Abstract

Aims

A significant proportion of cardiac surgical patients develop critical post-operative complications. We aimed to investigate the association of pre-operative 25-hydroxyvitamin D (25(OH)D) levels with major cardiac and cerebrovascular events (MACCE) in cardiac surgical patients.

Methods and results

From January 2010 to August 2011, we consecutively measured circulating 25(OH)D in 4418 operated patients. Of the study cohort, 38.0% had deficient 25(OH)D values (<30 nmol/L) and additional 32.3% had insufficient values (30–49.9 nmol/L), whereas only 3.1% had values >100 nmol/L. The incidence of MACCE was 11.5%. In multivariable-adjusted logistic regression models, the odds ratio of MACCE at deficient, inadequate, and high 25(OH)D levels was 2.23 [95% confidence interval (CI): 1.31–3.79], 1.73 (95% CI: 1.01–2.96) and 2.34 (95% CI: 1.12–4.89), respectively, compared with 25(OH)D levels of 75–100 nmol/L. A U-shaped association with circulating 25(OH)D was also present for duration of mechanical ventilatory support and intensive care unit stay. Multivariable-adjusted 6- and 12-month mortality were higher in patients with deficient 25(OH)D levels compared with patients with 25(OH)D levels of 75–100 nmol/L.

Conclusion

Deficient 25(OH)D levels are prevalent in cardiac surgical patients in Central Europe and are independently associated with the risk of MACCE. Further research should clarify the potential of vitamin D supplements in reducing cardiovascular risk in vitamin D-deficient patients and also the mechanisms leading to adverse effects on the cardiovascular system in the small group of patients with 25(OH)D levels >100 nmol/L.

Trial registration information: Clinicaltrials.gov identifier number: NCT01552382.

Introduction

Low cardiac output syndrome (LOS), myocardial infarction (MI), and in-hospital death are critical post-operative complications in cardiac surgery, affecting 8–12% of operated patients.1 Approximately 1–3% of patients also develop a potentially devastating post-operative stroke.2

Determination of factors that can influence the aforementioned major cardiac and cardiovascular events (MACCEs) is of paramount importance. Recent evidence suggests that vitamin D deficiency [e.g. circulating 25-hydroxyvitamin D levels (25(OH)D) values <30 nmol/L] is independently associated with total and cardiovascular mortality,3 fatal stroke,4 sudden cardiac death,5,6 and death due to heart failure.5 More importantly, vitamin D-supplemented heart failure patients show improved survival.7

EuroSCORE is a key predictor of MACCEs.8 Some EuroSCORE factors such as age, sex, chronic pulmonary disease, MI, and poor kidney function are also related to vitamin D status.9–12 Deficient circulating 25(OH)D levels are widespread across the world.13 In Europe, the prevalence of vitamin D deficiency is ∼15% in adolescents,14 20–40% in general practices,15,16 and up to 75% and more in nursing home residents.17

The effect of pre-operative 25(OH)D status on clinical outcome in cardiac surgery is currently not known. We therefore aimed to investigate in a cohort of cardiac surgical patients the association of pre-operative 25(OH)D levels with MACCE.

Methods

Patients and study design

Since January 2010, we have measured serum 25(OH)D levels as a cardiovascular risk marker at our institution (geographic latitude: 52°N). Measurements were consecutively performed in fasting pre-operative blood samples in all outward patients hospitalized for cardiac surgery. All samples were collected on the last day before cardiac surgery between January 2010 and August 2011. Patients with heart transplants, pacemaker/defibrillator implants and patients under 18 years of age were excluded. In 473 cardiac surgical patients with 25(OH)D measurements data sets were incomplete, e.g. baseline or outcome parameters were missing. Altogether, a total of 4418 patient samples could be included in the present analysis. The vast majority of patients were Caucasians. The investigation was performed according to the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) Statement for cohort studies (www.strobe-statement.org). The study was approved by the local ethics committee and was registered at clinicaltrials.gov as NCT01552382.

Data collection

The database consists of pre-, peri-, and post-operative data that were prospectively collected in all cardiac surgical patients at our institution. Besides 25(OH)D, 38 additional parameters were retrieved for each patient. Among them, 23 patient and surgical characteristics [age, sex, body mass index BMI), height, weight, left ventricular ejection fraction, smoking, concomitant diagnoses such as myocardial infarction, stroke, hypertension, diabetes, chronic obstructive pulmonary disease (COPD), haemofiltration, peripheral arterial occlusive disease (PAOD) stage II or higher, previous thoracic surgery, urgency of surgery, on-pump/off-pump surgery, type of surgery, use of medications such as diuretics, aspirin, ACE-inhibitors, statins, and clopidogrel), six biochemical parameters (calcium, creatinine, glucose, triglycerides, total cholesterol, and C-reactive protein), six major event categories (in-hospital death, post-operative MI, LOS, stroke, 6-, and 12-month mortality), and three other outcome parameters (duration of ventilatory support, intensive care unit stay, and in-hospital stay) were assessed. Data quality was controlled by checking diagnoses with data of the hospitalizing physician. Outcome parameters were checked with the data of the medical controlling of our clinic.

Primary endpoint

The primary endpoint was the rate of MACCEs, defined as in-hospital death, MI, LOS, or stroke. This composite endpoint was a priori chosen because vitamin D deficiency is independently associated with total and cardiovascular mortality,3 fatal stroke,4 MI,9 and sudden cardiac death,5,6 the latter being particularly influenced by LOS. Myocardial infarction was considered to have occurred in cases of new persistent ST-segment changes. Low cardiac output syndrome was defined as a cardiac index ≤2.5 L/min/m2 or mixed venous oxygen blood saturation (SvO2) ≤50% requiring high-dose inotropic support and/or the need of mechanical circulatory support. A stroke was considered present when a clinically manifest motoric, sensory, or cognitive neurological deficit was recorded due to a cerebrovascular event. All events were assessed until discharge.

Secondary endpoints

Secondary endpoints were the duration of mechanical ventilatory support, intensive care unit (ICU) stay, in-hospital stay, and 6- and 12-month mortality.

Biochemical analyses

Circulating 25(OH)D levels were analysed by the autoanalyzer Liaison (DiaSorin, Stillwater, MN, USA). The Liaison assay reveals very similar 25(OH)D results compared with the liquid chromatography tandem mass spectrometry method,18 which is considered the gold standard. The measuring range lies between 10 and 375 nmol/L. Values below 10 nmol/L were considered 9.9 nmol/L. Calcium, creatinine, glucose, triglycerides, cholesterol, and C-reactive protein were measured using the Architect Autoanalyzer (Abbott, Wiesbaden, Germany). Glomerular filtration rate (eGFR) was estimated using the creatinine-based modification of diet in renal disease formula.

Statistics

Categorical variables were summarized as frequencies and percentages; continuous variables were summarized as means and SDs. We used the χ2 test and analysis of variance, respectively, to assess group-specific differences in categorical variables and continuous variables. We tested normal distribution of the data using the Kolmogorov–Smirnov test. Normal distribution was considered if probability values were >0.05. Non-normally distributed data were logarithmically transformed before analysis.

