Advertisement

Screening and Treatment of Vitamin D Deficiency on Hospital Admission: Is There a Benefit for Medical Inpatients?

      Abstract

      Background

      Whether screening and treatment of vitamin D deficiency improves clinical outcomes in medical inpatients remains unclear. We performed a systematic search and summarized the evidence from observational and randomized, controlled studies (RCTs) on the influence of vitamin D deficiency and its treatment on patient outcomes.

      Methods

      The MEDLINE and EMBASE databases were searched in January 2015 using the terms “vitamin D deficiency,” “vitamin D,” “mortality,” “inpatient,” “length of stay,” “hospitalization,” “accidental falls,” “quality of life,” “activities of daily living,” and “intensive care units.”

      Results

      Of 239 potential studies, 29 observational studies and 5 RCTs were included. Observational studies revealed an association between vitamin D deficiency and adverse patient outcomes, such as mortality, higher incidence and severity of infections, and increased duration of hospital stay, in the inpatient and geriatric patient population. The RCT data, however, are inconclusive regarding beneficial effects of vitamin D supplementation for reduction of mortality and length of hospital stay, as well as fall prevention.

      Conclusions

      Despite evidence of associations in observational studies, high-quality evidence from interventional trials on the benefits of vitamin D supplementation in inpatients is currently lacking. On the basis of the available evidence, general vitamin D screening and supplementation for the medical inpatient population in an acute care setting cannot be recommended.

      Keywords

      Clinical Significance
      • Within 29 observational and 5 randomized trials, associations of vitamin D deficiency and increased risk for adverse outcomes were found.
      • Interventional trials did not find a benefit of vitamin D supplementation for preventing these outcomes.
      • General vitamin D screening and supplementation for the medical inpatient population can currently not be recommended.
      Vitamin D deficiency is a problem of rising global importance.
      • Mithal A.
      • Wahl D.A.
      • Bonjour J.P.
      • et al.
      Global vitamin D status and determinants of hypovitaminosis D.
      This worldwide worsening trend and new findings concerning the paracrine actions of vitamin D have led to renewed interest in this vitamin and its metabolites. Over the last few years a number of epidemiologic, observational, and randomized, controlled trials (RCTs) evaluating vitamin D deficiency and the effects of vitamin D supplementation have been conducted. In addition to its well-known influence on calcium metabolism and its role in bone health, associations with nonskeletal diseases, such as malignant, cardiovascular, autoimmune, and infectious diseases, have been suggested.

      Vitamin D Metabolism

      Humans derive their vitamin D from 2 sources: exposure to sunlight and diet. During exposure to the sun, ultraviolet B radiation converts 7-dehydrocholesterol in the epidermis to previtamin D3, which is converted into vitamin D3. Dietary vitamin D is found naturally in foods such as oily fish and in artificially fortified foods. Vitamin D from the diet and sunlight is metabolized in the liver to 25-hydroxyvitamin [25(OH)D] and in the kidneys by the 25-hydroxyvitamin D-1-α-hydroxylase to its active form, 1,25-dihydroxyvitmamin [1,25(OH)D].
      • Holick M.F.
      Vitamin D deficiency.
      Vitamin D is known for its role in calcium homeostasis and bone metabolism and is regulated by phosphorus, calcium, parathyroid hormone, and fibroblast growth factor-23 levels. Vitamin D mediates its actions through a nuclear receptor in cells. Recent studies show that most cells in the human body express vitamin D receptors, with some expressing the enzyme D-1-α-hydroxylase. Allegedly, 1,25(OH)D controls more than 200 genes. Thereby, it seems to be involved in cellular proliferation, differentiation, apoptosis, angiogenesis, insulin and renin production, and in stimulating macrophage cathelicidin production.
      • Bouillon R.
      • Bischoff-Ferrari H.
      • Willett W.
      Vitamin D and health: perspectives from mice and man.
      • Liu P.T.
      • Stenger S.
      • Li H.
      • et al.
      Toll-like receptor triggering of a vitamin D-mediated human antimicrobial response.
      • Nagpal S.
      • Na S.
      • Rathnachalam R.
      Noncalcemic actions of vitamin D receptor ligands.
      • DeLuca H.F.
      Overview of general physiologic features and functions of vitamin D.

      Method of Measurement

      Usually, vitamin D status is determined through measurement of serum 25(OH)D levels. It is the most important circulating metabolite of vitamin D, has a long half-life of 2 to 3 weeks, and represents vitamin D from both dietary and cutaneous sources. Circulating levels of 1,25(OH)2D are less reliable, usually normal or can even be elevated in a vitamin D-deficient state.
      • Holick M.F.
      High prevalence of vitamin D inadequacy and implications for health.
      However, conflicting evidence suggests that the bioavailable fraction of 25(OH)D is a better indicator of the vitamin D status.
      • LeFevre M.L.
      US Preventive Services Task Force. Screening for vitamin d deficiency in adults: U.S. Preventive Services Task Force recommendation statement.
      There is controversy regarding the best methodology for measuring 25(OH)D in the serum. Multiple testing methods are available, including competitive protein binding, immunoassay, high-performance liquid chromatography, and combined high-performance liquid chromatography and mass spectrometry. However, studies evaluating the quality of these methods are mostly inconclusive and also report variability between assay methods as well as between different laboratories using the same method.
      • LeFevre M.L.
      US Preventive Services Task Force. Screening for vitamin d deficiency in adults: U.S. Preventive Services Task Force recommendation statement.
      Currently there is no consensus on a definition of vitamin D deficiency or of the optimal vitamin D status, in part because its ideal level varies depending on different outcomes. The United States Endocrine Society guideline defines vitamin D deficiency as 25(OH)D <20 ng/mL (50 nmol/L) and vitamin D insufficiency as 25(OH)D between 21 and 29 ng/mL.
      • Holick M.F.
      • Binkley N.C.
      • Bischoff-Ferrari H.A.
      • et al.
      Evaluation, treatment, and prevention of vitamin D deficiency: an Endocrine Society clinical practice guideline.

      Screening for Vitamin D Deficiency

      There is ongoing controversy regarding who would profit from vitamin D screening. The Endocrine Society Clinical Practice Guideline recommends screening only in individuals at risk (eg, patients suffering from osteoporosis, chronic kidney disease, or malabsorption syndromes) and in whom a quick optimization of vitamin D status could be expected.
      • Holick M.F.
      • Binkley N.C.
      • Bischoff-Ferrari H.A.
      • et al.
      Evaluation, treatment, and prevention of vitamin D deficiency: an Endocrine Society clinical practice guideline.
      The clinical guidelines recently published by the US Preventive Services Task Force also do not recommend screening for vitamin D deficiency in asymptomatic adults.
      • LeFevre M.L.
      US Preventive Services Task Force. Screening for vitamin d deficiency in adults: U.S. Preventive Services Task Force recommendation statement.
      • LeBlanc E.S.
      • Zakher B.
      • Daeges M.
      • Pappas M.
      • Chou R.
      Screening for vitamin d deficiency: a systematic review for the U.S. Preventive Services Task Force.
      However, especially in the primary care setting, screening for vitamin D is often part of routine laboratory work, although there is little to no evidence supporting this practice.

      Vitamin D Intake

      The dietary allowance recommended by the Institute of Medicine for vitamin D is 600 IU/d for persons aged 1-70 years and 800 IU/d for adults aged >71 years. These recommendations are based on required serum levels for bone health, owing largely to scant evidence on the nonskeletal benefits of vitamin supplementation.
      • Institute of Medicine Committee to Review Dietary Reference Intakes for Vitamin D, Calcium
      The National Academies Collection: reports funded by National Institutes of Health.

      Vitamin D Supplementation

      For adults with vitamin D deficiency, the Endocrine Society Clinical Practice Guideline suggests supplementation with 50,000 IU of vitamin D2 or D3 once weekly for 8 weeks (or 6000 IU daily) to achieve a blood level >30 ng/mL, followed by maintenance therapy of 1500-2000 IU/d. In obese patients and patients with malabsorption syndromes or those taking medications affecting vitamin D metabolism, a dose that is 2-3 times higher is suggested.
      • Holick M.F.
      • Binkley N.C.
      • Bischoff-Ferrari H.A.
      • et al.
      Evaluation, treatment, and prevention of vitamin D deficiency: an Endocrine Society clinical practice guideline.
      Higher “bolus” supplementation (eg, 500,000 IU annually) is not recommended owing to an observed increase in falls.
      • Sanders K.M.
      • Stuart A.L.
      • Williamson E.J.
      • et al.
      Annual high-dose oral vitamin D and falls and fractures in older women: a randomized controlled trial.
      Because vitamin D2 and D3 show different affinities for the vitamin D receptor, the effect of differences in their doses on vitamin D status has been examined. Studies examining this question have not reached a conclusive answer, although some studies suggest that the 2 forms are equivalent at low doses but that D3 is more effective at high doses.
      • Institute of Medicine Committee to Review Dietary Reference Intakes for Vitamin D, Calcium
      The National Academies Collection: reports funded by National Institutes of Health.
      • Brouwer-Brolsma E.M.
      • Bischoff-Ferrari H.A.
      • Bouillon R.
      • et al.
      Vitamin D: do we get enough? A discussion between vitamin D experts in order to make a step towards the harmonisation of dietary reference intakes for vitamin D across Europe.
      • Heaney R.P.
      • Recker R.R.
      • Grote J.
      • Horst R.L.
      • Armas L.A.
      Vitamin D(3) is more potent than vitamin D(2) in humans.
      • Holick M.F.
      • Biancuzzo R.M.
      • Chen T.C.
      • et al.
      Vitamin D2 is as effective as vitamin D3 in maintaining circulating concentrations of 25-hydroxyvitamin D.
      As discussed above, vitamin D, in particular its long-term effects on chronic diseases such as cancer and cardiovascular and autoimmune diseases, has been the focus of attention over the last few years. The aims of this review were to examine the effects of vitamin D deficiency and vitamin D supplementation on different short-term clinical outcomes in medical inpatients, and to discuss the benefits of screening and treatment of vitamin D deficiency in an acute care setting.

      Methods

      Three investigators electronically searched the biomedical literature in January 2015, using the databases MEDLINE and EMBASE. We included the Medical Subject Heading terms and free text words “vitamin D deficiency,” “vitamin D,” “mortality,” “inpatient,” “length of stay,” “hospitalization,” “accidental falls,” “quality of life,” “activities of daily living,” and “intensive care units.” We found additional articles by reviewing the lists of references cited in the trials, systematic reviews, and meta-analyses retrieved through our search. The search strategy is summarized in Figure, whereas the complete search strategy is available from the first author.

      Selection of Studies

      Two researchers independently chose trials for analysis according to title and abstract. On the one hand we searched for observational studies looking at the effect of vitamin D levels on different short-term clinical outcomes for the medical inpatient; on the other hand we searched for RCTs investigating the impact of vitamin D supplementation on defined clinical outcome measures. We also included studies focusing on geriatric and critically ill patients, as well as studies that included patients in rehabilitation units. Trials with a mixed population of medical and surgical patients also were included. We excluded surgical, pediatric, and psychiatric patients and patients seen in outpatient settings. Case reports, reviews, and meta-analysis also were excluded, but the references cited in these publications were reviewed, and studies meeting our criteria were selected.

      Endpoints

      The primary endpoint for this analysis was all-cause mortality. Secondary endpoints include infections, disease severity, length of stay (LOS), and other health outcomes.

      Results

      The results of our analysis are summarized according to endpoint in Table. The complete results of our analysis can be found in Appendices 1 and 2 (available online).
      TableSummary of Findings, Categorized by Patient Type and Outcome
      OutcomeResults
      SignificantNot Significant
      Observational studies
       Medical inpatients (total 13 trials)
      The same study may have studied more than one clinical outcome.
      Mortalityn = 9 trials
      • 5 of 9 trials show a significant association between vitamin D deficiency and higher mortality
      Leow et al
      • Leow L.
      • Simpson T.
      • Cursons R.
      • Karalus N.
      • Hancox R.J.
      Vitamin D, innate immunity and outcomes in community acquired pneumonia.
      : OR 13.5, P = .002

      Lange et al
      • Lange N.
      • Litonjua A.A.
      • Gibbons F.K.
      • Giovannucci E.
      • Christopher K.B.
      Pre-hospital vitamin D concentration, mortality, and bloodstream infection in a hospitalized patient population.
      : ≤15 ng/mL: OR 1.45, P < .0001; 15-30 ng/mL: OR 1.30, P = .003

      Amrein et al
      • Amrein K.
      • Quraishi S.A.
      • Litonjua A.A.
      • et al.
      Evidence for a U-shaped relationship between prehospital vitamin D status and mortality: a cohort study.
      : <10 ng/mL: OR 2.01, P < .0001; 10-19.9 ng/mL: OR 1.89, P < .0001; 20-29.9 ng/mL: OR 1.34, P = .002

      Remmelts et al
      • Remmelts H.H.
      • van de Garde E.M.
      • Meijvis S.C.
      • et al.
      Addition of vitamin D status to prognostic scores improves the prediction of outcome in community-acquired pneumonia.
      : OR 2.949

      Khalili et al
      • Khalili H.
      • Talasaz A.H.
      • Salarifar M.
      Serum vitamin D concentration status and its correlation with early biomarkers of remodeling following acute myocardial infarction.
      : Association survival and Vitamin D concentration: r = 0.216, P < .001
      Abdelfatah et al
      • Abdelfatah M.
      • Nayfe R.
      • Moftakhar B.
      • et al.
      Low vitamin D level and impact on severity and recurrence of Clostridium difficile infections.
      : Severe deficiency: 2.5%; insufficiency: 1%; pre-deficiency: 3.1%, P = .43

      Anty et al
      • Anty R.
      • Tonohouan M.
      • Ferrari-Panaia P.
      • et al.
      Low levels of 25-hydroxy vitamin D are independently associated with the risk of bacterial infection in cirrhotic patients.
      : Association not significant (log–rank P = .34)

      Pletz et al
      • Pletz M.W.
      • Terkamp C.
      • Schumacher U.
      • et al.
      Vitamin D deficiency in community-acquired pneumonia: low levels of 1,25(OH)2 D are associated with disease severity.
      : 25(OH)D and fatal outcome: χ2/1 = 0.2374, P = .6261

      Lai et al
      • Lai L.
      • Qian J.
      • Yang Y.
      • et al.
      Is the serum vitamin D level at the time of hospital-acquired acute kidney injury diagnosis associated with prognosis?.
      : High 25(OH)D level: 45.8%; low 25(OH)D level: 47.9%, P = .554 by the log–rank test
      LOSn = 2 trials
      • 1 of 2 trials show significant association between vitamin D deficiency and longer LOS
      Pletz et al
      • Pletz M.W.
      • Terkamp C.
      • Schumacher U.
      • et al.
      Vitamin D deficiency in community-acquired pneumonia: low levels of 1,25(OH)2 D are associated with disease severity.
      : Lower 25(OH)D levels: associated with longer hospitalization: F1,290 = 17.69 P < .001
      Khalili et al
      • Khalili H.
      • Talasaz A.H.
      • Salarifar M.
      Serum vitamin D concentration status and its correlation with early biomarkers of remodeling following acute myocardial infarction.
      : Deficiency: 7.2 ± 2.6 d; normal vitamin D: 7.5 ± 2.9 d, P = .542
      Infectionsn = 8 trials
      • 7 of 8 trials show significant association between low 25(OH)D and higher incidence and severity of infections
      Lange et al
      • Lange N.
      • Litonjua A.A.
      • Gibbons F.K.
      • Giovannucci E.
      • Christopher K.B.
      Pre-hospital vitamin D concentration, mortality, and bloodstream infection in a hospitalized patient population.
      : ≤15 ng/mL: OR 1.29, P = .01; 15-30 ng/mL: OR 1.01, P = .9

      Abdelfatah et al
      • Abdelfatah M.
      • Nayfe R.
      • Moftakhar B.
      • et al.
      Low vitamin D level and impact on severity and recurrence of Clostridium difficile infections.
      : Severe complicated CDI: severe deficiency: 17.5%; insufficiency: 18.2%; pre-deficiency: 14%; control: 4%, P = .04

