Vitamin D kinetics in the acute phase of critical illness: A prospective observational study

Journal of Critical Care 43 (2018) 294–299

Contents lists available at ScienceDirect

Journal of Critical Care journal homepage: www.jccjournal.org

Vitamin D kinetics in the acute phase of critical illness: A prospective observational study☆ Tomasz Czarnik, MD, PhD a,⁎, Aneta Czarnik, MD b, Ryszard Gawda, MD a, Maciej Gawor, MD a, Maciej Piwoda, MD a, Maciej Marszalski, MD a, Magdalena Maj, MD a, Olimpia Chrzan, MD a, Rahim Said, MD a, Maja Rusek-Skora, MD a, Marta Ornat, MD a, Kamil Filipiak, MD a, Jakub Stachowicz, MD a, Robert Kaplon, PhD c, Miroslaw Czuczwar, MD, PhD d a

Department of Anesthesiology and Critical Care, PS ZOZ Wojewodzkie Centrum Medyczne w Opolu, Aleja Witosa 26, 45-418 Opole, Poland Department of Endocrinology, Szpital Wojewodzki w Opolu, ul. Kosnego 53, 45-372 Opole, Poland c Department of Operations Research, Wroclaw University of Technology, Wybrzeze Wyspianskiego 27, 50-370 Wroclaw, Poland d 2nd Department of Anesthesiology and Critical Care, Medical University of Lublin, ul. Staszica 16, 20-081 Lublin, Poland b

a r t i c l e

i n f o

Available online xxxx Keywords: Vitamin D Pleiotropy Intensive care Mortality Kinetics Critical illness

a b s t r a c t Purpose: The objective of this study was to assess the vitamin D kinetics in critically ill patients by performing periodic serum vitamin D measurements in short time intervals in the initial phase of a critical illness. Materials and methods: We performed vitamin D serum measurements: at admission and then in 12-hour time intervals. The minimum number of vitamin D measurements was 4, and the maximum was 8 per patient. Results: A total of 363 patients were evaluated for participation, and 20 met the inclusion criteria. All patients had an initial serum vitamin D level between 10.6 and 39 ng/mL. Nineteen patients had vitamin D levels between 10 and 30 ng/mL, which means that they had vitamin D insufficiency or deficiency, and only one patient had a normal vitamin D serum plasma level. We observed that the median of the vitamin D level decreases until the fourth measurement then stabilizes around the 4th and 5th measurement and then appears to increase unevenly. The highest drop is at the very beginning. Conclusions: The vitamin D serum level is changeable in the initial phase of a critical illness. We hypothesize that the serum vitamin D concentration can mirror the severity of illness. © 2017 Elsevier Inc. All rights reserved.

1. Introduction The widely recognized role of vitamin D (25-hydroxyvitamin D; 25(OH)D3) in the human body is usually linked to extracellular calcium metabolism by intestinal absorption regulation and skeletal mineralization [1-3]. However, its function is much more complex. The non-classic (pleiotropic) vitamin D mechanisms of action rely on its binding to the nuclear receptor (VDR) and the formation of a heterodimer with the retinoid X receptor. The abovementioned interaction regulates the transcription of DNA into RNA by binding to genomic sequences called vitamin D response elements (VDREs) in many tissues and subsequently mediates the regulation of cell proliferation, differentiation, apoptosis, angiogenesis, hormone secretion, membrane stabilization, anti-inflammatory action, blood pressure regulation, blood sugar control and, finally, regulation of innate and adaptive immunity [1,3-8]. ☆ Work should be attributed to the Department of Anesthesiology and Critical Care, PS ZOZ Wojewodzkie Centrum Medyczne w Opolu, Aleja Witosa 26, 45-418, Opole, Poland. ⁎ Corresponding author. E-mail address: [email protected] (T. Czarnik).

http://dx.doi.org/10.1016/j.jcrc.2017.09.179 0883-9441/© 2017 Elsevier Inc. All rights reserved.

