Changes in the Calcium-Parathyroid Hormone-Vitamin D Axis and Prognosis for Critically Ill Patients: A Prospective Observational Study
PLoS ONE 8(9): e75441. doi:10.1371/journal.pone.0075441
Jieyu Hu equal contributor, Zuojie Luo equal contributor mail, Xiaoqin Zhao, Qiang Chen, Zhaoyan Chen, Hua Qin, Yingfen Qin, Xinghuan Liang, Yingjun Suo
Objective: Vitamin D deficiency is prevalent in critically ill patients and may contribute to suboptimal clinical outcomes, but little is known about alterations of the calcium-parathyroid hormone (PTH)-vitamin D axis and prognosis in these individuals.
Methods: A prospective observational study was conducted on 216 patients admitted to a university-affiliated, tertiary-care medical intensive care unit(MICU) between June 2011 and December 2012. Serum levels of 25-hydroxyvitamin D, ionised calcium and intact PTH were determined within 24 h of MICU admission. The primary end point was all-cause hospital mortality within 90-days of admission.
Results: 95 patients (44%) showed 25-hydroxyvitamin D deficiency. Patients deficient in vitamin D showed significantly higher Acute Physiology and Chronic Health Evaluation II (APACHE II) score, rate of positive blood culture, incidence of multiple organ dysfunction syndrome, and 90-day mortality rate than did patients with vitamin D insufficiency or sufficiency (P<0.05), as well as lower levels of serum IgG. 25-Hydroxyvitamin D deficiency was identified as an independent risk factor for mortality (OR = 3.018, 95%CI 1.329–6.854, P = 0.008). Hypovitaminosis D in PTH-responders was associated with higher mortality than was the same condition in non-responders (P<0.05).
Conclusions: These results suggest that vitamin D deficiency is prevalent among MICU patients, suggesting a significant derangement of the calcium-PTH-vitamin D axis in critically ill patients. Vitamin D deficiency is an independent risk factor for 90-day mortality, and hypovitaminosis D in PTH-responders is associated with higher mortality than is the same condition in non-responders.
PDF is attached at the bottom of this page
See also VitaminDWiki
- Chance of dying within 1 month of entering hospital is 45 percent higher if low vitamin D – July 2013
- 20 percent of Critically ill in India had undetectable vitamin D, and 62 percent died – Jan 2013
- Hypomagnesemia (low Magnesium) is very prevalent in ICU – 1999 Magnesium mentioned in the study on this page
- Will lots of vitamin D help in ICU – RCT protocol Nov 2012
- Plasma exchange cut vitamin D levels in half – June 2012
- ICU Acinetobacter baumannii infection (Iraqibacter) is 2.5X more likely if low vitamin D – Oct 2013
- 540000 IU before ICU raised vitamin D by 25 ng in 2 days – March 2011
- Being vitamin D deficient in the ICU is not good - Feb 2012
- ICU survival increased with vitamin D single 540K IU loading dose - JAMA Sept 2014 - has the following chart
Discussion
Our results demonstrate a high prevalence of hypovitaminosis D in the critically ill patient population, with vitamin D deficiency observed in 44% of the patients in this study. Vitamin D deficiency was associated with higher disease severity and hospital mortality. It was also associated with higher incidence of positive blood culture and of MODS, as well as lower Ig G levels, all of which may worsen the prognosis of ICU patients.
Although our patients deficient in vitamin D fared worse than did our patients with insufficient or sufficient vitamin D for nearly all outcomes examined, the deficient group nevertheless did have a shorter ICU stay and fewer days on the ventilator than did the sufficient group. This is in contrast to recent work by Matthews 10 showing that the deficient group stayed longer in the ICU. This discrepancy may be due to the different type of study population: our study examined all-cause MICU patients, whereas Matthews et al. examined specifically surgical ICU patients. The discrepancy may also be due to high mortality among our patients deficient in vitamin D soon after MICU admission, resulting in overall shorter ICU stay and fewer days on the ventilator.