      Anty et al
      • Anty R.
      • Tonohouan M.
      • Ferrari-Panaia P.
      • et al.
      Low levels of 25-hydroxy vitamin D are independently associated with the risk of bacterial infection in cirrhotic patients.
      : Infected: significantly more severe deficiency than noninfected (P = .019)

      Quraishi et al
      • Quraishi S.A.
      • Litonjua A.A.
      • Moromizato T.
      • et al.
      Association between prehospital vitamin D status and hospital-acquired bloodstream infections.
      : <10 ng/mL: OR 1.95, P < .005; <20 ng/mL: OR 1.68, P < .002

      Ginde et al
      • Ginde A.A.
      • Camargo Jr., C.A.
      • Shapiro N.I.
      Vitamin D insufficiency and sepsis severity in emergency department patients with suspected infection.
      : Severe sepsis at baseline: OR 0.26, P = .006

      SOFA score >2 at baseline: OR 0.34, P = .049

      Severe sepsis at 24 h: OR 0.28, P = .005

      Wang et al
      • Wang W.J.
      • Gray S.
      • Sison C.
      • et al.
      Low vitamin D level is an independent predictor of poor outcomes in Clostridium difficile-associated diarrhea.
      : adjusted OR 5.6, P = .011

      Van der Wilden et al
      • van der Wilden G.M.
      • Fagenholz P.J.
      • Velmahos G.C.
      • Quraishi S.A.
      • Schipper I.B.
      • Camargo Jr., C.A.
      Vitamin D status and severity of Clostridium difficile infections: a prospective cohort study in hospitalized adults.
      : adjusted OR 0.92, P = .008
      Abdelfatah et al
      • Abdelfatah M.
      • Nayfe R.
      • Moftakhar B.
      • et al.
      Low vitamin D level and impact on severity and recurrence of Clostridium difficile infections.
      : RCDI: severe deficiency: 12.5%; insufficiency: 16.1%; pre-deficiency: 15.6%; control: 17.5%, P = .55

      Pletz et al
      • Pletz M.W.
      • Terkamp C.
      • Schumacher U.
      • et al.
      Vitamin D deficiency in community-acquired pneumonia: low levels of 1,25(OH)2 D are associated with disease severity.
      : Association between 25-(OH) and CURB, P = .325
      ICU admissionn = 1 trial
      • Significant association between vitamin D deficiency and higher rate of ICU admissions
      Remmelts et al
      • Remmelts H.H.
      • van de Garde E.M.
      • Meijvis S.C.
      • et al.
      Addition of vitamin D status to prognostic scores improves the prediction of outcome in community-acquired pneumonia.
      : ICU: 34.9 nmol/L [IQR, 23.8-46.3], non-ICU: 48.3 nmol/L (IQR, 30.8–68.4), P = .04
      Not applicable
       ICU patients (total 10 trials)
      The same study may have studied more than one clinical outcome.
      Mortalityn = 10 trials
      • 6 of 10 trials show significant association between vitamin D deficiency and increased mortality
      Amrein et al
      • Amrein K.
      • Zajic P.
      • Schnedl C.
      • et al.
      Vitamin D status and its association with season, hospital and sepsis mortality in critical illness.
      : Hospital mortality: deficiency: HR 1.63; insufficiency: HR 1.01

      Braun et al
      • Braun A.
      • Chang D.
      • Mahadevappa K.
      • et al.
      Association of low serum 25-hydroxyvitamin D levels and mortality in the critically ill.
      : In-hospital mortality: ≤15 ng/mL: OR 1.72, P < .0001; 15-30 ng/mL: OR 1.34, P = .04

      30-d mortality: ≤15 ng/mL: OR 1.69, P < .0001; 15-30 ng/mL: OR 1.36, P = .25

      90-d mortality: ≤15 ng/mL: OR 1.67, P < .0001; 15-30 ng/mL: OR 1.51, P = .001

      365-d mortality: ≤15 ng/mL: OR 1.72, P < .0001; 15-30 ng/mL: OR 1.44, P = .001

      McKinney et al
      • McKinney J.D.
      • Bailey B.A.
      • Garrett L.H.
      • Peiris P.
      • Manning T.
      • Peiris A.N.
      Relationship between vitamin D status and ICU outcomes in veterans.
      : Survival rates: vitamin D sufficient = 69%, vitamin D deficient = 44% (χ2 = 8.21, P = .004); RR of death after admission to ICU = 1.81

      Hu et al
      • Hu J.
      • Luo Z.
      • Zhao X.
      • et al.
      Changes in the calcium-parathyroid hormone-vitamin D axis and prognosis for critically ill patients: a prospective observational study.
      : Hospital mortality (r = −0.276, P < .001).

      Venkatram et al
      • 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.
      : Deficiency: OR 8.7, P = .0469; insufficiency: OR 4.3, P = .2081

      Moromizato et al
      • Moromizato T.
      • Litonjua A.A.
      • Braun A.B.
      • Gibbons F.K.
      • Giovannucci E.
      • Christopher K.B.
      Association of low serum 25-hydroxyvitamin D levels and sepsis in the critically ill.
      :

      In-hospital mortality: OR 1.62, P = .02

      30-d mortality: OR 1.55, P = .04

      90-d mortality: OR 1.63, P = .01
      Amrein et al
      • Amrein K.
      • Zajic P.
      • Schnedl C.
      • et al.
      Vitamin D status and its association with season, hospital and sepsis mortality in critical illness.
      : ICU mortality: deficiency: HR 1.30; insufficiency: HR 1.61

      Aygencel et al
      • Aygencel G.
      • Turkoglu M.
      • Tuncel A.F.
      • Candir B.A.
      • Bildaci Y.D.
      • Pasaoglu H.
      Is vitamin D insufficiency associated with mortality of critically ill patients?.
      : OR 1.8, P = .328.

      Su et al
      • Su L.X.
      • Jiang Z.X.
      • Cao L.C.
      • et al.
      Significance of low serum vitamin D for infection risk, disease severity and mortality in critically ill patients.
      : Vitamin D groups and mortality:

      28-d mortality: 1st = 0.6%, 2nd = 2.6%, 3rd = 3.8%, 3rd = 11.5%, 4th =11.5%, 5th = 4.5%

      90-d mortality: 1st = 0.6%, 2nd = 4.5%, 3rd = 6.4%, 4th = 18.6%, 5th = 7.7%

      → No significant difference between the 5 groups

      Higgins et al
      • Higgins D.M.
      • Wischmeyer P.E.
      • Queensland K.M.
      • Sillau S.H.
      • Sufit A.J.
      • Heyland D.K.
      Relationship of vitamin D deficiency to clinical outcomes in critically ill patients.
      :
      • Sufficient vs deficient: HR 0.89
      • Insufficient vs deficient: HR 1.23
      • Sufficient vs insufficient: HR 0.724
      Cecchi et al
      • Cecchi A.
      • Bonizzoli M.
      • Douar S.
      • et al.
      Vitamin D deficiency in septic patients at ICU admission is not a mortality predictor.
      : significant association univariate analysis (OR 0.995, 95% CI 0.991-0.998, P = .004), but not significant in a multivariate logistic regression model
      LOSn = 3 trials
      • 1 of 3 trials show significant association between vitamin D deficiency and LOS
      McKinney et al
      • McKinney J.D.
      • Bailey B.A.
      • Garrett L.H.
      • Peiris P.
      • Manning T.
      • Peiris A.N.
      Relationship between vitamin D status and ICU outcomes in veterans.
      : RR for a ≥3-d ICU stay = 2.0
      Amrein et al
      • Amrein K.
      • Zajic P.
      • Schnedl C.
      • et al.
      Vitamin D status and its association with season, hospital and sepsis mortality in critical illness.
      : Deficiency: 15.1 d; insufficiency: 15.0 d; sufficiency: 13.5 d, P = .743

      Venkatram et al
      • Hu J.
      • Luo Z.
      • Zhao X.
      • et al.
      Changes in the calcium-parathyroid hormone-vitamin D axis and prognosis for critically ill patients: a prospective observational study.
      : Deficiency = 7.2 ± 2.6 d (mean ± SD); normal vitamin D = 7.5 ± 2.9 d, P = .542
      Blood culture positivityn = 2 trials
      • 1 of 2 trials show significant association between vitamin D deficiency and higher rates of blood culture positivity
      Hu et al
      • Hu J.
      • Luo Z.
      • Zhao X.
      • et al.
      Changes in the calcium-parathyroid hormone-vitamin D axis and prognosis for critically ill patients: a prospective observational study.
      : Higher positive blood culture rate (P = .015)
      Amrein et al
      • Amrein K.
      • Zajic P.
      • Schnedl C.
      • et al.
      Vitamin D status and its association with season, hospital and sepsis mortality in critical illness.
      : Deficiency: 22.2%; insufficiency: 29.2%; sufficiency: 21.2%, P = .641
      Disease severityn = 5 trials
      • 3 of 5 trials show significant association between vitamin D deficiency and increased disease severity
      Hu et al
      • Hu J.
      • Luo Z.
      • Zhao X.
      • et al.
      Changes in the calcium-parathyroid hormone-vitamin D axis and prognosis for critically ill patients: a prospective observational study.
      : Higher APACHE II Score (P < .001), iPTH level (P < .001), incidence of MODS (P = .013)

      Relationship vitamin D and APACHE II score: r = −0.325, P < .001

      Moromizato et al
      • Moromizato T.
      • Litonjua A.A.
      • Braun A.B.
      • Gibbons F.K.
      • Giovannucci E.
      • Christopher K.B.
      Association of low serum 25-hydroxyvitamin D levels and sepsis in the critically ill.
      : Vitamin D levels and risk of sepsis: ≤15 ng/mL: OR 1.51, P = .001; 15-30 ng/mL: OR 1.18, P = .1

      Aygencel et al
      • Aygencel G.
      • Turkoglu M.
      • Tuncel A.F.
      • Candir B.A.
      • Bildaci Y.D.
      • Pasaoglu H.
      Is vitamin D insufficiency associated with mortality of critically ill patients?.
      :

      SOFA Score: insufficiency: 8 (6-11); sufficiency: 6 (4-9), P = .005

      APACHE II Score: insufficiency: 24 (19-28); sufficiency: 19 (16-25), P = .006

      Sepsis/septic shock: insufficiency: 57 (41%); sufficiency: 15 (24.2%), P = .026

      Organ dysfunction: insufficiency: 116 (84%); sufficiency: 44 (71%), P = .042
      Su et al
      • Su L.X.
      • Jiang Z.X.
      • Cao L.C.
      • et al.
      Significance of low serum vitamin D for infection risk, disease severity and mortality in critically ill patients.
      : Correlation of vitamin D with APACHE II: Pearson coefficient = −0.124 (P = .124); SAPS/SAPS II: Pearson coefficient = −0.039 (P = .629); and SOFA scores: Pearson coefficient: −0.12 (P = .221)

      Higgins et al
      • Higgins D.M.
      • Wischmeyer P.E.
      • Queensland K.M.
      • Sillau S.H.
      • Sufit A.J.
      • Heyland D.K.
      Relationship of vitamin D deficiency to clinical outcomes in critically ill patients.
      : Max SOFA scores: deficiency: 9.7 ± 4.2; insufficiency: 8.4 ± 4.2; sufficiency: 8.4 ± 3.3. P = .12

      Delta SOFA scores: sufficient: 1.3 ± 1.7; insufficient: 1.6 ± 2.1; deficient: 2.1 ± 2.3. P = .29
       Geriatric patients (total 4 trials)
      Mortalityn = 1 trial
      • Significant association between vitamin D deficiency and increased mortality
      Annweiler et al
      • Annweiler C.
      • Pochic S.
      • Fantino B.
      • et al.
      Serum vitamin D concentration and short-term mortality among geriatric inpatients in acute care settings.
      : Full adjusted logistic regression: OR 0.65, P = .029

      Stepwise backward model: OR 0.87, P = .029
      Not applicable
      LOSn = 2 trials
      • 2 of 2 trials show significant association between vitamin D deficiency and longer LOS
      Hélard et al
      • Helard L.
      • Mateus-Hamdan L.
      • Beauchet O.
      • Annweiler C.
      Hypovitaminosis D in geriatric acute care unit: a biomarker of longer length of stay.
      : Univariate and multiple linear regression models: P = .043 for adjusted model; β = −0.08, P = .015 for backward model.

      Inverse correlation: r = −0.14, P = .028

      Beauchet et al
      • Beauchet O.
      • Launay C.
      • de Decker L.
      • Fantino B.
      • Kabeshova A.
      • Annweiler C.
      Who is at risk of long hospital stay among patients admitted to geriatric acute care unit? Results from a prospective cohort study.
      : <8 d LOS: 24.1%, P = .009; 8-13 d LOS: 37%, P < .001; >13 d LOS: 49.7%, P = .014

      Significant association in univariate and multiple linear regression models of LOS with vitamin D deficiency (P < .001 for all models).
      Not applicable
      No. of acute diseasesn = 1 trial
      • Significant association between vitamin D deficiency and higher no. of acute diseases
      Significant:

      Sutra del Galy et al
      • Sutra Del Galy A.
      Vitamin D insufficiency and acute care in geriatric inpatients.
      : Adjusted β = 0.489, P = .03; in backward stepwise linear regression: β = 0.470, P = .03
      Not applicable
       Patients undergoing rehabiliation (total 2 trials)
      Functional outcomen = 2 trials
      • 0 of 2 trials show significant association between vitamin D deficiency and better functional outcome
      Not applicableKiebzak et al
      • Kiebzak G.M.
      • Moore N.L.
      • Margolis S.
      • Hollis B.
      • Kevorkian C.G.
      Vitamin D status of patients admitted to a hospital rehabilitation unit: relationship to function and progress.
      : No correlation with FIM efficiency or unadjusted change in total FIM score during hospitalization: r = .19, P = .066

      Pellicane et al
      • Pellicane A.J.
      • Wysocki N.M.
      • Mallinson T.R.
      • Schnitzer T.J.
      Prevalence of 25-hydroxyvitamin D deficiency in the acute inpatient rehabilitation population and its effect on function.
      : Association between total FIM efficiency scores and 25(OH)D levels: Higher than median (25.7 ng/mL) = 2.92 ± 1.37; lower than median = 2.411 ± 1.40; P = .110
      Randomized controlled trials
       ICU patients (total 1 trial)
      The same study may have studied more than one clinical outcome.
      Mortalityn = 1 trial
      • No significant effect of vitamin D supplements on hospital mortality or 6-mo mortality
      • Incidental finding in subgroup (severe vitamin D deficiency)
      Not applicableAmrein et al
      • Amrein K.
      • Schnedl C.
      • Holl A.
      • et al.
      Effect of high-dose vitamin D3 on hospital length of stay in critically ill patients with vitamin D deficiency: the VITdAL-ICU randomized clinical trial.
      : Main analysis : Intervention vs placebo: HR 0.78, P = .09

      Subgroup analysis severe vitamin D deficiency

      Intervention vs placebo: HR for hospital mortality 0.51, 95% CI 0.31-0.84, P = .009; HR for 6-mo mortality, 0.50, 95% CI 0.31-0.79, P = .003
      LOSn = 1 trial
      • No significant effect of vitamin D supplements on LOS
      Not applicableAmrein et al
      • Amrein K.
      • Schnedl C.
      • Holl A.
      • et al.
      Effect of high-dose vitamin D3 on hospital length of stay in critically ill patients with vitamin D deficiency: the VITdAL-ICU randomized clinical trial.
      : Length of hospital stay: Intervention: 20.1 d; placebo: 19.3 d, P = .98