Dysregulated vitamin D pleiotropy caused by vitamin D deficiency, which is currently linked with cancer, autoimmune, infectious and cardiovascular diseases, is well recognized in intensive care patients [1]. Vitamin D deficiency is linked to morbidity and life-threatening organ dysfunction due to a dysregulated host response to infection (sepsis) in the intensive care unit (ICU) [9-15]. However, a few studies did not confirm such dependencies, and it seems that this is still an area of uncertainty [16,17]. Several studies performed in ICUs report a relationship between an extremely low serum vitamin D concentration (severe deficiency) and mortality in intensive care patients [11,18-28]. However, most of those trials were retrospective, and the vitamin D concentration was measured only once (at a certain point of time, usually at admission). We still do not know if vitamin D is a modifiable factor, which, if properly corrected, could substantially influence the patient outcome or another simple marker of the poor condition of the patient. Typically, the course of the early resuscitation phase of a critical illness is dynamic, potentially influencing vitamin D kinetics (serum level changes in time). There are only a few prospective observational or interventional clinical trials studying vitamin D kinetics in critically ill patients [26,29-34]. The methodology of these trials is

T. Czarnik et al. / Journal of Critical Care 43 (2018) 294–299

heterogeneous, which makes any comparisons or generalizations very difficult or even impossible to perform. The main objective of this study was to assess the vitamin D kinetics in critically ill patients by performing periodic serum vitamin D measurements in short time intervals (every 12 h) in the initial phase of a critical illness. Defining the typical vitamin D kinetic profile could potentially be helpful for planning supplementation regimens and future randomized prospective trial designs to adequately study the relationship between vitamin D deficiency correction and mortality rate reduction in critically ill patients. We hypothesized that the serum vitamin D levels are unstable in the initial, acute phase of a critical illness. 2. Materials and methods This was a prospective observational study conducted from September 2015 to September 2016 in a single, eleven-bed, medical/surgical ICU. Written informed consent was obtained from the patients' relatives. The study was approved by the Regional Ethics Committee in Opole, Poland (protocol number: 214/2015; the date of approval: 25/ 03/2015), it was registered before the recruitment of participants (clinicaltrials.gov NCT02414386) and was carried out according to the principles of the Declaration of Helsinki. The current presentation should be treated as a preliminary report for a larger study on serum vitamin D level measurements in critically ill patients undergoing continuous renal replacement therapy with regional citrate anticoagulation. Our aim was to study the vitamin D serum levels and kinetics in a homogenous intensive care population, which should be treated as a representative control cohort in the ICU. We included in the study consecutive critically ill patients with vitamin D levels above 10 ng/mL at admission and the coexistence of respiratory and circulatory failure. We defined respiratory failure as a need for invasive mechanical ventilation and circulatory failure as a need for inotrope and/or vasopressor administration. Patients who met any of the following criteria were excluded: acute liver failure, acute kidney injury treated with renal replacement therapy, hypercalcemia at admission (total calcium plasma level N 10.6 mg/dL; total ionized calcium plasma level N 1.35 mmol/L), parathyroid gland disease at admission, serum vitamin D level b 10 ng/mL at admission, end-stage renal disease, admission from another ICU or readmission, age younger than 18 years, or lack of consent from relatives. The vitamin D serum measurement procedure strictly adhered to the standard defined by the Central Hospital Laboratory of PS ZOZ Wojewodzkie Centrum Medyczne w Opolu. Blood samples were taken from an arterial line, central venous line or by direct peripheral venous puncture and were collected in ethylenediamine tetraacetic acid (EDTA) tubes. Blood samples were protected from light exposure, transported to the hospital laboratory within 30 min, then centrifuged at 3500 rpm for 10 min and processed by laboratory technicians. The vitamin D serum level was measured using an electrochemiluminescence binding assay on Cobas e411 or Cobas 6000 immunoassay analyzers (Roche Diagnostics GmbH, Mannheim, Germany). The coefficient of variation (the amount of variability relative to the mean) of that method is estimated to be 0.8%–5.8% [35]. Consecutive patients admitted to the ICU were assessed in terms of the study participation (inclusion and exclusion criteria). In every patient included in the study group, blood samples were collected within 12 h of admission. In the majority of patients, the first vitamin D serum level was measured with the first laboratory diagnostic tests performed at the time of admission to the ICU. If the first vitamin D serum level was b10 ng/mL (severe vitamin D deficiency), the patient was excluded from the study. The next vitamin D serum levels were taken in 12hour time intervals (twice daily, at 6 am and 6 pm). The minimum acceptable number of vitamin D measurements was 4, and the maximum was 8 per patient. All demographic data (date, name, hospital documentation number, sex, age, diagnosis at admission, comorbidities, Therapeutic Intervention Scoring System (TISS-28), Sequential Organ Failure

