Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=iann20

Download by: [E-Library Insel], [Ville Pettilä] Date: 25 January 2016, At: 07:45

Annals of Medicine

ISSN: 0785-3890 (Print) 1365-2060 (Online) Journal homepage: http://www.tandfonline.com/loi/iann20

Vitamin D deficiency at admission is notassociated with 90-day mortality in patients withsevere sepsis or septic shock: ObservationalFINNAKI cohort study

Tero I. Ala-Kokko, Shivaprakash J. Mutt, Sara Nisula, Juha Koskenkari, JanneLiisanantti, Pasi Ohtonen, Meri Poukkanen, Jouko J. Laurila, Ville Pettilä,Karl-Heinz Herzig & The FINNAKI Study Group

To cite this article: Tero I. Ala-Kokko, Shivaprakash J. Mutt, Sara Nisula, Juha Koskenkari, JanneLiisanantti, Pasi Ohtonen, Meri Poukkanen, Jouko J. Laurila, Ville Pettilä, Karl-Heinz Herzig & TheFINNAKI Study Group (2016): Vitamin D deficiency at admission is not associated with 90-daymortality in patients with severe sepsis or septic shock: Observational FINNAKI cohort study,Annals of Medicine, DOI: 10.3109/07853890.2015.1134807

To link to this article: http://dx.doi.org/10.3109/07853890.2015.1134807

Published online: 22 Jan 2016.

Submit your article to this journal

Article views: 4

View related articles

View Crossmark data

ANNALS OF MEDICINE, 2016http://dx.doi.org/10.3109/07853890.2015.1134807

RESEARCH ARTICLE

Vitamin D deficiency at admission is not associated with 90-day mortalityin patients with severe sepsis or septic shock: Observational FINNAKIcohort study

Tero I. Ala-Kokkoa,b, Shivaprakash J. Muttc,d*, Sara Nisulae*, Juha Koskenkaria,b, Janne Liisananttia,b,Pasi Ohtonena,b, Meri Poukkanenf, Jouko J. Laurilaa,b, Ville Pettilae and Karl-Heinz Herzigc,d; The FINNAKI StudyGroup

aDepartment of Anaesthesiology, University of Oulu, Oulu, Finland; bDivision of Intensive Care Medicine, Medical Faculty, Oulu UniversityHospital, Medical Research Center Oulu, Research Group of Surgery, Anesthesiology and Intensive Care, University of Oulu, Oulu, Finland;cInstitute of Biomedicine and Biocenter of Oulu, University of Oulu, Oulu, Finland; dMedical Research Center Oulu, Oulu University Hospital,Oulu, Finland; eDivision of Anaesthesia and Intensive Care Medicine, Department of Surgery, Intensive Care Units, Helsinki UniversityCentral Hospital, Helsinki, Finland; fDepartment of Anaesthesia and Intensive Care, Lapland Central Hospital, Rovaniemi, Finland

ABSTRACTIntroduction Low levels of vitamin D have been associated with increased mortality in patientsthat are critically ill. This study explored whether vitamin D levels were associated with 90-daymortality in severe sepsis or septic shock.Methods Plasma vitamin D levels were measured on admission to the intensive care unit (ICU) in aprospective multicentre observational study.Results 610 patients with severe sepsis were included; of these, 178 (29%) had septic shock.Vitamin D deficiency (550 nmol/L) was present in 333 (55%) patients. The 90-day mortality did notdiffer among patients with or without vitamin D deficiency (28.3% vs. 28.5%, p¼ 0.789). Diabeteswas more common among patients deficient compared to those not deficient in vitamin D (30% vs.18%, p50.001). Hospital-acquired infections at admission were more prevalent in patients with avitamin D deficiency (31% vs. 16%, p50.001). A multivariable adjusted Cox regression modelshowed that low vitamin D levels could not predict 90-day mortality (550 nmol/L: hazard ratio (HR)0.99 (95% CI: 0.72–1.36), p40.9; and525 nmol/L: HR 0.44 (95% CI: 0.22–0.87), p¼ 0.018).Conclusions Vitamin D deficiency detected upon ICU admission was not associated with 90-daymortality in patients with severe sepsis or septic shock.

KEY MESSAGES

� In severe sepsis and septic shock, a vitamin D deficiency upon ICU admission was not associatedwith increased mortality.

� Compared to patients with sufficient vitamin D, patients with deficient vitamin D morefrequently exhibited diabetes, elevated C-reactive protein levels, and hospital-acquiredinfections upon ICU admission, and they more frequently developed acute kidney injury.

ARTICLE HISTORY

Received 21 August 2015Revised 15 December 2015Accepted 16 December 2015Published online 21 January2016

KEYWORDSSevere sepsis; septic shock;vitamin D; critical illness

Introduction

Among patients that are critically ill, the significance

of vitamin D levels at hospital admission is largely

unknown. Vitamin D has been suggested to play

an important role in modulating the innate

immune response by inhibiting the production of pro-

inflammatory cytokines (1,2). It is unclear, however,

whether hypovitaminosis might affect outcome in

patients with severe sepsis.

