Vitamin D (25-hydroxyvitamin D) deficiency (<25 nmol/L or <10 ng/mL) or insufficiency (25–75 nmol/L or 10–30 ng/mL)1 affects >1 billion people worldwide, spanning age groups, ethnicities, and gender.2 In the United States, more than one-third of adults,3,4 more than half of all hospitalized patients,5 and >95% of critically ill patients have a vitamin D deficiency.6
The multifaceted implications of vitamin D deficiency have long been recognized. For example, in a discussion on rickets at the Annual Meeting of the British Medical Association in 1888, WB Cheadle stated, “…the first point upon which I would venture to insist is that rickets is not to be regarded as a mere affection of the bones …. but it is something far more than this; the disease affects not only bones but muscles and ligaments, mucous membrane and skin, the blood and the nervous system.”7
There is now better understanding of the ways vitamin D contributes to vital processes aside from its “traditional” role in skeletal maintenance. Vitamin D receptors are found in most of the immune system cells, including macrophages, B and T lymphocytes, and neutrophils.8 Vitamin D also regulates antimicrobial peptide gene expression, thus improving innate immunity.6,9,10 As might therefore be expected, vitamin D deficiency is associated with infections, particularly respiratory infections.9,11,12
Low vitamin D concentrations are also associated with inflammation, especially in patients with cardiac disease.13 Inflammation increases the risk of cardiovascular disease and accompanies about 80% of all sudden cardiac deaths.14,15 It is thus not surprising that multiple investigations support a link between vitamin D deficiency and cardiovascular risk.16–18
The summative effects of vitamin D were evaluated in a prospective cohort from the Third National Health and Nutrition Examination Survey (NHANES III), a program designed to assess the health and nutritional status in the United States by combining interviews and physical examinations. NHANES III demonstrated an association between vitamin D deficiency and all-cause mortality.4,19
Postsurgical patients are particularly susceptible to cardiovascular and infectious complications, both of which appear to be aggravated by vitamin D deficiency, which itself is common in surgical patients. Compelling evidence thus suggests that patients with vitamin D deficiency may have worse outcomes than those with optimal concentrations, but this has not previously been studied in surgical patients. We therefore tested the primary hypothesis that noncardiac surgical patients with lower perioperative serum vitamin D concentrations are more likely to experience a composite outcome of all-cause in-hospital mortality, in-hospital cardiovascular morbidity, and in-hospital infectious morbidity.
With IRB approval, written informed consent was waived for this retrospective cohort analysis of 3509 adult patients. Analysis was restricted to patients who had noncardiac surgery at the Cleveland Clinic Main Campus between 2005 and 2011 and had at least one 25-hydroxyvitamin D measurement in the period starting 3 months before the procedure date to 1 month after, at our institution’s laboratory. We excluded pediatric patients and patients with ASA physical status >4.
Vitamin D concentrations were obtained from the Laboratory Medicine registry. Other data were obtained from the Cleveland Clinic Perioperative Health Documentation System, a registry in which data are prospectively collected.
The aim of the study was to evaluate the relationship between serum vitamin D concentration and all-cause in-hospital mortality, in-hospital cardiovascular morbidity, and serious in-hospital infections (definitions in Appendix 1). We did not analyze the 3 outcomes as a collapsed composite (“any-versus-none”) or analyze them separately. Rather, we assessed the “common effect” or “global” odds ratio (OR) of serum vitamin D concentration across the above 3 in-hospital outcomes, by using a multivariate (i.e., multiple outcomes) generalized estimating equation (GEE) model with unstructured covariance matrix to simultaneously capture the complete information on each outcome for a patient and to adjust for the within-patient correlation among the outcomes.20 The model was fit with adjustment for potential confounding variables listed in the Table 1 including the demographic, medical history variables, and type and duration of surgery. We assumed linear relationship between logit of the probability of having outcome and the continuous independent variables. To visualize the association of the composite outcome and the vitamin D concentration, we plotted the estimated probability (on the logit scale) of having in-hospital mortality/morbidity as a function of vitamin D concentration by using a multivariate GEE model incorporating a smooth term for vitamin D concentration. In addition, we also plotted the common effect OR (99% confidence interval [CI]) of each quintile of vitamin D concentration compared with the lowest quintile.