We graphically evaluated the unadjusted association between pre-operative 25(OH)D levels and MACCE using restricted cubic spline function. Multiple logistic regression analysis was carried out to assess the independent relationship between pre-operative 25(OH)D category and MACCE. According to published data,19–21 we used the following cut-off values for classifying vitamin D status: risk of deficiency (<30 nmol/L), risk of inadequacy (30–49.9 nmol/L), borderline status (50–74.9 nmol/L), adequacy (75–100 nmol/L), and potentially harmful (>100 nmol/L, to convert nanomolar to nanogram per millilitre divide by 2.496). The group with adequate vitamin D status was used as the reference group. We performed unadjusted analyses and estimated age- and sex-adjusted models, as well as multivariable models to examine the association between vitamin D status and the incidence of MACCE. Inclusion in the multivariable models was based on a priori determination of potential confounders of the association between 25(OH)D levels and MACCE. Covariates used for adjustment in multivariable models included concomitant diagnoses (previous cardiac surgery, previous myocardial infarction, diabetes mellitus, COPD, hypertension, previous stroke, haemofiltration, and PAOD), smoking, medications (diuretics, ACE-inhibitors, statins, clopidogrel, and aspirin), type of surgery [coronary artery bypass grafting (CABG), valve surgery, combined CABG and valve surgery, aortic surgery, on-pump/off-pump surgery), urgency of surgery, BMI, left ventricular ejection fraction, the logistic EuroSCORE, kidney function (eGFR in mL/min/1.73 m2), blood lipids, blood glucose, and inflammatory processes (C-reactive protein in mg/dL). We calculated absolute (incidence) rates and odds ratios (ORs) and corresponding 95% confidence intervals (CIs). In sensitivity analyses, we tested the reliability of our cut-off values by dividing the subgroup of patients with deficient 25(OH)D levels further into patients with severe deficiency (<15 nmol/L) and less severe deficiency (15–29.9 nmol/L).

For evaluating the association of 25(OH)D categories with secondary endpoints (duration of mechanical ventilatory support, ICU stay, and in-hospital stay), we used a two-factor analysis of covariance with the aforementioned patients and surgery characteristics as covariates. We performed multivariable-adjusted Cox regression analysis, including adjustment for season of blood sampling, to assess hazard ratios of 6- and 12-month mortality. Patients with incomplete follow-up were censored. All P values are reported two-sided. We considered a P-value of <0.05 as statistically significant. Analyses were performed using the statistical software package ‘Predictive Analysis SoftWare’ (PASW), version 18 (Chicago, IL, USA) and R 2.15.0 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Patient characteristics

Figure 1 illustrates the percentages of patients in different pre-operative 25(OH)D categories. Of the study cohort, 38.0% had deficient 25(OH)D values and additional 32.3% had insufficient values, whereas only 3.1% had values >100 nmol/L. Blood drawing in winter was most frequently associated with deficient 25(OH)D levels, whereas blood drawing in spring or summer was more often associated with adequate or high 25(OH)D levels (Table 1). Patients with vitamin D deficiency were more likely to be female, had a higher BMI and suffered more often from diabetes compared with patients who had adequate 25(OH)D concentrations. In addition, these patients had a higher prevalence of previous MI, were taking diuretics and ACE-inhibitors more often, and had higher blood lipid, glucose and EuroSCORE values. However, higher EuroSCORE values were also observed in patients with 25(OH)D values >100 nmol/L compared with patients with adequate 25(OH)D values.

Table 1

Characteristics of the study cohorta, broken down by 25-hydroxyvitamin D category

 25(OH)D: <30 nmo/L (n = 1680) 25(OH)D: 30–49.9 nmol/L (n = 1428) 25(OH)D: 50–74.9 nmol/L (n = 848) 25(OH)D: 75–100 nmol/L (n = 323) 25(OH)D: >100 nmol/L (n = 139) P value 
Gender (%, men) 59.1 70.5 76.9 76.2 69.1 <0.001 
Age (years) 68.1 ± 11.7 68.5 ± 10.6 67.8 ± 10.7 68.2 ± 10.3 68.0 ± 10.7 0.684 
Body mass index (kg/m227.9 ± 6.4 27.7 ± 6.0 27.2 ± 4.4 26.8 ± 4.3 26.7 ± 4.5 0.001 
Smokers (%) 35.9 38.3 38.8 37.7 34.5 0.325 
Left ventricular ejection fraction (%) 56 ± 13 56 ± 11 57 ± 12 56 ± 12 56 ± 12 0.114 
Glomerular filtration rate (mL/min/1.73 m273.6 ± 25.3 74.8 ± 22.4 76.0 ± 22.6 72.5 ± 24.0 71.8 ± 25.1 0.051 
EuroSCORE (logistic) 9.6 ± 11.6 8.8 ± 10.4 8.2 ± 10.2 7.8 ± 10.4 10.6 ± 10.5 0.002 
Season of blood drawing 
 Winter (%) 39.1 27.7 20.6 17.3 17.3 <0.001 
 Spring (%) 32.4 29.7 31.2 41.0 35.8 <0.001 
 Summer (%) 13.7 25.6 32.6 34.5 35.2 <0.001 
 Autumn (%) 14.7 16.9 15.5 7.2 11.7 <0.001 
Type of surgery 
 CABG (%) 41.7 37.9 35.9 43.5 28.8 0.001 
 Valve surgery (%) 31.9 33.6 34.4 32.1 33.1 0.730 
 Combined CABG and valve surgery (%) 14.5 16.3 16.3 13.9 18.7 0.382 
 Aortic surgery (%) 4.6 5.7 7.4 5.2 11.5 0.002 
 Others (%) 7.3 6.5 6.0 5.3 7.9 0.520 
 On-pump surgery (%) 77.3 77.4 78.8 76.9 78.4 0.915 
Urgency of surgery 
 Elective (%) 89.1 90.9 91.9 93.2 89.2 0.067 
 Urgency (%) 5.7 4.6 4.6 2.5 7.2 0.080 
 Emergency/ultima ratio (%) 5.2 4.5 3.3 4.3 3.6 0.430 
Concomitant diagnoses 
 Previous cardiac surgery (%) 8.1 8.0 7.5 9.3 15.8 0.059 
 Myocardial infarction (%) 18.6 18.1 12.7 13.9 15.8 0.001 
 Diabetes mellitus (%) 29.5 24.4 18.0 21.3 17.3 <0.001 
 Chronic obstructive pulmonary disease (%) 11.2 9.8 9.2 6.5 11.5 0.082 
 Stroke (%) 3.0 2.5 2.4 0.9 1.4 0.210 
 Renal replacement therapy (%) 2.0 1.5 1.2 2.2 5.0 0.023 
 Peripheral arterial occlusive disease (%) 9.5 9.7 7.8 9.3 12.2 0.399 
 Hypertension (%) 76.7 78.6 75.1 73.2 76.4 0.172 
Biochemical parameters 
 Calcium (mmol/L) 2.39 ± 0.13 2.39 ± 0.12 2.39 ± 0.12 2.39 ± 0.11 2.40 ± 0.11 0.590 
 C-reactive protein (mg/dL) 1.12 ± 2.63 0.97 ± 2.41 0.80 ± 2.03 1.17 ± 2.92 1.12 ± 2.03 0.019 
 Glucose (mg/dL) 123 ± 49 118 ± 45 113 ± 35 114 ± 39 114 ± 33 <0.001 
 Triglycerides (mg/dL) 164 ± 119 149 ± 86 134 ± 70 136 ± 75 124 ± 50 <0.001 
 Total cholesterol (mg/dL) 197 ± 48 194 ± 45 186 ± 45 184 ± 39 182 ± 40 <0.001 
Medications 
 Diuretics (%) 51.9 47.2 41.8 46.3 44.6 <0.001 
 Angiotensin-converting enzyme inhibitors (%) 51.2 51.7 44.9 49.1 40.3 0.002 
 Clopidogrel (%) 10.7 8.0 7.9 8.3 11.5 0.646 
 Aspirin (%) 52.7 50.2 52.5 50.3 51.1 0.045 
 Statins (%) 54.5 54.8 56.9 53.8 55.7 0.802 
 25(OH)D: <30 nmo/L (n = 1680) 25(OH)D: 30–49.9 nmol/L (n = 1428) 25(OH)D: 50–74.9 nmol/L (n = 848) 25(OH)D: 75–100 nmol/L (n = 323) 25(OH)D: >100 nmol/L (n = 139) P value 
Gender (%, men) 59.1 70.5 76.9 76.2 69.1 <0.001 
Age (years) 68.1 ± 11.7 68.5 ± 10.6 67.8 ± 10.7 68.2 ± 10.3 68.0 ± 10.7 0.684 
Body mass index (kg/m227.9 ± 6.4 27.7 ± 6.0 27.2 ± 4.4 26.8 ± 4.3 26.7 ± 4.5 0.001 
Smokers (%) 35.9 38.3 38.8 37.7 34.5 0.325 
Left ventricular ejection fraction (%) 56 ± 13 56 ± 11 57 ± 12 56 ± 12 56 ± 12 0.114 
Glomerular filtration rate (mL/min/1.73 m273.6 ± 25.3 74.8 ± 22.4 76.0 ± 22.6 72.5 ± 24.0 71.8 ± 25.1 0.051 
EuroSCORE (logistic) 9.6 ± 11.6 8.8 ± 10.4 8.2 ± 10.2 7.8 ± 10.4 10.6 ± 10.5 0.002 
Season of blood drawing 
 Winter (%) 39.1 27.7 20.6 17.3 17.3 <0.001 
 Spring (%) 32.4 29.7 31.2 41.0 35.8 <0.001 
 Summer (%) 13.7 25.6 32.6 34.5 35.2 <0.001 
 Autumn (%) 14.7 16.9 15.5 7.2 11.7 <0.001 
Type of surgery 
 CABG (%) 41.7 37.9 35.9 43.5 28.8 0.001 
 Valve surgery (%) 31.9 33.6 34.4 32.1 33.1 0.730 
 Combined CABG and valve surgery (%) 14.5 16.3 16.3 13.9 18.7 0.382 
 Aortic surgery (%) 4.6 5.7 7.4 5.2 11.5 0.002 
 Others (%) 7.3 6.5 6.0 5.3 7.9 0.520 
 On-pump surgery (%) 77.3 77.4 78.8 76.9 78.4 0.915 
Urgency of surgery 
 Elective (%) 89.1 90.9 91.9 93.2 89.2 0.067 
 Urgency (%) 5.7 4.6 4.6 2.5 7.2 0.080 
 Emergency/ultima ratio (%) 5.2 4.5 3.3 4.3 3.6 0.430 
Concomitant diagnoses 
 Previous cardiac surgery (%) 8.1 8.0 7.5 9.3 15.8 0.059 
 Myocardial infarction (%) 18.6 18.1 12.7 13.9 15.8 0.001 
 Diabetes mellitus (%) 29.5 24.4 18.0 21.3 17.3 <0.001 
 Chronic obstructive pulmonary disease (%) 11.2 9.8 9.2 6.5 11.5 0.082 
 Stroke (%) 3.0 2.5 2.4 0.9 1.4 0.210 
 Renal replacement therapy (%) 2.0 1.5 1.2 2.2 5.0 0.023 
 Peripheral arterial occlusive disease (%) 9.5 9.7 7.8 9.3 12.2 0.399 
 Hypertension (%) 76.7 78.6 75.1 73.2 76.4 0.172 
Biochemical parameters 
 Calcium (mmol/L) 2.39 ± 0.13 2.39 ± 0.12 2.39 ± 0.12 2.39 ± 0.11 2.40 ± 0.11 0.590 
 C-reactive protein (mg/dL) 1.12 ± 2.63 0.97 ± 2.41 0.80 ± 2.03 1.17 ± 2.92 1.12 ± 2.03 0.019 
 Glucose (mg/dL) 123 ± 49 118 ± 45 113 ± 35 114 ± 39 114 ± 33 <0.001 
 Triglycerides (mg/dL) 164 ± 119 149 ± 86 134 ± 70 136 ± 75 124 ± 50 <0.001 
 Total cholesterol (mg/dL) 197 ± 48 194 ± 45 186 ± 45 184 ± 39 182 ± 40 <0.001 
Medications 
 Diuretics (%) 51.9 47.2 41.8 46.3 44.6 <0.001 
 Angiotensin-converting enzyme inhibitors (%) 51.2 51.7 44.9 49.1 40.3 0.002 
 Clopidogrel (%) 10.7 8.0 7.9 8.3 11.5 0.646 
 Aspirin (%) 52.7 50.2 52.5 50.3 51.1 0.045 
 Statins (%) 54.5 54.8 56.9 53.8 55.7 0.802 