Previously, large-scale, retrospective, observational

laboratory and administrative registry studies have

shown an association between low pre-admission

serum 25-hydroxyvitamin D levels and mortality (3–5),

acute kidney injury (AKI) (6) and sepsis development (7)

in patients that were critically ill. Also, in a small,

prospective study with 100 patients that were critically

ill and required surgery, vitamin D levels within 24 h of

intensive care unit (ICU) admission were inversely related

CONTACT Tero I. Ala-Kokko tero.ala-kokko@ppshp.fi Department of Anaesthesiology, Division of Intensive Care Medicine, Oulu University Hospital, P.O.BOX 21, 90029 OUH, Finland*These authors contributed equally to this work.

� 2016 Taylor & Francis

Dow

nloa

ded

by [

E-L

ibra

ry I

nsel

], [

Vill

e Pe

ttilä

] at

07:

45 2

5 Ja

nuar

y 20

16

to 90-day mortality, even after adjusting for covariates

(8). In another prospective study with 201 patients,

however, vitamin D levels were not an independent risk

factor for mortality (9).

The role of vitamin D in sepsis is also controversial.

Retrospective registry studies with pre-admission vita-

min D sampling have suggested that low vitamin D

levels were associated with mortality in sepsis (10,11). In

contrast, a small prospective study showed that vitamin

D deficiency in patients with sepsis did not predict

mortality (12). Similarly, in a case control study, serum

vitamin D, measured soon after admission to the ICU,

was not associated with the development of acute lung

injury, hospital mortality, or one-year mortality among

patients with sepsis (13). Other studies suggested that a

vitamin D deficiency might predispose patients to

infection and/or increase sepsis severity. In a small

study among patients with a suspected infection, low

vitamin D levels measured in the emergency department

were associated with greater sepsis severity (14).

Similarly, in a retrospective analysis, pre-hospital meas-

urements of low vitamin D levels were associated with

the development of community-acquired pneumonia

and sepsis (15).

These conflicting results among studies on vitamin D

deficiency in patients that were critically ill led us to

conduct the present study, which was a substudy of the

prospective, multicentre observational FINNAKI study

(16). We aimed to study the association between vitamin

D levels measured at ICU admission and 90-day mortality

in a cohort of patients with severe sepsis or septic shock.

Materials and methods

This substudy of the prospective, observational, multi-

centre FINNAKI study was conducted in 17 Finnish ICUs,

between 1 September 2011 and 1 February 2012 (16).

The FINNAKI study included all emergency admissions

and elective postoperative admissions of adult patients

(age418 years) that were expected to stay in the ICU for

more than 24 h. Patients were excluded from the study

when they required intermediate care or chronic dialysis;

when the admission was elective, and the expected ICU

stay was less than 24 h; when a patient was readmitted

or transferred from another hospital, and they had

participated in the study for 5 days prior to the current

admission; when patients had inadequate language skills

or no permanent residence in Finland; or when patients

were organ donors.

In the present study, we measured plasma vitamin D

concentrations from blood samples drawn on the

admission day. All patients had presented with severe

sepsis or septic shock within 24 h of ICU admission.

Subjects from two ICUs were excluded, due to incom-

plete data on vasoactive treatment; patients were also

excluded when they had insufficient admission plasma

samples. Figure 1 illustrates the study flow chart.

The Ethics Committee of Helsinki University Hospital

approved the study protocol and the use of deferred

consent (DNRO 18/13/03/02/1010). Written consent was

obtained from all patients or their proxies.

Severe sepsis and septic shock were defined accord-

ing to the criteria of the American College of Chest

Physicians/Society of Critical Care Medicine (17). We

defined and staged AKI according to criteria from the

Kidney Disease: Improving Global Outcome (KDIGO);

these criteria included both serum creatinine (SCr) and

urine output data (18). The KDIGO classification defined

AKI as an increase in SCr by �26.5 mmol/L within 48 h, or

an increase in SCr to �1.5-fold higher than the baseline

level, or a urine output less than 0.5 mL/kg/h, measured

over 6 h. Organ failure was defined in terms of the

Sequential Organ Failure Assessment score (SOFA); a

score �3 indicated organ failure (19,20). Infections

present at admission into the ICU were considered

community-acquired, when they had already manifested

upon admission to the hospital. An infection that

manifested 448 h after hospital admission was defined

as hospital-acquired (21).

The following patient data were collected: age, sex,

body mass index, pre-existing comorbidities, chronic

medication, Acute Physiology and Chronic Health

Evaluation II (APACHE II) scores, daily SOFA scores,

type of admission, fluid balance at admission, source and

focus of infection, C-reactive protein (CRP) levels, and

platelet and leukocyte counts at admission. ICU, hospital,

and 90-day mortalities were recorded. Data on 90-day

survival were acquired from the Finnish Population

Register Centre. Information was not available on

whether patients had used vitamin D supplements at

baseline.