We assessed the heterogeneity of the vitamin D effect across the components of the in-hospital outcomes by testing the vitamin D-by-outcome interaction in a separate “distinct-effect” GEE model, which enables adjustment for the correlation among the component outcomes.20 The heterogeneity test compares the OR (log scale) for vitamin D concentration among the individual outcomes of interest. Significant heterogeneity, especially in opposite directions, would suggest that the individual ORs be given more importance than the global OR.21,22 Although the associations proved to be homogeneous among the individual outcomes (vitamin D-by-outcome interaction, P = 0.73), we nonetheless report associations between perioperative vitamin D concentration and specific in-hospital outcomes by using a GEE model, adjusting for the same potential confounders. A Bonferroni correction was used to adjust for testing multiple outcomes; thus, 98.3% CI were reported, and the significance criterion for the 3 primary outcomes was P < 0.017 (i.e., 0.05/3).
Hypotensive episodes may be due to vasopressor resistance in patients with lower vitamin D concentrations. We descriptively displayed the relationship between vitamin D concentration and intraoperative hemodynamic characteristics, including heart rate, arterial blood pressure, transfusions, and use of vasopressors (dobutamine, dopamine, epinephrine, norepinephrine, phenylephrine, and vasopressin) by reporting the summary statistics for each quintile of vitamin D concentration. Intraoperative hemodynamic data were acquired from our electronic anesthesia record-keeping system, which continuously records minute-by-minute data from physiologic monitors throughout anesthesia. Arterial blood pressure in patients with arterial catheters was recorded each minute; otherwise, pressure was measured oscillometrically and recorded at 1- to 5-minute intervals.
Sample Size Considerations
Available power for the study was assessed post hoc by comparing low versus normal concentrations of vitamin D (<25 ng/mL and ≥25 ng/mL) on the set of in-hospital outcomes by using a SAS macro developed for designs with multiple binary correlated end points (“multibinpow”) by using 1000 simulations.a With 3500 patients (1750/group), we had >90% power at the 0.05 significance level to detect a common effect OR of 0.7 or stronger for patients with vitamin D ≥25 ng/mL vs <25 ng/mL, by using the observed incidences of 1.65% for mortality, 6.69% for cardiovascular morbidity, and 8.82% for infectious morbidity (patients with vitamin D <25 ng/mL), and assuming a compound symmetric correlation structure with a between-outcome correlation of 0.05. Our study is more highly powered because vitamin D levels were analyzed as a continuous exposure variable.
SAS version 9.3 (SAS Institute, Cary) and R version 2.12.0 (The R Foundation for Statistical Computing, Vienna, Austria) were used for all statistical analyses.
Vitamin D levels were measured in 3509 patients; the median 25-hydroxyvitamin D concentration was 23.5 [1st–3rd quartiles: 14.6–33.6] ng/mL. The median [1st–3rd quartiles] difference between the vitamin D observation date and the date of surgery was 13 [−4, 46] days (i.e., 13 days before surgery). Vitamin D measurements of 1022 (29%) patients were obtained postoperatively; 924 and 727 of them were obtained >2 days and >1 week after surgery, respectively. Table 1 shows the summary statistics of baseline characteristics by the quintiles of serum vitamin D concentration. The Charlson comorbidity index is a summative score based on the concurrent clinical conditions in an individual patient.23
Higher vitamin D concentrations were associated with decreased odds of in-hospital mortality/morbidity (P = 0.003). The corresponding common effect OR of vitamin D level across individual in-hospital mortality/morbidity was estimated as 0.93 (95% CI, 0.88–0.97) for a 5-unit increase in vitamin D level (Table 2). The estimated probability of in-hospital mortality/morbidity decreases as vitamin D concentration increases in a rather linear fashion when all data points are plotted, though 97.4% of the values were below 60 ng/mL (Fig. 1).
