The effects of vitamin D on all-cause mortality in different diseases: an evidence-map and umbrella review of 116 randomized controlled trials

Abstract

PurposeTo conduct a solid evidence by synthesizing meta-analyses and updated RCTs about the effects of vitamin D on all-cause mortality in different health conditions.MethodsData sources: Pubmed, Embase, Web of Science, the Cochrane Library, Google Scholar from inception until 25th April, 2022. Study selection: English-language, meta-analyses and updated RCTs assessing the relationships between vitamin D and all-cause mortality. Data synthesis: Information of study characteristics, mortality, supplementation were extracted, estimating with fixed-effects model. A Measurement Tool to Assess Systematic Reviews, Grading of Recommendations Assessment, Development and Evaluation, and funnel plot was used to assess risk of bias. Main outcomes: All-cause mortality, cancer mortality, cardiovascular disease mortality.ResultsIn total of 27 meta-analyses and 19 updated RCTs were selected, with a total of 116 RCTs and 149, 865 participants. Evidence confirms that vitamin D reduces respiratory cancer mortality (RR, 0.56 [95%CI, 0.33 to 0.96]). All-cause mortality is decreased in patients with COVID-19 (RR, 0.54[95%CI, 0.33 to 0.88]) and liver diseases (RR, 0.64 [95%CI, 0.50 to 0.81]), especially in liver cirrhosis (RR, 0.63 [95%CI, 0.50 to 0.81]). As for other health conditions, such as the general health, chronic kidney disease, critical illness, cardiovascular diseases, musculoskeletal diseases, sepsis, type 2 diabetes, no significant association was found between vitamin D and all-cause mortality.ConclusionsVitamin D may reduce respiratory cancer mortality in respiratory cancer patients and all-cause mortality in COVID-19 and liver disorders' patients. No benefits showed in all-cause mortality after vitamin D intervention among other health conditions. The hypothesis of reduced mortality with vitamin D still requires exploration.Systematic review registrationhttps://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=252921, identifier: CRD42021252921.