This thesis explores innovative approaches to improving the estimation of the risk of atherosclerotic cardiovascular disease using association studies. The first part focuses on the performance of traditional risk factors in multi-ethnic populations closely reflecting patient populations encountered in clinical practice, including the distribution and differential impact of risk factors. In patients with hypertension from six ethnic groups, family history of cardiovascular disease was independently associated with self-reported coronary/peripheral vascular disease or stroke, confirming its potential prognostic role in multi-ethnic populations. This association was particularly prominent in individuals of African Surinamese background. Among individuals of South Asian background living in the Netherlands, a longitudinal analysis revealed that a glucose “delta” (difference in blood glucose between fasting and after a glucose load) was associated with 5-year incidence of diabetes similarly to (but not more strongly than) fasting or post-load glucose. An unexpectedly high prevalence of negative glucose delta (post-absorptive glucose lowering) was observed. The second part focuses on potential detection of early atherosclerotic damage using non-invasive markers of endothelial function obtained from the sublingual microcirculation. A small endothelial glycocalyx size, estimated by the Perfused Boundary Region (PBR), was associated with female sex and diabetes but not with other cardiovascular risk factors or self-reported cardiovascular disease. Reproducibility of PBR, vascular density, and red blood cell filling from the sublingual microcirculation was moderate to poor, and a large sample size was estimated to be needed to detect short-term changes caused by acute stimuli (a meal or cigarette smoking).