The associations were homogeneous among the 3 individual in-hospital outcomes (vitamin D-by-outcome interaction, P = 0.73). Among the individual outcomes, vitamin D level was significantly associated only with decreased odds of having cardiovascular morbidity (OR: 0.92 [98.3% CI, 0.84–1.00] for a 5-unit increase in vitamin D, P = 0.01, Table 2). The raw incidence by quintiles of the vitamin D concentration is provided (Appendix 2).
Although patients who had their vitamin D concentration measured after surgery were different than other patients (Appendix 3), the association between vitamin D concentration and postoperative outcomes was not different (interaction P value = 0.92). Among 1022 patients who had their vitamin D concentration measured postoperatively, the common effect OR of vitamin D concentration across individual in-hospital mortality/morbidity was estimated as 0.92 (95% CI, 0.82–1.03) for a 5-unit increase in vitamin D concentration.
In addition, we found that the odds versus patients with vitamin D <13 ng/mL (i.e., 1st quintile) were significantly lower in patients with vitamin D 13–20, 20–27, 27–36, and >36 ng/mL (i.e., 2nd–5th quintiles); the corresponding estimated odds ratios were 0.65 (99% CI, 0.43–0.98), 0.53 (0.35–0.80), 0.44 (0.28–0.70), and 0.49 (0.31–0.78), respectively. However, there was no statistically significant difference among individual quintiles >13 ng/mL (Fig. 2).
We also observed that mean arterial blood pressure and systolic and diastolic blood pressures were descriptively similar among the groups divided by the quintiles of vitamin D level. Patients in the lowest vitamin D group, on average, were more likely to receive a vasopressor, to receive more red blood cells and colloid, to receive less crystalloid, and to have faster heart rates during surgery (Table 3). However, none of the differences was clinically important.
The importance of vitamin D is apparent in that it is one of the few compounds that is both absorbed and produced by the human body. Vitamin D has been associated with multiple human processes for >100 years,7 though the impact of vitamin D insufficiency and deficiency in biological functions and disease processes is continually being discovered. In our study, higher vitamin D concentrations were associated with decreased odds of in-hospital mortality/morbidity.
The common effect OR for our composite outcome (compared with vitamin D concentrations <13 ng/mL, the lowest quintile of our study population) was inversely related to serum vitamin D concentration. These results are comparable with studies reviewed in a meta-analysis that found that mortality was inversely related to vitamin D concentration in nonoperative settings.24 In that summative study, mortality decreased with higher vitamin D concentrations up to 87.5 nmol/L (35 ng/mL), whereas higher concentrations provided no apparent additional benefit.
Our study showed a similar trend: all concentration ranges >13 ng/mL showed a statistically significant decrease in common effect OR compared with the lowest range. Although there was no statistically significant difference among individual groups >13 ng/mL, nonetheless it is reasonable to conclude that there is concentration dependence after evaluating Figure 2, which shows a fairly linear decrease in the estimated common effect ORs as vitamin D concentration increases from the 2nd quintile (13–20 ng/mL) to the 4th quintile (27–36 ng/mL). Our study expands on previous investigations by calculating the marginal decrease of events associated with increasing vitamin D concentrations, but more importantly, our study differs in being perioperative. For each 5 ng/mL increase in vitamin D concentration, there was a reduction of the corresponding common effect OR for the composite outcome (OR 0.93, 95% CI, 0.88–0.97). This remained consistent over the range from 4 to 44 ng/mL, which encompassed 90% of the observed concentrations.
Among the individual components of our composite, only cardiovascular outcomes were significantly associated with decreased odds (OR 0.92, 95% CI, 0.84–1.00), though there were strong trends for infection and mortality. It has been shown that low vitamin D concentrations are associated with an increased probability of these minor outcomes by decreasing innate host defenses.25 That may have been the case in our patients, but the infection component of our composite was restricted to major infections and thus excluded minor infections such as urinary tract, superficial surgical site, and minor nosocomial infections. Lack of significance on mortality may be related to the relatively low incidence of the event (1.2%).