295

Assessment Score (SOFA), additional laboratory tests) were recorded into the hospital's electronic database and stored in the electronic form. After the recruitment process, patient data were extracted from the electronic database, the patient identification was blinded, and the data were transferred to the statistician for statistical analysis. We summarized the patients' descriptive statistics, including the mean, median, interquartile range (25th to 75th percentile), and ranges. Some of these measures were illustrated on a box-plot. Linear and nonlinear mixed effect models were used to investigate (a) the pattern of changes in the levels of vitamin D over time, (b) how baseline covariates, such as age, TISS-28, and SOFA, affect the average response, and (c) patient specific-effects. All computations were performed in R ver. 3.3.2 (R Core Team, 2016) using the lmr4 package (Bates et al., 2015) and ggplot2 package (Wickham, 2016) [36-38]. 3. Results A total of 363 patients were evaluated for participation in the trial. After the initial evaluation, 343 patients were excluded from the study. Exclusion reasons were the following: no circulatory failure, no respiratory failure, vitamin D measurement was not performed, endstage renal disease, acute kidney injury treated with renal replacement therapy, admission from another ICU or readmission, acute liver failure, age b 18. The study flow chart is depicted in Fig. 1. In 127 (35%) patients with coexisting respiratory and circulatory failure, vitamin D serum levels were measured. In 95 patients, vitamin D serum plasma level was b10 ng/mL in the first measurement and, in 44 patients, it was b3 ng/mL in the first measurement, which is an undetectable serum concentration. In 32 patients, the vitamin D serum plasma level was N10 ng/mL (20 patients in the range of 10–20 ng/mL, 11 patients in the range of 20–30 ng/mL, and 1 patient with N30 ng/mL). In 7 patients initially included in the trial, the minimum number of 4 vitamin D measurements was not reached. In 5 patients with vitamin D serum plasma levels of N 10 ng/mL who were initially included in the trial, acute kidney injury was subsequently diagnosed, and they underwent continuous renal replacement therapy with regional citrate anticoagulation; these patients were also excluded. Finally, 20 patients met the inclusion criteria and were included in the study for a vitamin D kinetics evaluation. The baseline demographics of the study group are depicted in Table 1. All patients in the study group had an initial serum vitamin D level between 10.6 and 39 ng/mL. Nineteen patients in the study group had serum plasma vitamin D levels between 10 and 30 ng/mL, which means that they had vitamin D insufficiency or deficiency, and only one patient in the study group had a normal vitamin D serum plasma level. Absolute numbers for vitamin D serum levels are depicted in Table 2. None of patients in the study group received potential vitamin D metabolism modifiers. The only exception was the patient with septic shock received supplemental dose of steroids but the risk of vitamin D metabolism modification was assessed as low. Fig. 2 summarizes the observations at each time of measurement using summary statistics. The median of the vitamin D level decreases until the fourth measurement. It stabilizes around the 4th and 5th measurement and then appears to increase unevenly. In the large majority of the time points, the distributions are skewed. The variability (according to the interquartile range) of the observations is not constant over time, but there is no apparent trend. One patient always had relatively high levels of vitamin D, which are depicted as outliers. Based on the evidence presented so far, we start by fitting a model with three random effects: intercept, slope, and a quadratic term, obtaining LL = − 347.96 (log likelihood), AIC = 715.91 and df = 10 (degrees of freedom). We compare this model with a model based on a linear trend (LL = − 365.36, AIC = 742.72, df = 6), a model based on an inverse relationship (LL = − 357.69, AIC = 727.39, df = 6) and a flexible semiparametric model, i.e., a restricted cubic spline (LL = −348.23, AIC = 716.47, df = 10). Likelihood ratio tests (LRTs)

296

T. Czarnik et al. / Journal of Critical Care 43 (2018) 294–299

Fig. 1. Study flow chart.

were employed to compare the nested models (all p-values b 0.001), whereas the Akaike Information Criterion was used to compare the non-nested models. As a result, the first model was chosen. It is worth Table 1 Descriptive statistics of patients and measurement characteristics.

