To the Editor:

Arabi et al. report in the recent systematic review and meta-analysis of the effect of vitamin D supplementation on hemoglobin concentration that supplementation with vitamin D had no significant effect on hemoglobin and ferritin levels, while positive effects on transferrin saturation and iron status were observed [1]. There is concern that the studies in this meta-analysis included highly variable patient cohorts and vitamin D dosing, acute and chronic treatment strategies, and not all patients had anemia.

The potential beneficial impact of vitamin D on hemoglobin in patients with anemia was not separately examined. Vitamin D supplementation may provide a safe, simple and cost-effective therapy for prevention and/or treatment of anemia. There is concern that the conclusion of this meta-analysis “… vitamin D had no significant effect on hemoglobin levels” is not meaningful based on the wide diversity of studies included. Why would vitamin D impact on hemoglobin if hemoglobin levels are normal?

Did the authors consider assessment of the total vitamin D dose in each of the studies, and separate analysis of the trials in anemic patients? The Table below confirms the wide variability in total dose and duration in these randomized clinical trials, which makes it challenging to discern an impact of vitamin D on hemoglobin in anemic patients.

It has clearly been documented that high-dose vitamin D is required for hepcidin suppression, which is required for optimal increase in endogenous erythropoiesis and subsequent increase in hemoglobin in anemic patients who have high hepcidin levels and anemia of inflammation or chronic kidney disease [2,3,4,5,6,7]. We would like to highlight the very important finding in Fig. 2 (effect of vitamin D on hemoglobin) in 2.3.6 “Critically ill patients”, that the placebo cohort mean hemoglobin was 6.7 g/dL vs. 11.25 g/dL in the high-dose (500,000 units total, 100,000 units daily for 5 days) vitamin D intervention cohort.

It should also be noted that in Table 1 for the Smith study in mechanically ventilated critically ill adults, the duration of the vitamin D intervention was listed as “4 weeks”, but the study drug was administered in 5 equal doses over 5 days, not 4 weeks. There are additional errors in that Table regarding drug dosing, duration and year of publication, all corrected in the Table below.

Table 1 Wide variability in Vitamin D dose and duration, type of patients enrolled, and whether anemia was present or not in randomized clinical trials