The only composite component that individually showed significance was cardiovascular complications. Low vitamin D concentrations are associated with increased arterial stiffness and endothelial dysfunction in human blood vessels, which may explain the independent association with cardiovascular mortality.26,27 In addition, animal studies show that vitamin D deficiency causes deregulated stimulation of the renin-angiotensin system, leading to hypertension.28
The relationship between vitamin D levels and the odds of having a common effect outcome clearly shows an association, but causation cannot be claimed. There is, however, the possibility of reverse causation. Patients with cardiac problems may spend more time indoors due to limitations on physical activity, in turn diminishing sun exposure and decreasing vitamin D concentrations. Consistent with this theory, previous studies demonstrated that physical activity is closely related to serum vitamin D concentrations.29,30 Lower vitamin D concentrations are associated with frailty, characterized by decreased strength and endurance measures in older adults. In frail individuals, low vitamin D concentrations have an additive interaction on all-cause mortality.31 Therefore, regardless of directionality, lower vitamin D concentrations are associated with worse outcomes, as our study illustrates.
Optimal concentrations of serum vitamin D for various populations have been discussed extensively in recent years, resulting in a wide range of (generally increasing) target concentrations. The most common currently accepted definition of vitamin D insufficiency is a 25-hydroxyvitamin D concentration between 10 to 30 ng/mL, with concentrations <10 ng/mL, indicating deficiency.1 With the use of these definitions, <20% of our study patients were considered vitamin D deficient. However, our first, second, and middle quintiles completely encompass the insufficiency range. Thus, >60% of our study population was either vitamin D deficient or insufficient (Table 3). That each of our 3 lowest quintiles was associated with serious perioperative complications suggests that current guidelines are generally appropriate.
In a previous study,32 we were unable to demonstrate an association between serum vitamin D concentrations and outcomes after cardiac surgery. There are several potential explanations for this difference. Perhaps, the most obvious is that cardiac surgery involves much more tissue injury and inflammation than most noncardiac surgery; it is possible that vitamin D effects are simply overwhelmed by the insult of cardiac surgery. A second possibility is that hemodilution during cardiopulmonary bypass per se reduces vitamin D concentrations; however, the decrease is transient and resolves within 24 hours.33 The former explanation thus seems more likely.
Analysis of baseline demographics (Table 1) in our population indicates that the fraction of women increased with vitamin D concentration, as did age. Though a previous investigation reported the opposite trend,34 the more recent NHANES III investigation found that vitamin D concentrations were similar in men and women and as a function of age.35 Surgical patients may differ from the NHANES III population; furthermore, our data are more recent. The higher concentrations of vitamin D we observed in Caucasians are consistent with previous studies.34,35 Lower concentrations of vitamin D observed in diabetics are similarly consistent with previous studies, which also show insulin resistance.36,37
An advantage of our study is that we analyzed the 3 outcomes as a composite by using a multivariate analysis that adjusts for the correlations between the outcomes and improves the power as compared with the traditional methods, including analyzing the outcomes as a collapsed composite or separately. There are nonetheless several limitations inherent to retrospective analysis of data. The timing of the vitamin D concentration measurements varied over a range starting 3 months before surgery to 1 month afterwards, although 71% of the samples were obtained before surgery. Though seasonal variations may occur, an individual’s vitamin D level remains relatively stable over a 1-year period.38
A post hoc analysis limited to postoperative vitamin D concentrations shows a similar common effect OR 0.92 for each 5-unit increase in vitamin D concentration. We have no way of knowing the specific indications, preoperatively or postoperatively, for which the vitamin D levels were evaluated. We cannot deny possible selection bias in which patients were tested in this retrospective study. Indeed, patients with postoperative vitamin D measurements, compared with the preoperative measurement group, were sicker as indicated by physical status class, Charlson comorbidity index, and need for emergency surgery. Patients without vitamin D measurements were not included in the study, which theoretically may be healthier patients. Admittedly, the mean and median vitamin D levels may be different than other populations in other locations. Multiple variables were accounted for in the analysis to adjust for potential confounders to address this concern. We also do not know whether low vitamin D concentrations were treated or how effective the treatments might have been. It is thus possible that actual serum concentrations at the time of surgery were higher than the values used in our analysis. However, with varying vitamin D repletion protocols, it can take up to 9 months to reach adequate concentrations.39 It thus seems unlikely that acute treatment, even if provided, would much alter serum concentration in our patients. Due to the relative stability of 25-hydroxyvitamin D concentrations, we included both preoperative and postoperative measurements in our study. Only a small portion of noncardiac surgical patients had 25-hydroxyvitamin D measurements and were included in this study. We have no information as to why these patients had their vitamin D levels measured and thus how this patient selection may have influenced our results; results may differ in other settings and other populations.
