Vitamin D at hospital admission as an independent predictor of outcome of sepsis patients: Results of a secondary analysis from a “Need-Speed” trial prospective cohort

Background Early prognostic stratification of septic patients at hospital admission is challenging, especially in the elderly. Considering the well-known immunomodulatory effects of vitamin D and its common deficiency among elderly, we aimed to evaluate if vitamin D plasma levels (25(OH)D) at hospital admission could be a prognostic biomarker in septic patients. Methods Secondary analysis of data and samples from the multicenter Need-Speed trial, a multicenter study involving the emergency and internal medicine wards of five Italian hospitals. 1132 consecutive patients admitted to hospital with suspected sepsis were enrolled. 859 were confirmed to have sepsis at the end of the diagnostic work-up and 829 patients were included in the analysis. 25(OH)D at admission was measured using an automated chemiluminescence assay. Results Among the 829 patients included (median age 81 years [IQR: 72–87]), severe hypovitaminosis D was observed (median 11.2 ng/mL [IQR: 8.0–19.4]). At univariate analysis, baseline 25(OH)D levels were significantly lower in patients who died at 30 and 90 days and higher among patients discharged alive within 15 days (p < 0.05). Multivariate models confirmed 25(OH)D as an independent predictor of 90-day mortality and of discharge alive at 15 days (p < 0.05). Moreover, a 25(OH)D threshold of 12 ng/mL independently predicted both survival at 90 days (OR: 0.5961 [0.4068–0.8735]) and discharge alive at 15 days (OR: 1.6055 [1.1738–2.1960]). Conclusions Lower 25(OH)D levels were independently associated with poorer clinical outcomes in older patients with sepsis. Therefore, hypovitaminosis D assessment may provide prognostic value beyond validated risk stratification tools, warranting prospective validation.