Vitamin D deficiency in the critically ill

Vitamin D plays an essential role in calcium homeostasis and immune modulation and is required for the normal function of the musculoskeletal, cardiovascular and central nervous system (1). However, vitamin D deficiency is common in the general population as well as among critically ill (1,2).

Small amounts of vitamin D are derived from food. Another important source of vitamin D is 7-dehydrocholesterol in the skin that is converted to previtamin D3 by photosynthesis via ultraviolet B radiation (“sun exposure”). Previtamin D3 is converted to vitamin D3 and then hydroxylated in the liver to 25-hydroxyvitamin D (25(OH)D) and ultimately in the kidney to the biologically active 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) (1–3). An important function of 1,25(OH)2D3 is stimulation of calcium resorption from the gut (4). 25(OH)D is the accepted marker to evaluate vitamin D status (1,5). Various definitions of vitamin D deficiency ( < 25 to < 75 nmol/L) and different 25(OH)D assays are used in the literature (1,6,7). Therefore, it is hard to compare different studies and estimate the number of patients with true vitamin D deficiency. To correct for this variation, the international vitamin D standardization program was designed (8). Their protocol states that 25(OH)D concentrations below 30 nmol/L should be considered deficient (9). In the general population, vitamin D deficiency is especially common among children, young adults and elderly (1). It is estimated that one billion people worldwide have vitamin D deficiency or insufficiency (1). A recent study of Cashman et al. among 55,844 European individuals showed that 13.0% had serum 25(OH)D concentrations below 30 nmol/L and 40.4% below 50 nmol/L (10). Deficiency is most frequently due to poor vitamin D intake and decreased production of vitamin D as a consequence of insufficient sun exposure (1). Studies show that all-cause mortality appears to decrease as 25(OH)D levels increase, with optimal levels >75 nmol/L (11,12). In hospitalized patients, serum 25(OH)D levels below 50 nmol/L appear to be associated with adverse outcomes (13,14).

About 50–60% of patients admitted to the intensive care unit have vitamin D deficiency (2,4,15,16). This deficiency is presumably caused by decreased intake and absorption, increased losses and decreased production of vitamin D (2,4). In addition, end-organ resistance to vitamin D has been described (4). During critical illness there may also be an increase in tissue vitamin D requirement, resulting in an enhanced conversion of 25(OH)D to the active 1,25(OH)2D (17). This may result in even lower “vitamin D levels” as serum 25(OH)D is the accepted marker for vitamin D status. Several studies report low vitamin D levels to be a biomarker associated with sepsis severity, organ dysfunction and even mortality in critically ill patients (18–24).

Ala-Kokko et al. investigated the association between vitamin D levels at ICU admission and 90-day mortality in patients with severe sepsis or septic shock and reported their results in this journal. Deficiency was defined as 25(OH)D levels below 50 nmol/L. They reported no significant differences in 90-day mortality in patients with or without vitamin D deficiency (25). However, it is unclear whether results consider true vitamin D deficiency or whether low serum status of vitamin D resulted from an inflammatory response causing a shift from serum vitamin D into tissues. This would also explain why low serum vitamin D was significantly associated with elevated CRP levels and hospital-acquired infections at ICU admission, but not with 90-day mortality. Patients with lower CRP on admission and sufficient vitamin D status may be in earlier stages of the disease and the inflammatory response. Vitamin D levels may drop as sepsis progresses. This hypothesis was, however, not tested as vitamin D levels were only assessed on ICU admission.

Recently, Amrein et al. conducted a randomized controlled trial among critically ill patients comparing high dose vitamin D supplementation with placebo. They found a survival benefit in critically ill patients with severe vitamin D deficiency (<30 nmol/L) on ICU admission receiving vitamin D supplementation. Vitamin D was supplemented in high dosages (540,000 IU of oral cholecalciferol). These dosages are considered pharmacological rather than nutritional. In hospital mortality, 28-day mortality and 6-month mortality were significantly lower in the supplemented group, however, only in the predefined severe deficiency group. As in the total group no outcome effects were demonstrated that this may potentially imply that some patients also may have had some negative effects of the supplementation, although this cannot directly be concluded from the published data. For the total group, supplementation does not seem to be warranted, possibly only for those patients with severe deficiency (26).