Patient characteristic (n = 20) Age, years SOFA at admission TISS-28 at admission Vitamin D deficiency at admission, (n = 12) Vitamin D insufficiency at admission, (n = 7) Vitamin D normal level at admission, (n = 1) Primary diagnosis at admission (n = 20) Cardiac arrest, (n = 10) Respiratory failure, (n = 5) Cardiogenic shock, (n = 3) Septic shock, (n = 1) Multi trauma, (n = 1)

Mean Median IQR

Range

64.2 12.9 39.1 14.4 23.2

64.0 12.5 39.5 14.4 22.0

7.3 4.0 10.8 4.7 1.5

47.0–84.0 8.0–16.0 29.0–51.0 10.6–19.2 21.5–29.4

39.0

39.0

19.4 19.0 17.7 21.8 10.6

17.0 21.5 19.2 21.8 10.6

6.4 7.1 6.0

10.7–39.0 13.2–23.4 10.9–22.9

mentioning that the most flexible model gave an identical fit, but it has a more complicated structure, and thus was discarded. Finally, the baseline covariates included in the model turned out to be clearly insignificant. Hence, it was decided not to include them in further model testing and analysis. The estimated coefficients of the fixed effects: intercept (18.61, sd. err. = 1.56), slope (− 2.58, sd. err. = 0.35) and quadratic term (0.32, str. err. = 0.06) are highly significant (all p-values b 0.001). To decide which, if any, random effects are significant, we performed LRTs sequentially, starting from no random effect and ending with three random effects. The following values of the test statistic suggest that the intercept and slope are significant: 219, 23.8 and 1.8. Thus, the parameters were re-estimated, which yields: AIC = 711.7, logLik = −348.89, and the model explains almost 95% of the variance in the data (Table 3). The vector of fixed effect coefficients determines the shape of the curve that describes changes in the level of vitamin D for an average patient. For the first measurement, the average level is 18.57 ng/L. This value is changing over time, which is summarized by another two coefficients. Rather than consider a discrete rate of change in time, one might investigate the change of the average level of vitamin D per

T. Czarnik et al. / Journal of Critical Care 43 (2018) 294–299 Table 2 Absolute numbers for vitamin D serum levels

297

Table 3 Estimates, standard errors, and statistics for the model with two random effects.

Patient

M1

M2

M3

M4

M5

M6

M7

M8

Parameters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

10.9 12.2 16.2 39 18 10.6 19.2 17.7 21.6 14.9 14.2 13.2 22.9 22 21.8 23.4 14.5 21.5 29.4 10.7

8.4 3.4 7.1 35 16.6 15.8 13.7 15.8 20.8 9.9 19.5 9.8 22 18.1 22.3 24.7 9.6 16.9 32.5 4.6

5.1 3.2 7.8 32.5 13.7 13.8 12.6 15 16.8 6.6 18.2 11.2 22 14.6 23 20.9 9.4 14.4 25.9 5.7

5.4 5 10.3 38.5 12.1 13.2 9.4 9.7 10.4 8.7 17.1 9.7 19.7 14 20.4 19.8 6.5 16.2 25.3 6.2

5.8 6.6 . 42.5 8 . 9.1 8.9 14.1 4.6 10.5 11.1 22.9 15.1 13.6 17.7 5.8 19.3 24.5 5.2

6.7 4.3 . 42.8 . . 8.2 12.3 9 . 15.4 13.7 23.7 14.7 . 16 8.2 . 25.6 7.2

6.6 9 . 40 . . 11 10.7 11.1 . 13.2 16.2 21.1 12.1 . 22.4 . . 29.8 8.2

8.9 . . . . . 10.5 8.8 11.2 . 11.9 . 26.6 12.3 . . . . 31.9 .