      Length of ICU stay: Intervention: 9.6 d; placebo: 10.7, P = .38
       Geriatric patients (total 4 trials)
      The same study may have studied more than one clinical outcome.
      No. of fallsn = 3 trial
      • 1 of 2 trials show a significant effect of vitamin D supplementation on the no. of falls per person
      Bischoff et al
      • Bischoff H.A.
      • Stahelin H.B.
      • Dick W.
      • et al.
      Effects of vitamin D and calcium supplementation on falls: a randomized controlled trial.
      : Vitamin D plus calcium treatment: 62% reduction in falls, r = −0.96, P < .0002
      Latham et al
      • Latham N.K.
      • Anderson C.S.
      • Lee A.
      • et al.
      A randomized, controlled trial of quadriceps resistance exercise and vitamin D in frail older people: the Frailty Interventions Trial in Elderly Subjects (FITNESS).
      :

      Vitamin D vs placebo:
      • RR (95% CI): 1.12 (0.79-1.59)
      • HR (95% CI): 1.14 (0.80-1.62)
      Burleigh et al
      • Burleigh E.
      • McColl J.
      • Potter J.
      Does vitamin D stop inpatients falling? A randomised controlled trial.
      : No significant difference in no. of falls
      No. of fallersn = 2
      • 0 of 2 trials show a significant effect of vitamin D supplementation on no. of fallers
      Not applicableBischoff et al
      • Bischoff H.A.
      • Stahelin H.B.
      • Dick W.
      • et al.
      Effects of vitamin D and calcium supplementation on falls: a randomized controlled trial.
      : Vitamin D vs placebo: RR 0.7

      Burleigh et al
      • Burleigh E.
      • McColl J.
      • Potter J.
      Does vitamin D stop inpatients falling? A randomised controlled trial.
      : No. of fallers: vitamin D vs placebo: RR 0.82, χ2 P = .263
      Physical health, muscle powern = 2
      • 0 of 2 trials show a significant effect of vitamin D supplementation on physical health and/or muscle power
      Not applicableLatham et al
      • Latham N.K.
      • Anderson C.S.
      • Lee A.
      • et al.
      A randomized, controlled trial of quadriceps resistance exercise and vitamin D in frail older people: the Frailty Interventions Trial in Elderly Subjects (FITNESS).
      : Physical performance measures, including maximum strength at 3 mo (active vs placebo, median = 12.5 kg [95% CI, 10.5-13.9] vs 11.5 kg [95% CI, 9.0-13.9])

      Corless et al
      • Corless D.
      • Dawson E.
      • Fraser F.
      • et al.
      Do vitamin D supplements improve the physical capabilities of elderly hospital patients?.
      :

      Independence score: r = 0.26, P > .1

      Muscle strength score: r = 0.12, P > .3
      P values are statistically significant at < .05.
      25(OH)D = 25-hydroxyvitamin D; APACHE = acute physiology and chronic health evaluation; CDI = Clostridium difficile infection; CI = confidence interval; CURB = new-onset confusion, urea >7 mmol/L, respiratory rate ≥30 breaths per minute, systolic or diastolic blood pressure <90 mm Hg or ≤60 mm Hg, respectively (pneumonia/LRTI risk scoring system); FIM = functional independence measure; HR = hazard ratio; ICU = intensive care unit; iPTH = intact parathyroid hormone; IQR = interquartile range; LOS = length of stay; MODS = multiorgan dysfunction syndrome; OR = odds ratio; RCDI = recurrence Clostridium difficile infection; RR = relative risk; SOFA = sequential organ failure assessment.
      The same study may have studied more than one clinical outcome.

      Observational Studies: Trials Including Medical Inpatients

      We found 13 trials with a study population consisting of medical inpatients.

      Primary Endpoint: Mortality

      Of the 9 trials that investigated the effect of vitamin D deficiency on mortality, only 3 looked at mortality as the primary endpoint.
      • Leow L.
      • Simpson T.
      • Cursons R.
      • Karalus N.
      • Hancox R.J.
      Vitamin D, innate immunity and outcomes in community acquired pneumonia.
      • Lange N.
      • Litonjua A.A.
      • Gibbons F.K.
      • Giovannucci E.
      • Christopher K.B.
      Pre-hospital vitamin D concentration, mortality, and bloodstream infection in a hospitalized patient population.
      • Amrein K.
      • Quraishi S.A.
      • Litonjua A.A.
      • et al.
      Evidence for a U-shaped relationship between prehospital vitamin D status and mortality: a cohort study.
      They reported a significant association between vitamin D deficiency and increased all-cause mortality. Two additional studies also found a significant relationship: one of them in a subgroup of patients with community-acquired pneumonia (CAP),
      • Remmelts H.H.
      • van de Garde E.M.
      • Meijvis S.C.
      • et al.
      Addition of vitamin D status to prognostic scores improves the prediction of outcome in community-acquired pneumonia.
      and the other in patients hospitalized for acute myocardial infarction.
      • Khalili H.
      • Talasaz A.H.
      • Salarifar M.
      Serum vitamin D concentration status and its correlation with early biomarkers of remodeling following acute myocardial infarction.
      Four studies did not find a significant correlation between vitamin D deficiency and mortality. All studied a specific subgroup of medical inpatients: those with Clostridium difficile infections,
      • Abdelfatah M.
      • Nayfe R.
      • Moftakhar B.
      • et al.
      Low vitamin D level and impact on severity and recurrence of Clostridium difficile infections.
      hepatic cirrhosis,
      • Anty R.
      • Tonohouan M.
      • Ferrari-Panaia P.
      • et al.
      Low levels of 25-hydroxy vitamin D are independently associated with the risk of bacterial infection in cirrhotic patients.
      CAP,
      • Pletz M.W.
      • Terkamp C.
      • Schumacher U.
      • et al.
      Vitamin D deficiency in community-acquired pneumonia: low levels of 1,25(OH)2 D are associated with disease severity.
      or acute kidney injury.
      • Lai L.
      • Qian J.
      • Yang Y.
      • et al.
      Is the serum vitamin D level at the time of hospital-acquired acute kidney injury diagnosis associated with prognosis?.

      Secondary Endpoints: Infections and Other Health Outcomes

      Eight trials assessed the effect of vitamin D deficiency either on the rate of infections or their severity. Two reported a higher incidence of bloodstream infections in vitamin D-deficient patients.
      • Lange N.
      • Litonjua A.A.
      • Gibbons F.K.
      • Giovannucci E.
      • Christopher K.B.
      Pre-hospital vitamin D concentration, mortality, and bloodstream infection in a hospitalized patient population.
      • Quraishi S.A.
      • Litonjua A.A.
      • Moromizato T.
      • et al.
      Association between prehospital vitamin D status and hospital-acquired bloodstream infections.
      Quraishi et al
      • Quraishi S.A.
      • Litonjua A.A.
      • Moromizato T.
      • et al.
      Association between prehospital vitamin D status and hospital-acquired bloodstream infections.
      looked at hospital-acquired bloodstream infections in particular. One study found an association between vitamin D deficiency and higher sepsis severity, and reported a worse prognosis in septic patients.
      • Ginde A.A.
      • Camargo Jr., C.A.
      • Shapiro N.I.
      Vitamin D insufficiency and sepsis severity in emergency department patients with suspected infection.
      Three trials studied patients with Clostridium difficile infections: 2 reported a significantly higher severity of infection in vitamin D-deficient patients, compared with sufficient patients.
      • Abdelfatah M.
      • Nayfe R.
      • Moftakhar B.
      • et al.
      Low vitamin D level and impact on severity and recurrence of Clostridium difficile infections.
      • Wang W.J.
      • Gray S.
      • Sison C.
      • et al.
      Low vitamin D level is an independent predictor of poor outcomes in Clostridium difficile-associated diarrhea.
      The remaining study found worse progress of disease, with fewer resolutions of infection being associated with deficiency.
      • van der Wilden G.M.
      • Fagenholz P.J.
      • Velmahos G.C.
      • Quraishi S.A.
      • Schipper I.B.
      • Camargo Jr., C.A.
      Vitamin D status and severity of Clostridium difficile infections: a prospective cohort study in hospitalized adults.
      Pletz et al
      • Pletz M.W.
      • Terkamp C.
      • Schumacher U.
      • et al.
      Vitamin D deficiency in community-acquired pneumonia: low levels of 1,25(OH)2 D are associated with disease severity.
      reported a significant association between lower 1,25(OH)D levels and increased pneumonia severity, but the same association with 25(OH)D was not significant. The incidence of bacterial infections in cirrhotic patients has been found to not be associated with vitamin D levels.
      • Anty R.
      • Tonohouan M.
      • Ferrari-Panaia P.
      • et al.
      Low levels of 25-hydroxy vitamin D are independently associated with the risk of bacterial infection in cirrhotic patients.
      Only 1 study looked at intensive care unit (ICU) admission rates and found a significant inverse relationship with vitamin D levels.
      • Remmelts H.H.
      • van de Garde E.M.
      • Meijvis S.C.
      • et al.
      Addition of vitamin D status to prognostic scores improves the prediction of outcome in community-acquired pneumonia.
      An investigation of a correlation between vitamin D levels and length of hospital stay showed contradicting results.
      • Khalili H.
      • Talasaz A.H.
      • Salarifar M.
      Serum vitamin D concentration status and its correlation with early biomarkers of remodeling following acute myocardial infarction.
      • Pletz M.W.
      • Terkamp C.
      • Schumacher U.
      • et al.
      Vitamin D deficiency in community-acquired pneumonia: low levels of 1,25(OH)2 D are associated with disease severity.

      Observational Studies: Trials Including Intensive Care Unit, Geriatric, and Rehabilitation Patients

      We found 10 trials with a population of ICU patients, often with mixed medical and surgical patients. For geriatric patients we found 4 trials, whereas rehabilitation patients had been studied in 2 trials that were included in this review.

      Primary Endpoint: Mortality

      All 10 studies including ICU patients examined the association between vitamin D deficiency and mortality and reported contradictory results. Six reported a significant relationship with increased mortality,
      • Amrein K.
      • Zajic P.
      • Schnedl C.
      • et al.
      Vitamin D status and its association with season, hospital and sepsis mortality in critical illness.
      • Braun A.
      • Chang D.
      • Mahadevappa K.
      • et al.
      Association of low serum 25-hydroxyvitamin D levels and mortality in the critically ill.
      • McKinney J.D.
      • Bailey B.A.
      • Garrett L.H.
      • Peiris P.
      • Manning T.
      • Peiris A.N.
      Relationship between vitamin D status and ICU outcomes in veterans.
      • Hu J.
      • Luo Z.
      • Zhao X.
      • et al.
      Changes in the calcium-parathyroid hormone-vitamin D axis and prognosis for critically ill patients: a prospective observational study.
      • 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.
      • Moromizato T.
      • Litonjua A.A.
      • Braun A.B.
      • Gibbons F.K.
      • Giovannucci E.
      • Christopher K.B.
      Association of low serum 25-hydroxyvitamin D levels and sepsis in the critically ill.
      although Amrein et al found that this relationship held good only for in-hospital mortality and not for ICU mortality. Four studies, of which one studied a subgroup of patients with severe sepsis, found no significant correlation.
      • Aygencel G.
      • Turkoglu M.
      • Tuncel A.F.
      • Candir B.A.
      • Bildaci Y.D.
      • Pasaoglu H.
      Is vitamin D insufficiency associated with mortality of critically ill patients?.
      • Su L.X.
      • Jiang Z.X.
      • Cao L.C.
      • et al.
      Significance of low serum vitamin D for infection risk, disease severity and mortality in critically ill patients.
      • Higgins D.M.
      • Wischmeyer P.E.
      • Queensland K.M.
      • Sillau S.H.
      • Sufit A.J.
      • Heyland D.K.
      Relationship of vitamin D deficiency to clinical outcomes in critically ill patients.
      • Cecchi A.
      • Bonizzoli M.
      • Douar S.
      • et al.
      Vitamin D deficiency in septic patients at ICU admission is not a mortality predictor.
      The only study that included a geriatric population and had mortality as an endpoint showed a significant association between vitamin D deficiency and short-term mortality in geriatric patients.
      • Annweiler C.
      • Pochic S.
      • Fantino B.
      • et al.
      Serum vitamin D concentration and short-term mortality among geriatric inpatients in acute care settings.

      Secondary Endpoints: Disease Severity and Other Health Outcomes

      Hu et al
      • Hu J.
      • Luo Z.
      • Zhao X.
      • et al.
      Changes in the calcium-parathyroid hormone-vitamin D axis and prognosis for critically ill patients: a prospective observational study.
      and Aygencel et al,
      • Aygencel G.
      • Turkoglu M.
      • Tuncel A.F.
      • Candir B.A.
      • Bildaci Y.D.
      • Pasaoglu H.
      Is vitamin D insufficiency associated with mortality of critically ill patients?.
      whose study included ICU patients, report a significant relationship between higher disease severity and vitamin D deficiency, with deficient patients also facing a higher risk of multiorgan dysfunction syndrome. Moromizato et al
      • Moromizato T.
      • Litonjua A.A.
      • Braun A.B.
      • Gibbons F.K.
      • Giovannucci E.
      • Christopher K.B.
      Association of low serum 25-hydroxyvitamin D levels and sepsis in the critically ill.
      found a significantly higher risk for sepsis in deficient patients. Two of five studies found no significant correlation between vitamin D levels and disease severity.
      • Aygencel G.
      • Turkoglu M.
      • Tuncel A.F.
      • Candir B.A.
      • Bildaci Y.D.
      • Pasaoglu H.
      Is vitamin D insufficiency associated with mortality of critically ill patients?.
      • Su L.X.
      • Jiang Z.X.
      • Cao L.C.
      • et al.
      Significance of low serum vitamin D for infection risk, disease severity and mortality in critically ill patients.
      • Higgins D.M.
      • Wischmeyer P.E.
      • Queensland K.M.
      • Sillau S.H.
      • Sufit A.J.
      • Heyland D.K.
      Relationship of vitamin D deficiency to clinical outcomes in critically ill patients.
      Two of three studies did not find a significant relationship between vitamin D levels and LOS
      • van der Wilden G.M.
      • Fagenholz P.J.
      • Velmahos G.C.
      • Quraishi S.A.
      • Schipper I.B.
      • Camargo Jr., C.A.
      Vitamin D status and severity of Clostridium difficile infections: a prospective cohort study in hospitalized adults.
      • Hu J.
      • Luo Z.
      • Zhao X.
      • et al.
      Changes in the calcium-parathyroid hormone-vitamin D axis and prognosis for critically ill patients: a prospective observational study.
      ; only one study reports a significantly higher LOS in patients with vitamin D deficiency.
      • Braun A.
      • Chang D.
      • Mahadevappa K.
      • et al.
      Association of low serum 25-hydroxyvitamin D levels and mortality in the critically ill.
      Two studies evaluated blood culture positivity in relationship to vitamin D status, again with opposing results.
      • van der Wilden G.M.
      • Fagenholz P.J.
      • Velmahos G.C.
      • Quraishi S.A.
      • Schipper I.B.
      • Camargo Jr., C.A.
      Vitamin D status and severity of Clostridium difficile infections: a prospective cohort study in hospitalized adults.
      • McKinney J.D.
      • Bailey B.A.
      • Garrett L.H.
      • Peiris P.
      • Manning T.
      • Peiris A.N.
      Relationship between vitamin D status and ICU outcomes in veterans.
      Two studies report significantly longer LOS in geriatric patients with vitamin D deficiency.
      • Helard L.
      • Mateus-Hamdan L.
      • Beauchet O.
      • Annweiler C.
      Hypovitaminosis D in geriatric acute care unit: a biomarker of longer length of stay.
      • Beauchet O.
      • Launay C.
      • de Decker L.
      • Fantino B.
      • Kabeshova A.
      • Annweiler C.
      Who is at risk of long hospital stay among patients admitted to geriatric acute care unit? Results from a prospective cohort study.
      Del Galy et al
      • Sutra Del Galy A.
      Vitamin D insufficiency and acute care in geriatric inpatients.
      found a significant inverse correlation between a number of acute diseases and vitamin D levels in geriatric patients. Two studies on the relationship between vitamin D levels and functional outcome measures among patients in a rehabilitation setting did not document a positive correlation between the two.
      • Aygencel G.
      • Turkoglu M.
      • Tuncel A.F.
      • Candir B.A.
      • Bildaci Y.D.
      • Pasaoglu H.
      Is vitamin D insufficiency associated with mortality of critically ill patients?.
      • Kiebzak G.M.
      • Moore N.L.
      • Margolis S.
      • Hollis B.
      • Kevorkian C.G.
      Vitamin D status of patients admitted to a hospital rehabilitation unit: relationship to function and progress.
      • Pellicane A.J.
      • Wysocki N.M.
      • Mallinson T.R.
      • Schnitzer T.J.
      Prevalence of 25-hydroxyvitamin D deficiency in the acute inpatient rehabilitation population and its effect on function.

