Vitamin D and health: genome-wide analysis and insights from Mendelian randomisation.

Inadequate vitamin D intake has been implicated as a risk factor for many important diseases such as diabetes, cancer and cardiovascular diseases. However, much of the evidence for the health effects of vitamin D comes from observational studies, and conclusions based on these studies may be affected by methodological problems such as bias and confounding. In this study, we will use a novel approach based on genetic information (so-called Mendelian Randomisation) and our aim is to see if the associations between vitamin D and the risk of various diseases can be replicated. If replication is found, this would give further support for a causal association between vitamin D and the outcome of interest. We will use data from the nationwide 1958 British birth cohort, in which over 7000 participants have been followed-up over four decades and detailed information has been collected on genetic markers and a wide-range of health related factors, including vitamin D status, lifestyle, diet, and indicators of disease risk.

Inadequate vitamin D intake has been implicated as a risk factor for diabetes, cancer and cardiovascular diseases. Much of the evidence for health effects comes from observational studies and may be affected by problems of bias and confounding. The use of vitamin D prophylaxis in intervention studies is limited by concerns of possible toxicity. We propose to carry out a genome-wide analysis on 25(OH)D in the British 1958 cohort study - the first such study in a general population. Further, after identification of genetic variants associated with vitamin D metabolism, we will proceed to apply a mendelian randomization approach to obtaining evidence as to the effects of vitamin D. In mendelian randomization, genetic variants are used as proxy markers for variations in environmentally modifiable exposures (here vitamin D intake and levels). If evidence for an association between vitamin D and disease risk (suggested by other types of studies) can be obtained using mendelian randomization, this would provide evidence for a causal relation. We will evaluate method limitations and the use multiple indicators (located in varying parts of the human genome) as a way of overcoming some of these problems. The key aim of the project is to provide information on an approach, which could provide a safe and cost-effective way of evaluating the influences of vitamin D on health.