Fixed effects Intercept Slope Quadratic Random effects Intercept (sd) Slope (sd) Residual (sd)

M1, M2, M3… – Consecutive measurements of vitamin D serum plasma levels [ng/mL].

small change of time. It can be derived by calculating the first derivative, which yields a linear function of time, i.e., −2.49 + 0.61t. Substituting the time for the consecutive values (the first measurement at t = 0), one gets the following rate of changes: − 2.49, −1.88, − 1.27, −0.66, −0.05, 0.56, 1.17, 1.78. It is clear that the highest drop in the level of vitamin D is at the very beginning. Subsequently, a minor decline appears, which might be considered a stabilization; then, the final phase reflects slight growth. Fig. 3 illustrates these behaviors (blue line with a dot). What further emerges from the analysis is the patient effect, which is significant not only for the average level of vitamin D but also for the rate of change. It means that each patient's intercept differs from the average value of 18.57 ng/L by the random effect of 7 ng/L. As seen in Fig. 2, there is a clear shift of an average-level curve to a patient-level curve. Such variability is attributed to a different initial level of vitamin D. Similarly, each patient's slope (rate of change) differs from the average value of −2.49 ng/L by the random effect of 0.72 ng/L. These discrepancies translate into a level of curvature (see Fig. 3 and compare the shape of the blue line with dots to the orange lines with triangles). The estimated correlation between intercepts and slopes is 0.18. Although the relationship is weak, it might indicate that a higher initial level of vitamin D results in a smaller decline.

Estimate

Std. error

t

p-Value

18.57 −2.49 0.31

1.62 0.35 0.05

11.45 −7.15 6.31

b0.001 b0.001 b0.001

7.00 0.72 2.21

4. Discussion According to the widely used definition, a serum vitamin D level of b20 ng/mL is defined as a deficiency, a level between 20 and 30 ng/mL as an insufficiency, and a level N 30 ng/mL as a normal value [7,39,40]. This prospective, single-center observational study revealed that the typical intensive care population of patients is extremely prone to vitamin D deficiency in the acute phase of a critical illness. In a group of 127 patients, 115 (90.5%) of them had vitamin D deficiency (serum plasma level b 20 ng/mL), 95 (75%) of them had severe vitamin D deficiency (serum plasma level b 10 ng/mL) and 44 (34.6%) patients in this group had undetectable serum plasma levels of vitamin D (b3 ng/mL). Eleven patients (8.7%) had vitamin D insufficiency (serum plasma level between 20 and 30 ng/mL). Finally, only one patient had a normal vitamin D serum level at admission: 39 ng/mL. According to summary statistics of the vitamin D kinetics in the study group, we observed that the vitamin D serum levels are unstable during critical illness. One of the most likely reasons for such an observation could be a hypothesis that the serum vitamin D concentration mirrors the severity of illness. The initial drop in the serum vitamin D level, followed by a period of stabilization and, finally, an increase could be consistent with the typical course of a critical illness (critical instability in the beginning, steady state and then stabilization). The relative instability of the vitamin D serum concentration could also colligate with a decrease in circulating albumin and vitamin D binding protein (VDBP), which typically occurs after fluid resuscitation in the ICU (the post-dilution effect), decreased synthesis of VDBP, renal wasting of vitamin D, interstitial extravasation caused by increased vascular permeability following an inflammatory reaction to critical illness, lack of exposure to sunlight, malnutrition, decreased renal production of 1,25(OH)D3, and increased tissue conversion of 25(OH)D3 to 1,25(OH)D3 [1,29]. Another reason for the relative instability of the

Fig. 2. Box plots repeated for each measurement. It summarizes both the distribution of the level of vitamin D (minimum, maximum and outliers, lower and upper quartiles, median) and changes over time.

298

T. Czarnik et al. / Journal of Critical Care 43 (2018) 294–299

Fig. 3. Repeated profile plot of the level of vitamin D for each patient. Consecutive values within observed values (grey lines with squares), estimated values for an average patient (blue lines with dots) and estimated values for each patient (orange lines with triangles) are connected by line segments. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