CABG, coronary artery bypass grafting.

aMean ± SD or percentage of observations when appropriate.

Figure 1

Numbers and percentages of cardiac surgical patients in different 25-hydroxyvitamin D categories.

Figure 1

Numbers and percentages of cardiac surgical patients in different 25-hydroxyvitamin D categories.

In multivariable models, blood drawing between autumn and spring, female sex, higher BMI, diabetes, and a history of MI remained independently associated with deficient 25(OH)D levels (see Supplementary material online, Table S1). Season of blood drawing was also significantly associated with high 25(OH)-D levels, the probability being lowest if blood samples were collected in autumn or winter.

Primary endpoint

Overall, the risk of MACCE was 11.5%. Figure 2 illustrates the unadjusted relationship between pre-operative 25(OH)D levels and the composite outcome parameter MACCE. Risk was high in patients with deficient 25(OH)D levels and also in patients with 25(OH)D values >100 nmol/L. In Table 2, the OR of MACCE is given by subgroup of 25(OH)D category. Compared with the reference group, in those patients who had 25(OH)D levels in the deficiency range or >100 nmol/L the OR of MACCE was approximately twice as high in the unadjusted model. Results did not change substantially in multivariable-adjusted models. Sensitivity analyses also demonstrate that results did not change substantially when patients with deficient 25(OH)D levels were further divided into subgroups with severe and less severe deficiency in the fully adjusted model of MACCE (OR for severe 25(OH)D deficiency = 2.42 (95% CI: 1.39–4.26) and for less severe 25(OH)D deficiency = 2.14 (95% CI: 1.29–3.57). Incidence rates for the different components of MACCE were all lowest in the subgroup of patients with 25(OH)D levels of 75–100 nmol/L (see Supplementary material online, Table S2). Multivariable-adjusted OR for LOS was significantly higher and OR for in-hospital mortality tended to be higher in patients with deficient 25(OH)D levels compared with adequate 25(OH)D levels (see Supplementary material online, Table S2). We also imputed data of the 473 patients with missing clinical data in additional statistical analyses. Results did not differ substantially compared with the analysis which was restricted to patients with a complete data set (data not shown).

