Should socioeconomic factors be considered as traditional risk factors for cardiovascular disease, as confounders, or as risk modifiers?