vitamin D serum concentration could be the coefficient of variation for laboratory method used in this study. We already know that low vitamin D status is common in critically ill patients (estimated prevalence of 40–99%) [18,41,42]. However, we still do not know if the initial (pre-disease) vitamin D serum level influences the subsequent course of critical illness and its outcome. It is worthwhile mentioning that a higher initial level of vitamin D might result in a smaller decline, as our analysis reveals. Finally, we still do not know if vitamin D supplementation during the initial phase of a critical illness could potentially influence the outcome, especially in patients with severe deficiency. Nair et al. performed a randomized study to determine the effect of two doses of intramuscular cholecalciferol on serial, serum vitamin D levels. They found that correction of a vitamin D deficiency is possible in critically ill patients, but no statistically significant difference in mortality and hospital length of stay was observed between the groups. The relationship between vitamin D repletion and improved outcomes was not assessed. The authors concluded that it is imperative to investigate the biological and clinical impacts of vitamin D repletion in critically ill patients in adequately powered interventional trials [34]. Quraishi et al. performed a randomized, placebo-controlled trial to compare the changes in vitamin D status in septic ICU patients who were treated with placebo versus cholecalciferol. They found that high-dose cholecalciferol supplementation improves the vitamin D levels in patients with sepsis and septic shock. The authors concluded that larger trials are needed to assess whether optimizing the vitamin D status improves sepsis-related clinical outcomes [33]. Amrein et al. found, in the largest study so far, that in critically ill patients with vitamin D deficiency, the administration of an oral high dose of vitamin D compared with a placebo did not reduce the hospital length of stay, hospital mortality and 6-

month mortality. Lower hospital mortality was observed in the severe vitamin D deficiency subgroup of patients. However, the authors concluded that this finding should be considered a hypothesis, and it requires further investigation. That study was the only one revealing the trend towards a favorable outcome in a subpopulation of patients with severe vitamin D deficiency who were treated with high-dose oral vitamin D [31]. We report a few limitations. The first limitation is the small size of the study group. We do not know if our observations could be generalized to a larger population of patients who are treated in ICUs. The second limitation is the observational nature of the trial. A prospective, interventional, randomized, placebo-controlled trial studying the relationship between a vitamin D supplementation regimen and the outcome in critically ill patients with severe vitamin D deficiency (serum vitamin D level b 10 ng/mL) would be more informative. Finally, the last limitation is that serum vitamin D levels were measured in only 35% of the patients admitted to the ICU. However, our intention was to perform the measurements in a homogenous group of patients, which are the most representative of the general intensive care population, therefore respiratory and circulatory failures were chosen as typical features of a critical illness. 5. Conclusions The intensive care population of patients is extremely prone to severe vitamin D deficiency. We observed that the vitamin D serum levels are changeable during critical illness. Large, randomized, placebo-controlled trials are needed to assess whether a correction of severe vitamin D deficiency in the acute phase of a critical illness could influence