      Randomized, Controlled Trials

      Amrein et al
      • Amrein K.
      • Schnedl C.
      • Holl A.
      • et al.
      Effect of high-dose vitamin D3 on hospital length of stay in critically ill patients with vitamin D deficiency: the VITdAL-ICU randomized clinical trial.
      conducted an RCT in 475 critically ill patients, looking at the effects of supplementation with high-dose vitamin D3 on the LOS and mortality (both in-hospital mortality and at mortality at 6 months). There was no significant difference in these endpoints between the intervention and control groups. Although it was not part of the initial hypothesis, they performed a subgroup analysis of severely vitamin D-deficient patients (defined as ≤12 ng/mL) and reported significantly lower in-hospital mortality in the severely deficient patients in the intervention group.
      The other RCTs that were reviewed focused mainly on the prevention of falls in patients. Latham et al
      • Latham N.K.
      • Anderson C.S.
      • Lee A.
      • et al.
      A randomized, controlled trial of quadriceps resistance exercise and vitamin D in frail older people: the Frailty Interventions Trial in Elderly Subjects (FITNESS).
      examined the impact of 2 different interventions—a resistance exercise program and treatment with calciferol—on physical function and self-reported measures of physical function in an RCT that included 243 patients in the geriatric rehabilitation unit. Neither of the interventions showed a significant reduction in risk of falls or an improvement in physical health. The resistance exercise intervention even seemed to be harmful, with more fatigue and a higher incidence of musculoskeletal injuries. Another RCT performed by Bischoff-Ferrari et al in 2003
      • Bischoff H.A.
      • Stahelin H.B.
      • Dick W.
      • et al.
      Effects of vitamin D and calcium supplementation on falls: a randomized controlled trial.
      examined the impact of adding 400 IU vitamin D to 600 mg calcium on falls and musculoskeletal function over a period of 12 weeks in patients in long-stay geriatric care units. The combination of calcium and vitamin D reduced the number of falls per person by 49%. Corless et al
      • Corless D.
      • Dawson E.
      • Fraser F.
      • et al.
      Do vitamin D supplements improve the physical capabilities of elderly hospital patients?.
      conducted an RCT in geriatric patients that examined the effect of 9000 IU vitamin D2 over a period of up to 40 weeks on activities of daily living. During these 40 weeks no significant difference in changes in muscle power or independence scores could be detected between the 2 groups. Another RCT performed by Burleigh et al
      • Burleigh E.
      • McColl J.
      • Potter J.
      Does vitamin D stop inpatients falling? A randomised controlled trial.
      that included patients in an acute geriatric ward examined the effects of 800 IU vitamin D added to calcium vs 1200 mg of calcium alone on falls during hospitalization. The authors reported no significant difference in number of falls among patients in the 2 groups, as well as no improvements in functional measures.

      Discussion

      Our search strategy and inclusion criteria resulted in the inclusion of a total of 34 trials that examined the effect of vitamin D deficiency and its treatment on a variety of medical outcomes. Our review revealed inconsistent findings. With regard to the relationship between vitamin D deficiency and mortality, the results varied between the different studies and populations. Multiple studies with medical inpatients found a significantly higher incidence or severity of infection in patients with vitamin D deficiency. However, the comparability of the findings of the different studies is limited by differences in the patient subpopulations (eg, patients with Clostridium difficile infections or CAP) studied. There was also no consensus among the different trials that examined LOS as an endpoint. Only in the geriatric population did all reviewed trials report a significant association between vitamin D deficiency and longer LOS. No study of general medical inpatients has examined the impact of vitamin D deficiency on functional outcome measures. This aspect has only been studied in a rehabilitation setting. We also found no studies examining quality of life as an endpoint.
      There were only 5 RCTs that evaluated the effects of vitamin D supplementation on the outcome measures in question (ie, mortality, functional outcome measures), with 1 of them including patients in the ICU, whereas the remaining 4 included patients in geriatric rehabilitation units. The majority of them did not find a significant association between vitamin D supplementation and the particular outcome studied. These trials may be limited by an inadequate number of patients, or heterogeneity in the chosen duration and dose of the treatment, which might have been too short or too low.
      We also observed that the assay method and definition of vitamin D deficiency varied across the trials we reviewed. The results also may have been influenced by the known variability in specificity and sensitivity between the individual methods of measurement, as well as by the variability between different laboratories. The dosage and duration of vitamin D supplementation in the different RCTs ranged from a single loading dose of 540,000 IU to 400 IU daily over a period of 4 months. We found these interventions to be chosen quite randomly, without reference to official recommendations.
      An important limitation of this review was the heterogeneity of the reviewed trials, in terms of the populations as well as the endpoints studied. This, of course, limits the comparability and generalization of these results. Our search did not retrieve any trials providing an overview of adverse outcomes in medical inpatients, although it was our intention to retrieve and review such studies. Even though vitamin D deficiency and its impact on health issues is widely discussed and extensive research has been conducted in this field, we discovered that the findings are often limited to a very specific patient population and cannot be generalized for medical inpatients presenting to the emergency department.
      Several large trials investigating the effects of vitamin D supplementation on nonskeletal outcomes such as cancer, cardiovascular disease, and mortality in the general population are currently underway. The findings of these studies may fill an important knowledge gap. However, the question remains as to whether screening and supplementation of a pre-existing deficiency could improve additional acute medical outcome measures that are yet to be explored in detail and thus reduce health care costs by reducing the use of hospital resources. Although the findings we summarized are inconsistent and often lack statistical significance, they suggest a possible association of vitamin D deficiency with short-term adverse outcomes. As with all observational studies, one has to examine the question of causality. Is vitamin D deficiency the reason for a worse outcome or merely a marker for higher disease severity and poorer disease progression? Well-designed RCTs with adequate statistical power are needed to confirm or disprove these hypotheses. We therefore see the need for more observational, as well as randomized controlled studies on these topics.
      In conclusion, despite evidence of associations in observational studies, high-quality evidence from interventional trials on the benefits of vitamin D supplementation in patients in an acute care setting is currently lacking. On the basis of the available evidence, general vitamin D screening and supplementation for the medical inpatient population presenting at the emergency department cannot be recommended.

      Acknowledgments

      The authors thank Heidrun Janka from the University Library, Basel for her help with the literature search. Assistance in the preparation of this manuscript was provided by Prasad Kulkarni, PhD, CMPP, of Asclepius Medical Communications LLC, Ridgewood, NJ, and was funded by the author.

      Appendix

      Appendix 1Observational Studies Evaluating the Impact of Vitamin D Deficiency on Clinical Outcomes
      First Author, Y, JournalTitlePopulation (n)Definition of 25(OH)D Deficiency and Method of Vitamin D MeasurementEndpoints and Follow-Up Period (FU)Key FindingsBias/Limitations
      Medical inpatients
       Karin Amrein, 2014, Journal of Clinical Endocrinology and MetabolismEvidence for a U-Shaped Relationship Between Prehospital Vitamin D Status and Mortality: A Cohort StudyAdults hospitalized for acute care (n = 24,094)
      • Severe deficiency: <10 ng/mL
      • Deficiency: 10-19.9 ng/mL
      • Insufficiency: 20-29.9 ng/mL
      • Method: Chemiluminescence assay, radioimmunoassay, or liquid chromatography-mass spectroscopy
      • All-cause 90-d mortality
      • 30-d, 365-d, and in-hospital mortality
      • Follow-up = 30, 90, 365 d
      Preadmission 25(OH)D levels <30 ng/mL and ≥60 ng/mL were significantly associated with an increased risk of 90- and 365-d mortality after hospital admission
      • Adjusted associations with all-cause 90-d mortality:
        • <10 ng/mL: OR 2.01; 95% CI, 1.75-2.45; P < .0001
        • 10-19.9 ng/mL: OR 1.89; 95% CI, 1.75-2.45; P < .0001
        • 20-29.9 ng/mL: OR 1.34; 95% CI, 1.09-1.45, P = .002
        • 30-49.9 ng/mL: OR 1 (referent)
        • 50-59.9 ng/mL: OR 0.94; 95% CI, 0.69-1.26; P = .67
        • 60-69.9 ng/mL: OR 1.52; 95% CI, 1.03-2.25; P = .037
        • ≥70 ng/mL: OR 1.69; 95% CI, 1.09-2.61; P = .018
      • Adjusted associations with all-cause 365-d mortality:
        • <10 ng/mL: OR 2.58; 95% CI, 2.26-2.96; P < .0001
        • 10-19.9 ng/mL: OR 2.0; 95% CI, 1.79-2.23; P < .0001
        • 20-29.9 ng/mL: OR 1.39; 95% CI, 1.24-1.56; P < .0001
        • 30-49.9 ng/mL: OR 1 (referent)
        • 50-59.9 ng/mL: OR 1.09; 95% CI, 0.88-1.35: P = .45
        • 60–69.9 ng/mL: OR 1.33; 95% CI, 0.96-1.82; P = .083
        • ≥70 ng/mL or greater: OR 1.66; 95% CI, 1.18-2.34; P = .004
      (Adjusted for age, gender, race, patient type, Deyo-Charlson index, season)
      • Selection bias: only patients with 25(OH)D measured before hospitalization were included. This limits the generalizability to all hospitalized patients
      • Ascertainment bias: not every patient had 25(OH)D measured before hospitalization → Decreased generalizability
      • Retrospective, observational study design (limited causality)
      • No adjustment for PTH levels and albumin measurements (possible correlation between low albumin and low Vitamin D)
       Nancy Lange, 2013, American Journal of MedicinePre-hospital Vitamin D Concentration, Mortality, and Bloodstream Infection in a Hospitalized Patient PopulationHospitalized patients (medical (n = 14,190) + surgical) with serum 25(OH)D measured between 7 and 365 d before hospital admission (total n = 23,603)
      • Primary analysis: deficiency: <15 ng/mL
      • Secondary analysis: deficiency: <20 ng/mL
      • Method: Between 1993 and 2007, the chemiluminescence assay, the radioimmunoassay, or mass spectroscopy was used at different times as a 25(OH)D assay method
      • 30-d all-cause mortality after hospital admission
      • All cause in-hospital mortality and bloodstream infections
      • Follow-up = not specified
      Statistically significant association between lower pre-hospital 25(OH)D <15 ng/mL and higher mortality as well as more bloodstream infections
      • Adjusted association with 30-d mortality:
        • ≤15 ng/mL: OR 1.45; 95% CI, −1.21, 1.74; P < .0001
        • 15-30 ng/mL: OR 1.30; 95% CI, −1.10, 1.54; P = .003
        • ≥30 ng/mL: OR 1.00
      • Adjusted association with positive blood cultures:
        • ≤15 ng/mL: OR 1.29; 95% CI, 1.06-1.57; P = .01
        • 15-30 ng/mL: OR 1.01; 95% CI, 0.83-1.21; P = .9
        • ≥30 ng/mL: OR 1.00
      (adjusted for age, gender, race, Deyo-Charlson index, season, patient type [surgical vs medical], creatinine, blood urea nitrogen, hematocrit, time between 25(OH)D draw and hospital admission)
      • Retrospective cohort study
      • Information bias: accuracy of exposure, covariate, and outcome measures between the vitamin D groups may differ
      • Ascertainment bias: the patient cohort under study had vitamin D measured for unknown reasons
      • Limited generalizability
      • 25(OH)D levels are not drawn at the time of hospital admission
       Adit A. Ginde, 2011, Academic Emergency MedicineVitamin D Insufficiency and Sepsis Severity in Emergency Department (ED) Patients With Suspected InfectionPatients at ED with hospital admission planned for suspected infection (n = 81)
      • Insufficiency: <75 nmol/L
      • Method: Liquid chromatography–tandem mass spectrometry
      • Association of Vitamin D deficiency with higher infection severity
      • Severe sepsis: sepsis + acute dysfunction of ≥1 organ system
      • Illness severity: Acute Physiology Age Chronic Health Evaluation (APACHE) II, and Sepsis-related Organ Failure Assessment (SOFA) scores
      • Follow-up = clinical and laboratory measures assessed at 2 time points: baseline (in the ED) and 24 (±2) h later
      Significant association of vitamin D deficiency with all markers of illness severity at baseline and 24-h time points except with the APACHE II score

      Baseline:
      • Severe sepsis:
        • <75 nmol/L = 39% (n = 61); ≥75 nmol/L = 4% (n = 24)
        • OR 0.26; 95% CI, 0.07-0.96; P = .006
      • SOFA-Score >2:
        • <75 nmol/L: 28% (n = 44); ≥75 nmol/L = 3% (n = 18)
        • OR 0.34; 95% CI, 0.09–1.34; P = .049
      • 24h: Severe sepsis:
        • <75 nmol/L = 43% (n = 67); ≥75 nmol/L = 5 (n = 29)
        • OR 0.28; 95% CI, 0.08-1.01; P = .005 Two or more organ system dysfunction:
        • <75 nmol/L: 32% (n = 50); >75 nmol/L = 3% (n = 18)
        • OR 0.30; 95% CI, 0.07-1.25; P = .02
      • APACHE II score >25:
        • <75 nmol/L = 12% (n = 19); ≥75 nmol/L = 0
        • OR not calculable, P = .06
      • SOFA score >2:
        • <75 nmol/L = 40% (n = 63); ≥75 nmol/L = 5% (n = 29)
        • OR 0.35; 95% CI, 0.10-1.24; P = .02
      Significant association between vitamin D insufficiency and worsening clinical course (measured by increase in SOFA score from baseline to 24 h)
      • <75 nmol/L = 29% (n = 45); ≥75 nmol/L = 3% (n = 18)
      • OR 0.34; 95% CI, 0.85-1.36; P = .04
      • Reduced risk of detection bias due to blinding of outcome assessors to vitamin D status and objective of the study
      • Pilot study, exploratory nature, observational design (cannot prove causal relationships)
      • Modest sample size: underpowered to perform robust multivariable analyses or other inferential analyses
      • Possible residual confounding even after adjustment for age
       Sadeq A. Quraishi, 2013, American Journal of Clinical NutritionAssociation Between Prehospital Vitamin D Status and Hospital-Acquired Bloodstream InfectionsLaboratory and administrative data from the electronic medical records of patients hospitalized in 2 different hospitals (n = 2135)
      • Deficiency: <20 ng/mL
      • Method: Chemiluminescence assay, radioimmunoassay, or liquid chromatography mass, 7-365 d before hospitalization
      • Association between prehospital serum 25(OH)D concentrations and risk of hospital-acquired bloodstream infections (HABSI: >48 h after hospital admission)
      • Follow-up = not specified
      Prehospital 25(OH)D <10 ng/mL was associated with a significant increase in odds of HABSI in hospitalized adults 25(OH)D-HABSI association seems to be preserved when 25(OH)D is obtained within 90 d of admission

