Comment 60% reduction in risk? (Score 5, Informative) 478
I believe that Vitamin D might protect against some cancers.
However, I do not agree that Vitamin D deficiency can be responsible for about 60% cancers.
Here are my reasons why:
1) The process of carcinogenesis (initiation of the first DNA mutation/ adduct required to form cancer to the stage of clinically overt disease) in most cases takes more than 4 years. This clinical trial is only 4 years and too premature to reach to conclusions.
2) I have yet to read the paper, but it is necessary to know whether this trial was truly randomized meaning that the those who got the Vitamin D pill and those who got the placebo were similar to each other in all other ways. It is possible that if it is not randomized, a healthier cohort of people chose to take Vitamin D for a long time.
3) It is also important to know how they treated those people who dropped out of taking the Vitamin D pills. It is possible that unhealthier people dropped out and then we were comparing all subjects in the placebo group to the "healthier" people in the Vitamin D group.
4) A risk reduction of 60% (= relative risk of 0.4) is epidemiologically very strong and if that was the case, we would have already found such a role of Vitamin D much earlier (like 30 years before or so). There is something called Bradford Hill's criteria for causation in epidemiology which has strength of association as one of the criteria. The rationale for that is if we had a confounder which is actually responsible for the effect, we would have known it before because it is more likely to have a stronger effect. The same principle goes here. We do not know anything that could prevents so many types of cancer with such great attributable fraction. The magnitude of effects of like 2.5 or reduction of risk to 0.4 were the strengths we used to see in the papers of 1970s. Hence I think there could be some issues with the study design and data analysis of this study if they found such a great magnitude of effect.
Having said that I think that Vitamin D might prevent many cancers, but I expect a lower magnitude of the effect.
However, I do not agree that Vitamin D deficiency can be responsible for about 60% cancers.
Here are my reasons why:
1) The process of carcinogenesis (initiation of the first DNA mutation/ adduct required to form cancer to the stage of clinically overt disease) in most cases takes more than 4 years. This clinical trial is only 4 years and too premature to reach to conclusions.
2) I have yet to read the paper, but it is necessary to know whether this trial was truly randomized meaning that the those who got the Vitamin D pill and those who got the placebo were similar to each other in all other ways. It is possible that if it is not randomized, a healthier cohort of people chose to take Vitamin D for a long time.
3) It is also important to know how they treated those people who dropped out of taking the Vitamin D pills. It is possible that unhealthier people dropped out and then we were comparing all subjects in the placebo group to the "healthier" people in the Vitamin D group.
4) A risk reduction of 60% (= relative risk of 0.4) is epidemiologically very strong and if that was the case, we would have already found such a role of Vitamin D much earlier (like 30 years before or so). There is something called Bradford Hill's criteria for causation in epidemiology which has strength of association as one of the criteria. The rationale for that is if we had a confounder which is actually responsible for the effect, we would have known it before because it is more likely to have a stronger effect. The same principle goes here. We do not know anything that could prevents so many types of cancer with such great attributable fraction. The magnitude of effects of like 2.5 or reduction of risk to 0.4 were the strengths we used to see in the papers of 1970s. Hence I think there could be some issues with the study design and data analysis of this study if they found such a great magnitude of effect.
Having said that I think that Vitamin D might prevent many cancers, but I expect a lower magnitude of the effect.