Definition of vitamin D deficiency

No consensus exists regarding optimal plasma levels of

vitamin D, or the best cut-off values in critically ill

conditions. We chose cut-off values for defining vitamin

D sufficiency (75 nmol/L; to convert nmol/L to ng/mL,

divide by 2.496) and vitamin D deficiency (50 nmol/L),

according to the US Endocrine Society clinical practice

guidelines (22). In our analyses, we chose another cut-off

value of 525 nmol/L to define a severe vitamin D

deficiency. This low range was previously shown to be

associated with illness severity among patients that were

critically ill (23).

2 T. I. ALA-KOKKO ET AL.

Dow

nloa

ded

by [

E-L

ibra

ry I

nsel

], [

Vill

e Pe

ttilä

] at

07:

45 2

5 Ja

nuar

y 20

16

Measurement of vitamin D plasma levels

Plasma vitamin D levels were measured with the

25-Hydroxy vitamin D EIA kit (Immunodiagnostic

Systems GmbH, Frankfurt, Germany), according to the

manufacturer’s instructions. Plasma samples were stored

at �70 �C, and they were thawed once for analysis. All

analyses were performed at the same time. To assess

accuracy, in each 96 plate certified 25-hydroxyvitamin D

amounts from the National Institute of Standards and

Technology, USA; Standard Reference Material-972

(SRM-972) were included (24). The inter-assay variations

(CVs) ranged from 5.9% to 8.7% for the different control

and certified materials.

Statistical analysis

Summary statistics for continuous or ordinal variables

are expressed as medians with 25th and 75th percentiles

Figure 1. Flow chart for the selection of patients included in the study.

ANNALS OF MEDICINE 3

Dow

nloa

ded

by [

E-L

ibra

ry I

nsel

], [

Vill

e Pe

ttilä

] at

07:

45 2

5 Ja

nuar

y 20

16

or as mean values with standard deviations (SD).

Comparisons between groups were performed with an

independent-samples t-test. Categorical variables were

analysed with Pearson’s chi-square test.

The impact of the vitamin D level on mortality was not

linear. Therefore, two different multivariable, adjusted,

Cox regression models were built to assess the impact of

a vitamin D deficiency on 90-day mortality. In model I,

the samples were divided into two groups of vitamin D

levels (550 vs.�50 nmol/L); in model II, the samples

were divided into four groups of vitamin D levels (525,

25 to550, 50 to575 and�75 nmol/L). Both models were

adjusted for the APACHE II-score, multiorgan failure

during 0–4 days, diabetes, use of systemic corticoster-

oids, length of hospital stay before ICU admission and

AKI. The results of the Cox regression models are

presented as the hazard ratio (HR) with 95% confidence

interval (95% CI). Kaplan–Meier survival curves were

evaluated with log-rank tests.

Two-tailed p values are reported, and a p value less

than 0.05 was considered statistically significant.

Analyses were performed with SPSS software for

Windows (IBM Corp., Released 2012, IBM SPSS Statistics

for Windows, Version 21.0, IBM Corp., Armonk, NY).

Results

Out of 5353 patients admitted to the ICU, 610 were in

severe sepsis or septic shock. Sixty-two percent were

male (Table 1). The median APACHE II and SAPS II scores

at admission were 23 and 42, respectively, and one-third

of the patients had septic shock at admission. The main

source of infection (76%) was community-acquired, and

pneumonia was the most common (54%) focus of

infection. Out of 610 patients, 315 (51.6%) also had

multiple organ failure at admission. The median (25–

75th percentile) vitamin D concentration was 48.1 nmol/

L (35.7–61.2) at admission. At ICU admission, more than

half (333/610, 54.6%) of the study patients had vitamin D

deficiency, with serum levels below 50 nmol/L. Sixty-two

percent (62.5%) of patients that were vitamin D-deficient

was male, and the mean age was 62.4 years (Table 1).

The numbers of patients with vitamin D levels525 nmol/

L, 25 to550 nmol/L, 50 to575 nmol/L and�75 nmol/L

were 56 (9%), 277 (45%), 218 (36%) and 59 (10%),

respectively.

Patients with vitamin D deficiency were younger, had

significantly higher CRP levels at admission, and had

diabetes more frequently than patients with sufficient

vitamin D serum levels (Table 1). Patients with vitamin D

deficiency had higher maximum SOFA scores than

patients with sufficient vitamin D levels, but they had

similar numbers of failing organs. Hospital-acquired

infections (HAIs) were also significantly more frequent

among patients with deficient levels than those with

sufficient levels of vitamin D. The development of AKI

and the need for renal replacement therapy was more

frequent in the vitamin D-deficient group (Table 2).

Otherwise, the groups were not different in the severity

of sepsis, organ dysfunctions, or demographics. There

was no difference in the fluid balance between groups

on the day of vitamin D sampling.

