We appreciate Rings interest in our recent study  reporting a significant association between the use of vitamin D and a lower risk of death from cardiovascular disease in an observational cohort of 242 haemodialysis patients. The problems raised by Ring derived from the non-randomized, uncontrolled design of the study, and the small number of events during the mean follow-up period of 61±23 months. We recognized the limitations of this study, and discussed this in detail in the report. By carefully reading the discussion section, it will be apparent that most of his questions are addressed.
We do not believe that our finding was the result of inappropriate use of statistics applied to a small group of subjects, because there are at least two similar observations made in much larger cohorts. A report by the Japanese Society for Dialysis Therapy  showed that 1-year survival was better in haemodialysis patients treated with vitamin D than those not on such medication based on the analysis of 77 486 Japanese patients. Teng et al.  found a significant survival difference between two groups, one treated with paricalcitol and the other treated with calcitriol, in a cohort of more than 60 000 haemodialysis patients in the US. Although these studies were not randomized and, controlled, it is likely that mortality risk has a strong association with use of vitamin D regardless of ethnic groups.
However, these results do not directly indicate that vitamin D treatment has a favourable effect on survival. We carefully stated in our report that the use of alfacalcidol was associated with reduced risk for cardiovascular death. The observation supports the view that vitamin D treatment may be effective in preventing cardiovascular death, but it is not the proof of this hypothesis.
We agree with Ring that many factors should be clarified and controlled to draw a solid conclusion as to whether the close association between the use of vitamin D and mortality risk is the effect of vitamin D, or a mere association confounded by some factors related to such medication. Clearly, we need a well-designed, controlled prospective clinical trial.
Conflict of interest statement. None declared.