Published on March 30, 2016
March 30, 2016
In December 2011, the National Multiple Sclerosis Society convened a two-day meeting of 40 MS experts from around the world. They were neurologists, epidemiologists, clinical trial specialists, laboratorians like me, molecular biologists, geneticists, and statisticians. They were asked to put their collaborative heads together to develop a research protocol that could provide the highest level of evidence as to whether taking vitamin D can help to prevent MS. There were no vitamin D cynics in this group of bright, enthusiastic individuals. But despite everyone’s best efforts, nothing came of the meeting.
How do you “prove” that if people take vitamin D, their risk of developing MS will be lower, the way sunshine and higher serum 25(OH)D levels in populations are related to lower risk of MS? The answer to this question starts from an appreciation of what is called the evidence pyramid.
At the wide, low end of the evidence pyramid is what was once known as expert opinion, i.e. what the esteemed professor says. This is a wide base because there are a lot of experts and everyone has an opinion. Opinion is not proof of anything. Next up, is evidence from the laboratory, be it cellular biology or animal research; which generate scientific data relatively easily, but which are hard to translate to clinical practice. Higher on the pyramid still, are case histories, or case reports that present the experience of individual patients. Those lack a context, since you would not know what would happen without, say, the vitamin D.
Still higher up the pyramid is the important field of epidemiology, the study of disease in populations. Epidemiology is powerful; it put anti-smoking and seatbelt use into public policy. Epidemiology demonstrates relationships, but it is not absolute proof that a lack of vitamin D causes multiple sclerosis. In a court of law, epidemiology would still be called circumstantial evidence.
Yes, some will argue that a set of considerations known as the Hill criteria will amount to “proof” in epidemiology. This idea stems from work of Bradford Hill who “proved”, through logic, that smoking is a cause of lung cancer. However, the same kind of analysis has been presented to me relating vitamin D to MS, and that relationship is much less powerful than smoking to lung cancer, and much tougher to prove. I am absolutely on the side of vitamin D, but I have also been in on serious, mainstream efforts to “prove it”.