Urinary sodium excretion study attracts controversy

May 25 2016

By Eleanor McDermid

A pooled analysis in The Lancet showing a U-shaped association between urinary sodium excretion and cardiovascular disease (CVD) has drawn criticism.

A statement from American Heart Association president Mark Creager (Dartmouth-Hitchcock Medical Center) described the study as "flawed" and reiterated the Association's call for "the vast majority of persons around the globe to consume less sodium".

The analysis by Andrew Mente (Hamilton Health Sciences, Ontario, Canada) and colleagues differs from previous research in that it addresses the CVD impact of salt intake according to the presence or absence of hypertension.

The research incorporates data from 133,118 participants of the PURE, ONTARGET, TRANSCEND and EPIDREAM trials, and suggests that high salt intake is linked to adverse CVD outcomes only in people with hypertension, whereas low salt intake increases CVD risk irrespective of hypertension status.

Salt intake was estimated based on fasting morning urine samples, extrapolated to 24-hour excretion by means of the Kawasaki formula. In a press statement, Francesco Cappuccio (World Health Organization Collaborating Centre for Nutrition, University of Warwick, UK) described the approach as a "flawed assessment" and also highlighted the issue of reverse causality, given the entire study population was drawn from clinical trials involving high-risk people.

He also noted the "statistical sin" of splitting participants into normotensive and hypertensive categories, a division he described as "biologically meaningless".

The researchers report that among patients with high urinary sodium excretion of at least 7 g/day, those with hypertension had a significant 23% increased risk of death and CVD events relative to those with moderate 4-5 g/day excretion. However, this association was not observed among patients without hypertension over a median follow-up of 4.2 years.

But patients with low sodium excretion, of less than 3 g/day, had an increased CVD risk relative to those with moderate excretion regardless of hypertension status, with risk increases of 34% and 26% for those with and without hypertension, respectively.

The associations persisted after accounting for blood pressure, and were also evident for death from any cause.

In a linked commentary, Eoin O'Brien (University College Dublin, Ireland) argues that the "provocative" findings should be given due consideration. In particular, he notes that, while the method of estimating salt intake is not appropriate on an individual basis, it has been validated against 24-hour urine collection and is therefore suitable for assessing average salt intake in large cohorts.

He says that "we must acknowledge that given the dependency of so many physiological systems on the sodium cation, it should come as no surprise that a low-salt-for-all policy would benefit some and disadvantage others".

"So rather than allowing contrary evidence to dispel the positive efforts that have been made to reduce the salt content of foods, we must now direct our efforts to formulating a policy that will benefit the majority in society without [compromising] the minority."

Licensed from medwireNews with permission from Springer Healthcare Ltd. ©Springer Healthcare Ltd. All rights reserved. Neither of these parties endorse or recommend any commercial products, services, or equipment.