistical Analyses
Categorical variables are given as frequencies and percentages, continuous variables as mean and standard deviation or as median (25th-75th percentile interquartile range – IQR), when appropriate, according to data distribution.
The comparison between placebo group and vitamin D3 group was made with the two-sample t-test for continuous variables, and by contingency tables and Fisher’s exact test for categorical variables.
The comparison between vitamin D metabolites on day 0 and day 7 was made with the Wilcoxon two-sample test.
Moreover, the increase at day 7 (Δ value) for each vitamin D metabolite ( 25(OH)D, 1,25(OH)D, bioavailable 25(OH)D, bioavailable 1,25(OH)D ) was calculated as a percentage of the value on day 0 and, after normalization by logarithm conversion, a generalized linear regression model was applied to compare the entity of the increase between the four variables.
Hospital mortality, 28-day mortality, and 6-month mortality were compared in the two groups by contingency tables and Fisher’s exact test.
In the treatment group, with the aim of recognizing the impact of the increase of the plasmatic values of vitamin D metabolites, in order to estimate the odds ratio for 28-day mortality and 6-month mortality, univariate logistic regressions were also applied to the four vitamin D variables measured on day 7, adjusting the analysis with the corresponding value on day 0.
Odds ratios are reported with a95% interval confidence (CI 95%). All tests were two-sided, and a p value of <0.05 was indicative of statistical significance. Data handling and analyses were done with SAS 9.4 software (SAS Institute Inc, Cary, NC, USA).
References
1. Quraishi SA, De Pascale G, Needleman JS, Nakazawa H, Kaneki M, Bajwa EK, Camargo CA, Jr., Bhan I, (2015) Effect of Cholecalciferol Supplementation on Vitamin D Status and Cathelicidin Levels in Sepsis: A Randomized, Placebo-Controlled Trial. Critical care medicine 43: 1928-1937