–More questions raised about broad efforts to restrict salt; AHA condemns study
A large new analysis offers more evidence that broad salt restriction doesn’t benefit most people and may even harm some people. The study did find that salt restriction may be beneficial to the minority of people with high blood pressure who also consume high levels of salt. Though by no means a definitive study, the paper published online in the Lancet adds to growing concerns that population-wide efforts to restrict salt may be misguided. However, the American Heart Association said “the public should not be confused by the flawed study” and said it stands by its support for low-salt diets.
Average sodium consumption in the United States is 3,400 mg per day. The American Heart Association recommends that sodium levels be no higher than 1,500 mg/day, while some other organizations recommend sodium levels below 2,300 mg.
The new analysis combines data from more than 133,000 hypertensive and normotensive people who participated in four large prospective studies. Salt consumption was estimated by measuring sodium from a urine sample. For both hypertensive and normotensive people low levels of salt consumption below 3,000 mg per day were associated with an increased risk of cardiovascular disease and death. However, at the highest levels of salt consumption (more than 7,000 mg per day) there was an increased risk in the people with hypertension but no increased risk in people without hypertension.
“While our data highlights the importance of reducing high salt intake in people with hypertension, it does not support reducing salt intake to low levels,” said the lead author of the study, Andrew Mente, (McMaster University), in a McMaster press release. “Our findings are important because they show that lowering sodium is best targeted at those with hypertension who also consume high sodium diets.”
The authors recommend that population-wide efforts to lower salt consumption should only be undertaken in areas like central Asia or China where the average sodium intake is very high.
The Debate Goes On
Speaking on behalf of the American Heart Association, Elliott Antman (Brigham & Women’s Hospital), said he strongly disagreed with the methods of the study along with its conclusions:
“It is disappointing to see that a flawed methodological approach to the important question of the relationship between sodium intake and cardiovascular events is perpetuated in this latest publication. Use of spot urine specimens obtained first in the morning are not a good surrogate for the gold standard, which is serial 24 hour urine measurements. Thus from the very start there is an unreliable estimation of the 24 hour urinary sodium at baseline — that is, at entry into the studies that are pooled in this report. Furthermore, it is unrealistic to expect that a measurement of urinary sodium at entry into the studies that are pooled can provide a reliable estimate of the sodium intake during the course of followup of the studies. Thus we cannot have confidence in the relationships between sodium and outcomes that are reported in this paper. We cannot use the information in this paper to guide public policy. The AHA maintains its previous statements of the concerns about the approach taken in such analyses and continues to recommend less than 1500 mg of sodium per day for all sources.”
Andrew Mente and another study author, Martin O’Donnell, offered the following response to Antman:
“In developing public policy, large population studies are required to describe the relationship between sodium (salt) intake and health outcomes in the general population. While repeated 24-hour urine collections may be the reference standard for measuring sodium intake in individuals or small groups, they are completely impractical and arguably unnecessary for large population studies. Use of a fasting morning urine to estimate sodium intake in populations has been shown to provide reliable estimates (compared to 24-hour urines), and accepted by leading medical journals (Lancet, NEJM, JAMA) and the WHO as a practical approach to measuring intake in populations. The generalisability of studies using fasting urines is superior to 24-hour urine collection, as the latter is associated with high rates of non-completion. Use of a single measurement of a risk factor in populations is common in medical research. Major advances in our understanding of the importance of blood pressure, diabetes and cholesterol have been based on single measurements of these risk factors in large epidemiologic studies (e.g. Framingham, INTERHEART)-consider the consequences of ignoring those studies!
A recommendation for low sodium intake (<1.5g/day) in the entire population should be based on definitive proof of benefit, rather than merely attempting to disprove studies that contradict. We are unaware of any studies demonstrating that sodium intake of <1.5g/day is significantly associated with lowest cardiovascular rates in the general population (compared to moderate), but are aware of numerous studies (including those using 24-hour urine collections) that report a higher risk in those with low sodium intake. A meta-analyses of these previous studies report moderate sodium intake to be associated with lowest cardiovascular risk, even when studies using ‘spot’ urine measurements are excluded. The current recommendation for low sodium intake is based on an assumed cardiovascular benefit, inferred from small blood pressure trials. Current evidence, from international studies conducted over the past 6 years, suggest that those assumptions may be flawed, and guidelines need to be updated to reflect contemporary evidence.”
Previous Coverage of the Salt Debate:
- Salt, Science, And The American Heart Association’s Double Standard
- Get Rid of Sugar, Not Salt, Say Authors
- An Expert’s Perspective: Why Salt Is Not Like Tobacco And Why Guidelines Are Tricky
- Salt Report From IOM Sparks Much Heat, Only A Little Light
- Why Guidelines Are Bad For Science
- Why Guidelines Should Be Waged Like War
- New Studies Fuel The Debate Over Sodium