The identification of risk factors for cardiovascular diseases (CVD) using longitudinal cohort studies has been a major advance in medicine during the second half of the past century and, stimulating the development of risk-reduction strategies and treatments, has contributed to the decline of age-standardised cardiovascular mortality over the past few decades, at least in high-income countries [1,2]. CVD risk factors have been classified as non-modifiable, such as family history, older age and male sex; and modifiable, such as systolic blood pressure, low-density lipoprotein cholesterol level, tobacco smoking, diabetes and elevated body-mass index [3,4].