To the extent that vitamin D supplementation was instituted and effective, it would reduce the apparent association of low concentrations with adverse outcomes. However, the effect of supplementation of vitamin D levels on outcomes remains questionable.40 Furthermore, treatment was presumably most likely in patients with the lowest concentrations, which would further reduce the apparent concentration dependence of our findings. But even with the possibility of treatment, our results remain highly statistically significant and show clinically important concentration dependence over the entire range of observed values.
Adjustments in the multivariate analysis were made for known confounders for the components of common effect such as baseline demographics and intraoperative variables. But surely, there remain unknown factors that could have substantial impact on the results. Data for this study were collected from the Cleveland Clinic Perioperative Health Documentation System, exclusively representing the Clinic’s Main Campus hospitals. However, the Cleveland Clinic is a tertiary care hospital, drawing patients from around the country and around the world, and the results can presumably at least be generalized to other tertiary centers. Geography is an important factor in determining average vitamin D concentrations. Cleveland is not among the sunniest cities, making it difficult to generalize average vitamin D results. However, the concentration dependence we observed is presumably valid.
In summary, vitamin D concentrations were associated with a composite of in-hospital death, serious infections, and serious cardiovascular events. However, vitamin D levels were presumably not obtained randomly; therefore, a large randomized trial of preoperative vitamin D supplementation and postoperative outcomes is justified.
Name: Alparslan Turan, MD.
Contribution: This author helped design and conduct the study, analyze the data, and write the manuscript.
Attestation: Alparslan Turan has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
Name: Brian D. Hesler, MD.
Contribution: This author helped design and conduct the study, analyze the data, and write the manuscript.
Attestation: Brian Hesler has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Jing You, MS.
Contribution: This author helped design the study, analyze the data, and write the manuscript.
Attestation: Jing You has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Leif Saager, MD.
Contribution: This author helped design the study, analyze the data, and write the manuscript.
Attestation: Leif Saager has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Martin Grady, MD.
Contribution: This author helped design and conduct the study, analyze the data, and write the manuscript.
Attestation: Martin Grady has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Ryu Komatsu, MD.
Contribution: This author helped conduct the study and write the manuscript.
Attestation: Ryu Komatsu has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Andrea Kurz, MD.
Contribution: This author helped design the study and write the manuscript.
Attestation: Andrea Kurz has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Daniel I. Sessler, MD.
Contribution: This author helped design and conduct the study and write the manuscript.