To our knowledge, ours is the first study to report a positive correlation between serum 25(OH)D concentration and Ig G levels in an ICU population. This relationship contrasts with the inverse relationship observed in chronic diseases that often involve vitamin D deficiency, such as systemic lupus erythematosus (SLE) 15 and cystic fibrosis 16. In fact, cellular studies have shown that 1,25(OH)2D, the active form of vitamin D, decreases B cell proliferation, plasma cell differentiation and Ig G secretion 17. Our observational study does not allow us to discern whether serum 25(OH)D levels are causally linked to Ig G levels. Further studies are needed to examine the correlation between these two factors in both critical illness and chronic disease.
Although serum levels of 25(OH)D are widely used as an indicator of vitamin D status, consensus is lacking about the cut-off values for defining deficiency and sufficiency. Serum 25(OH)D levels are inversely associated with PTH levels until the former reach 30–40 ng/ml, at which point PTH levels begin to level off 18, 19 and intestinal calcium absorption is maximal 20. Therefore most experts suggest defining vitamin D sufficiency as serum 25(OH)D ≥30 ng/ml and deficiency as <20 ng/ml 1. Our study adopted these thresholds to define vitamin D status in critically ill patients.
Hypovitaminosis D in critically ill patients is multifactorial and may arise from limited sunlight exposure, poor nutritional status, aging, obesity, liver failure, chronic kidney disease, interaction with medications, abnormal gastrointestinal function and effects of fluid resuscitation 21. Lee et al. 4 speculate that tissues require greater amounts of vitamin D during critical illness, leading to increased conversion of 25(OH)D into the active form 1,25(OH)2D, thereby lowering serum levels of 25(OH)D.
We observed a direct association between 25(OH)D deficiency and hospital mortality within 90 days of MICU admission. This association may have multiple causes, given the pleiotropic actions of vitamin D in
- immunity,
- endothelial/mucosal function,
- glucose metabolism and
- calcium homeostasis.
In fact, vitamin D deficiency may help explain many of the varied morbidities frequently observed among critically ill patients, including
- systemic inflammatory response syndrome,
- sepsis,
- organ failure and
- metabolic dysfunction 4.
In particular, the immunomodulatory actions of vitamin D may explain its observed effects on the prognosis of critically ill patients. In our study, rates of positive blood culture and of MODS were higher among patients with vitamin D deficiency than among those with vitamin D sufficiency; deficient patients also tended to have higher CRP levels. These findings are consistent with recent evidence that vitamin D enhances the innate immune response by inducing production of cathelicidin (LL-37), an endogenous antimicrobial peptide produced by macrophages and neutrophils 22. This peptide can fight against a broad spectrum of infectious agents, including Gram-negative and -positive bacteria, fungi and mycobacteria 23. Vitamin D has also been found to down-regulate production of proinflammatory cytokines such as interleukin (IL)-1, IL-2, IL-6, IL-8, IL-12, interferon-γ and tumor necrosis factor-α, as well as T helper 1 cells and B lymphocytes in the adaptive immune system 24, 25. At the same time, vitamin D up-regulates production of anti-inflammatory cytokines IL-4, IL-5 and IL-10, shifting the overall immune phenotype to a T helper 2 subtype; and it promotes expression of T-regulatory cells, which turn off the adaptive immune response 26. These cellular and molecular studies suggest that vitamin D deficiency dysregulates the innate immunity system and compromises the ability of critically ill patients to down-regulate the adaptive immune response. These effects may explain the association between vitamin D deficiency and increased mortality observed in our cohort.
Vitamin D is well known as a key participant in the calcium-PTH axis 1, which is responsible for maintaining calcium homeostasis, yet how the axis changes during critical illness is poorly understood. We found hypocalcaemia in 89.4% of our patients, at the upper end of the prevalence range of 15–88% reported for adult patients in the ICU 27, 28. While hypocalcaemia can be caused by several morbidities frequently found in the ICU, such as sepsis, burns, pancreatitis and rhabdomyolysis, it can also result from vitamin D insufficiency and deficiency. Normally the body avoids hypocalcaemia by boosting secretion of PTH, which increases renal calcium re-absorption and calcium release from the skeleton through bone resorption. The hormone also indirectly increases 1,25(OH)D levels, thereby increasing intestinal calcium absorption. In our cohort, serum 25(OH)D levels did not correlate with levels of ionised or total calcium, suggesting that the hypocalcaemia in our patients had multiple, complex causes. Hypovitaminosis D accounted for only a fraction of hypocalcaemia cases, suggesting that vitamin D supplementation by itself would not correct the problem.
