A Confounding Mess
In this week’s issue of the Annals of Internal Medicine (unfortunately you need a subscription to see the article) Dr. Andrew Gray MD, and Mark Bolland PhD write an interesting editorial called Vitamin D: A Place in the Sun. In it they discuss the current flood of claims for health benefits of vitamin D. When analyzing a finding such as the link between Vitamin D levels and a disease state, one of the most difficult aspects is to control for “confounding variables.” The definition of a confounding variable in The Free Dictionary online is, “interference by a third variable so as to distort the association being studied between two other variables, because of a strong relationship with both of the other variables.” Low vitamin D levels are associated with such potential confounding variables as obesity, sedentary lifestyle, and lack of sunlight exposure. These conditions and others are potentially associated with depression, diabetes, and many of the other conditions for which low vitamin D levels has been suggested as a risk factor Hence the concern about confounding variables not accounted for in the association of these conditions with vitamin D. This is elegantly stated in these authors middle paragraph of their editorial:
“Although some evidence from preclinical studies suggests that components of the vitamin D system might impact favorably on some diseases,4 it seems intuitively unlikely that a singlehormone could play a substantial role in preventing or ameliorating the diverse range of diseases that have been linked to low levels of vitamin D. A more plausible and prosaic explanation for thefindings of the observational vitamin D studies is the presence of common confounders. Vitamin D levels are directly related to sunlight exposure and physical activity, and inversely related to adiposity.5-6 It is likely that less healthy individuals, who are more likely to subsequently experience morbid events, will be heavier, less active, and more sunlight-deprived than healthier ones and therefore have lower levels of 25(OH)D. This notion is supported by previous studies7and by the study of Llewellyn et al,2 in which several indices of poor health were more commonly observed at baseline in those with lower levels of vitamin D. Thus, low vitamin D levels may simply be a marker for lower health status rather than a cause of it. Healthiness is difficult to measure and adjust for, as illustrated by the disparate results of observational studies and randomized controlled trials (RCTs) of postmenopausal hormone therapy and cardiovasculardisease or antioxidant treatment and cardiovascular disease and cancer.8-11 We should therefore treat the data from observational studies of vitamin D with caution.”
Although low vitamin D levels are associated with multiple conditions it will be interesting to see if supplementation of vitamin D lowers the risk of these conditions or whether only improving the confounding conditions, like more exercise, lower levels of obesity, more sunlight, etc. works to prevent the medical conditions. It will sure be nice if just taking those nice small, easy to swallow yellow capsules works, but as for me I remain skeptical.