Sodium restriction is a cornerstone of heart failure management, but many people would be surprised to learn that there is no good supporting evidence for the practice. In the 2009 heart failure guidelines sodium restriction in heart failure received a Class I recommendation (recommended), but this was based only on expert consensus (a C level of evidence of C). More recently the 2013 guidelines downgraded the recommendation to Class IIa (reasonable) based on the same level of evidence. Now a new observational study raises further questions about sodium restriction in heart failure.
In a paper published in JACC: Heart Failure, researchers from Rush University Medical Center sought to assess the impact of sodium restriction, which was tracked by serial food frequency questionnaires, on outcomes in 833 heart failure patients who participated in the HF Adherence and Retention Trial (HART). No evidence to support sodium restriction was found; in fact, sodium restriction was associated with significant increase in the risk of death or heart failure hospitalization. However, any firm interpretation is impossible due to the retrospective and observational nature of the study.
The authors write that their “findings support further downgrade of the ACCF/AHA sodium restriction recommendation in patients with chronic HF to class IIb (efficacy less well established, conflictive evidence), and press the need for multicenter randomized trial to definitively address the role of sodium restriction in HF management.”
Heart Failure Experts Decry Absence of Evidence
In an accompanying editorial comment, Scott Hummel and Matthew Konerman (University of Michigan) acknowledge the study’s limitations but recommend that the current guidelines “should be taken with a grain of salt.” They write: “Considering the challenge this restriction poses for patients, it is even more important to clarify whether sodium restriction is beneficial at all.”
In a second editorial comment, Clyde Yancy (Northwestern University) points out that the “traditional logic” supporting sodium restriction “has been so clear that to call for proof seemed heretical. Yet, an exhaustive search of available literature does not provide evidence.” “The current study,” he continues, “adds to the momentum to call the question but does not provide answers.”
Asked to comment on the paper, several additional experts decried the current situation and called for more data. Milton Packer (Baylor University, Dallas) sent the following statement:
I am not certain what we are supposed to do with these data. A retrospective analysis of the association of reported (not confirmed) sodium intake and various outcomes is not a particularly useful piece of evidence. Should we dismiss the findings entirely? Should the findings lead us to question our core recommendations regarding sodium restriction? How is one supposed to know? These data do not help clarify the situation at all. We have had no basis for any of our recommendations regarding sodium restriction during the past 50 years, although these recommendations have changed a great deal (for no good reason). After this report, we still have no basis for any of our recommendations regarding sodium restriction. We were ignorant before; we are not any smarter now. Did we really need this report to tell us that we lack evidence for our recommendations regarding dietary sodium in patients with heart failure?
Salim Yusuf (McMaster University), said he fully concurred with the authors and commenters “that we need definitive randomized trials of whether extreme lowering of Na is helpful, safe or useless.” He continued:
The growing body of evidence from epi studies is consistent in not showing any clinical benefit with lower Na intakes below 4 gms per day in several different populations ( healthy free living eg the PURE study, those with vascular disease eg in ONTARGET, in diabetics and now in heart failure ).
Surely guidelines cannot keep on ignoring the collective weight of evidence that very low Na consumption is associated with worse outcomes. Of course epidemiologic studies cannot necessarily indicate whether the excess events with low Na is causal. The only way forward is to conduct a series of RCTs to examine the impact of extreme lowering of Na in several different populations where lowering Na is being recommended.
While such trials are underway, it would be prudent to stop recommending extreme lowering of Na (say below 3 gms of Na per day) as one cannot exclude harm and there is no evidence of benefit.
The issue is no longer who is right, but what is right.