Table 2

Adjusted odds ratio (OR) for MACCE by subgroup of serum 25-hydroxyvitamin D concentration

 n MACCE, n (%) Model 1, OR (95% CI) Model 2, OR (95% CI) Model 3, OR (95% CI) Model 4, OR (95% CI) 
25-Hydroxyvitamin D 
 <30 nmol/L 1680 241 (14.3) 2.29 (1.46–3.61) 2.24 (1.42–3.54) 2.21 (1.31–3.74) 2.23 (1.31–3.79) 
 30–49.9 nmol/L 1428 147 (10.3) 1.57 (0.99–2.50) 1.55 (0.97–2.47) 1.71 (1,00–2.92) 1.73 (1.01–2.96) 
 50–74.9 nmol/L 848 78 (9.2) 1.39 (0.85–2.27) 1.41 (0.86–2.31) 1.62 (0.92–2,83) 1.65 (094–2.91) 
 75–100 nmol/L 323 22 (6.8) 1.0 (reference) 1.0 (reference) 1.0 (reference) 1.0 (reference) 
 >100 nmol/L 139 21 (15.1) 2.44 (1.30–4.61) 2.45 (1.30–4.64) 2.31 (1.11–4.79) 2.34 (1.12–4.89) 
 n MACCE, n (%) Model 1, OR (95% CI) Model 2, OR (95% CI) Model 3, OR (95% CI) Model 4, OR (95% CI) 
25-Hydroxyvitamin D 
 <30 nmol/L 1680 241 (14.3) 2.29 (1.46–3.61) 2.24 (1.42–3.54) 2.21 (1.31–3.74) 2.23 (1.31–3.79) 
 30–49.9 nmol/L 1428 147 (10.3) 1.57 (0.99–2.50) 1.55 (0.97–2.47) 1.71 (1,00–2.92) 1.73 (1.01–2.96) 
 50–74.9 nmol/L 848 78 (9.2) 1.39 (0.85–2.27) 1.41 (0.86–2.31) 1.62 (0.92–2,83) 1.65 (094–2.91) 
 75–100 nmol/L 323 22 (6.8) 1.0 (reference) 1.0 (reference) 1.0 (reference) 1.0 (reference) 
 >100 nmol/L 139 21 (15.1) 2.44 (1.30–4.61) 2.45 (1.30–4.64) 2.31 (1.11–4.79) 2.34 (1.12–4.89) 

Model 1: unadjusted data; model 2: adjusted for age and sex; model 3: adjusted for as in model 2 and for all concomitant diagnoses and medications listed in Table 1, smoking, type of surgery, on-pump/off-pump surgery, urgency of surgery, BMI, left ventricular ejection fraction, blood glucose and lipids, and the logistic EuroSCORE; model 4: adjusted for as in model 3 and for kidney function (eGFR) and inflammatory processes (C-reactive protein).

Figure 2

Spline function graph of the unadjusted relationship between pre-operative 25-hydroxyvitamin D concentration and probability of the composite adverse outcome. Top and bottom lines represent the 95% confidence interval of the relationship.

Figure 2

Spline function graph of the unadjusted relationship between pre-operative 25-hydroxyvitamin D concentration and probability of the composite adverse outcome. Top and bottom lines represent the 95% confidence interval of the relationship.

Secondary endpoints

Data on mechanical ventilatory support, ICU stay, and in-hospital stay are presented in Table 3, broken down by 25(OH)D category. Duration of mechanical ventilatory support and ICU stay also showed a U-shaped association with vitamin D status. In detail, patients with deficient 25(OH)D levels and with levels >100 nmol/L had a prolonged duration of mechanical ventilatory support as well as prolonged ICU stay, whereas duration of mechanical ventilatory support and ICU stay was lowest in patients with 25(OH)D levels of 75–100 nmol/L. In-hospital stay was not related to vitamin D status. Multivariable-adjusted 6- and 12-month mortality were higher in patients with deficient pre-operative 25(OH)D levels compared with adequate pre-operative 25(OH)D levels (see Supplementary material online, Table S2).

Table 3

Ventilatory support, intensive care unit stay, and in-hospital stay by subgroup of 25-hydroxyvitamin D concentration

 <30 nmol/L (n = 1680) 30–49.9 nmol/L (n = 1428) 50–74.9 nmol/L (n = 848) 75–100 nmol/L (n = 323) >100 nmol/L (n = 139) P valuea,b 
Ventilatory support (h) 42.1 ± 143.8 33.2 ± 114.5 30.9 ± 114.7 23.9 ± 82.0 54.4 ± 166.7 0.03 
ICU stay (h) 94.3 ± 209.3 77.8 ± 192.7 72.2 ± 166.3 71.8 ± 192.5 117.6 ± 259.2 0.016 
In-hospital stay (days) 14.9 ± 10.0 14.6 ± 9.3 14.2 ± 8.2 14.8 ± 12.5 16.1 ± 10.6 0.710 
 <30 nmol/L (n = 1680) 30–49.9 nmol/L (n = 1428) 50–74.9 nmol/L (n = 848) 75–100 nmol/L (n = 323) >100 nmol/L (n = 139) P valuea,b 
Ventilatory support (h) 42.1 ± 143.8 33.2 ± 114.5 30.9 ± 114.7 23.9 ± 82.0 54.4 ± 166.7 0.03 
ICU stay (h) 94.3 ± 209.3 77.8 ± 192.7 72.2 ± 166.3 71.8 ± 192.5 117.6 ± 259.2 0.016 
In-hospital stay (days) 14.9 ± 10.0 14.6 ± 9.3 14.2 ± 8.2 14.8 ± 12.5 16.1 ± 10.6 0.710 

aAnalysis of covariance was used to compare logarithmically transformed dependent variables between the 25(OH)D categories.

bIncluded covariates: all variables listed in Table 1 except for season of blood drawing.

Discussion

This investigation could demonstrate that in cardiac surgical patients circulating 25(OH)D levels below 30 nmol/L were independently associated with poor in-hospital outcome and increased post-operative mortality. Whereas the risk of MACCE was lowest at circulating 25(OH)D levels of 75–100 nmol/L, it rose again at values >100 nmol/L. However, the latter group was very small. Generally, data are in gross agreement with earlier results of a U-shaped or inverse J-shaped association between 25(OH)D and fatal outcome.10,16,20,22 Nonetheless, to the best of our knowledge this study demonstrates for the first time a significant increase in cardiovascular morbidity and mortality at 25(OH)D levels >100 nmol/L. Our results concur with a recent meta-analysis of prospective cohort studies on circulating 25(OH)D levels and total mortality in the general population,22 suggesting optimal 25(OH)D levels around 75–87.5 nmol/L.

Several beneficial vitamin D effects on the cardiovascular system have been identified and summarized. 23 Despite these findings the measurement of circulating 25(OH)D in the clinical setting is still hotly debated and has been questioned.24 However, it is noteworthy that recent studies reported a high prevalence of adverse outcomes, including high mortality rates, in critically ill patients with deficient 25(OH)D levels.25,26 Our results demonstrate that deficient 25(OH)D levels are very prevalent in cardiac surgical patients in Central Europe and are associated with a two-fold higher risk of MACCE compared with adequate 25(OH)D levels. It should also be noted that vitamin D was already demonstrated to reduce falls and fractures in elderly people27,28 and that the daily dose for fracture prevention should be 800–2000 international units. Justification for the prevention and treatment of vitamin D deficiency needs only one proven benefit and these are (at least) the beneficial effects on musculoskeletal health. Likely evidence does also exist for beneficial vitamin D effects on total mortality.29,30 In line with our data, vitamin D administration is most effective in those patients who have deficient 25(OH)D levels.11,31 Nevertheless, future research is still needed to clarify causality of the observed association between deficient 25(OH)D concentrations and the risk of MACCE.

The composite endpoint MACCE consists of components that are also important outcome parameters in patients with cardiovascular disease not receiving cardiac surgery.3–6,9 In a meta-analysis of prospective cohort studies,32 an inverse association between circulating 25(OH)D ranging from 20 to 60 nmol/L and the risk of cardiovascular disease has been demonstrated. The dose–response association indicates an up to two-fold higher relative risk of cardiovascular disease at circulating 25-OHD levels of 20 nmol/L compared with levels of 60 (to 75) nmol/L. However, this meta-analysis also recognized the need to further clarify the association between 25(OH)D >60 nmol/L and cardiovascular disease risk.