T. Czarnik et al. / Journal of Critical Care 43 (2018) 294–299

outcomes or if the vitamin D levels can only be used as a simple tool for monitoring the condition of the patient. Sources of financial support for the work Department of Anesthesiology and Critical Care, PS ZOZ Wojewodzkie Centrum Medyczne w Opolu, Aleja Witosa 26, 45-418, Opole, Poland (institutional). Conflicts of interest None. Author contributions TC, AC designed the trial. TC, AC, RG, MC supervised the conduct of the trial and data collection. TC drafted the manuscript, takes responsibility for the paper as a whole. TC, RG, MG, MP, MM, MM, OC, RS, MRS, MO, KF, JS undertook recruitment of participants. TC managed the data, including quality control. RK provided statistical advice on study design and analyzed the data. TC, RG, RK, MC contributed substantially to the final version of the manuscript. References [1] Quraishi SA, Camargo CA. Vitamin D in acute stress and critical illness. Curr Opin Clin Nutr Metab Care 2012;15(6):625–34. [2] Bendik I, Friedel A, Roos FF, Weber P, Eggersdorfer M. Vitamin D: a critical and essential micronutrient for human health. Front Physiol 2014;5:248. [3] Amrein K, Christopher KB, McNally JD. Understanding vitamin D deficiency in intensive care patients. Intensive Care Med 2015;41:1961–4. [4] Lee P, Eisman JA, Center JR. Vitamin D deficiency in critically ill patients. N Engl J Med 2009;360(18):1912–4. [5] McKinney TJ, Patel JJ, Benns MV, Nash NA, Miller KR. Vitamin D status and supplementation in the critically ill. Curr Gastroenterol Rep 2016;18:18. [6] Rosen CJ. Vitamin D insufficiency. N Engl J Med 2011;364:248–54. [7] Holick MF. Vitamin D deficiency. N Engl J Med 2007;357:266–81. [8] Paul G, Brehm JM, Alcom JF, Holguin F, Aujla SJ, Celedon JC. Vitamin D and asthma. Am J Respir Crit Care Med 2012;185(2):124–32. [9] Rech MA, Hunsaker T, Rodriguez J. Deficiency in 25-hydroxyvitamin D and 30-day mortality in patients with severe sepsis and septic shock. Am J Crit Care 2014; 23(5):72–9. [10] Moromizato T, Litonjua AA, Braun AB, Gibbons FK, Giovannucci E, Christopher KB. Association of low serum 25-hydroxyvitamin D levels and sepsis in the critically ill. Crit Care Med 2014;42:97–107. [11] de Haan K, Groeneveld AB, de Geus HR, Egal M, Struijs A. Vitamin d deficiency as a risk factor for infection, sepsis and mortality in the critically ill: systematic review and meta-analysis. Crit Care 2014;18:660. [12] Kempker JA, Tangpricha V, Ziegler TR, Martin GS. Vitamin D in sepsis: from basic science to clinical impact. Crit Care 2012;16:316. [13] De Pascale G, Vallecoccia MS, Schiattarella A, Di Gravio V, Cutuli SL, Bello G, et al. Clinical and microbiological outcome in septic patients with extremely low 25hydroxyvitamin D level at initiation of critical care. Clin Microbiol Infect 2016; 22(5):456. [14] Upala S, Sanguankeo A, Permpalung N. Significant association between vitamin D deficiency and sepsis: a systematic review and meta-analysis. BMC Anesthesiol 2015;15:84. [15] Parekh D, Patel JM, Scott A, Lax S, Dancer RCA, D'Souza V, et al. Vitamin D deficiency in human and murine sepsis. Crit Care Med 2017;45:282–9. [16] Ala-Kokko TI, Mutt SJ, Nisula S, Koskenkari J, Liisanantti J, Ohtonen P, et al. Vitamin D deficiency at admission is not associated with 90-day mortality in patients with severe sepsis and septic shock: observational FINNAKI cohort study. Ann Med 2016; 48:67–75. [17] Cecchi A, Bonizzoli M, Douar S, Mangini M, Paladini S, Gazzini B, et al. Vitamin D deficiency in septic patients at ICU admission is not a mortality predictor. Minerva Anestesiol 2011;77(12):1184–9.