The crude 90-day mortalities for patients with defi-

cient and those with sufficient vitamin D levels were

29.2% vs. 29.7%, p¼ 0.789. Figure 2 illustrates the

Kaplan–Meier survival analysis for 90-day mortality,

stratified according to vitamin D deficiency. Table 3

presents the univariate analysis for 90-day mortality in

this study population. In the multivariable, adjusted Cox

regression model I, low vitamin D levels (550 nmol/L)

had a HR of 0.99 (95% CI: 0.72–1.36, p40.9). In the

multivariable, adjusted Cox model II, the HRs for

different vitamin D levels were compared to a sufficient

level of �75 nmol/L. Thus, levels525 nmol/L had a HR of

0.44 (95% CI: 0.22–0.87, p¼ 0.018); levels between 25

and 550 nmol/L had a HR of 0.78 (95% CI: 0.48–1.28,

p¼ 0.78); and levels between 50 and 575 nmol/L had a

HR of 0.64 (95% CI: 0.39–1.05, p¼ 0.078).

Discussion

In this study, patients with vitamin D deficiency had

diabetes, elevated CRP levels at admission, HAIs, and

development of AKI more frequently than patients with

sufficient vitamin D. However, plasma vitamin D levels

were not associated with 90-day mortality.

To date, the present study was the largest prospect-

ive, multicentre study that evaluated a patient cohort

with severe sepsis or septic shock for plasma vitamin D

levels at ICU admission. More than half of these patients

had a vitamin D deficiency. This prevalence of vitamin D

deficiency was similar to those reported in previous

studies (10,13). Earlier retrospective studies have shown

that low vitamin D levels were associated with mortality;

however, in those studies, the time period between

vitamin D sampling and ICU admission was variable,

ranging from 7 days to 1 year (3,5,6). In contrast, this

study found that a vitamin D deficiency, measured at

admission, was not associated with mortality. Our results

were consistent with those from a case control study by

Barnett et al. (12) and a prospective observational study

by Cecchi et al. (13). Those studies also examined

mortality among patients with severe sepsis or septic

shock and, similar to our study, they found no differ-

ences in mortality among patients with low or high

vitamin D levels.

4 T. I. ALA-KOKKO ET AL.

Dow

nloa

ded

by [

E-L

ibra

ry I

nsel

], [

Vill

e Pe

ttilä

] at

07:

45 2

5 Ja

nuar

y 20

16

We also found that, compared to patients with

sufficient vitamin D, patients with vitamin D deficiency

exhibited a higher frequency of diabetes, more pro-

nounced inflammatory reaction, and a predisposition

for HAIs. These findings suggested that vitamin D

deficiency may be a surrogate marker for poor health

and a risk factor for severe infections. Consistent

with this interpretation, previous data showed that

Table 1. Demographic and clinical data of patients categorized according to levels of vitamin D. Data expressed as meanwith standard deviation (SD) or as median (25–75th percentile) or number and percentage when appropriate.

Parameter All n¼ 610Deficient (550 nmol/L)

(n¼ 333)Non-deficient (�50 nmol/L)

(n¼ 277) p

Age (years) 62.4 (15.8) 61.2 (15.0) 65.0 (16.5) 0.004BMI (kg/m2) 28.0 (9.2) 28.5 (6.9) 27.5 (11.4) 0.18Sex (male) 381 (63) 209 (63) 172 (63) 0.87Co-morbidities

COPD 77 (13) 44 (13) 33 (12) 0.71Hypertension 322 (53) 181 (55) 141 (51) 0.42Atherosclerosis 80 (13) 38 (12) 42 (15) 0.19Cardiac insufficiency 63 (11) 28 (9) 35 (13) 0.12Diabetes 150 (25) 100 (30) 50 (18) 0.001Chronic renal insufficiency 42 (7) 22 (7) 20 (7) 0.87Chronic steroid medication 72 (12) 34 (10) 38 (14) 0.17Statin medication 171 (28) 88 (27) 83 (30) 0.32

APACHE II 23 (18–30) 24 (19–30) 23 (18–29) 0.21SAPS II 43 (34–54) 43 (34–56) 42 (34–52) 0.29SOFA adm. 8 (6–10) 8 (6–10) 8 (6–10) 0.19SOFA max. 9 (7–11) 9 (7–11) 8 (6–10) 0.034Operative adm. 142 (23) 82 (25) 60 (22) 0.40Emergency adm. 598 (99) 327 (99) 271 (98) 0.20Hosp. stay before ICU admission (days) 0.0 (0.0–2.0) 1.0 (0.0–3.0) 0.0 (0.0–1.0) 0.037Blood culture positive 135 (22) 71 (30) 64 (31) 0.76Septic shock on adm. 178 (29) 100 (30) 78 (28) 0.61Septic shock during days 1–5 in the ICU 477 (78) 262 (79) 215 (78) 0.75MOF on admission 315 (52) 180 (54) 135 (49) 0.19Fluid balance on adm. (L) 1071 (�54–2884) 1037 (�126–2867) 1084 (1.5–2923) 0.55CRP on adm. (mg/L) 162 (61–270) 168 (65–294) 156 (54–247) 0.002Leukocyte count (�103/mL) 12.8 (8.4–17.5) 12.8 (8.6–18.1) 12.6 (8.3–17.5) 0.72Platelet count (�109/mL) 17.7 (109–262) 189 (100–271) 166 (113–251) 0.089Source of infection*

