Salt Intake and Mortality.

Feng J. He,G. A. MacGregor

DOI: https://doi.org/10.1093/ajh/hpu095

2014-01-01

American Journal of Hypertension

Abstract:To the Editor: In a recent meta-analysis of cohort studies, “Compared With Usual Sodium Intake, Low- and Excessive-Sodium Diets Are Associated With Increased Mortality: A Meta-Analysis,” Graudal et al. claimed that there was a U-shaped association between salt intake and all-cause mortality, with intakes of 12.363g/day both being associated with higher mortality.1 However, their conclusion is invalid because of the severe methodological flaws of the studies included. Two recent articles from the Science Advisory of the American Heart Association,2 along with several other articles, have provided detailed analysis of the methodological issues in cohort studies (e.g., reverse causality, errors in salt assessment, residual confounding).2 It is surprising that Graudal et al. did not mention any of these articles and the inherent problem of estimating individuals’ usual salt intake when it varies hugely from day to day. It is extraordinary that Graudal et al. continue to quote the meta-analysis by Taylor et al. as outcome trial evidence, in spite of the fact that the meta-analysis has already been withdrawn from The Cochrane Library, largely because of the retraction of the trial in heart failure (http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD009217.pub2/abstract). Graudal et al. seemed to be unaware of or chose to ignore this. Taylor et al.’s meta-analysis was published simultaneously in The Cochrane Library and the American Journal of Hypertension. Surely, the American Journal of Hypertension should also retract this article. Graudal et al. tried to ascribe the higher mortality with a lower salt intake to the elevated renin-aldosterone activity, sympathetic nerve activation, and lipid abnormalities. They quoted their own meta-analysis, which included a large number of trials with an acute salt loading to abrupt and severe salt restriction (e.g., from 20 to <1g/day for only a few days). Such metabolic studies are not relevant to public health recommendations for modest salt reduction over a prolonged period of time. Indeed, recent meta-analyses of randomized trials have demonstrated that, with longer-term modest salt reduction, there was only a small increase in plasma renin activity and no significant change in the sympathetic nervous activity or lipids. Graudal et al. again failed to quote these more relevant articles. As usual, Alderman, a coauthor, has once again failed to declare that he has worked over many years as a consultant to the Salt Institute.3 As editor-in-chief of the American Journal of Hypertension, this could be viewed as a very serious conflict of interest. The totality of evidence, including epidemiological studies, animal studies, randomized trials, and outcome studies all show the substantial benefits of salt reduction. The findings from Graudal et al.’s meta-analysis with methodological flaws should not divert us from reducing population salt intake. Most countries are now adopting a policy of reducing salt intake by persuading the food industry to reformulate food with less salt, as is occurring successfully in the United Kingdom, with an accompanying fall in population blood pressure and mortality from stroke and ischemic heart disease.4 The World Health Organization has recommended salt reduction as 1 of the top 3 priority actions to tackle the global noncommunicable disease crisis. A reduction in salt from the current intake of 9–12g/day to the recommended level of <5–6g/day will have major beneficial effects on health along with major cost savings in all countries around the world.