In line with earlier results,16 only a small percentage of patients had 25(OH)D levels beyond 100 nmol/L. Nevertheless, the increased risk of MACCE in this subgroup needs particular consideration. At present, possible mechanisms of adverse cardiovascular 25(OH)D effects at levels >100 nmol/L, if any, are unclear. In our study, summer blood drawing was an independent predictor of high 25(OH)D levels. Summer blood drawing is also related to enhanced calcium absorption efficiency from the gut.33 Excess calcium intake—and thus a high amount of absorbed calcium—have been made responsible for an enhanced risk of incident MI,34 probably induced by a transient rise in serum calcium and subsequent vascular calcification.35 Interestingly enough, excess vitamin D has also been made responsible for several documented and unforeseeable deaths.36 However, hypercalcaemia, which is the hallmark of vitamin D intoxication, does not occur unless 25(OH)D levels exceed 375 nmol/L.37 Note that fasting serum calcium levels did not differ significantly between 25(OH)D categories. Alternative explanations are also possible: high circulating 25(OH)D levels sometimes reflect low availability of the active vitamin D hormone 1,25-dihydroxyvitamin D (1,25(OH)2D),38 which has important actions on the cellular and subcellular level.39 Thus, in some cases high 25(OH)D levels may indicate deficient instead of excess vitamin D action. Higher 25(OH)D levels have also been reported in individuals with the APOE ɛ 4 gene variant.40 Since this gene variant is associated with an increased cardiovascular disease risk, high 25(OH)D levels may probably only be indicative of an increased risk of MACCE but not causally related to it.

Our study has several strengths. First, we were able to recruit a large number of patients within a relatively small period of time. This substantially reduces the risk of results being biased by progress in cardiac surgical procedures and intensive care medicine. Second, due to the short follow-up, we could reliably assess circulating 25(OH)D levels at the time when an event occurred or the patient was censored. Thus, with respect to the primary endpoint the short follow-up released us from the need to perform adjustments for season of blood drawing. Third, we were able to perform multiple adjustments for various demographically and clinically relevant data, including important surgery-related variables. Finally, the independent and U-shaped association of 25(OH)D with clinically important secondary endpoints such as duration of mechanical ventilatory support and ICU stay further underlines the assumption of a causal relationship between circulating 25(OH)D levels and clinical outcome.

Our study has the limitation that, due to the lack of 1,25(OH)2D measurements, we were unable to completely resolve the question of whether excess vitamin D or deficient vitamin D action is responsible for the excess rate of MACCE in patients with 25(OH)D levels >100 nmol/L. Moreover, the study was largely restricted to Caucasians. However, a U-shaped association between 25(OH)D and total mortality has also been reported in mixed ethnic groups.10

In conclusion, deficient circulating 25(OH)D levels are very prevalent in cardiac surgical patients and independently associated with the risk of MACCE. Our data add to the accumulating evidence that vitamin D may have important beneficial effects on the cardiovascular system. Moreover, our data support the assumption that the target range should be 75–100 nmol/L. Nevertheless, randomized controlled trials are needed to assess causality. Clarification of the mechanisms leading to adverse effects on the cardiovascular system in the small group of patients with high 25(OH)D levels is also needed.

Supplementary material

Supplementary material is available at European Heart Journal online.

Authors’ contribution

Study concept and design: A.Z.; acquisition of data: A.Z., J.K., J.D.; analysis and interpretation of data: A.Z., J.B.; drafting the manuscript: A.Z.; critical revision of the manuscript for important intellectual content: C.K., J.F.G.

Conflict of interest: A.Z. received speaker honoraria from DiaSorin, Germany and Abbott, Germany.