Another recent study performed by Lee et al. investigated the relation between enhanced bone resorption and mortality rate in critically ill patients (27). Vitamin D deficiency may lead to a decrease in calcium absorption. Compensatory mechanisms are activated to maintain calcium homeostasis; the secretion of parathyroid hormone increases, leading to increased renal calcium reabsorption and bone resorption (4). Furthermore, accelerated bone loss has been shown in critically ill patients. This is presumably caused by systemic inflammation, generalized tissue catabolism and immobilization (28–30). Increased bone resorption is associated with increased mortality in the general population (31,32). However, a relationship between enhanced bone resorption and mortality in critically ill patients had not been investigated. Lee et al. conducted a retrospective study to investigate the relationship between preadmission bisphosphonate use and clinical outcome of critically ill patients. Bisphosphonates inhibit bone resorption. A significant in-hospital mortality reduction was shown in patients receiving bisphosphonates before ICU admission [mortality rate ratio (MRR) 0.32, 95% CI 0.10–0.97, p = 0.048]. Patients receiving vitamin D or vitamin D and bisphosphonates before ICU admission also had a significant survival benefit (MMR 0.49, 95% CI 0.38–0.62, p < 0.01; and MRR 0.38, 95% CI 0.20–0.71, p < 0.01) (27).

It remains difficult to draw solid conclusions regarding the predictive value of vitamin D status for mortality and the effect of supplementation in critically ill patients. Ala-Kokko and colleagues performed the largest prospective trial of vitamin D status in septic patients to date (25). Their results are, however, contradictory with previously performed trials and confirmation of their findings by further research is necessary to draw firm conclusions.

Amrein and coworkers showed a survival benefit of supplementation in severely deficient patients (26). However, in their trial critically ill patients with heterogeneous admission diagnoses were studied and it is not possible to extract from the data available whether these benefits may be extrapolated and are relevant for patients with severe sepsis or septic shock.

Furthermore, the relation between deficiency, mortality rates and the value of supplementation may be more complex than one might initially consider. Vitamin D supplementation may be beneficial even if vitamin D status is not related to mortality risk. Nutritional supplementation of vitamin D is unlikely to affect mortality rates. However, high dose vitamin D supplementation may cause beneficial pharmacological effects as it is reported that high levels of vitamin D may be necessary for optimal immune function (33).

However, when low plasma levels are only indicative of the severity of disease or when only redistribution plays an important role, it is unlikely that supplementation will be helpful, as the actual availability for tissues may not be low at all, and only risks associated with hypervitaminosis will arise. Hypercalcemia or vitamin D toxicity results when serum vitamin D concentrations are greater than 400–500 nmol/L (34). In addition, other studies show that all-cause mortality increases with vitamin D concentrations greater than 120 nmol/L (27,35). Moreover, low levels may also be due to an adaptive response that could be potentially beneficial during critical illness.

As long as we do not know how exactly vitamin D supplementation may interfere with the pathophysiological processes involved in severe sepsis and septic shock and no clear association with baseline deficiency and mortality can be shown, vitamin D supplementation for these patients should be considered experimental. Without the results of specific randomized trials addressing vitamin D supplementation in sepsis and septic shock we probably only should aim to provide the recommended daily allowances for healthy persons in critically ill patients and consider pharmacological dosages experimental with an unknown safety profile.

Disclosure statement

Dr. Koekkoek declares not to have any conflicts of interest. Dr. van Zanten reported that he has received honoraria for advisory board meetings, lectures, and travel expenses from Abbott, Baxter, BBraun, Danone-Nutricia, Fresenius Kabi, and Nestlé-Novartis.