CAI 424 (76) 212 (69) 212 (84) 50.001HAI 137 (24) 96 (31) 41 (16) 50.001

Infection focus 0.046Lungs 297 (54) 153 (51) 144 (58)Abdominal 143 (26) 85 (28) 58 (24)Urinary 31 (6) 17 (6) 14 (6)Soft tissue and skin 44 (8) 31 (10) 13 (5)CNS 17 (3) 8 (3) 9 (4)Endocarditis 6 (1) 1 (0.3) 5 (2)Catheter related 1 (0.2) 0 (0) 1 (0.4)Other foreign body 1 (0.2) 0 (0) 1 (0.4)Focus unknown 8 (1.5) 6 (2) 2 (0.8)

OutcomeICU mortality 67 (11) 43 (13) 24 (9) 0.12Hospital mortality 131 (22) 79 (24) 52 (19) 0.1790 days mortality 178 (29) 99 (30) 79 (29) 0.79

APACHE II, Acute Physiology and Chronic Health Evaluation Score; BMI, body mass index; CAI, community-acquired infection; HAI, hospital-acquired infection; SAPS II, Simplified Acute Physiology Score; SOFA adm., Sequential Organ Failure Assessment Score on admission to the ICU;SOFA max., Maximum Sequential Organ Failure Assessment Score achieved during the ICU stay; COPD, chronic obstructive pulmonary disease.

*Source of infection data is missing from 49 cases (7.5% in the vitamin D deficient group and 8.6% in the non-deficient group).

Table 2. Organ failures and ICU support treatments used categorized as vitamin D plasma levels. Results presentedas numbers and percentages.

Parameter All n¼ 610 Deficient (550 nmol/L) (n¼ 333) Non-deficient (�50 nmol/L) (n¼ 277) p

MOF days 0–4 366 (60) 208 (63) 158 (57) 0.19AKI during ICU 332 (54) 196 (59) 136 (49) 0.018Need for RRT 101 (17) 65 (20) 36 (13) 0.037Need for MV 411 (67) 229 (69) 182 (66) 0.44Need for vasoactives 465 (76) 251 (75) 214 (77) 0.63Sepsis steroid 194 (33) 114 (35) 80 (29) 0.16

MOF, multiorgan failure; AKI, acute kidney injury; RRT, renal replacement therapy; MV, mechanical ventilation.

ANNALS OF MEDICINE 5

Dow

nloa

ded

by [

E-L

ibra

ry I

nsel

], [

Vill

e Pe

ttilä

] at

07:

45 2

5 Ja

nuar

y 20

16

hypovitaminosis was associated with certain chronic

illnesses (25,26), obesity (27,28), the winter season (29),

and all-cause mortality in the general population (30),

and with the severity of illness among patients that were

critically ill (23). Another interpretation of our findings

might be that the inflammatory response to disease led

to lower plasma vitamin D levels, but not to a deficiency,

per se. Previous studies have, however, demonstrated

that vitamin D deficiency contributed to the patient risk

for severe infections and sepsis (6,7,15,31). Based on the

present study design, we did not measure baseline

vitamin D levels before the emergence of sepsis. Vitamin

D has been suggested to act as an immunomodulator.

According to that theory, patients in a high pro-

inflammatory state, such as those with severe sepsis,

would benefit the most from vitamin D supplementa-

tion. In our study population, plasma CRP levels were

elevated in patients with a vitamin D deficiency. In

contrast, in an earlier study, plasma vitamin D levels in

healthy volunteers did not correlate with cytokine levels

following an administration of endotoxin (32). Also, in a

Figure 2. Kaplan–Meier survival curve for 90 days, according to the vitamin D level. Vitamin D (25-hydroxyvitamin D) deficiency wasdefined as a plasma level less than 50 nmol/L. The solid line shows the group with vitamin D levels �50 nmol/L; the dotted line showsthe group with levels less than 50 nmol/L (p¼ 0.68, according to a log-rank test).

Table 3. Univariate analysis for 90-day mortality.

ParameterSurvivors(n¼ 432)

Non-survivors(n¼ 178) p

Age (years) 60.7 (15.6) 69 (14.7) 50.001Male gender 278 (64) 103 (58) 0.14BMI (kg/m2) 28.5 (10.1) 26.8 (6.6) 0.014APACHE II 22.6 (7) 29.2 (9.0) 50.001Chronic renal insufficiency 24 (6) 18 (10) 0.044Diabetes 115 (27) 35 (20) 0.067Vitamin D concentration

(nmol/L)48.2 (36.0–60.6) 46.9 (34.4–62.7) 0.43

Vitamin D deficient 234 (54) 99 (56) 0.79Emergency admission 422 (98) 176 (99) 0.23Operative admission 107 (25) 35 (20) 0.17MOF on admission 197 (45) 118 (66) 50.001Septic shock on admission 117 (27) 61 (34) 0.076Blood culture positive 98 (23) 37 (21) 40.9Lung infection 218 (51) 94 (53) 0.039Abdominal 103 (24) 45 (25) 0.57Septic shock during ICU stay 322 (75) 155 (87) 0.01MOF during days 0–4 226 (52) 140 (79) 50.001AKI 210 (49) 122 (69) 50.001Need for RRT days 0–4 49 (11) 48 (27) 50.001

Data expressed as mean with standard deviation (SD) or as median (25–75thpercentile) or number and percentage when appropriate. RRT, renalreplacement therapy; AKI, acute kidney injury; MOF, multiorgan failure;APACHE II, Acute Physiology and Chronic Health Evaluation Score; BMI,body mass index.