References

1
Massoudy
P
Thielmann
M
Lehmann
N
Marr
A
Kleikamp
G
Maleszka
A
Zittermann
A
Körfer
R
Radu
M
Krian
A
Litmathe
J
Gams
E
Sezer
O
Scheld
H
Schiller
W
Welz
A
Dohmen
G
Autschbach
R
Slottosch
I
Wahlers
T
Neuhäuser
M
Jöckel
KH
Jakob
H
Prognostic impact of previous percutaneous coronary interventions on coronary artery bypass graft surgery: a multicenter analysis
J Thorac Cardiovasc Surg
 , 
2009
, vol. 
137
 (pg. 
840
-
845
)
2
Møller
CH
Penninga
L
Wetterslev
J
Steinbrüchel
DA
Gluud
C
Clinical outcomes in randomized trials of off- vs. on-pump coronary artery bypass surgery: systematic review with meta-analyses and trial sequential analyses
Eur Heart J
 , 
2008
, vol. 
29
 (pg. 
2601
-
2616
)
3
Dobnig
H
Pilz
S
Scharnagl
H
Renner
W
Seelhorst
U
Wellnitz
B
Kinkeldei
J
Boehm
BO
Weihrauch
G
Maerz
W
Independent association of low serum 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D levels with all-cause and cardiovascular mortality
Arch Intern Med
 , 
2008
, vol. 
168
 (pg. 
1340
-
1349
)
4
Pilz
S
Dobnig
H
Fischer
JE
Wellnitz
B
Seelhorst
U
Boehm
BO
März
W
Low vitamin D levels predict stroke in patients referred to coronary angiography
Stroke
 , 
2008
, vol. 
39
 (pg. 
2611
-
2613
)
5
Pilz
S
März
W
Wellnitz
B
Seelhorst
U
Fahrleitner-Pammer
A
Dimai
HP
Boehm
BO
Dobnig
H
Association of vitamin D deficiency with heart failure and sudden cardiac death in a large cross-sectional study of patients referred for coronary angiography
J Clin Endocrinol Metab
 , 
2008
, vol. 
93
 (pg. 
3927
-
3935
)
6
Drechsler
C
Pilz
S
Obermayer-Pietsch
B
Verduijn
M
Tomaschitz
A
Krane
V
Espe
K
Dekker
F
Brandenburg
V
März
W
Ritz
E
Wanner
C
Vitamin D deficiency is associated with sudden cardiac death, combined cardiovascular events, and mortality in haemodialysis patients
Eur Heart J
 , 
2010
, vol. 
31
 (pg. 
2253
-
2261
)
7
Gotsman
I
Shauer
A
Zwas
DR
Hellman
Y
Keren
A
Lotan
C
Admon
D
Vitamin D deficiency is a predictor of reduced survival in patients with heart failure; vitamin D supplementation improves outcome
Eur J Heart Fail
 , 
2012
, vol. 
14
 (pg. 
357
-
366
)
8
Rodés-Cabau
J
Deblois
J
Bertrand
OF
Mohammadi
S
Courtis
J
Larose
E
Dagenais
F
Déry
JP
Mathieu
P
Rousseau
M
Barbeau
G
Baillot
R
Gleeton
O
Perron
J
Nguyen
CM
Roy
L
Doyle
D
De Larochellière
R
Bogaty
P
Voisine
P
Nonrandomized comparison of coronary artery bypass surgery and percutaneous coronary intervention for the treatment of unprotected left main coronary artery disease in octogenarians
Circulation
 , 
2008
, vol. 
118
 (pg. 
2374
-
2381
)
9
Giovannucci
E
Liu
Y
Hollis
BW
Rimm
EB
25-Hydroxyvitamin D and risk of myocardial infarction in men: a prospective study
Arch Intern Med
 , 
2008
, vol. 
168
 (pg. 
1174
-
1180
)
10
Melamed
ML
Michos
ED
Post
W
Astor
B
25-Hydroxyvitamin D levels and the risk of mortality in the general population
Arch Intern Med
 , 
2008
, vol. 
168
 (pg. 
1629
-
1637
)
11
Lehouck
A
Mathieu
C
Carremans
C
Baeke
F
Verhaegen
J
Van Eldere
J
Decallonne
B
Bouillon
R
Decramer
M
Janssens
W
High doses of vitamin D to reduce exacerbations in chronic obstructive pulmonary disease: a randomized trial
Ann Intern Med
 , 
2012
, vol. 
156
 (pg. 
105
-
114
)
12
Pilz
S
Iodice
S
Zittermann
A
Grant
WB
Gandini
S
Vitamin D status and mortality risk in chronic kidney disease: a meta-analysis of prospective studies
Am J Kid Dis
 , 
2011
, vol. 
58
 (pg. 
374
-
382
)
13
Mithal
A
Wahl
DA
Bonjour
JP
Burckhardt
P
Dawson-Hughes
B
Eisman
JA
El-Hajj Fuleihan
G
Josse
RG
Lips
P
Morales-Torres
J
IOF Committee of Scientific Advisors (CSA) Nutrition Working Group
Global vitamin D status and determinants of hypovitaminosis D
Osteoporos Int
 , 
2009
, vol. 
20
 (pg. 
1807
-
1820
)
14
González-Gross
M
Valtueña
J
Breidenassel
C
Moreno
LA
Ferrari
M
Kersting
M
De Henauw
S
Gottrand
F
Azzini
E
Widhalm
K
Kafatos
A
Manios
Y
Stehle
P
HELENA Study Group
Vitamin D status among adolescents in Europe: the Healthy Lifestyle in Europe by Nutrition in Adolescence study
Br J Nutr
 , 
2012
, vol. 
107
 (pg. 
755
-
764
)
15
Zittermann
A
The estimated benefits of vitamin D for Germany
Mol Nutr Food Res
 , 
2010
, vol. 
54
 (pg. 
1164
-
1171
)
16
Durup
D
Jørgensen
HL
Christensen
J
Schwarz
P
Heegaard
AM
Lind
B
A Reverse J-Shaped Association of All-Cause Mortality with Serum 25-Hydroxyvitamin D in General Practice, the CopD Study
J Clin Endocrinol Metab
 , 
2012
, vol. 
97
 (pg. 
2644
-
2652
)
17
Pilz
S
Dobnig
H
Tomaschitz
A
Kienreich
K
Meinitzer
A
Friedl
C
Wagner
D
Piswanger-Sölkner
C
März
W
Fahrleitner-Pammer
A
Low 25-hydroxyvitamin D is associated with increased mortality in female nursing home residents
J Clin Endocrinol Metab
 , 
2012
, vol. 
97
 (pg. 
E653
-
E657
)
18
Farrell
CJ
Martin
S
McWhinney
B
Straub
I
Williams
P
Herrmann
M
State-of-the-art vitamin D assays: a comparison of automated immunoassays with liquid chromatography-tandem mass spectrometry methods
Clin Chem
 , 
2012
, vol. 
58
 (pg. 
531
-
542
)
19
IOM (Institute of Medicine)
Dietary Reference Intakes for Calcium and Vitamin D
 , 
2011
Washington, DC
The National Academies Press
20
Michaëlsson
K
Baron
JA
Snellman
G
Gedeborg
R
Byberg
L
Sundström
J
Berglund
L
Arnlöv
J
Hellman
P
Blomhoff
R
Wolk
A
Garmo
H
Holmberg
L
Melhus
H
Plasma vitamin D and mortality in older men: a community-based prospective cohort study
Am J Clin Nutr
 , 
2010
, vol. 
92
 (pg. 
841
-
848
)
21
Bischoff-Ferrari
HA
Giovannucci
E
Willett
WC
Dietrich
T
Dawson-Hughes
B
Estimation of optimal serum concentrations of 25-hydroxyvitamin D for multiple health outcomes
Am J Clin Nutr
 , 
2006
, vol. 
84
 (pg. 
18
-
28
)
22
Zittermann
A
Iodice
S
Pilz
S
Grant
WB
Bagnardi
V
Gandini
S
Vitamin D deficiency and mortality risk in the general population: a meta-analysis of prospective cohort studies
Am J Clin Nutr
 , 
2012
, vol. 
95
 (pg. 
91
-
100
)
23
Pilz
S
Tomaschitz
A
März
W
Drechsler
C
Ritz
E
Zittermann
A
Cavalier
E
Pieber
TR
Lappe
JM
Grant
WB
Holick
MF
Dekker
JM
Vitamin D, cardiovascular disease and mortality
Clin Endocrinol
 , 
2011
, vol. 
75
 (pg. 
575
-
584
)
24
Sattar
N
Welsh
P
Panarelli
M
Forouhi
NG
Increasing requests for vitamin D measurement: costly, confusing, and without credibility
Lancet
 , 
2012
, vol. 
379
 (pg. 
95
-
96
)
25
Lee
P
Eisman
JA
Center
JR
Vitamin D deficiency in critically ill patients
N Engl J Med
 , 
2009
, vol. 
360
 (pg. 
1912
-
1914
)
26
Braun
AB
Gibbons
FK
Litonjua
AA
Giovannucci
E
Christopher
KB
Low serum 25-hydroxyvitamin D at critical care initiation is associated with increased mortality
Crit Care Med
 , 
2012
, vol. 
40
 (pg. 
63
-
72
)
27
Bischoff-Ferrari
HA
Dawson-Hughes
B
Staehelin
HB
Orav
JE
Stuck
AE
Theiler
R
Wong
JB
Egli
A
Kiel
DP
Henschkowski
J
Fall prevention with supplemental and active forms of vitamin D: a meta-analysis of randomised controlled trials
BMJ
 , 
2009
, vol. 
339
 pg. 
b3692
 
28
Bischoff-Ferrari
HA
Willett
WC
Orav
EJ
Lips
P
Meunier
PJ
Lyons
RA
Flicker
L
Wark
J
Jackson
RD
Cauley
JA
Meyer
HE
Pfeifer
M
Sanders
KM
Stähelin
HB
Theiler
R
Dawson-Hughes
B
A pooled analysis of vitamin D dose requirements for fracture prevention
N Engl J Med
 , 
2012
, vol. 
367
 (pg. 
40
-
49
)
29
Autier
P
Gandini
S
Vitamin D supplementation and total mortality: a meta-analysis of randomized controlled trials
Arch Intern Med
 , 
2007
, vol. 
167
 (pg. 
1730
-
1737
)
30
Bjelakovic
G
Gluud
LL
Nikolova
D
Whitfield
K
Wetterslev
J
Simonetti
RG
Bjelakovic
M
Gluud
C
Vitamin D supplementation for prevention of mortality in adults
Cochrane Database Syst Rev
 , 
2011
, vol. 
7
 pg. 
CD007470
 
31
Stockton
KA
Mengersen
K
Paratz
JD
Kandiah
D
Bennell
KL
Effect of vitamin D supplementation on muscle strength: a systematic review and meta-analysis
Osteoporos Int
 , 
2011
, vol. 
22
 (pg. 
859
-
871
)
32
Wang
L
Song
Y
Manson
JE
Pilz
S
März
W
Michaëlsson
K
Lundqvist
A
Jassal
SK
Barrett-Connor
E
Zhang
C
Eaton
CB
May
HT
Anderson
JL
Sesso
HD
Circulating 25-hydroxy-vitamin D and risk of cardiovascular disease: a meta-analysis of prospective studies
Circ Cardiovasc Qual Outcomes
 , 
2012
, vol. 
5
 (pg. 
819
-
829
)
33
Zittermann
A
Scheld
K
Stehle
P
Seasonal variations in vitamin D status and calcium absorption do not influence bone turnover in young women
Eur J Clin Nutr
 , 
1998
, vol. 
52
 (pg. 
501
-
506
)
34
Bolland
MJ
Avenell
A
Baron
JA
Grey
A
MacLennan
GS
Gamble
GD
Reid
IR
Effect of calcium supplements on risk of myocardial infarction and cardiovascular events: meta-analysis
BMJ
 , 
2010
, vol. 
341
 pg. 
c3691
 
