ISSN 1758-2024 Neurodegen. Dis. Manage. (2011) 1(3), 179–182 10.2217/NMT.11.26 © 2011 Future Medicine Ltd Over the last decade there has been increasing interest in the role of vitamin D insufficiency in the genesis of autoimmune diseases, including multiple sclerosis (MS). The notion of possible links between vitamin D and MS originally arose from ecological patterns: the latitudinal gradient in MS [1], with higher prevalence and incidence with increasing distance from the equator [2]; and temporal patterns of increasing MS incidence over the last 50 years or so [3]. Latitude is highly correlated with levels of ambient UV radiation (UVR), particularly in the UVB wavelengths that are essential to the endogenous synthesis of vitamin D [4]. Following on from these ecological observations, individual-level case–control [5] and cohort [6] studies have indeed demonstrated that higher vitamin D status, measured as the serum concentration of 25-hydroxyvitamin D (25[OH]D), is associated with decreased MS risk. The effect is relatively modest, with a 7% decrease in odds of a first demyelinating event (FDE) for every 10 nmol/l increase in 25(OH)D concentration in the Ausimmune Study [5]. In the US military cohort study, there was evidence of a threshold effect, with decreased risk only at levels of >99.1 nmol/l, but this has not been consistently observed [6]. Given that the recent US Institute of Medicine report [101] recommends 25(OH)D levels of >50 nmol/l as ‘vitamin D adequacy’, a threshold of nearly 100 nmol/l is high, and similar to the point of inflexion (i.e., increasing risk at levels greater than ~100 nmol/l) of the reported U-shaped relationship between 25(OH)D concentration and all-cause mortality [7,8]. Vitamin D supplementation to these levels or higher may not be without risk. Observational epidemiological studies generally use a single 25(OH)D level to represent vitamin D status, often without determining the extent to which this reflects UVR exposure or diet-/supplementderived vitamin D. Average vitamin D status over an undefined period preceding disease onset might be assumed to be the important risk factor [9] – with the single measurement assumed to approximate