Moderator's view: Salt, cardiovascular risk, observational research and recommendations for clinical practice.

In observational studies, blood pressure (BP), cholesterol and nutritional status biomarkers, including sodium intake, coherently show a J- or U-shaped relationship with health outcomes. Yet these data may reflect a stable sodium intake or a reduced intake due to comorbidities or intercurrent disease, or an intentional decrease in salt intake. Adjusting for comorbidities and risk factors may fail to eliminate confounding. For cholesterol and BP, we base our recommendations for prevention and treatment on interventional (experimental) studies. For sodium, we lack the perfect large-scale trial we would need, but substantial circumstantial information derived from interventional studies cannot be ignored. The objection that modelling the risk of salt excess for cardiovascular disease events based on the effect of salt intake on BP is unjustified fails to consider a recent meta-analysis showing that, independently of the intervention applied, intensive BP-lowering treatment (average BP 133/76 mmHg), compared with the less intensive treatment (140/81 mmHg), is associated with a 14% risk reduction for major cardiovascular events. In this knowledge context, inertia, i.e. awaiting the 'mother trial', is not justified. While recognizing that this trial may still be needed and that actual data, rather than modelled data, are the ideal solution, for now, the World Health Organization recommendation of reducing salt intake to <2 g/day of sodium (5 g/day of salt) in adults stands.