      Adjusted association between 25(OH)D and bloodstream infections:
      • <10 ng/mL: OR 1.95; 95% CI, 1.22-3.12; P < .005
      • <20 ng/mL: OR 1.68; 95% CI, 1.22-2.33; P < .002
      • >20 ng/mL: OR 1.00 (referent)
      (Not confounded by assay type, year of hospital admission, creatinine, blood urea nitrogen, white blood cell count, hematocrit, or time between 25(OH)D draw and hospital admission)
      • Ascertainment bias: vitamin D measured for unknown reasons that may have been absent in other hospitalized patients → limited generalizability to all hospitalized patients
      • Observational design (limited proof of causality)
      • Possible residual confounding even after adjustment: no adjustment for use of glucocorticoids, sun exposure, immobilization, alcohol intake, smoking status, or genetics
      • 25(OH)D not drawn at the time of hospital admission
       Mohamed Abdelfatah, 2015, Journal of Investigative MedicineLow Vitamin D Level and Impact on Severity and Recurrence of Clostridium difficile InfectionsPatients hospitalized at a single institution with C. difficile infection (CDI) and with a positive C. difficile toxin assay (n = 271)
      • Severe deficiency: <10 ng/mL
      • Deficiency: <20 ng/mL
      • Insufficiency: 10-19.9 ng/mL
      • Pre-deficiency: 20-29.9 ng/mL
      • Control: ≥30 ng/mL
      • Method: Not specified
      • Recurrence Clostridium difficile infections (RCDI): after resolution of the initial CDI episode for at least 10 d and after discontinuation of the CDI therapy within 90 d
      • Severity of CDI and 30-d mortality in relation to vitamin D level
      • Follow-up = 90 d
      Significant association with more severe CDI
      • Severe deficiency: 7 of 40 patients (17.5%); insufficiency: 17 of 93 (18.2%); pre-deficiency: 9 of 64 (14%); control: 3 of 74 (4%); P = .04
      No significant association with increased 30-d mortality
      • Severe deficiency: 1 of 40 patients (2.5%); insufficiency: 1 of 93 (1%); pre-deficiency: 2 of 64 (3.1%); control: 2 of 74 (2.7%); P = .43
      No significant difference in recurrent CDI between the different degrees of deficiency
      • Severe deficiency: 5 of 40 patients (12.5%); insufficiency: 15 of 93 (16.1%); pre-deficiency: 10 of 64 (15.6%); control: 13 of 74 (17.5%); P = .55
      Normal vitamin D level may have a protective effect against severe CDI
      • Retrospective design
      • Small sample size
      • The timing of the blood samples drawn for vitamin D levels and CDI diagnoses may not have been standardized for all patients
       Wallace J. Wang, 2014, Therapeutic Advances in GastroenterologyLow Vitamin D Level Is an Independent Predictor of Poor Outcomes in Clostridium Difficile-Associated DiarrheaPatients with manifestations of CDAD (Clostridium difficile-associated diarrhea) and a positive C. difficile toxin assay (n = 62)
      • Low levels: 25(OH)D levels <21 ng/mL
      • Method: Not specified
      • Impact of low vitamin D levels in CDAD resolution (classified as having resolved, recurred or expired)

      • Follow-up = 30 d after discharge
      Association of vitamin D levels with 30-d status

      Vitamin D level and age independently associated with 30-d status (recurred/expired vs resolved). Patients with hypovitaminosis D had a nearly 5 times greater likelihood of nonresolution of CDAD than patients with normal vitamin D levels
      • Adjusted OR 4.75; 95% CI, 1.18–19.1; P = .028
      • After backward elimination: Adjusted OR 5.6; 95% CI, 1.49-21.12; P = .011
      • Small sample size
      • Patients considered resolved based on history obtained via a telephone survey, without follow-up stool studies
      • High mortality rate in this cohort of hospitalized patients with multiple comorbidities
      • Analysis of patients with CDI only, limits generalizability
       Gwendolyn M. van der Wilden, 2013, Journal of Parenteral and Enteral NutritionVitamin D Status and Severity of Clostridium difficile Infections: A Prospective Cohort Study in Hospitalized AdultsPatients with confirmed CDI (n = 100)
      • Severe deficiency: <10 ng/mL
      • Deficiency: 10-19.9 ng/mL Insufficiency: 20-29.9 ng/mL
      • Normal: ≥30 ng/mL
      • Method: Liquid chromatography–tandem mass spectrometry
      • Investigate whether vitamin D status is associated with the severity of CDI in hospitalized patients
      • Follow-up = until hospital discharge
      25(OH)D levels are associated with CDI severity (Simple: Group A; Severe: Group B):
      • Normal: Group A, n (%): 19 (27), Group B: 5 (17); P = .31
      • Insufficiency: Group A: 24 (34), Group B : 9 (31); P = .79
      • Deficiency: Group A: 22 (31), Group B: 12 (41), P = .32
      • Severe deficiency: Group A: 6 (9), Group B: 3 (10), P = .76
      • In multivariable regression model 25(OH)D levels (adjusted OR 0.92; 95% CI, 0.87-0.98; P = .008) and recent hospitalization were significantly associated with CDI severity
      • Stepwise regression analysis did not change the association of 25(OH)D3 with CDI severity: adjusted OR 0.93; 95% CI, 0.88-0.98
      Study suggests that every 1-ng/mL increase in 25(OH)D3 was associated with an 8% decrease in the risk of CDI severity
      • Blood samples for vitamin D status assessments drawn after the diagnosis of CDI → reverse causation cannot be ruled out
      • Single center study
      • Adjusted for the burden of chronic disease by, not not for other important confounders and covariates
      • Analysis of patients with CDI only, limits generalizability
       Rodolphe Anty, 2014, Clinical and Translational GastroenterologyLow Levels of 25-Hydroxy Vitamin D are Independently Associated with the Risk of Bacterial Infection in Cirrhotic PatientsHospitalized patients with cirrhosis, irrespective of the underlying etiology (n = 88)
      • Severe deficiency: <10 ng/mL
      • Deficiency: 10-20 ng/mL
      • Method: Direct competitive chemiluminescence immunoassay
      Comparison of the 25-OH vitamin D serum level in patients with and without bacterial infection
      • Follow-up = median 212 (range, 66-344 d)
      • A severe deficiency in vitamin D was significantly more frequent in infected patients compared with noninfected patients (P = .019)
      • Logistic regression analysis: severe deficiency in 25(OH)D level was associated with the risk of infection, independently of the Child-Pugh score and CRP level
      • Prediction of mortality: Cox regression model (including the Child-Pugh score, the presence of an infection, and the presence of a severe deficiency in 25-OH vitamin D): only the presence of an infection was significantly associated with the mortality (relative risk = 3.24 [1.2-8.76], P = .02)
      • Monocentric study
      • Limited sample size: did not allow detection of a difference in mortality
      • Analysis of patients with hepatic cirrhosis only, limits generalizability
       Lingyun Lai, 2013, PLoS OneIs the Serum Vitamin D Level at the Time of Hospital- Acquired Acute Kidney Injury Diagnosis Associated with Prognosis?AKI patients (n = 200),

      surgical + medical controls: critically ill patients without AKI (n = 13) and age- and gender-matched healthy subjects for comparison (n = 17) (total n = 230)
      • Deficiency: <37.5 nmol/L
      • Insufficiency: <75 nmol/L
      • Sufficiency: ≥75 nmol/L
      • 25-(OH)D:
        • Low <30.5 nmol/L
        • High >30.5 nmol/L
      • 1,25- (OH)2D:
        • Low <44.3 pmol/L
        • High >44.3 pmol/L
      • Method:
        • 25 (OH)D: enzyme-linked immunosorbant assay (ELISA)
        • 1,25-(OH): enzyme immunoassay
      • Possible association between the low vitamin D status with AKI and the 90-d overall mortality rate • AKI was determined using the Risk, Injury, Failure, Loss, and End-stage kidney (RIFLE) classification criteria • Follow-up = 90 d• Significant increase of the degree of 1,25 (OH)D with increased AKI severity. This significant correlation could not be showed with 25(OH)D levels

      1,25-(OH)2D: Risk = 72.6 ± 69.4 pmol/L; Injury = 53.7 ± 32.7 pmol/L; Failure = 42.2 ± 29.3 pmol/L; P = .042

      25-(OH)D: Risk = 33.3 ± 13.9 nmol/L; Injury = 41.0 ± 19.5 nmol/L; Failure = 32.6 ± 10.5 nmol/L; P = .083

      • Neither 25-(OH)D nor 1,25-(OH)2 D level at the time of AKI diagnosis correlated significantly with 90-d all-cause mortality.

      25 (OH)D: High: 45.8% (n = 22), Low: 47.9% (n = 23)

      Survival curve: no significant difference between those 2 groups according to the Kaplan-Meier plot (P = .554 by the log-rank test)

      1,25 (OH)D High: 43.8% (n = 21); Low: 50.0% (n = 24)

      Survival curve also showed no significant difference between those 2 groups according to the Kaplan-Meier plot (P = .147 by the log-rank test)

      Subgroup analysis with septic patients: survival curve: no significant difference between high and low 25-(OH)D level (P = .840 by the log-rank test) or between high and low 1,25-(OH)2D level (P = .877 by the log-rank test) according to the Kaplan-Meier plot
      • Small, single-center, observational cohort study
      • No adjustment for potential influencing variables
      • The causes of AKI are very complex and multiplex, potentially confounding one another
      • Analysis of patients with AKI only, limits generalizability
       Hilde H. F. Remmelts, 2012, Clinical Infectious DiseasesAddition of Vitamin D Status to Prognostic Scores Improves the Prediction of Outcome in Community-Acquired PneumoniaData used from adult patients who participated in a previous RCT (patients with confirmed CAP) (n = 272)
      • Deficiency: <50 nmol/L
      • Insufficiency: 50-75 nmol/L
      • Sufficiency: >75 nmol/L
      • Method: TOTAL Liaison chemiluminescence assay
      • Impact of vitamin D status on clinical outcome in CAP patients
      • Contribution of vitamin D status to the prognostic accuracy of other biomarkers and commonly used prognostic scores
      • Hypothesis: low 25-(OH) D level is related to adverse outcome, due to the lack of immunomodulatory activities in case of deficiency
      • Follow-up = 30 d after hospital admission
      • Significant association between Vitamin D deficiency and higher rate of ICU admission (ICU: 34.9 nmol/L [IQR, 23.8-46.3]; non-ICU: 48.3 nmol/L [IQR, 30.8-68.4]; P = .04)
      • Significant association with Vitamin D deficiency and lower 30-d survival (Dead: 25.8 nmol/L [IQR, 19.8-40.1]; Alive: 48.8 nmol/L [IQR, 32.4–68.9], P < .01)
      • Statistically significant inverse correlation between vitamin D levels and the composite endpoint mortality/ICU admission: 50-75 nmol/L: Adjusted OR 0.510; 95% CI, 0.068-3.807; <50 nmol/L: Adjusted OR 2.949; 95% CI, 0.640-13.598
      • Adding vitamin D status to predictors of 30-d mortality, such as cortisol level; PSI score, and CURB-65 score, individually, improved all prognostic accuracies. Best prognostic accuracy: the combination of vitamin D status and PSI score
      • Cortisol + vitamin D: AUC = 0.74; 95% CI, 0.63-0.85. Regression coefficient (for vitamin D status) = 1.782, P (for vitamin D status) = .021
      • CURB-65 score + vitamin D: AUC = 0.84; 95% CI, 0.76-0.91, Regression coefficient (for vitamin D status) = 2.268, P (for vitamin D status) = 0.032
      • PSI score + vitamin D: AUC = 0.83, 95% CI 0.71-0.94, regression coefficient (for vitamin D status) = 1.832, P (for vitamin D status) = 0.021
      • Small sample size, only small number of patients reached endpoints
      • Observational design of the study, not able to establish a causal relationship between vitamin D deficiency and adverse outcome.
      • The presence of unmeasured confounding cannot be ruled out
      • Analysis limited to patients with CAP, limits generalizability
       Mathias W. Pletz, 2014, Respiratory ResearchVitamin D Deficiency in Community-Acquired Pneumonia: Low Levels of 1,25(OH)2 D Are Associated With Disease SeverityRandom sample of patients with confirmed pneumonia enrolled into the large prospective CAPNETZ cohort study (n = 300)
      • Deficiency: <20 ng/mL
      • Insufficiency: 21-29 ng/mL
      • Sufficiency: 30-100 ng/mL
      • Method: 25-OH-chemiluminescent immunoassay 1,25-OH-1,25-Dihydroxy Vitamin D EIA kit (immunoextraction followed by quantitation by enzyme immunoassay)
      • Identify associations between disease severity and/or specific respiratory pathogens in community-acquired pneumonia (CAP) and the levels of 25-OH and 1,25-OH2
      • Follow-up = 30 ± 2 d and 180 d after enrollment
      • Significant association between pneumonia severity (measured by CURB-score) and 1,25-OH2 serum levels (P = .011), but not with 25-OH levels (P = .325)
      • Hospitalization rate: 25-(OH)D levels significantly lower for patients who needed hositalization (12.8 ± 7.8 vs 16.2 ± 10.0 ng/mL); 1,25-(OH2)D levels significantly lower for patients who needed to be hospitalized: 44.3 ± 28.2 vs 58.2 ± 30.0 pg/mL
      • Significant impact of 25(OH)D (F1,290 = 17.69 P < .001) and 1,25-OH2 serum concentrations (F1,290 = 46.16, P < .001) on duration of hospitalization
      • No significant correlation between 25(OH)D levels (χ2/1 = 0.2374; P = .6261)/1,25(OH)D2 levels (χ2/1 = 0.1964; P = .6576) and fatal outcome. This might be due to a limited number of deaths in this cohort (n = 13)
      • Retrospective, observational study design
      • Lack of a healthy control group
      • Analysis limited to patients with CAP, limits generalizability
       Leong Leow, 2011, RespirologyVitamin D, Innate Immunity and Outcomes in Community-Acquired PneumoniaAll adult patients with CAP admitted to the Respiratory Medicine service (n = 112)
      • Deficiency: <50 nmol/L
      • Severe deficiency: <30 nmol/L
      • Sufficiency: ≥50 nmol/L
      • Method: Competitive electro chemiluminescence assay kit
      • 30-d mortality
      • Follow-up = 30 d
      • Severe deficiency associated with higher 30-d mortality, independent of patient age, sex, comorbidities, the systemic inflammatory response and other prognostic factors, including CURB65 scores: OR 13.5; 95% CI, 2.6-69.1; P = .002
      • Relationship between 25-(OH) D concentrations and mortality was not linear. Increased risk of mortality only in those with severe deficiency
      • Observational study design, cannot establish causal association
      • Relatively small sample size
      • Analysis limited to patients with CAP, limits generalizability
       Hossein Khalili, 2012, Clinical Research CardiologySerum Vitamin D Concentration Status and Its Correlation With Early Biomarkers of Remodeling After Acute Myocardial InfarctionAll patients with first episode of acute STEMI admitted to the Tehran Heart Center (n = 139)
      • Severe deficiency: ≤12.5 nmol/L
      • Moderate deficiency: 12.5-25 nmol/L
      • Mild deficiency: 25-35 nmol/L
      • Method: Radioimmunoassay
      • Association of 25(OH)D with in-hospital survival rate and early cardiac remodeling biomarkers
      • Follow-up = during hospitalization
      • The results show a statistically significant relationship between vitamin D levels and patients' survival rate (r = 0.216, P < .001), as well as a significant inverse relationship between serum MMP-9 and level of 25(OH) D (r = −0.271; P = .013) in patients after AMI
      • There was no statistically significant association between vitamin D levels and LOS (deficiency = 7.2 ± 2.6 d; normal vitamin D = 7.5 ± 2.9 d; P = .542)
      • Observational study, limited sample size
      • Analysis of patients with STEMI only, resuls may not be generalizable to all patients
      Intensive care unit (ICU) patients
       Karin Amrein, 2014, Critical CareVitamin D Status and Its Association With Season, Hospital and Sepsis Mortality in Critical IllnessSurgical and nonsurgical critically ill patients (n = 655 (36.3% medical)Severe deficiency: <12 ng/mL; Deficiency: ≤19.9 ng/mL; Insufficiency: 20-29.9 ng/mL
      • Method:
        • June 2009 enzyme immunoassay
        • After 2009: IDS-iSYS (based on chemiluminescence technology)
      • Prevalence of vitamin D deficiency and insufficiency
      • The association of ICU mortality and hospital mortality with 25(OH)D status, adjusted for covariates like baseline Simplified Acute Physiology Score II (SAPS II), age, gender, glomerular filtration rate (GFR), and C-reactive protein (CRP)
      • Subgroup analysis in patients with available blood culture results during hospital stay
      • Follow-up = duration of hospital stay
      • All-cause hospital mortality: In adjusted Cox regression analysis significantly higher in patients with vitamin D deficiency compared with patients with insufficient and sufficient levels (Deficiency: HR 1.63; 95% CI, 0.93-2.86; Insufficiency: HR 1.01; 95% CI, 0.52-1.96). Even more distinct results when hospital mortality rates were compared between month-specific 25(OH)D tertiles (Low tertile: HR 2.05; 95% CI, 1.31-3.22; Intermediate tertile: HR 1.92; 95% CI, 1.21-3.06)
      • Association with hospital mortality in Cox regression analysis with 25(OH)D as a continuous variable: HR 0.979; 95% CI, 0.962-0.997
      • ICU Mortality: No significant correlation between different Vitamin D levels (Deficiency: HR 1.30; 95% CI, 0.61-2.75; Insufficiency: HR 1.61; 95% CI, 0.70-3.67; P = .344), as well as tertiles (low tertile: HR 1.70; 95% CI, 0.98-2.96; Intermediate tertile: HR 1.6; 95% CI, 0.95-2.91)
      • Association with ICU mortality in Cox regression analysis with 25(OH) D as a continuous variable: HR 0.997; 95% CI, 0.975-1.018
      • ICU LOS: Significant negative correlation to vitamin D levels: median (IQR): Deficiency: 6.9 (9.8); Insufficiency: 4.9 (7.7); Sufficiency: 5.2 (6.5); P = .004
      • Hospital LOS: not significantly associated: Deficiency: 15.1 (19.1); Insufficiency: 15.0 (18.7); Sufficiency: 13.5 (16.8); P = .743
      • Blood culture positivity: No significant correlation: Deficiency: 22.2%; Insufficiency: 29.2%; Sufficiency: 7 (21.2%); P = .641
      • Possible selection bias
      • Ascertainment bias: vitamin D measured for unknown reasons, no data on patients who did not have 25(OH)D measured
      • Retrospective observational single-center design, limited generalizability
      • Relatively small sample size, insufficient statistical power to detect weaker associations
      • No details on 1,25(OH)2D levels, premorbid health status, comorbidities, the use of vitamin D supplements, the type of infection, albumin levels, liver dysfunction and therefore lower vitamin D binding protein, which is associated with lower 25(OH)D levels
       Andrea Braun, 2011, Critical Care MedicineAssociation of Low Serum 25-Hydroxyvitamin D Levels and Mortality in the Critically IllCritically ill-medical + surgical (n = 2 399)
      • Deficiency: <15 ng/mL
      • Insufficiency: 16-29 ng/mL
      • Method: Not specified
      • Determine the relationship between preadmission 25(OH)D deficiency and mortality after ICU admission
      • Follow-up = 30, 90, and 356 d after ICU
      Preadmission 25(OH)D deficiency was a significant predictor of 30-, 90-, and 365-d post-ICU day mortality, as well as in-hospital mortality, and remained a significant predictor of survival after multivariable adjustments for relevant comorbidities
      • 30-d mortality:
        • ≤15 ng/mL: OR 1.69; 95% CI, 1.28-2.17; P < .0001
        • 15-30 ng/mL: OR 1.36; 95% CI, 1.03-1.79; P = .25
        • ≥30 ng/mL: OR 1.00 (referent)
      • 90-d mortality:
        • ≤15 ng/mL: OR 1.67; 95% CI, 1.28-2.17; P < .0001
        • 15-30 ng/mL: OR 1.51; 95% CI, 1.19–1.93; P = .001
        • ≥30 ng/mL: OR 1.00 (referent)
      • 356-d mortality:
        • ≤15 ng/mL: OR 1.72; 95% CI, 1.35-2.19; P < .0001
        • 15-30 ng/mL: OR 1.44; 95% CI, 1.15-1.79; P = .001
        • ≥30 ng/mL: OR 1.00
      • In-hospital mortality:
        • ≤15 ng/mL: OR 1.72; 95% CI, 1.27-2.33; P < .0001
        • 15-30 ng/mL: OR 1.34; 95% CI, 1.01-1.79; P = .04
        • ≥30 ng/mL: OR 1.00 (referent)
      (adjustment for age, sex, race, Deyo-Charlson index, season, type [surgical vs medical], and sepsis)
      • Possible selection bias: Only patients who had their vitamin D statuses investigated before hospitalization
      • Significant differences in the total ICU population and the 25(OH)D cohort: decreased generalizability of to all critically ill patients
      • Possible residual confounders: no adjustment for BMI, immobilization, lack of sun exposure, and smoking status
      • Vitamin D levels measured before admission, may have changed since then
      • No data available on vitamin D supplementations
       A. Cecchi, 2011, Minerva AnestesiologicaVitamin D Deficiency in Septic Patients at ICU Admission Is Not a Mortality PredictorPatients admitted to the ICU with a diagnosis of major trauma (n = 72) and severe sepsis/septic shock (n = 98) (total n = 170)
      • Deficiency: <20 ng/mL
      • Method: Radioimmunoassay within first 24 h
      • Evaluate correlation of vitamin D levels with admission pathology and outcome
      • Follow-up = during ICU stay (septic patients = 6 [3-15] d)
      • Patients with sepsis/septic shock: significant association between vitamin D deficiency and mortality in univariate analysis (OR 0.995; 95% CI, 0.991-0.998; P = .004), but not significant in a multivariate logistic regression model. The lack of significance in multivariate analysis: may be due to limited sample size
      • Limited sample size
      • Observational study design
       JD McKinney, 2011, Journal of American Medical Directors AssociationRelationship Between Vitamin D Status and ICU Outcomes in VeteransAll patients admitted to the ICU with a serum 25(OH)D level drawn within 1 mo of ICU admission (n = 136)
      • Deficiency: <20 ng/mL
      • Method: Automated chemiluminescent method
      • Length of ICU stay (dichotomized as >3 d or <3 d) and survival (up to 6 mo after ICU admission)
      • Follow-up = 6 mo
      • Significant association of vitamin D deficiency with increased ICU LOS:
        • ICU LOS ≥3 d: Sufficiency: 29%; Insufficiency: 58% [χ2(1) = 11.38; P = .001], RR = 2.0
      • Significant association of vitamin D deficiency and survival rates: Sufficiency: 69%; Deficiency: 44% (χ2 8.21; P = .004), RR of death after admission to ICU = 1.81
      • Vitamin D levels only drawn in small number of veterans, results may not be generalizable
      • Possible residual confounders
      • Retrospective, observational study design
       David Higgins, 2012, Journal of Parenteral and Enteral NutritionRelationship of Vitamin D Deficiency to Clinical Outcomes in Critically Ill PatientsMixed medical-surgical ICU (n = 196)Deficiency: ≤30 nmol/L

