Criteris d’inclusió (els han de complir tots) 1. Diabetis mellitus tipus 2 (E11, E11.9) 2. Obesitat (E66.8, E66.9) o índex de massa corporal ? 22 kg/m2 3. No estar en tractament amb antidiabètics (codis ATC: A10) 4. HbA1c ?7% i ?10% 5. Edat: ?10 anys i ?17 anys Criteris d?exclusió (qualsevol) 1. Diabetis mellitus tipus 1(E10.9) 2. Diabetis mellitus tipus 2 (E11, E11.9) diagnosticada fa més d?un any 3. Haver tingut alguna vegada alguna prescripció de sitagliptina, vildagliptina, alogliptina, saxagliptina, exenatide o liraglutide (codis ATC: A10BH i A10BX). 4. Tractament crònic amb corticosteroides orals (codi ATC: H02) 5. Hipertiroïdisme en tractament (E05.9) 6. HIV (Z21) 7.Neoplàsia (C18.9, C50.9, C25.9, C61, C34, C20, C73, C43.9,C96.9, C95.9, C91.1, C92.0, C92.1, C64) 8. Hepatitis B o C (B18.2, B16.9)

ABSTRACT: Body iron stores have been suggested to play a role in the development of cardiovascular disease (CVD) through its pro-oxidant properties. However, epidemiological studies on iron status and the risk of cardiovascular disease have showed contradictories results.
OBJECTIVE: To investigate the association between iron status and the risk of CVD in a primary health care sample.
METHODS: Design: Prospective cohort study; Study population: The study will use the Information System for the Development of Research in Primary Care (SIDIAP database). Cohort definition: The analysis will be restricted to subjects aged between 35 and 74 years for whom serum ferritin measurements at baseline (January 1 2006 to December 31 2008) are available and who did not have known cardiovascular disease. Follow-up: Participants recruited in 2006 will be followed for up to 6 years and will have the longest follow-up. Participants entering the study during the year 2008 will be followed for 3-4 years, thus having the shortest follow-up. Consequently, we expect a median follow-up of 5 years. Primary outcome: composite endpoint of cardiovascular death, non-fatal myocardial infarction and non-fatal stroke.
STATISTICAL ANALYSIS: Hazards ratios (95% confidence intervals) for CVD endpoints with quintiles of ferritin after multivariable adjustment will be estimated using survival analysis with age as time scale and individual starting follow-up times (age at recruitment) treated as staggered entries. To test for non-linear relationships in all cardiovascular models, we will use restricted B-splines with knots at 20th, 50th and 80th percentiles of each serum ferritin distribution.
EXPECTED RESULTS: We will find that high levels of serum ferritin will be associated with an increase of CVD.