35
Zittermann
A
Pilz
S
Börgermann
J
Gummert
JF
Calcium supplementation and vitamin D: a trigger for adverse cardiovascular events?
Future Cardiol
 , 
2011
, vol. 
7
 (pg. 
725
-
727
)
36
Ziegler
R
Minne
H
Raue
F
Paar
G
Delling
G
Observations in vitamin D and dihydrotachysterol poisoning
Dtsch Med Wochenschr
 , 
1975
, vol. 
100
 (pg. 
415
-
416
419–423. [Article in German]
37
Jones
G
Pharmacokinetics of vitamin D toxicity
Am J Clin Nutr
 , 
2008
, vol. 
88
 (pg. 
582S
-
586S
)
38
Zittermann
A
Cautious interpretation of the reverse J-shaped association between circulating 25-hydroxyvitamin D and total mortality is necessary
J Clin Endocrinol Metab
 , 
2012
 
[Letter to the Editor] Published online 14 June
39
Zittermann
A
Vitamin D in preventive medicine – are we ignoring the evidence?
Br J Nutr
 , 
2003
, vol. 
89
 (pg. 
552
-
572
)
40
Huebbe
P
Nebel
A
Siegert
S
Moehring
J
Boesch-Saadatmandi
C
Most
E
Pallauf
J
Egert
S
Müller
MJ
Schreiber
S
Nöthlings
U
Rimbach
G
APOE ɛ4 is associated with higher vitamin D levels in targeted replacement mice and humans
FASEB J
 , 
2011
, vol. 
25
 (pg. 
3262
-
3270
)

Comments

4 Comments
Alternate explanations for the observed U-shaped 25(OH)D-mortality rate relation
17 February 2013
William B. Grant

The paper by Zittermann and colleagues reported a U-shaped relation for serum 25-hydroxyvitamin D [25(OH)D]-probability of major cardiac and cardiovascular events (MACCEs) within a year of surgery for those who underwent cardiac surgery.1 The authors argued that this finding was a correct interpretation of the data and consistent with other studies. However, the higher mortality rate with highest serum 25(OH)D concentration was only statistically significant in two of the four studies cited, for all-cause mortality rate in Denmark (Ref. 16), and for cancer in the U.S. but not for cardiovascular disease (Ref. 20). However, Ref. 16 had no information on health conditions and cause of death for those included in the study. The "J-shape" observed for hazard ratio with respect to serum 25(OH)D concentration had a hazard ratio near 2.5 for 25(OH)D concentration <10 nmol/l, near zero for 50-90 nmol/l, and near 1.5 near 150 nmol/l. Only 5% of those in the study had serum 25(OH)D concentrations >100 nmol/l, and it is likely that they were taking extra vitamin D supplements due to an underlying vitamin D-deficiency disease.

This idea leads to an alternative explanations for the finding of higher MACCE with higher serum 25(OH)D concentrations found in Ref. 1, namely that those with higher serum 25(OH)D concentrations were advised by their physician to take extra vitamin D supplements due to having a vitamin D-deficiency disease. Inspection of the data for concomitant diagnoses in Table 1 indicates that those with serum 25(OH)D level >100 nmol/l had the highest rate of concomitant diagnoses for five of the eight categories. Thus, given the fact that the role of vitamin D in reducing risk of cardiovascular disease has been known from studies in Germany since at least 2008, it is very likely that these patients were given higher doses of supplementary vitamin D.

This hypothesis is supported by two observational studies from the United States. For elderly men, frailty status was found strictly inversely correlated with serum 25(OH)D concentration.2 However, for elderly women, there was a U-shaped relation, with higher frailty rates for both low and high serum 25(OH)D concentrations.3 The likely reason for the different findings is that elderly women in the United States are much more likely to be diagnosed with osteoporosis and advised by their physician to take extra vitamin D supplements than are men.

A possible way to test this hypothesis is to examine the medical records of the 139 patients in the highest serum 25(OH)D category to see whether they were advised to take high vitamin D doses.

Another possible explanation is that serum 25(OH)D concentrations changed with time after surgery, especially for those originally who had >100 nmol/l. It has been reported that the observed beneficial effect of higher serum 25(OH)D concentration on all-cause mortality rate decreases with increasing follow-up time.4 To test whether this effect played a role in the results presented in Ref. 1, it would be worthwhile to compare the MACCEs for 6 and 12 months after surgery.

References

1.Zittermann A, Kuhn J, Dreier J, Knabbe C, Gummert JF, B?rgermann J. Vitamin D status and the risk of major adverse cardiac and cerebrovascular events in cardiac surgery. Eur Heart J 2013 Jan 12. [Epub ahead of print]

2. Ensrud KE, Blackwell TL, Cauley JA, Cummings SR, Barrett-Connor E, Dam TT, Hoffman AR, Shikany JM, Lane NE, Stefanick ML, Orwoll ES, Cawthon PM; Osteoporotic Fractures in Men Study Group. Circulating 25-hydroxyvitamin D levels and frailty in older men: the osteoporotic fractures in men study. J Am Geriatr Soc 2011;59:101-106.

3. Ensrud KE, Ewing SK, Fredman L, Hochberg MC, Cauley JA, Hillier TA, Cummings SR, Yaffe K, Cawthon PM; Study of Osteoporotic Fractures Research Group. Circulating 25-hydroxyvitamin D levels and frailty status in older women. J Clin Endocrinol Metab 2010;95:5266-5273.

4. Grant WB. Effect of follow-up time on the relation between prediagnostic serum 25-hydroxyitamin D and all-cause mortality rate. Dermatoendocrinol 2012;4:198-202.

Conflict of Interest:

I receive funding from Bio-Tech Pharmacal (Fayetteville, AR), and the Sunlight Research Forum (Veldhoven) and have received funding from the UV Foundation (McLean, VA), the Vitamin D Council (San Luis Obispo, CA), and the Vitamin D Society (Canada).

Submitted on 17/02/2013 7:00 PM GMT
Future studies are needed
17 February 2013
Armin Zittermann (with Jan F. Gummert, Jochen Boergermann)

A U-shaped or inverse J-shaped association between circulating 25OHD and clinical outcome is well known, not only from cohort studies focusing on all-cause mortality and cancer but also from investigations on allergy risk and tuberculosis (1,2). However, the present study (3) is positively the first investigation demonstrating a significant U-shaped association between 25OHD and cardiovascular disease. In response to Grant's comment, we randomly checked 100 reports from the referring physicians as well as 100 discharge summaries for comments on supplementary vitamin D use. We re -examined letters and reports from the beginning, middle, and end of the inclusion period. None of the documents listed supplementary vitamin D use. Although a bias of supplementary vitamin D use on our study results cannot be ruled out with absolute certainty, it is noteworthy that a large population-based US study (4) reported a significant U-shaped association between circulating 25OHD and all-cause mortality, at least in women, even after adjustments were made for various covariates including supplementary vitamin D. Due to the suggested protective vitamin D effects on clinical outcome, high preoperative doses of supplementary vitamin D should decrease and not increase the risk of MACCE. Thus, other mechanisms may at least in part explain the aforementioned U-shaped association. Recently, one of us published a re-analysis of earlier data from our institution in an E-letter (5), demonstrating that high 25OHD levels (>125 nmol/l) may sometimes reflect low availability of the vitamin D hormone 1,25(OH)2D. Results are in line with experimental data demonstrating that high circulating 25OHD levels can result in low renal 1alpha-hydroxylase activity (6). High 25OHD levels may also stimulate 24-hydroxylase, which degrades 1,25(OH)2D and 25OHD to their 24-hydroxylated inactive forms (2). In the future, it is necessary to pursue several lines of evidence to clarify the underlying mechanisms of any potential harmful vitamin D effects at circulating 25OHD levels above 100 nmol/l to 125 nmol/l.