      Insufficiency: >30-60 nmol/L

      Sufficiency: >60 nmol/L

      Method: Radioimmunoassay
      • Hypothesis: vitamin D deficiency is associated with poor patient outcomes, particularly longer ICU stay and greater risk of infection
      • Diagnosis of ICU-acquired infection = infection present after 48 h of ICU admission
      • Suspected infection = presence of a new positive culture result or initiation of new antibiotics after 48 h of ICU admission
      • Follow-up = 28 d
      • No significant association between baseline 25(OH)D status and 28-d all-cause mortality in a cox proportional hazard model (controlled for age, sex, body mass index, and APACHE II):
        • Sufficient vs deficient: adjusted HR 0.89; 95% CI, 0.37-2.24
        • Insufficient vs deficient: adjusted HR 1.23; 95% CI, 0.61-2.49
        • Sufficient vs insufficient: adjusted HR 0.724; 95% CI, 0.33-1.63
        • P = .46
      • Sufficient levels of 25(OH)D were significantly associated with a shorter time-to-alive ICU discharge:
        • Sufficient vs deficient: adjusted HR 2.11; 95% CI, 1.27-3.51
        • Insufficient vs deficient: adjusted HR 1.59; 95% CI, 1.05-2.40
        • Sufficient vs insufficient: adjusted HR 1.33; 95% CI, 0.86-2.06
        • P = .01
      • SOFA scores were higher in patients with vitamin D deficiency, but this association did not reach statistical significance
        • Max SOFA scores: deficient: 9.7 ± 4.2; insufficient: 8.4 ± 4.2; sufficient: 8.4 ± 3.3; P = .12
        • Delta SOFA scores: sufficient: 1.3 ± 1.7; insufficient: 1.6 ± 2.1; deficient: 2.1 ± 2.3; P = .29
      • Nonsignificant association with vitamin D deficiency and higher infection rate
        • Rate of ≥1 probable infection: OR 3.20; 95% CI, 0.784-13.07; P = .11
        • Pneumonia: sufficient: 5%; insufficient: 10%; deficient: 16%
        • Culture-positive pneumonia: OR 4.24; 95% CI, 0.50–38.8; P = .21
        • Probable pneumonia: OR 3.07; 95% CI, 0.763–12.36; P = .11
      • Relatively small sample size and confounding variables that cannot be accounted for in multivariate analysis
      • Study possibly underpowered
      • Observational study design
       Gulbin Aygencel, 2013, Critical Care Research and PracticeIs Vitamin D Insufficiency Associated with Mortality of Critically Ill Patients?Patients consecutively admitted to the ICU (n = 201)Insufficiency: <20 ng/mL

      Sufficiency: ≥20 ng/mL

      Method: Radioimmunoassay
      • Evaluate vitamin D status in critically ill patients and demonstrate the relationship between vitamin D status and mortality
      • Follow-up = 30 d
      • Significant correlation between ICU-mortality and vitamin D insufficiency in univariate analysis: insufficiency: 60 (43%); sufficiency: 16 (26%); P = .027
      • Vitamin D insufficiency was not an independent risk factor for mortality in logistic regression analysis (OR 1.8; 95% CI, 0.554-5.863; P = .328), thus vitamin D insufficiency may be a helper but not a real risk factor for mortality
      • Significant association between vitamin D levels and illness severity on admission:
        • SOFA Score: insufficiency: 8 (6-11); sufficiency: 6 (4-9); P = .005
        • APACHE II Score: insufficiency: 24 (19-28); sufficiency: 19 (16-25); P = .006
        • Sepsis/septic shock: insufficiency: 57 (41%); sufficiency: 15 (24.2%); P = .026
        • Organ dysfunction: insufficiency: 116 (84%); sufficiency: 44 (71%); P = .042
      • Association between vitamin D levels and ICU LOS (in days) not significant: insufficient: 9 (5-14); deficient: 8 (5-13); P = .355
      • Observational, single-center study with a small sample size; therefore, study results may not be generalizable
       Jieyu Hu, 2013; PLoS OneChanges in the Calcium-Parathyroid Hormone-Vitamin D Axis and Prognosis for Critically Ill Patients: A Prospective Observational StudyPatients admitted to the medical intensive care unit (MICU) who stayed more than 48 h in the MICU (n = 216)Deficiency: ≤19.9 ng/mL

      Insufficiency: 20–29.9 ng/mL

      Sufficiency: ≥30 ng/mL

      Method: High-performance liquid chromatography
      • Hospital mortality within 90 d of ICU admission
      • Assess the effects of changes in calcium-PTH-vitamin D axis on disease severity and prognosis
      • Follow-up = 90 d
      • Vitamin D deficiency was associated with significantly higher incidence of positive blood culture (P = .015), and of MODS (P = .013), as well as a higher APACHE II Score (r = −0.325; P < .001) and 90-d mortality (r = −0.276; P < .001)
      • Small sample size
      • Observational study design, causative link cannot be concluded
      • No data on levels of 1,25(OH)2D, vitamin D binding protein (DBP), or magnesium (= possible confounders)
       Long-Xiang Su, 2013, Chinese Medical JournalSignificance of Low Serum Vitamin D for Infection Risk, Disease Severity and Mortality in Critically Ill PatientsAll patients hospitalized in the ICU (surgical + respiratory + emergency) (total n = 206) (50 healthy controls + 51 ICU controls + 105 ICU patients with sepsis)Deficiency not defined
      • Patients divided into 5 groups:
        • 1st: >20 ng/mL
        • 2nd: 15-20 ng/mL
        • 3rd: 10-15 ng/mL
        • 4th: 5-10 ng/mL
        • 5th: <5 ng/mL
      Method: High performance liquid chromatography and tandem mass spectrometry
      • Study evaluated vitamin D for diagnosis, relevance to sepsis, its value in disease severity and prognostic assessment of ICU cases. 28- and 90-d mortality
      • Follow-up = 90 d
      • No significant correlation of vitamin D with APACHE II (Pearson coefficient = −0.124, P = .124), SAPS/SAPS II (Pearson coefficient = −0.039, P = .629), or SOFA scores (Pearson coefficient: −0.12, P = .221): Vitamin D levels did not indicate severity of disease
      • No significant difference in different groups of vitamin D levels and 28-d (1st = 0.6%, 2nd = 2.6%, 3rd = 3.8%, 3rd = 11.5, 4th = 11.5, 5th = 4.5%) or 90-d mortality (1st = 0.6%, 2nd = 4.5%, 3rd = 6.4%, 4th = 18.6%, 5th = 7.7%)
      • Only the APACHE II and SAPS scores are independent risk factors to deaths caused by sepsis
      • Observational study design
      • Small sample size
       Sindhaghatta Venkatram, 2011, Critical CareVitamin D Deficiency Is Associated With Mortality in the Medical Intensive Care UnitAll patients admitted to the MICU with available levels of 25(OH)D (n = 437)
      • Deficiency: <19.9 ng/dL Insufficiency: 20-29.9 ng/dL
      • Normal: >30 ng/dL
      • Method: Liquid chromatography–tandem mass spectrometry
      • Hospital mortality
      • Duration of mechanical ventilation and MICU LOS
      • Follow-up = not mentioned
      • 25(OH)D deficiency is associated with increased hospital mortality
        • Deficiency: OR 8.7; 95% CI, 1.03-72.8; P = .0469
        • Insufficiency: OR 4.3; 95% CI, −0.4-40.9; P = .2081
        • Mean 25(OH)D level for survivors: 27.9 ± 9.7 ng/dL, for nonsurvivors: 9.7 ± 4.7 ng/dL (P < .0001)
        • With a prevalence rate of 77.8%: PPV for mortality with 25(OH)D level <10 ng/dL = 83.64%; NPV = 29.52%
      • 25 (OH)D levels were not significantly associated with days on a ventilator (deficiency: 6.9 ± 6.0; insufficiency: 5.9 ± 6.0; normal: 6.4 ± 5.1; P = .77) and ICU length of stay (deficiency: 4.3 ± 4.5; insufficiency: 3.7 ± 3.9; normal: 4.2 ± 3.7, P = .54)
      • Retrospective single-center study
      • 25(OH)D levels not sequentially sampled
      • 25(OH)D measurment on admission probably because of preadmission deficiency
      • MICU-Study: may not be generalizable
      • PTH and 1,25(OH)D2 levels not available (possible confounders)
       Takuhiro Moromizato, 2013, Critical Care MedicineAssociation of Low Serum 25-Hydroxyvitamin D Levels and Sepsis in the Critically IllMedical + surgical ICU patients (n = 209)Deficiency: ≤15 ng/mL