299

[18] Leaf DE, Croy HE, Abrahams SJ, Raed A, Waikar SS. Cathelicidin antimicrobial protein, vitamin D, and risk of death in critically ill patients. Crit Care 2015;19:80. [19] Hu J, Luo Z, Zxao X, Chen Q, Chen Z, Qin H, et al. Changes in the calcium-parathyroid hormone-vitamin D axis and prognosis for critically ill patients: a prospective observational study. Plos One 2013;8(9):e75441. [20] Amrein K, Quraishi SA, Litonjua AA, Gibbons FK, Pieber TR, Camargo CA, et al. Evidence for a U-shaped relationship between prehospital vitamin D status and mortality: a cohort study. J Clin Endocrinol Metab 2014;99(4):1461–9. [21] Quraishi SA, Bittner EA, Blum L, Caitlin M, McCarthy BA, Bhan I, et al. Prospective study of vitamin D status at initiation of care in critically ill surgical patients and risk of 90-day mortality. Crit Care Med 2014;42(6):1365–71. [22] Braun A, Chang D, Mahadevappa K, Gibbons FK, Liu Y, Giovannucci E, et al. Association of low serum 25-hydroxyvitamin D levels and mortality in the critically ill. Crit Care Med 2011;39(4):671–7. [23] Amrein K, Litonjua AA, Moromizato T, Quraishi SA, Gibbons FK, Pieber TR, et al. Increases in pre-hospital serum 25(OH)D concentrations are associated with improved 30-day mortality after hospital admission: a cohort study. Clin Nutr 2016; 35(32):514–21. [24] Braun AB, Gibbons FK, Litonjua AA, Giovannucci E, Christopher KB. Low serum 25hydroxyvitamin D at critical care initiation is associated with increased mortality. Crit Care Med 2012;40(1):63–72. [25] Venkatram S, Chilimuri S, Adrish M, Salako A, Patel M, Diaz-Fuentes G. Vitamin D deficiency is associated with mortality in the medical intensive care unit. Crit Care 2011;15:R292. [26] Moraes SB, Friedman G, Wawrzeniak IC, Marques LS, Nagel FM, Lisboa TC, et al. Vitamin D deficiency is independently associated with mortality among critically ill patients. Clinics (Sao Paulo) 2015;70(5):326–32. [27] Amrein K, Zajic P, Schnedl C, Waltensdorfer A, Fruhwald S, Holl A, et al. Vitamin D status and its association with season, hospital and sepsis mortality in critical illness. Crit Care 2014;18:R47. [28] Matthews LR, Ahmed Y, Wilson KL, Griggs DE, Danner OK. Worsening severity of vitamin D deficiency is associated with increased length of stay, surgical intensive care unit cost, and mortality rate in surgical intensive care unit patients. Am J Surg 2012; 204(1):37–43. [29] Nair P, Lee P, Reynolds C, Nguyen ND, Myburgh J, Eisman JA, et al. Significant perturbation of vitamin D-parathyroid-calcium axis and adverse clinical outcomes in critically ill patients. Intensive Care Med 2013;39(2):267–74. [30] Krishnan A, Ochola J, Mundy J, Jones M, Kruger P, Duncan E, et al. Acute fluid shifts influence the assessment of serum vitamin D status in critically ill patients. Crit Care 2010;14:R216. [31] Amrein K, Schnedl C, Holl A, Riedl R, Chrisopher KB, Pachler C, et al. Effect of highdose vitamin D3 on hospital length of stay in critically ill patients with vitamin D deficiency: the VITdAL-ICU randomized clinical trial. JAMA 2014;312(15):1520–30. [32] Venkatesh B, Davidson B, Robinson K, Pascoe R, Appleton C, Jones M. Do random estimations of vitamin D3 parathyroid hormone reflect the 24-h profile in the critically ill? Intensive Care Med 2012;38:177–9. [33] Quraishi SA, De Pascale G, Needleman JS, Nakazawa H, Kaneki M, Bajwa EK, et al. Effect of cholecalciferol supplementation on vitamin D status and cathelicidin levels in sepsis: a randomized, placebo-controlled trial. Crit Care Med 2015;43:1928–37. [34] Nair P, Venkatesh B, Lee P, Kerr S, Hoechter DJ, Dimeski G, et al. A randomized study of a single dose of intramuscular cholecalciferol in critically ill adults. Crit Care Med 2015;43:2313–20. [35] van Gammeren AJ, van Gool N, de Groot MJ, Cobbaert CM. Analytical performance evaluation of the Cobas 6000 analyzer – special emphasis on trueness verification. Clin Chem Lab Med 2008;46(6):863–71. [36] R Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2016(URL https://www.R-project.org/). [37] Bates D, Maechler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. J Stat Softw 2015;67(1):1–48. [38] Wickham H. ggplot2: elegant graphics for data analysis. New York: Springer-Verlag; 2016. [39] Venkatesh B, Nair P. Hypovitaminosis D and morbidity in critical illness: is there proof beyond reasonable doubt? Crit Care 2014;18:138. [40] Ross AC, Manson JE, Abrams SA, Aloia JF, Brannon PM, Clinton SK, et al. The 2011 report on dietary reference intakes for calcium and vitamin D from the Institute of Medicine: what clinicians need to know. J Clin Endocrinol Metab 2011;96(1):53–8. [41] Brook K, Camargo CA, Christopher KB, Quraishi SA. Admission vitamin D status is associated with discharge destination in critically ill surgical patients. Ann Intensive Care 2015;5(1):23. [42] Kvaran RB, Sigurdsson MI, Skarphedinsdottir SJ, Sigurdsson GH. Severe vitamin D deficiency is common in critically ill patients at a high northern latitude. Acta Anaesthesiol Scand 2016;60(9):1289–96.