6 T. I. ALA-KOKKO ET AL.

Dow

nloa

ded

by [

E-L

ibra

ry I

nsel

], [

Vill

e Pe

ttilä

] at

07:

45 2

5 Ja

nuar

y 20

16

recent, small, randomized controlled study that included

67 patients with either severe sepsis or septic shock, a

single intravenous dose of calcitriol did not affect the

cytokine levels or mortality, although it increased IL-10

mRNA expression (33). These mixed results were not

surprising, because the role of vitamin D in the immune

system is certainly complex.

To date, only one randomized, placebo-controlled

study has examined the effect of oral, high-dose vitamin

D supplementation on mortality in patients that

were critically ill and had deficient vitamin D levels

(550 nmol/L) (34). Overall, that study reported no

benefit to the clinical status or mortality with vitamin

D supplementation. However, among patients with

severe vitamin D deficiency (�30 nmol/L), supplementa-

tion with vitamin D significantly lowered hospital

mortality, but not 28-day, or 6-month mortality.

However, in contrast to our cohort, only 8% of patients

in that study cohort had sepsis at admission. The Amrein

study was not limited to patients with sepsis; therefore,

those results may not be extrapolated to this patient

population. Interestingly in our Cox model with vitamin

D classified into four classes (model II) patients with

vitamin D values525 nmol/L were in lower risk to die

than those with vitamin D values�75 nmol/L. This result

might be due to relatively small sample size but might

be also a reflection of a nonlinear effect of vitamin D on

mortality. This further strengthens our conclusion that

low vitamin D values do not increase mortality within

severe sepsis patients. Furthermore, we observed rela-

tively low 90-day mortality (30%) due to severe sepsis

and septic shock; thus, it could be argued that, in our

patients, vitamin D played less of a biological role than

other factors and treatment effects.

Our study also had some inherent limitations. First,

fluid resuscitation and subsequent haemodilution may

have affected the levels of vitamin D (35). In our cohort,

however, the fluid balance on the admission day was

not different between patients with low or high vitamin

D levels. Second, the study was observational in nature;

therefore, the results could only show an association,

not a causal relationship. Third, confounding factors

might exist that were not accounted for in the

multivariate regression analysis. Furthermore, only a

single plasma vitamin D measurement was performed

at admission.

The strengths of this study included its large size and

its prospective nature. Our cohort included consecutive,

mixed patients critically ill with severe sepsis, from 15

different ICUs. To date, this was the largest study to

assess the association between admission vitamin D

levels and mortality in patients with severe sepsis or

septic shock. We also implemented a fixed 90-day

mortality endpoint, which is considered clinically

relevant.

Conclusions

In conclusion, vitamin D deficiency at ICU admission was

not associated with 90-day mortality in patients critically

ill with severe sepsis or septic shock.

Acknowledgements

This study was financially supported by the Medical ResearchCentre of Oulu, University of Oulu. We appreciated the expertassistance provided by the research nurses, Sinikka Salkio andSari Sutinen, in organizing the logistics of the samples. We alsoacknowledge the clinical investigators of the FINNAKI studygroup for collecting the data and providing care for the studypatients. The FINNAKI study group; Central Finland CentralHospital: Raili Laru-Sompa, Anni Pulkkinen, Minna Saarelainen,Mikko Reilama, Sinikka Tolmunen, Ulla Rantalainen, MarjaMiettinen. East Savo Central Hospital: Markku Suvela, KatrinePesola, Pekka Saastamoinen, Sirpa Kauppinen. HelsinkiUniversity Central Hospital: Ville Pettila, Kirsi-Maija Kaukonen,Anna-Maija Korhonen, Sara Nisula, Suvi Vaara, Raili Suojaranta-Ylinen, Leena Mildh, Mikko Haapio, Laura Nurminen, SariSutinen, Leena Pettila, Helina Laitinen, Heidi Syrja, KirsiHenttonen, Elina Lappi, Hillevi Boman. Jorvi Central Hospital:Tero Varpula, Paivi Porkka, Mirka Sivula Mira Rahkonen, AnneTsurkka, Taina Nieminen, Niina Prittinen. Kanta-Hame CentralHospital: Ari Alaspaa, Ville Salanto, Hanna Juntunen, TeijaSanisalo. Kuopio University Hospital: Ilkka Parviainen, AriUusaro, Esko Ruokonen, Stepani Bendel, Niina Rissanen,Maarit Lang, Sari Rahikainen, Saija Rissanen, Merja Ahonen,Elina Halonen, Eija Vaskelainen. Lansi Pohja Central Hospital:Jorma Heikkinen, Timo Lavander, Kirsi Heinonen, Anne-MariJuopperi. Middle Ostrobothnia Central Hospital: TadeuszKaminski, Fiia Gaddnas, Tuija Kuusela, Jane Roiko. SatakuntaHospital District: Vesa Lund, Paivi Tuominen, Pauliina Perkola,Riikka Tuominen, Marika Hietaranta, Satu Johansson. SouthKarelia Central Hospital: Seppo Hovilehto, Anne Kirsi, PekkaTiainen, Tuija Myllarinen, Pirjo Leino, Anne Toropainen.Tampere University Hospital: Anne Kuitunen, Ilona Leppanen,Markus Levoranta, Sanna Hoppu, Jukka Sauranen, JyrkiTenhunen, Atte Kukkurainen, Samuli Kortelainen, Simo Varila.Turku University Hospital: Outi Inkinen, Niina Koivuviita, JuttaKotamaki, Anu Laine. Oulu University Hospital: Tero Ala-Kokko,Jouko Laurila, Sinikka Salkio. Vaasa Central Hospital: Simo-Pekka Koivisto, Raku Hautamaki, Maria Skinnar.