Magnesium is also a major participant in the calcium-PTH axis, and in critical patients, nutritional deficiency and organ dysfunction can lead to magnesium deficiency. This may impair PTH response or result in target organ resistance to PTH, leading in turn to hypocalcaemia. Thus, magnesium deficiency may explain the hypocalcaemia in many of our patients, but we cannot be sure because magnesium levels were not recorded.
Among our patients with secondary hyperparathyroidism, levels of ionised calcium were 25% lower than the normal range observed in patients with sufficient vitamin D. Such reduced levels may reflect severe deficiency of circulating 25(OH)D, age-related impairment of renal 1α- hydroxylation, and compromised calcium absorption.
Approximately 40% of our patients with hypovitaminosis D showed a reduced response to PTH. This proportion is similar to the 60% reported by Nair et al. 29 The potential causes and mechanisms of impaired PTH response to hypovitaminosis D remain unclear. They may include abnormalities in the parathyroid calcium sensing receptor, age-related impairment of renal 1α- hydroxylation, abnormal function of the 1,25(OH)2D receptor, abnormalities in the FGF23-Klotho axis and other genetic abnormalities 14. Magnesium deficiency can also decrease the activity of magnesium-dependent enzymes, inhibiting PTH synthesis and regulation. Whatever the cause, impaired PTH response ironically appears to be associated with better prognosis for critically ill patients. In our study, hypovitaminosis D in PTH-responders was associated with higher APACHE II scores than was the same condition in non-responders. Similarly Nair et al. 29 found PTH-responders to have a higher Simplified Acute Physiology Score (SAPS) II at ICU admission, leading them to speculate that the lack of PTH response may in fact indicate vitamin D sufficiency in tissues. A study of PTH response in elderly with hypovitaminosis D 30 showed that the simultaneous presence of secondary hyperparathyroidism was associated with increased bone turnover and fracture risk, as well as shorter survival. We further found that hypovitaminosis Din PTH-responders was associated with higher 90-day mortality than was the same condition in non-responders.
We did not find a correlation between mortality and PTH in our study, whereas several studies have suggested a direct association between the two variables 31, 32, 33 that was independent of vitamin D status and renal function 31. Elevated PTH may increase mortality through its effects on cardiac muscle contractility and its ability to promote atherosclerosis and vascular calcification 34. Elevated PTH may also suppress the immune system, in particular by compromising leukocyte function, which should make patients more susceptible to infection 35. We speculate that the combined negative effects of hypovitaminosis D and elevated PTH are the most likely cause of increased mortality observed in our PTH-responders with hypovitaminosis D. In fact, the frequent co-occurrence of hypovitaminosis D and impaired PTH response among critically ill patients may reflect the body’s efforts to protect itself from the adverse effects of the altered calcium-PTH-vitamin D axis, and increased expression of calcium sensing receptor in parathyroid glands would be compatible with a more efficient control of PTH synthesis and secretion by low serum ionised calcium.
Our work has several potential limitations. First, the sample size was small, and we did not monitor 25(OH)D, iPTH, and ionised calcium levels over time. Second, our study was conducted in an MICU and cannot be generalised to cardiac, surgical, or other types of ICUs. Third, we did not collect data on levels of 1,25(OH)2D, vitamin D binding protein (DBP), or magnesium, so we cannot exclude them as possible confounders. Finally, although our data are consistent with an association between vitamin D deficiency and severity of illness and hospital mortality, we cannot conclude a causative link. Randomised controlled trials to evaluate whether vitamin D supplementation in critically ill patients can improve their clinical outcomes are warranted.
Despite its limitations, our study suggests the need for new lines of research. This study is, to our knowledge, the first to report a correlation between hypovitaminosis D and lower Ig G levels in an ICU population, which should be explored in greater detail in future work. We also found that hypovitaminosis D in PTH-responders is associated with higher APACHE II scores and mortality than was the same condition in non-responders. This raises the intriguing possibility that reduced PTH response in the presence of hypovitaminosis D is a protective mechanism, which should be explored further.
