RESUMO
We studied the association between admission serum 25-hydroxy vitamin D3 level and in-hospital mortality in a prospective cohort of critically ill patients admitted to the medical intensive care unit of a tertiary care referral center. Of the 180 patients enrolled, 129 were included. Vitamin D3 deficiency was observed in 37% (n = 48) and supra-physiological levels (≥250 nmol/L) in 15.5% (n = 20). Patients with supraphysiological vitamin D3 levels were grouped as outliers. There was no difference in mortality (p = 0.41) between vitamin D3 deficient (21/48) and non-deficient (36/81) patients in analysis with and without outliers. Patients with vitamin D3 ≥250 nmol/L had a significantly higher (p = 0.02) Simplified Acute Physiology Score (SAPS) II and mortality (p = 0.003) [mean (SD) 60.1 ± 17.1 and 75% (15/20), respectively] when compared with the rest [45.6 ± 18 and 38.5% (42/109), respectively]. The sensitivity, specificity and SAPS II independent odds ratio to predict mortality in patients with supraphysiological vitamin D3 levels were 26.3, 93.1 and 3.7% (95% confidence interval 1.2-11.4; p = 0.03), respectively. In conclusion, vitamin D3 deficiency in our cohort was not associated with mortality. A patient subset with supra-physiological vitamin D levels had higher illness severity scores and mortality. Extrinsic factors interfering with test results were ruled out. A biological hypothesis to explain this observation is proposed. Further clarification of mechanisms leading to this observation is warranted.