Attestation: Daniel I. Sessler has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
This manuscript was handled by: Sorin J. Brull, MD, FCARCSI (Hon).
a Mascha EJ. Power Calculations for Tests on a Vector of Binary Outcomes (MULTBINPOW). Cleveland Clinic Statistical Software Series 10 edn 2011. Available at: http://www.lerner.ccf.org/qhs/software/multbinpow.php
1. Rosen CJ. Clinical practice. Vitamin D insufficiency. N Engl J Med. 2011;364:248–54
2. Holick MF. Vitamin D deficiency. N Engl J Med. 2007;357:266–81
3. Bogunovic L, Kim AD, Beamer BS, Nguyen J, Lane JM. Hypovitaminosis D in patients scheduled to undergo orthopaedic surgery: a single-center analysis. J B1 Joint Surg Am. 2010;92:2300–4
4. Ginde AA, Scragg R, Schwartz RS, Camargo CA Jr. Prospective study of serum 25-hydroxyvitamin D level, cardiovascular disease mortality, and all-cause mortality in older U.S. adults. J Am Geriatr Soc. 2009;57:1595–603
5. Thomas MK, Lloyd-J1s DM, Thadhani RI, Shaw AC, Deraska DJ, Kitch BT, Vamvakas EC, Dick IM, Prince RL, Finkelstein JS. Hypovitaminosis D in medical inpatients. N Engl J Med. 1998;338:777–83
6. Jeng L, Yamshchikov AV, Judd SE, Blumberg HM, Martin GS, Ziegler TR, Tangpricha V. Alterations in vitamin D status and anti-microbial peptide levels in patients in the intensive care unit with sepsis. J Transl Med. 2009;7:28
7. Cheadle WB. A Discussion on Rickets, in the Section of Diseases of Children at the Annual Meeting of the British Medical Association, held in Glasgow, August, 1888. Br Med J. 1888;2:1145–48
8. Di Rosa M, Malaguarnera M, Nicoletti F, Malaguarnera L. Vitamin D3: a helpful immuno-modulator. Immunology. 2011;134:123–39
9. Gombart AF. The vitamin D-antimicrobial peptide pathway and its role in protection against infection. Future Microbiol. 2009;4:1151–65
10. Bartley J. Vitamin D: emerging roles in infection and immunity. Expert Rev Anti Infect Ther. 2010;8:1359–69
11. Ginde AA, Mansbach JM, Camargo CA Jr. Vitamin D, respiratory infections, and asthma. Curr Allergy Asthma Rep. 2009;9:81–7
12. Bartley J. Vitamin D, innate immunity and upper respiratory tract infection. J Laryngol Otol. 2010;124:465–9
13. Murr C, Pilz S, Grammer TB, Kleber ME, Meinitzer A, Boehm BO, Marz W, Fuchs D. Vitamin D deficiency parallels inflammation and immune activation, the Ludwigshafen Risk and Cardiovascular Health (LURIC) study. Clin Chem Lab Med. 2012;50:2205–12
14. Albert CM, Ma J, Rifai N, Stampfer MJ, Ridker PM. Prospective study of C-reactive protein, homocysteine, and plasma lipid levels as predictors of sudden cardiac death. Circulation. 2002;105:2595–9
15. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med. 1997;336:973–9
16. Wang TJ, Pencina MJ, Booth SL, Jacques PF, Ingelsson E, Lanier K, Benjamin EJ, D’Agostino RB, Wolf M, Vasan RS. Vitamin D deficiency and risk of cardiovascular disease. Circulation. 2008;117:503–11
17. 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;168:1174–80
18. Scragg R, Jackson R, Holdaway IM, Lim T, Beaglehole R. Myocardial infarction is inversely associated with plasma 25-hydroxyvitamin D3 levels: a community-based study. Int J Epidemiol. 1990;19:559–63
19. 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;168:1629–37
20. Legler JM, Lefkopoulou M, Ryan LM. Efficiency and power of tests for multiple binary outcomes. J Am Stat Assoc. 1995;90:680–93
21. Mascha EJ, Imrey PB. Factors affecting power of tests for multiple binary outcomes. Stat Med. 2010;29:2890–904
