Association between baseline vitamin D metabolite levels and long-term cardiovascular events in patients with rheumatoid arthritis from the CIMESTRA trial: protocol for a cohort study with patient-record evaluated outcomes

INTRODUCTION: Cardiovascular morbidity and mortality is increased in patients with rheumatoid arthritis (RA), and among these patients, the prevalence of hypovitaminosis D is high. Moreover, low vitamin D levels have been associated with increased cardiovascular risk in healthy subjects. OBJECTIVE: To evaluate the long-term risk of cardiovascular events in patients having low total 25-hydroxyvitamin D levels at baseline compared with patients with normal levels, in an efficiently treated, closed cohort of patients with an early diagnosis of RA. METHODS AND ANALYSIS: This study is a prospective, closed, blinded endpoint cohort study, based on secondary analyses from a previous randomised trial (CIMESTRA study; NCT00209859, approved September 1999) including 160 patients with an early diagnosis of RA from Danish University clinics. Primary outcome will be the proportion of patients with any cardiovascular event in the follow-up period, evaluated using systematic journal audits. Logistic regression models will test the hypothesis that there are more cardiovascular events in enrolled patients with a low level of vitamin D (< 50 nmol/L). Secondarily, Cox regression models, based on survival analysis, will determine the extent to which independent variables (including different levels of vitamin D at baseline) predict whether a cardiovascular event will occur, and also when this will be. ETHICS AND DISSEMINATION: All patients have received verbal and written information before enrolment, and have given written consent at baseline. To disseminate comprehension of factors of prognostic importance to cardiovascular outcome in RA, we will attempt to have a first draft ready no later than 1 year after the adjudication process has finished. If low vitamin D levels can predict cardiovascular events in RA, it is relevant to take into account in a prediction model, to be considered by patients, physicians and other decision-makers. TRIAL REGISTRATION NUMBER: The parental controlled trial is registered as NCT00209859.