Disclosure of interest

The authors declare that they have no conflict of interest.

References

1. Maruotti N, Cantatore FP. Vitamin D and the immune

system. J Rheumatol. 2010;37:491–5.2. Mutt SJ, Karhu T, Lehtonen S, Lehenkari P, Carlberg C,

Saarnio J, et al. Inhibition of cytokine secretion from

adipocytes by 1,25-dihydroxyvitamin D3 via the NF-kB

pathway. FASEB J. 2012;26:4400–7.

ANNALS OF MEDICINE 7

Dow

nloa

ded

by [

E-L

ibra

ry I

nsel

], [

Vill

e Pe

ttilä

] at

07:

45 2

5 Ja

nuar

y 20

16

3. Braun A, Chang D, Mahadevappa K, Gibbons FK, Liu Y,Giovannucci E, et al. Association of low serum 25-hydroxyvitamin D levels and mortality in the critically ill.Crit Care Med. 2011;39:671–7.

4. Venkatram S, Chilimuri S, Adrish M, Salako A, Patel M, Diaz-Fuentes G. Vitamin D deficiency is associated withmortality in the medical intensive care unit. Crit Care.2011;15:R292.

5. Braun AB, Gibbons FK, Litonjua AA, Giovannucci E,Christopher KB. Low serum 25-hydroxyvitamin D at criticalcare initiation is associated with increased mortality. CritCare Med. 2012;40:63–72.

6. Braun AB, Litonjua AA, Moromizato T, Gibbons FK,Giovannucci E, Christopher KB, et al. Association of lowserum 25-hydroxyvitamin D levels and acute kidney injuryin the critically ill. Crit Care Med. 2012;40:3170–9.

7. Moromizato T, Litonjua AA, Braun AB, Gibbons FK,Giovannucci E, Christopher KB. Association of low serum25-hydroxyvitamin D levels and sepsis in the critically ill.Crit Care Med. 2014;42:97–107.

8. Quraishi SA, Bittner EA, Blum L, McCarthy CM, Bhan I,Camargo CA Jr. Prospective study of vitamin D status atinitiation of care in critically ill surgical patients and risk of90-day mortality. Crit Care Med. 2014;42:1365–71.

9. Aygencel G, Turkoglu M, Tuncel AF, Candır BA, Bildacı YD,Pasaoglu H. Is vitamin D insufficiency associated withmortality of critically ill patients? Crit Care Res Pract.2013;2013:856747.

10. Amrein K, Zajic P, Schnedl C, Waltensdorfer A, Fruhwald S,Holl A. Vitamin D status and its association with season,hospital and sepsis mortality in critical illness. Crit Care.2014;24:R47.

11. Rech MA, Hunsaker T, Rodriguez J. Deficiency in 25-hydroxyvitamin D and 30-day mortality in patients withsevere sepsis and septic shock. Am J Crit Care.2014;23:e72–9.

12. Barnett N, Zhao Z, Koyama T, Janz DR, Wang CY, May AK,et al. Vitamin D deficiency and risk of acute lung injury insevere sepsis and severe trauma: a case-control study. AnnIntensive Care. 2014;4:5–14.

13. Cecchi A, Bonizzoli M, Douar S, Mangini M, Paladini S,Gazzini B, et al. Vitamin D deficiency in septic patients atICU admission is not a mortality predictor. MinervaAnestesiol. 2011;77:1184–9.

14. Ginde AA, Camargo CA Jr, Shapiro NI. Vitamin D insuffi-ciency and sepsis severity in emergency departmentpatients with suspected infection. Acad Emerg Med.2011;18:551–4.

15. Jovanovich AJ, Ginde AA, Holmen J, Jablonski K, Allyn RL,Kendrick J, et al. Vitamin D level and risk of community-acquired pneumonia and sepsis. Nutrients. 2014;6:2196–205.

16. Nisula S, Kaukonen KM, Vaara ST, Korhonen AM,Poukkanen M, Karlsson S, et al. Incidence, risk factorsand 90-day mortality of patients with acute kidney injuryin Finnish intensive care units: the FINNAKI study.Intensive Care Med. 2013;39:420–8.

17. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, KnausWA, et al. Definitions for sepsis and organ failure andguidelines for the use of innovative therapies in sepsis.The ACCP/SCCM Consensus Conference Committee.American College of Chest Physicians/Society of CriticalCare Medicine. Chest. 1992;101:1644–55.

18. Kidney Disease. KDIGO clinical practice guideline for acute

kidney injury. Kidney Int. 2012;2:1–138.19. Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A,

Bruining H, et al. The SOFA (Sepsis-related Organ Failure

Assessment) score to describe organ dysfunction/failure.

On behalf of the Working Group on Sepsis-Related

Problems of the European Society of Intensive Care

Medicine. Intensive Care Med. 1996;22:707–10.20. Brun-Buisson C, Meshaka P, Pinton P, Vallet B. EPISEPSIS: a

reappraisal of the epidemiology and outcome of severe

sepsis in French intensive care units. Intensive Care Med.

2004;30:580–8.21. Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance

definition of health care-associated infection and criteria

for specific types of infections in the acute care setting.

Am J Infect Control. 2008;36:309–32.22. Holick MF, Binkley NC, Bischoff-Ferrari HA, Gordon CM,

Hanley DA, Heaney RP, et al. Evaluation, treatment, and

prevention of vitamin D deficiency: an endocrine society

clinical practice guideline. J Clin Endocrinol Metab.

2011;96:1911–30.23. Nair P, Lee P, Reynolds C, Nguyen ND, Myburgh J, Eisman

JA, et al. Significant perturbation of vitamin D-parathyroid-

calcium axis and adverse clinical outcomes in critically ill

patients. Intensive Care Med. 2013;39:267–74.24. Wallace AM, Gibson S, de la Hunty A, Lamberg-Allardt C,

Ashwell M. Measurement of 25-hydroxyvitamin D in the

clinical laboratory: current procedures, performance char-

acteristics and limitations. Steroids. 2010;75:477–88.25. Muscogiuri G, Sorice GP, Ajjan R, Mezza T, Pilz S, Prioletta

A, et al. Can vitamin D deficiency cause diabetes and

cardiovascular diseases? Present evidence and future

perspectives. Nutr Metab Cardiovasc Dis. 2012;22:81–7.26. Huotari A, Herzig KH. Vitamin D and living in northern

latitudes-an endemic risk area for vitamin D deficiency.

Int J Circumpolar Health. 2008;67:164–78.27. Mutt SJ, Hypponen E, Saarnio J, Jarvelin MR, Herzig KH.

Vitamin D and adipose tissue-more than storage. Front

Physiol. 2014;5:228–36.28. Larose TL, Chen Y, Camargo CA Jr, Langhammer A,

Romundstad P, Mai XM. Factors associated with vitamin D

deficiency in a Norwegian population: the HUNT Study.

J Epidemiol Community Health. 2014;68:165–70.29. Voipio AJ, Pahkala KA, Viikari JS, Mikkila V, Magnussen CG,

Hutri-Kahonen N, et al. Determinants of serum 25(OH)D

concentration in young and middle-aged adults. The

cardiovascular risk in young Finns study. Ann Med.

2015;47:253–62.30. Schottker B, Jorde R, Peasey A, Thorand B, Jansen EH,

Groot L, et al. Vitamin D and mortality: meta-analysis of

individual participant data from a large consortium of

cohort studies from Europe and the United States. BMJ.

2014;348:g3656.31. Lange N, Litonjua AA, Gibbons FK, Giovannucci E,

Christopher KB. Pre-hospital vitamin D concentration,

mortality, and bloodstream infection in a hospitalized

patient population. Am J Med. 2013;126:640.e19–27.32. Kox M, van den Berg MJ, van der Hoeven JG, Wielders JP,

van der Ven AJ, Pickkers P. Vitamin D status is not

associated with inflammatory cytokine levels during

experimental human endotoxaemia. Clin Exp Immunol.

2013;171:231–6.

8 T. I. ALA-KOKKO ET AL.

Dow

nloa

ded

by [

E-L

ibra

ry I

nsel

], [

Vill

e Pe

ttilä

] at

07:

45 2

5 Ja

nuar

y 20

16

33. Leaf DE, Raed A, Donnino MW, Ginde AA, Waikar SS.Randomized controlled trial of calcitriol in severe sepsis.Am J Respir Crit Care Med. 2014;190:533–41.

34. Amrein K, Schnedl C, Holl A, Riedl R, Christopher KB,Pachler C, et al. Effect of high-dose vitamin D3 on hospitallength of stay in critically ill patients with vitamin D

deficiency: the VITdAL-ICU randomized clinical trial. JAMA.2014;312:1520–30.

35. Rousseau AF, Damas P, Janssens M, Kalin S, Ledoux D,Le Goff C, et al. Critical care and vitamin D statusassessment: what about immunoassays and calculatedfree 25OH-D? Clin Chim Acta. 2014;437:43–7.

ANNALS OF MEDICINE 9

Dow

nloa

ded

by [

E-L

ibra

ry I

nsel

], [

Vill

e Pe

ttilä

] at

07:

45 2

5 Ja

nuar

y 20

16