      Insufficiency: 15-30 ng/mL

      Sufficiency: ≥30 ng/mL

      Method: Chemiluminescence assay, the radioimmunoassay, or mass spectroscopy
      • Relationship between preadmission 25(OH)D deficiency and sepsis in critically ill patients
      • Determine the association between preadmission 25(OH)D deficiency and mortality in patients with sepsis • Follow-up = not specified
      • Increased risk of sepsis in patients with preadmission 25(OH)D deficiency:
        • ≤15 ng/mL: OR 1.51; 95% CI, 1.17-1.94; P = .001
        • 15-30 ng/mL: OR 1.18; 95% CI, 0.96-1.45; P = .1
      • Preadmission 25(OH)D deficiency remained a significant predictor of sepsis following multivariable adjustments for relevant comorbidities
      • Patients who are not 25(OH)D sufficient before hospital admission who develop sepsis have a significantly higher risk for mortality
        • In-hospital mortality : <30 ng/mL: OR 1.62; 95% CI, 1.07-2.46, P = .02
        • 30-d mortality: <30 ng/mL: OR 1.55; 95% CI, 1.02-2.34; P = .04
        • 90-d mortality: <30 ng/mL: OR 1.63; 95% CI, 1.11-2.39; P = .01
      • (adjusted for age, gender, race [white, non-white], type [surgical vs medical], and Deyo-Charlson index)
      • Retrospective, observational study, small sample size
      • Selection bias possible: vitamin D status investigated for a unknown reason that may be absent in other critically ill patients (limits generalizability)
      • Ascertainment bias: not every critically ill patient had vitamin D measured before hospitalization
      • No adjustment for calcium, PTH or albumin levels
      • Possible residual confounding
      • No information on vitamin D supplementation
      Geriatric patients
       Cédric Annweiler, 2010, Advances in TherapySerum Vitamin D Concentration and Short-Term Mortality Among Geriatric Inpatients in Acute Care SettingsAll patients admitted to the geriatric acute care unit (n = 399)Deficiency: ≤50 nmol/L

      Method: Radioimmunoassay
      • Association between serum 25(OH)D concentration and the occurrence of short-term mortality (ie, in-hospital mortality) among acute care geriatric inpatients
      • Follow-up = 10.8 ± 0.4 hospital d
      Significant association between increased serum 25(OH)D concentrations and a low short-term mortality rate in a geriatric acute care unit
      • Full adjusted logistic regression: OR 0.65; P = .029
      • Stepwise backward model: OR 0.87; P = .029
      (adjusted for: age, gender, BMI, supine systolic blood pressure, numbers of acute diseases, chronic diseases, and hospital days, serum albumin concentration, creatinine clearance, and season of hospital admission)
      • Use of a cross-sectional design: may limit the exploration of the association between serum 25(OH)D and short-term mortality, and does not allow any causal inference
       Olivier Beauchet, 2013, Journal of Nutrition, Health & AgingWho Is at Risk of Long Hospital Stay Among Patients Admitted to Geriatric Acute Care Unit? Results From a Prospective Cohort StudyPatients aged ≥75 y hospitalized in geriatric acute care unit (n = 531)
      • Deficiency: <25 nmol/L
      • Method: Radioimmunoassay
      • To confirm that vitamin D deficiency is associated with a longer LOS among older inpatients admitted to a geriatric acute care unit
      • Which combinations of clinical risk factors of longer LOS associated with vitamin D deficiency best predicted a longer LOS
      • Follow-up = not specified
      Significant association in univariate and multiple linear regression models of LOS with vitamin D deficiency (P < .001 for all models)
      • <8 d LOS: 24.1% (n = 40); P = .009
      • 8-13 d LOS: 37% (n = 71); P < .001
      • >13 d LOS: 49.7% (n = 86); P = .014
      Vitamin D deficiency, delirium and male gender are risk factors for a longer LOS
      • Study population probably unrepresentative of the general older population
      • Analyses adjusted for characteristics measured at baseline assessment, but not for their change during hospitalization
      • nonavailability of other biological measures (eg, serum albumin)
       Lise Hélard, 2013, Disease MarkersHypovitaminosis D in Geriatric Acute Care Unit: A Biomarker of Longer Length of StayInpatients of geriatric acute care unit (n = 253)
      • Deficiency: ≤20 ng/mL
      • Method: Radioimmunoassay
      • LOS
      • Follow-up = length of hospital stay (mean: 14.7 ± 8.1 d)
      Inpatients with vitamin D deficiency at time of admission had a LOS extended by approximately 3 d (<20 ng/mL: 15.2 ± 8.2 d; >20 ng/mL: 12.1 ± 7.0d, P = .017)
      • Adjusted model: P = .043
      • β = −0.08; 95% CI, −0.14, −0.02
      • Backward model: P = .015
      • Inverse correlation between serum 25(OH)D concentration and LOS: r = −0.14; P = .028
      Evidence of linear association between 25(OH)D concentration and LOS
      • Possible selection bias: study cohort restricted to Caucasian older inpatients hospitalized in a single geriatric acute care unit
      • Observational and retrospective design prevented inferring causality
      • Maybe not all confounding factors considered (ie, BMI), also other factors influencing LOS: beds available, health insurance guidelines, etc
       Aurélien Sutra del Galy, 2009, Letter to the EditorVitamin D Insufficiency and Acute Care in Geriatric InpatientsSubjects aged 75 y and older, hospitalized in a geriatric acute care unit (n = 278)
      • Insufficiency: ≤20 ng/mL
      • Method: Not specified
      • Association between low serum 25(OH)D concentrations and a greater number of acute diseases in geriatric inpatients admitted into acute care unit
      • Follow-up = Not specified
      Significant correlation between 25(OH)D insufficiency (independent of the numbers of chronic diseases and drugs taken per day) and number of acute diseases in frail elderly adults
      • Crude β = 0.498; P = .03
      • Adjusted β (age and sex) = 0.522; P = .02
      • Adjusted for all confounders: β = 0.489; P = .03
      • Backward stepwise linear regression: β = 0.470; P = .03
      • Retrospective cross-sectional design
      Patients undergoing rehabiliation
       Gary M. Kiebzak, 2007, American Journal of Physical Medicine & RehabilitationVitamin D Status of Patients Admitted to a Hospital Rehabilitation Unit: Relationship to Function and ProgressAmbulatory patients in a tertiary general hospital rehabilitation unit (n = 100)Deficiency: <20 nmol/L

      Method: Radioimmunoassay
      • Association between vitamin D status, baseline function, and short-term rehabilitation progress, using grip strength and the Functional Independence Measure (FIM) tool
      • Follow-up = not specified
      • 25(OH)D levels correlate significantly with total FIM score (r = 0.25; P < .011) and grip strength (r = 0.23; P < .021)
      • Multiple-regression model using variables that significantly correlated with admission FIM score (age, grip strength, and body weight): only serum 25(OH)D was a significant contributor (P = .02) to the multiple r value (multiple r = 0.42; P = .001)
      • No correlation with FIM efficiency or unadjusted change in total FIM score during hospitalization: r = 0.19; P = .066
      • 25(OH)D >41.3 nmol/L (median value): significantly higher FIM efficiency scores than 25(OH)D <41.3 nmol/L, but higher 25(OH)D correlated with significantly shorter LOS, which influenced the FIM efficiency calculation
      • Correlation between low 25(OH) and longer LOS: Spearman r = −0.235; P < .018
      • Coefficient of determination (r2) values: all <9% → suggests that 25(OH)D levels explained <9% of the variability in these measures
      • No direct comparison of rehabilitation patients with low 25(OH)D against high serum 25(OH)D (which was the original intent of the study), because too few patients with 25(OH)D >80.0 nmol/L
       Anthony J. Pellicane, 2011, Archives of Physical Medicine and RehabilitationPrevalence of 25 Hydroxyvitamin D Deficiency in the Acute Inpatient Rehabilitation Population and Its Effect on FunctionAcademic acute rehabilitation facility (n = 101)Deficiency: <10.0 ng/mL

      Method: Enzyme-linked immunosorbant assay kit
      • 25(OH)D level
      • Total/motor/cognitive FIM efficiency (FIM efficiency score = difference in discharge and admission FIM scores/length of stay)
      • Follow-up = during hospitalization (13.41 ± 6.8)
      • Total FIM efficiency: Sufficient = 2.96 ± 1.42; Insufficient = 2.29 ± 1.41; P = .039
      • Motor FIM efficiency: Sufficient = 2.50 ± 1.18; Insufficient = 1.98 ± 1.19; P = .056
      • Cognitive FIM efficiency: Sufficient = 0.46 ± 0.56; Insufficient = 0.32 ± 0.38; P = .150
      • After controlling for demographic and other clinical factors: 25(OH)D status (b = −0.38) (P = .22) was not a significant predictor of total FIM efficiency score (adjusted R2 = 0.23; F5,82 = 6.15; P = .0001)
      • Association between total FIM efficiency scores and 25 (OH)D levels: Higher than median (25.7 ng/mL) = 2.92 ± 1.37; lower than median = 2.411 ± 1.40; P = .110
      • This might be due to 25(OH)D levels not being low enough to impact functional performance
      • Small sample size, may not be representative of the broader population of rehabilitation patients
      • Mixed diagnostic population studied: not ideal for attempting to describe prevalence, may affected the ability to make firm conclusions about 25(OH)D level and its relation to function
      • No information on vitamin D supplementation
      P valuses are statistically significant at P < .05.
      1,25(OH)D2 = 1,25-dihydroxyvitmamin D; 25(OH)D = 25-hydroxyvitamin D; 95% CI = 95% confidence interval; AMI = acute myocardial infarction; APACHE = acute physiology and chronic health evaluation; BMI = body mass index; CAP = community-acquired infection; CDI = Clostridium difficile infection; CURB = new-onset confusion = urea >7 mmol/L = respiratory rate ≥30 breaths per minute = systolic or diastolic blood pressure <90 mm Hg or ≤60 mm Hg = respectively (pneumonia/LRTI risk scoring system); ED = emergency department; EMS = elderly mobility scale; FIM = functional independence measure; HABSI = hospital-acquired bloodstream infections; HR = hazard ratio; iPTH = intact parathyroid hormone; IQR = interquartile range; LOS = length of stay; (M)ICU = (medical) intensive care unit; MMP-9 = matrix metallopeptidase 9; MODS = multiorgan dysfunction syndrome; NPV = negative predicitive value; OR = odds ratio; PPV = positive predictive value; PSI = pneumonia severity index; PTH = parathyroid hormone; RCDI = recurrence Clostridium difficile infection; RCT = randomized, controlled trial; RR = relative risk; SOFA = sequential organ failure assessment; STEMI = ST-segment elevation myocardial infarction; TG = triglyceride; VLDL = very low density lipoprotein.
      Appendix 2Randomized, Controlled Trials Evaluating the Benefits of Vitamin D Supplementation
      First Author, Y, JournalTitlePopulation (Total = n, Intervention = i, Control = c)Study Intervention (I) Control (C)Definition of Deficiency Method of Vitamin D MeasurementEndpoints Follow-Up PeriodKey FindingsBias
      Karin Amrein, 2014, JAMAEffect of High-Dose Vitamin D3 on Hospital Length of Stay in Critically Ill Patients with Vitamin D Deficiency: The VITdAL-ICU Randomized Clinical TrialCritically ill adults (medical and surgical population) (n = 475, i = 237, c = 238)I: 540,000 IU of vitamin D3 in 45 mL of oleum arachidis

      C: 45 mL of oleum arachidis
      Deficiency: <20 ng/mL

      Severe deficiency: ≤12 ng/mL

      Method:
      • Chemiluminescence technology
      • 100 samples were also analyzed from frozen samples (−70°C) by liquid chromatography tandem–mass spectrometry (LC-MS/MS)
      • Length of hospital stay and mortality
      • Follow-up: 6 mo
      • Administration of high-dose vitamin D3 compared with placebo did not improve hospital length of stay, hospital mortality, or 6-mo mortality
        • Hospital LOS: Intervention: 20.1 d (IQR, 11.1-33.3); placebo: 19.3 d (IQR, 11.1-34.9); P = .98
        • ICU LOS: Intervention: 9.6 d (IQR, 4.2-17.8); placebo: 10.7 d (IQR, 4.9-21.9); P = .38
        • 6-mo mortality: Intervention: 35.0%; 95% CI, 29.0%-41.5%; placebo: 42.9%; 95% CI, 36.5%-49.4%; HR 0.78; 95% CI, 0.58-1.04; P = .09
      • Incidental finding in subgroup analysis in severe vitamin D deficiency: lower hospital mortality in intervention group
        • Hospital LOS: Intervention: 20.1 d (IQR, 12.9-39.1); Placebo: 19.0 d (IQR, 11.6-33.8); P = .40
        • ICU LOS: Intervention: 9.7 d (IQR, 4.2-17.3); placebo: 9.1 d (IQR, 4.1-20.1); P = .98
        • Hospital mortality: Adjusted HR 0.51; 95% CI, 0.31-0.84; P = .009
        • 6-mo mortality: Intervention: 34.7%; 95% CI, 25.4%-45.0%; placebo: 50.0%; 95% CI, 39.9%-60.1%; P for interaction = .12; Adjusted HR 0.50; 95% CI, 0.31-0.79; P = .003
      • No evident selection bias: patients randomly assigned to one group in a 1:1 ratio, using the Randomizer for Clinical Trials tool
      • Blinding: double-blinded, low risk of performance or selection bias
      Nancy K. Latham, 2003, Journal of the American Geriatrics SocietyA Randomized, Controlled Trial of Quadriceps Resistance Exercise and Vitamin D in Frail Older People: The Frailty Interventions Trial in Elderly Subjects (FITNESS)Patients admitted to geriatric rehabilitation units (n = 243, i1 = 120, i2 = 121, c1 = 123, c2 = 122)I1) Resistance exercise intervention: quadriceps exercise program using adjustable ankle cuff weights 3×/wk for 10 wk

      I2) Single oral dose:

      Six 1.25-mg calciferol tablets (300,000 IU)

      C1) Exercise control group: frequency- matched telephone calls and home visits from research physical therapists. inquiry about recovery, general advice on problems C2) Matching placebo tablets
      Deficiency: <12 ng/mL

      Method: Acetonitrile radioimmunoassay
      • Physical health according to the short-form health survey at 3 mo and falls over 6 mo
      • Effects on physical health at 3 and 6 mo (PCS and falls over 6 mo
      • Physical performance and self-reported measures of physical function, social activities, and mental health
      • Follow-up: 6 mo
      • No significant reduction of risk of falls:
        • Resistance exercise vs attention control: RR of a fall (95% CI): 0.96 (0.67-1.36); HR of time to first fall (95% CI): 0.97 (0.68-1.37)
        • Vitamin D vs placebo: RR (95% CI): 1.12 (0.79-1.59); HR (95% CI): 1.14 (0.80-1.62)
      • → No significant treatment differences for any of these variables
      • No significant improvement in physical component score (mean (95% CI):
        • Resistance exercise: 34 (32-36); Control: 35 (33-37); difference, mean (95% CI): −1 (−4, 1)
        • Vitamin D: 35 (33-37); Placebo: 35 (33-36); difference, mean (95% CI): 0 (−2, 3)
      • Modest effect of vitamin D supplementation in people with extremely low 25(OH)D levels
      • Resistance exercise intervention seems to be harmful, with increase of self-reported fatigue and with higher incidence of musculoskeletal injuries
      • Vitamin D did not alter any of the physical performance measures, including maximum strength at 3 mo: active vs placebo, median = 12.5 kg (95% CI, 10.5-13.9) vs 11.5 kg (95% CI, 9.0-13.9)
      • Random sequence generation: patients were allocated to 1 of the 4 treatment arms using a computerized central randomization scheme
      • Small chance of performance bias in the resistance exercise intervention group because only single blind, but attention control visits were used to minimize bias
      • Detection bias: research nurses who conducted follow-up visits were blinded to the assigned treatments
      • Outcome assessment was blinded
      Elizabeth Burleigh, 2007, Age and AgeingDoes Vitamin D Stop Inpatients Falling? A Randomised Controlled TrialPatients of general assessment and rehabilitation wards in an acute geriatric unit (n = 205, i = 101, c = 104)I: 800 IU cholecalciferol plus 1200 mg calcium carbonate once daily