References

1. Hyppoenen E, Berry DJ, Wjst M et al. Serum 25-hydroxyvitamin D and IgE - a significant but nonlinear relationship. Allergy 2009; 64:613-620

2. Nielsen NO, Skifte T, Andersson M et al. Both high and low serum vitamin D concentrations are associated with tuberculosis: a case-control study in Greenland. Br J Nutr 2010; 104:1487-91

3. Zittermann A, Kuhn J, Dreier J et al. Vitamin D status and the risk of major adverse cardiac and cerebrovascular events in cardiac surgery. Eur Heart J 2013 Jan 12 [Epub ahead of print]

4. Melamed ML, Michos ED, Post W et al. 25-hydroxyvitamin D levels and the risk of mortality in the general population. Arch Intern Med 2008; 168:1629-1637

5. Zittermann A. Cautious interpretation of the reverse J-shaped association between circulating 25-hydroxyvitamin D and total mortality is necessary. J Clin Endocrinol Metab [E-Letter to the article by Durup et al. 2012;97:2644-2652], published June 14, 2012

6. Vieth R, Fraser D, Kooh SW. Low dietary calcium reduces 25- hydroxycholecalciferol in plasma of rats. J Nutr 1987; 117:914-918

Conflict of Interest:

A.Z. received speaker honoraria from DiaSorin, Germany and Abbott, Germany. None of the co-authors have a conflict of interest to declare.

Submitted on 17/02/2013 7:00 PM GMT
Re Zittermann et al, Vitamin D status and the risk of major adverse cardiac and cerebrovascular events in cardiac surgery.
25 February 2013
Barbara J Boucher

It was interesting to see long-standing suspicions that poor vitamin D status could increase health risks supported by the study of vitamin D status pre-operatively on cardiac surgery mortality (1). However, the U shaped curve, suggesting that risk worsens with high or low status, is worrying because similar J or U shaped curves are being reported in increasing numbers of studies. Most nutritional factors, even water, have optimal benefits when ‘not too little, not too much, but just right ‘amounts are taken (the Goldilocks effect) (2). Increases in risk appeared at =75nmol/l in this study, i.e., within the normal range for healthy free-living humans in sunny climes (80-160 nmol/l] (3); suggesting unidentified confounders affecting more replete subjects - less well patients might be supplemented late in their illness, implying that chronic deficiency affects health outcomes more than short-term improvement, or, that supplementation included factors with adverse effects. This is a possible as vitamin D is given with multivitamins, with calcium and in fish-liver oils. Current concerns include possible adverse cardiovascular effects of calcium supplementation (4) and high-dose vitamin A antagonizes vitamin D [PubMed reveals 105 papers on this interaction], through receptor level interactions. Reduced lung cancer mortality in never-smokers with higher vs. lower vitamin D status, in those with serum retinyl esters <but not >7.0 µg/dl is in line with mechanistic data (5,6). Vitamin A is an essential nutrient but many westernized populations have vitamin A intakes = recommended amounts (7). Thus, all future studies on vitamin D should adjust for these potential confounders. Meanwhile, can the authors tell us whether adjustment for supplementation, or for total dietary vitamin A intakes, affects the outcomes in their study?

References

1. Zittermann A, Kuhn J, Dreier J, Knabbe C, Gummert JF, B?rgermann J. Vitamin D status and the risk of major adverse cardiac and cerebrovascular events in cardiac surgery. Eur Heart J 2013; Jan 12. [Epub ahead of print]

2. http://en.wikipedia.org/wiki/Goldilocks_principleGoldilocks effect

3. Heaney RP. Health is better at serum 25(OH)D above 30ng/mL. J Steroid Biochem Mol Biol. 2012; Sept. [Epub ahead of print].

4. Boucher BJ, Calcium supplements may increase the risk of cardiovascular events in postmenopausal women. Evid Based Med. 2012;17:16-7.

5. Carlberg C. Lipid soluble vitamins in gene regulation. Biofactors 1999;10:91-7. Review.

6. Cheng TY, Neuhouser ML Serum 25-hydroxyvitamin D, vitamin A, and lung cancer mortality in the US population: a potential nutrient-nutrient interaction. Cancer Causes Control. 2012; 23:1557-65

7. Flynn A, Hirvonen T, Mensink GB, Ock? MC, Serra-Majem L, Stos K, Szponar L, Tetens I, Turrini A, Fletcher R, Wildemann T. Intake of selected nutrients from foods, from fortification and from supplements in various European countries. Food Nutr Res 2009;12;53.doi:10.3402/fnr.v53i0.2038. [Epub ahead of print] From Barbara J Boucher MD, FRCP. Hon Professor Centre for Diabetes, Bart's & The London School of Medicine & Dentistry. Queen Mary University of London. Blizard Institute, 2-4 Newark Street London E12AT, UK

Conflict of Interest:

None declared

Submitted on 25/02/2013 7:00 PM GMT
Reply from the authors
27 February 2013
Armin Zittermann (with Jan F. Gummert, Jochen Boergermann)

We have read the thoughts of Dr. Boucher with great interest and agree with her concerns regarding the accumulating evidence of U-shaped associations between circulating 25OHD and clinical outcomes. Apart from the possibility that supplemental intake of vitamin D and/or other nutrients such as vitamin A may have biased our results (see also our response to a previous letter [1]) other explanations should be taken into account as well: In some cases, high 25OHD levels may only be an indicator for reduced levels of the vitamin D hormone 1,25-dihydroxyvitamin D (1,25[OH]2D), as demonstrated in an earlier e-letter by one of us [2]. High dietary calcium intake may be one candidate for this phenomenon, since calcium does not only suppress 1,25(OH)2D levels [3] but also increases the half-life of 25OHD [4]. Another candidate may be severe renal impairment, which is associated with suppressed 1,25(OH)2D levels and may in some cases lead to enhanced 25OHD levels, due to reduced substrate need of the renal enzyme 1-alpha-hydroylase. This assumption is supported by the fact that experimental data demonstrate markedly decreased circulating 25OHD levels in case of high renal 1-alpha- hydroxylase activ?ity and vice versa [5]. However, it is also possible that at levels above 100 nmol/l 25OHD itself does already result in clinically relevant adverse effects in some patients. It is probably incorrect if we use 25OHD levels which are observed in physically active, traditionally living populations as reference values for ageing populations with sedentary lifestyle and high cardiovascular risk profiles. As outlined by us earlier [6], calcium and vitamin D metabolism differs between healthy physically active people and individuals with sedentary lifestyle. Future studies have to clarify whether one or more of the aforementioned mechanisms can explain our findings in cardiac surgical patients.

References

1. Zittermann A, Gummert JF, Boergermann J. Future studies are needed. 2013 Eur Heart J, E-letter, published February 18, 2013

2. Zittermann A. Cautious interpretation of the reverse J-shaped association between circulating 25-hydroxyvitamin D and total mortality is necessary. J Clin Endocrinol Metab [E-Letter to the article by Durup et al. 2012;97:2644-2652], published June 14, 2012

3. Vieth R, Fraser D, Kooh SW. Low dietary calcium reduces 25- hydroxycholecalciferol in plasma of rats. J Nutr. 1987;117:914-918

4. Giovannucci E. Dietary influences of 1,25(OH)2 vitamin D in relation to prostate cancer: a hypothesis. Cancer Causes Control. 1998;9:567-582

5. Lips P. Interaction between vitamin D and calcium. Scand J Clin Lab Invest Suppl. 2012;243:60-64

6. Zittermann A, Pilz S, Boergermann J, Gummert JF. Calcium supplementation and vitamin D: A Trigger for Adverse Cardiovascular Events? Future Cardiol 2011;7:725-727

Conflict of Interest:

A.Z. received speaker honoraria from DiaSorin, Germany and Abbott, Germany. None of the co-authors have a conflict of interest to declare.

Submitted on 27/02/2013 7:00 PM GMT