22. Mascha EJ, Sessler DI. Statistical grand rounds: design and analysis of studies with binary- event composite endpoints: guidelines for anesthesia research. Anesth Analg. 2011;112:1461–71
23. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–83
24. 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;95:91–100
25. Youssef DA, Ranasinghe T, Grant WB, Peiris AN. Vitamin D’s potential to reduce the risk of hospital-acquired infections. Dermatoendocrinol. 2012;4:167–75
26. Al Mheid I, Patel R, Murrow J, Morris A, Rahman A, Fike L, Kavtaradze N, Uphoff I, Hooper C, Tangpricha V, Alexander RW, Brigham K, Quyyumi AA. Vitamin D status is associated with arterial stiffness and vascular dysfunction in healthy humans. J Am Coll Cardiol. 2011;58:186–92
27. 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;168:1340–9
28. Li YC, Qiao G, Uskokovic M, Xiang W, Zheng W, Kong J. Vitamin D: a negative endocrine regulator of the renin-angiotensin system and blood pressure. J Steroid Biochem Mol Biol. 2004;89-90:387–92
29. Lym YL, Joh HK. Serum 25-hydroxyvitamin D3 is related to fish intake and exercise in Korean adult men. Asia Pac J Clin Nutr. 2009;18:372–6
30. Scragg R, Camargo CA Jr. Frequency of leisure-time physical activity and serum 25-hydroxyvitamin D levels in the US population: results from the Third National Health and Nutrition Examination Survey. Am J Epidemiol. 2008;168:577–86
31. Smit E, Crespo CJ, Michael Y, Ramirez-Marrero FA, Brodowicz GR, Bartlett S, Andersen RE. The effect of vitamin D and frailty on mortality among non-institutionalized US older adults. Eur J Clin Nutr. 2012;66:1024–8
32. Turan A, Grady M, You J, Mascha EJ, Keeyapaj W, Komatsu R, Bashour CA, Sessler DI, Saager L, Kurz A. Low vitamin D concentration is not associated with increased mortality and morbidity after cardiac surgery. PLoS 1. 2013;8:e63831
33. Krishnan A, Ochola J, Mundy J, J1s M, Kruger P, Duncan E, Venkatesh B. Acute fluid shifts influence the assessment of serum vitamin D status in critically ill patients. Crit Care. 2010;14:R216
34. Zadshir A, Tareen N, Pan D, Norris K, Martins D. The prevalence of hypovitaminosis D among US adults: data from the NHANES III. Ethn Dis. 2005;15:S5–97
35. Ginde AA, Liu MC, Camargo CA Jr. Demographic differences and trends of vitamin D insufficiency in the US population, 1988-2004. Arch Intern Med. 2009;169:626–32
36. Scragg R, Sowers M, Bell CThird National Health and Nutrition Examination Survey. . Serum 25-hydroxyvitamin D, diabetes, and ethnicity in the Third National Health and Nutrition Examination Survey. Diabetes Care. 2004;27:2813–8
37. Liu E, Meigs JB, Pittas AG, McKeown NM, Economos CD, Booth SL, Jacques PF. Plasma 25-hydroxyvitamin D is associated with markers of the insulin resistant phenotype in nondiabetic adults. J Nutr. 2009;139:329–34
38. Major JM, Graubard BI, Dodd KW, Iwan A, Alexander BH, Linet MS, Freedman DM. Variability and reproducibility of circulating vitamin D in a nationwide U.S. population. J Clin Endocrinol Metab. 2013;98:97–104
39. Whiting SJ, Calvo MS. Correcting poor vitamin D status: do older adults need higher repletion doses of vitamin D3 than younger adults? Mol Nutr Food Res. 2010;54:1077–84
40. Hsia J, Heiss G, Ren H, Allison M, Dolan NC, Greenland P, Heckbert SR, Johnson KC, Manson JE, Sidney S, Trevisan MWomen’s Health Initiative Investigators. . Calcium/vitamin D supplementation and cardiovascular events. Circulation. 2007;115:846–54
41. Cohen J. Statistical Power Analysis for the Behavioral Science. 19882nd ed Hillsdale, NJ Lawrence Erlbaum Associates Inc.