In conclusion, vitamin D deficiency and hypocalcaemia are highly prevalent in the critically ill population, and hypovitaminosis D accounts for only a small number of hypocalcaemia cases. Critically ill patients often show significant dysregulation of the calcium-PTH-vitamin D axis, and vitamin D deficiency is an independent risk factor for prognosis in serious illness. Hypovitaminosis D in PTH-responders is associated with higher APACHE II score and mortality than is the same condition in non-responders. Our findings highlight the complex interactions between PTH and hypovitaminosis D, which merit further study. Our results also suggest that 25(OH)D and iPTH levels should be measured as part of the routine tests performed in the MICU. Future research should examine whether correction of 25(OH)D deficiency improves outcomes for ICU patients.
Tables of data not in the PDF are attached at the bottom of this page. The most interesting table follows
Table S2 Characteristics of patients stratified by vitamin D status
Variable a | Sufficient (N=63) | Insufficient (N=58) | Deficient (N=95) | P Value |
Age (yr) | 64 (33, 75) | 58 (47, 76) | 67 (54, 74) | 0.297 |
Male gender, N (%) | 36 (57.1) | 33 (56.9) | 51 (53.7) | 0.886 |
25(OH)D levels (ng/ml) | 35.1 (34.1, 40.6) | 22.9 (20.6, 27.2) | 14.3 (13.6, 17.1) | <0.001 |
APACHE II score | 19 (18, 20) | 22 (17, 26) | 25 (20, 28) | <0.001 |
Length of ICU stay (d) | 15.2 (10.3, 23.6) | 8.6 (6.9, 20.5) | 10.5 (6.8, 25.3) | 0.113 |
Time on ventilator (d) | 8 (3.4, 20) | 4.5 (2.8, 11) | 5.3 (3.3, 10.5) | 0.195 |
iPTH (pg/ml) | 42.1 (27.2, 81.2) | 95.8 (45.9, 121.3) | 95.6 (58.6, 151.2) | <0.001 |
Albumin-adjusted total calcium (mmol/L) | 1.98 (1.82, 2.11) | 1.99 (1.79, 2.09) | 1.91 (1.78, 2.04) | 0.324 |
Ionised calcium (mmol/L) | 0.91 (0.84, 0.99) | 0.88 (0.76, 0.96) | 0.82 (0.72, 0.91) | <0.001 |
Serum phosphate (mmol/L) | 1.01 (0.58, 1.27) | 0.98 (0.51, 1.13) | 0.83 (0.67, 1.07) | 0.75 |
WBC (×109) | 10.7 (8.6, 14.3) | 11.2 (7, 14.8) | 12.1 (8.0, 21.7) | 0.206 |
Hb (g/L) | 102 (78, 134) | 97 (91.3, 128.2) | 95 (81.3, 117) | 0.068 |
Serum albumin (g/L) | 31.2 (27.3, 35) | 29.6 (27.9, 33.4) | 28.3 (25.8, 31.8) | 0.132 |
Serum creatinine (μmol/L) | 74 (37, 101) | 61 (55, 107) | 68 (47, 101) | 0.829 |
Ig G (g/L) | 13.43 (9.71, 16.1) | 9.51 (7.32, 12.66) | 9.29 (7.03, 12.21) | <0.001 |
Lactate (mmol/L) | 2 (1.1, 3.3) | 2 (1.1, 3.2) | 2.3 (1.6, 3.7) | 0.061 |
CRP (mmol/L) | 69 (37, 124) | 124 (42, 186) | 99 (44, 200) | 0.207 |
Positive blood culture rate, N (%) | 2 (3.2) | 8 (13.8) | 18 (18.9) | 0.015 |
MODS, N (%) | 16 (25.4) | 25 (43.1) | 46 (48.4) | 0.013 |
90-day hospital mortality, N (%) | 10 (15.9) | 14 (24.1) | 38 (40) | 0.003 |
a Values are reported as median (interquartile range IQR 1, IQR3), unless noted otherwise.