There can be too much of a good thing, but research finds that too little can be harmful, too. Against some guidelines, a study suggests that too little sodium also can increase the risks of cardiovascular events, stroke and death.

The study, published in the Aug. 14 issue of the New England Journal of Medicine, also explored the effect of potassium on risks of death, cardiovascular events and stroke.

Martin O’Donnell, MB, PhD, of McMaster University in Hamilton, Ontario, and colleagues utilized 101,945 participants from the international Prospective Urban Rural Epidemiology (PURE) study from January 2003 through October 2013. They determined amounts of sodium and potassium intake based on excretion in urinalysis.

They found that the ideal sodium excretion level was between 4 and 5.99 g per day and that more or less than that increased risks. Seven g per day or more increased risk for major cardiovascular events (odds ratio 1.16), major stroke (odds ratio 1.29) and, significantly, death from cardiovascular-related causes (odds ratio 1.54). Most physicians have recommended reducing sodium intake, particularly in hypertensive or cardiac patients.

However, O’Donnell et al also found that less than 3g per day correlated with increased risks of major cardiovascular events (odds ratio 1.3), cardiovascular-related death (odds ratio 1.77), major stroke (odds ratio 1.37) or death from any cause (odds ratio 1.38) as well.

Meanwhile, the risks for death and/or cardiovascular events decreased with potassium excretion beyond 1.5 g or less.

Why?

Basic human biochemistry requires a certain level of sodium to function and below 3 g per day, the renin-angiotensin-aldosterone system is activated, they wrote. This finding was not at odds with previous studies, as patients with hypertension, diabetes or cardiovascular disease are already sensitive to sodium adjustment and by lowering those sodium levels, stresses on the whole system are lowered to a point.

O’Donnell et al noted that the reduction in risk with increased potassium intake appeared to be relative to either blood pressure effects or was a symptom of a healthier lifestyle impacting overall health in a positive way.

They wrote that their epidemiological study should not be interpreted as a recommendation for clinical practice.