      C: 1200 mg calcium carbonate only once daily
      Deficiency: not specified

      Method: Nichols Advantage Immunoassay

      Analyzer
      • Determine whether vitamin D supplementation has an effective role in falls prevention in older patients
      • Follow-up: the intervention was continued until time of patient discharge or death
      • No significant difference in number of fallers between the 2 groups: RR = 0.82; 95% CI, 0.59-1.16; χ2 P = .263
      • No significant difference in number of falls was shown between the two groups
      • Time to the first fall longer in treatment group → no statistical significance
      • No significant difference for improvement in functionality scores (Cannard, EMS, Barthel) was found between the groups
      • In this study the prescription 800 IU cholecalciferol additionally to 1.200 mg calcium did not prove effective in in preventing falls. This could be due to an insufficient treatment period (mean 30 d) and participant numbers
      • Randomization: random numbers table to treatment or control groups
      • Double-blinding: reduced chance of performance and detection bias
      • No indication for attrition bias
      Heike A. Bischoff, 2003, Journal of Bone and Mineral ResearchEffects of Vitamin D and Calcium Supplementation on Falls: A Randomized Controlled TrialPatients in long-stay geriatric care units in Switzerland (n = 122)I: vitamin D (400 IU) plus calcium group (600 mg) (n = 62)

      C: calcium group (600 mg) (n = 60)
      Deficiency: not specified

      Method: Radioimmunoassay
      • Number of falls per person during the treatment period
      • Musculoskeletal function and bone remodeling
      • Assessments took place at the beginning of treatment period and after 12 wk
      • Follow-up: 126 d
      • Vitamin D plus calcium treatment reduced the number of falls by 62% (r = −0.96; 95% CI, 37%-77%; P < .0002) and the number of falls per person by 49% (r = −0.68; 95% CI, 14%-71%; P = .01)
      • Smaller number of falls in the calcium + vitamin D group (P = .045)
      • No significant difference in the number of fallers between the groups (RR = 0.7; 95% CI, 0.3-1.5). This might be due to the small sample size
      • Effect was in favor of the calcium + vitamin D treatment, with a 30% lower risk to be a faller
      • Improved musculoskeletal function in calcium + vitamin D group: P = .0094
      • Randomization method unclear
      • Double-blinding reduced risk of performance and detection bias
      • Large drop-out rate may cause attrition bias
      Corless D, 1985, Age and AgeingDo Vitamin D Supplements Improve the Physical Capabilities of Elderly Hospital Patients?Patients in the geriatric wards, who had been inpatients for over 4 wk and were likely to stay for a further 8 wk + 6 patients from day hospital (n = 65, i = 32, c = 33)I: 9000 units vitamin D2

      C: Placebo: similar number of tablets of identical appearance containing lactose
      Low: <20 nmol/L

      Low normal: 21.1-40 nmol/L

      Normal: >40 nmol/L

      Method: Not specified
      Effect of oral vitamin D supplements on the ability of elderly hospital patients with low or low normal plasma 25(OH)D to perform basic activities of daily living

      Follow-up: 40 wk
      • No significant difference for the 2 groups could be showed in changes in “muscle power” and “independence scores,” either at 40 wk or in shorter intervals (12, 24 wk):
        • Independence score:
          • Baseline: Control (C): 1.89 ± 0.10; Treatment (T): 1.89 ± 0.10; 40 wk: C: −0.06 ± 0.11; T: −0.15 ± 0.26
        • Muscle strength score:
          • Baseline: C: 2.20 ± 0.10; T: 2.26 ± 0.09; 40 weeks: C: +0.01 ± 0.09; T: −0.05 ± 0.28
        • Unweighted total ADL:
          • Base-line: C: 23.6 ± 1.0; T: 23.8 ± 0.9; 40 weeks: C: +0.1 ± 1.2; T: −0.1 ± 3.0
        • Mean time taken(s)/test:
          • Baseline: C: 37.2 ± 3.2; T: 37.6 ± 4.0; 40 weeks: C: +0.6 ± 4.7; T: −4.7 ± 7.5
        • Mean number of tests unable to do:
          • Baseline: C: 4.9 ± 0.6; T: 5.0 ± 0.5; 40 weeks: C: 0.0 ± 0.6; T: −0.3 ± 1.5
        • Mental assessment score:
          • Baseline: C: 13.0 ± 0.9; T: 12.3 ± 0.6; 40 weeks: C: +1.2 ± 1.2: T: −1.4 ± 1.3
      • Subgroup analysis for baseline 25(OH)D levels ≤20 nmol/L:
        • Independence score: r = 0.26, P > .1; Muscle strength score: r = 0.12, P > .3
      • Randomization by a computer program into approximately equal numbers of patients
      • Double-blinding
      P valuses are statistically significant at P < .05.
      25(OH)D = 25-hydroxyvitamin D; 95% CI = 95% confidence interval; EMS = elderly mobility scale; HR = hazard ratio; ICU = intensive care unit; IQR = interquartile range; LOS = length of stay; OR = odds ratio; PCS = physical component score.

      References

        • Mithal A.
        • Wahl D.A.
        • Bonjour J.P.
        • et al.
        Global vitamin D status and determinants of hypovitaminosis D.
        Osteoporos Int. 2009; 20: 1807-1820
        • Holick M.F.
        Vitamin D deficiency.
        N Engl J Med. 2007; 357: 266-281
        • Bouillon R.
        • Bischoff-Ferrari H.
        • Willett W.
        Vitamin D and health: perspectives from mice and man.
        J Bone Min Res. 2008; 23: 974-979
        • Liu P.T.
        • Stenger S.
        • Li H.
        • et al.
        Toll-like receptor triggering of a vitamin D-mediated human antimicrobial response.
        Science. 2006; 311: 1770-1773
        • Nagpal S.
        • Na S.
        • Rathnachalam R.
        Noncalcemic actions of vitamin D receptor ligands.
        Endocr Rev. 2005; 26: 662-687
        • DeLuca H.F.
        Overview of general physiologic features and functions of vitamin D.
        Am J Clin Nutr. 2004; 80: 1689S-1696S
        • Holick M.F.
        High prevalence of vitamin D inadequacy and implications for health.
        Mayo Clin Proc. 2006; 81: 353-373
        • LeFevre M.L.
        US Preventive Services Task Force. Screening for vitamin d deficiency in adults: U.S. Preventive Services Task Force recommendation statement.
        Ann Intern Med. 2015; 162: 133-140
        • Holick M.F.
        • Binkley N.C.
        • Bischoff-Ferrari H.A.
        • et al.
        Evaluation, treatment, and prevention of vitamin D deficiency: an Endocrine Society clinical practice guideline.
        J Clin Endocrinol Metab. 2011; 96: 1911-1930
        • LeBlanc E.S.
        • Zakher B.
        • Daeges M.
        • Pappas M.
        • Chou R.
        Screening for vitamin d deficiency: a systematic review for the U.S. Preventive Services Task Force.
        Ann Intern Med. 2015; 162: 109-122
        • Institute of Medicine Committee to Review Dietary Reference Intakes for Vitamin D, Calcium
        The National Academies Collection: reports funded by National Institutes of Health.
        in: Ross A.C. Taylor C.L. Yaktine A.L. Del Valle H.B. Dietary Reference Intakes for Calcium and Vitamin D. National Academies Press, National Academy of Sciences, Washington, DC2011
        • Sanders K.M.
        • Stuart A.L.
        • Williamson E.J.
        • et al.
        Annual high-dose oral vitamin D and falls and fractures in older women: a randomized controlled trial.
        JAMA. 2010; 303: 1815-1822
        • Brouwer-Brolsma E.M.
        • Bischoff-Ferrari H.A.
        • Bouillon R.
        • et al.
        Vitamin D: do we get enough? A discussion between vitamin D experts in order to make a step towards the harmonisation of dietary reference intakes for vitamin D across Europe.
        Osteoporos Int. 2013; 24: 1567-1577
        • Heaney R.P.
        • Recker R.R.
        • Grote J.
        • Horst R.L.
        • Armas L.A.
        Vitamin D(3) is more potent than vitamin D(2) in humans.
        J Clin Endocrinol Metab. 2011; 96: E447-E452
        • Holick M.F.
        • Biancuzzo R.M.
        • Chen T.C.
        • et al.
        Vitamin D2 is as effective as vitamin D3 in maintaining circulating concentrations of 25-hydroxyvitamin D.
        J Clin Endocrinol Metab. 2008; 93: 677-681
        • Leow L.
        • Simpson T.
        • Cursons R.
        • Karalus N.
        • Hancox R.J.
        Vitamin D, innate immunity and outcomes in community acquired pneumonia.
        Respirology. 2011; 16: 611-616
        • Lange N.
        • Litonjua A.A.
        • Gibbons F.K.
        • Giovannucci E.
        • Christopher K.B.
        Pre-hospital vitamin D concentration, mortality, and bloodstream infection in a hospitalized patient population.
        Am J Med. 2013; 126: 640.e19-640.e27
        • Amrein K.
        • Quraishi S.A.
        • Litonjua A.A.
        • et al.
        Evidence for a U-shaped relationship between prehospital vitamin D status and mortality: a cohort study.
        J Clin Endocrinol Metab. 2014; 99: 1461-1469
        • Remmelts H.H.
        • van de Garde E.M.
        • Meijvis S.C.
        • et al.
        Addition of vitamin D status to prognostic scores improves the prediction of outcome in community-acquired pneumonia.
        Clin Infect Dis. 2012; 55: 1488-1494
        • Khalili H.
        • Talasaz A.H.
        • Salarifar M.
        Serum vitamin D concentration status and its correlation with early biomarkers of remodeling following acute myocardial infarction.
        Clin Res Cardiol. 2012; 101: 321-327
        • Abdelfatah M.
        • Nayfe R.
        • Moftakhar B.
        • et al.
        Low vitamin D level and impact on severity and recurrence of Clostridium difficile infections.
        J Investig Med. 2015; 63: 17-21
        • Anty R.
        • Tonohouan M.
        • Ferrari-Panaia P.
        • et al.
        Low levels of 25-hydroxy vitamin D are independently associated with the risk of bacterial infection in cirrhotic patients.
        Clin Transl Gastroenterol. 2014; 5: e56
        • Pletz M.W.
        • Terkamp C.
        • Schumacher U.
        • et al.
        Vitamin D deficiency in community-acquired pneumonia: low levels of 1,25(OH)2 D are associated with disease severity.
        Respir Res. 2014; 15: 53
        • Lai L.
        • Qian J.
        • Yang Y.
        • et al.
        Is the serum vitamin D level at the time of hospital-acquired acute kidney injury diagnosis associated with prognosis?.
        PLoS One. 2013; 8: e64964
        • Quraishi S.A.
        • Litonjua A.A.
        • Moromizato T.
        • et al.
        Association between prehospital vitamin D status and hospital-acquired bloodstream infections.
        Am J Clin Nutr. 2013; 98: 952-959
        • Ginde A.A.
        • Camargo Jr., C.A.
        • Shapiro N.I.
        Vitamin D insufficiency and sepsis severity in emergency department patients with suspected infection.
        Acad Emerg Med. 2011; 18: 551-554
        • Wang W.J.
        • Gray S.
        • Sison C.
        • et al.
        Low vitamin D level is an independent predictor of poor outcomes in Clostridium difficile-associated diarrhea.
        Therap Adv Gastroenterol. 2014; 7: 14-19
        • van der Wilden G.M.
        • Fagenholz P.J.
        • Velmahos G.C.
        • Quraishi S.A.
        • Schipper I.B.
        • Camargo Jr., C.A.
        Vitamin D status and severity of Clostridium difficile infections: a prospective cohort study in hospitalized adults.
        JPEN J Parenter Enteral Nutr. 2014; 39: 465-470
        • Amrein K.
        • Zajic P.
        • Schnedl C.
        • et al.
        Vitamin D status and its association with season, hospital and sepsis mortality in critical illness.
        Crit Care. 2014; 18: R47
        • Braun A.
        • Chang D.
        • Mahadevappa K.
        • et al.
        Association of low serum 25-hydroxyvitamin D levels and mortality in the critically ill.
        Crit Care Med. 2011; 39: 671-677
        • McKinney J.D.
        • Bailey B.A.
        • Garrett L.H.
        • Peiris P.
        • Manning T.
        • Peiris A.N.
        Relationship between vitamin D status and ICU outcomes in veterans.
        J Am Med Dir Assoc. 2011; 12: 208-211
        • Hu J.
        • Luo Z.
        • Zhao X.
        • 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: e75441
        • 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
        • Moromizato T.
        • Litonjua A.A.
        • Braun A.B.
        • Gibbons F.K.
        • Giovannucci E.
        • Christopher K.B.
        Association of low serum 25-hydroxyvitamin D levels and sepsis in the critically ill.
        Crit Care Med. 2014; 42: 97-107
        • Aygencel G.
        • Turkoglu M.
        • Tuncel A.F.
        • Candir B.A.
        • Bildaci Y.D.
        • Pasaoglu H.
        Is vitamin D insufficiency associated with mortality of critically ill patients?.
        Crit Care Res Pract. 2013; 2013: 856747
        • Su L.X.
        • Jiang Z.X.
        • Cao L.C.
        • et al.
        Significance of low serum vitamin D for infection risk, disease severity and mortality in critically ill patients.
        Chin Med J. 2013; 126: 2725-2730
        • Higgins D.M.
        • Wischmeyer P.E.
        • Queensland K.M.
        • Sillau S.H.
        • Sufit A.J.
        • Heyland D.K.
        Relationship of vitamin D deficiency to clinical outcomes in critically ill patients.
        JPEN J Parenter Enteral Nutr. 2012; 36: 713-720
        • Cecchi A.
        • Bonizzoli M.
        • Douar S.
        • et al.
        Vitamin D deficiency in septic patients at ICU admission is not a mortality predictor.
        Minerva Anestesiol. 2011; 77: 1184-1189
        • Annweiler C.
        • Pochic S.
        • Fantino B.
        • et al.
        Serum vitamin D concentration and short-term mortality among geriatric inpatients in acute care settings.
        Adv Ther. 2010; 27: 245-249
        • Helard L.
        • Mateus-Hamdan L.
        • Beauchet O.
        • Annweiler C.
        Hypovitaminosis D in geriatric acute care unit: a biomarker of longer length of stay.
        Dis Markers. 2013; 35: 525-529
        • Beauchet O.
        • Launay C.
        • de Decker L.
        • Fantino B.
        • Kabeshova A.
        • Annweiler C.
        Who is at risk of long hospital stay among patients admitted to geriatric acute care unit? Results from a prospective cohort study.
        J Nutr Health Aging. 2013; 17: 695-699
        • Sutra Del Galy A.
        Vitamin D insufficiency and acute care in geriatric inpatients.
        J Am Geriatr Soc. 2009; 57: 1721-1723
        • Kiebzak G.M.
        • Moore N.L.
        • Margolis S.
        • Hollis B.
        • Kevorkian C.G.
        Vitamin D status of patients admitted to a hospital rehabilitation unit: relationship to function and progress.
        Am J Phys Med Rehabil. 2007; 86: 435-445
        • Pellicane A.J.
        • Wysocki N.M.
        • Mallinson T.R.
        • Schnitzer T.J.
        Prevalence of 25-hydroxyvitamin D deficiency in the acute inpatient rehabilitation population and its effect on function.
        Arch Phys Med Rehabil. 2011; 92: 705-711
        • Amrein K.
        • Schnedl C.
        • Holl A.
        • et al.
        Effect of high-dose vitamin D3 on hospital length of stay in critically ill patients with vitamin D deficiency: the VITdAL-ICU randomized clinical trial.
        JAMA. 2014; 312: 1520-1530
        • Latham N.K.
        • Anderson C.S.
        • Lee A.
        • et al.
        A randomized, controlled trial of quadriceps resistance exercise and vitamin D in frail older people: the Frailty Interventions Trial in Elderly Subjects (FITNESS).
        J Am Geriatr Soc. 2003; 51: 291-299
        • Bischoff H.A.
        • Stahelin H.B.
        • Dick W.
        • et al.
        Effects of vitamin D and calcium supplementation on falls: a randomized controlled trial.
        J Bone Miner Res. 2003; 18: 343-351
        • Corless D.
        • Dawson E.
        • Fraser F.
        • et al.
        Do vitamin D supplements improve the physical capabilities of elderly hospital patients?.
        Age Ageing. 1985; 14: 76-84
        • Burleigh E.
        • McColl J.
        • Potter J.
        Does vitamin D stop inpatients falling? A randomised controlled trial.
        Age Ageing. 2007; 36: 507-513