Moderate Sodium Intake Tied to Lower CVD Events: PURE

Marlene Busko

August 13, 2014

HAMILTON, ON — Risk of death or major cardiovascular events over 3.7-years of follow-up in the massive  Prospective Urban Rural Epidemiology (PURE) study can be explained by a J-shaped curve, investigators report, at least when dietary sodium intake is estimated by a fasting morning-urine sodium test—a method that many criticize as ill-suited to the purpose[1]. The risk was lowest for people with a moderate sodium urinary excretion (a surrogate for estimated dietary sodium intake) of 3.00 to 5.99 g/day.

A related analysis[2], also based on more than 100 000 people who participated in PURE, found that the link between sodium excretion and blood pressure ranged from strong in participants with high sodium excretion to nonsignificant in those with low sodium excretion, and increased potassium appeared to be protective.

PURE is the largest study to examine sodium intake and clinical outcomes in a general population, Dr Martin O'Donnell (McMaster University, Hamilton, ON, and National University of Ireland, Galway) told heartwire . "It confirms prior observations that high salt intake increases the risk of cardiovascular events, [and] "it appears that the lowest risk is the consumption of salt in moderation rather than the sort of excessive lowering of salt intake [such as that recommended by American Heart Association (AHA) guidelines]," he said.

"In our cohort, less than 5% of the people were consuming what AHA guidelines recommend (<2.3 g or 1.5 g/day)," he noted. Clinicians and policy makers agree that a high salt intake is bad and increases the risk of CVD outcomes, he continued, "but whether excessive efforts need to be put out to try to reduce the entire population down to a level of sodium intake that very few people consume . . . there's a very big question mark over that."

The studies are published in the August 14, 2014 issue of the New England Journal of Medicine. As reported by heartwire , some of these analyses from PURE were presented by Dr Salim Yusuf (McMaster University) at a debate at the European Society of Hypertension 2013 Scientific Sessions and in an oral presentation at the European Society of Cardiology 2013 Scientific Sessions.

Estimated Dietary Sodium and Impact on BP, CVD Events

Whether excessive efforts [are needed] . . . to reduce the entire population down to a level of sodium intake that very few people consume . . . there's a very big question mark over that.

Lowering sodium intake might reduce hypertension, which affects an estimated one billion people worldwide and is a leading cause of death, stroke, MI, congestive heart failure, and chronic renal impairment, O'Donnell and colleagues write.

PURE is a large, international epidemiologic study that was designed to understand risk factors for CVD and diabetes. It enrolled adults aged 35 to 70 in 17 low-, middle-, and high-income countries on all continents.

The first study showed that for each 1-g increase in estimated sodium excretion, there was a 2.11-mm-Hg increase in systolic blood pressure and a 0.78-mm-Hg increase in diastolic blood pressure. The slope was steeper for people with higher sodium intake or hypertension or older age. Potassium excretion was inversely linked with systolic blood pressure.

The second study found that a composite outcome of death and major cardiovascular events occurred in 3.3% of the 101 945 participants during a mean follow-up of 3.7 years. Close to half of the participants (45.8%) had an estimated sodium excretion of 4.00 to 5.99 g/day (reference range). Fewer participants had an estimated sodium excretion of <3.00 g/day (10.6%), 3.00 to 3.99 g/day (20.7%), 6.00 to 6.99 g/day (12.1%), or >7.00 g/day (10.8%).

Compared with the reference range, those with the highest sodium excretion (>7.00 g/day) had a 15% increased risk and those with the lowest sodium excretion (<3.00 g/day) had a 27% increased risk of the composite outcome.

How Valid are the Sodium Estimates and J-Shaped Curves?

Asked to comment, Dr Lawrence J Appel (Welch Center for Prevention, Epidemiology and Clinical Research, Johns Hopkins University, Baltimore, MD), questioned the way dietary sodium was imputed from early-morning urine samples and was skeptical about the J-shaped curve.

The recently published Trials of Hypertension Prevention (TOPH) study, which he coauthored[3], "was the only observational study with multiple 24-hour urine collections [and] documented a direct progressive relationship [between estimated dietary sodium intake and CVD events and death] in the range [of sodium] usually consumed in the US," he said.

"I'm not at this point willing to say reduced sodium intake is harmful," he said. "Sample size does not always trump other factors."

You have to be careful about drawing inferences, much less changing policy.

The estimate of sodium intake is probably the biggest problem with this study, he maintains. The AHA has been very concerned about the validity of spot urine collections (used in the PURE study) as an accurate estimate of usual, long-term dietary intake, as described in a recent AHA Science Advisory, he notes[4]. "I know it's challenging, but we recommend 24-hour urines on multiple occasions, not spot collections."

Although repeated measures of sodium (and potassium) in a 24-hour urine sample would be ideal, this is not practical in large observational studies, O'Donnell counters. Instead, in PURE, they analyzed fasting morning urines and used a Kawasaki formula to estimate 24-hour sodium excretion.

These were not spot urines, he emphasizes, in the same way that a fasting blood glucose is not a random blood glucose. Moreover, they recently published a study showed that the intraclass correlation coefficient between estimated 24-hour urine sodium (from a fasting morning-urine sample) and actual 24-hour urine sodium was 0.71, which is better than the correlation between office blood pressure and 24-hour ambulatory blood pressure[5].

However, he agrees that this does not estimate usual sodium intake, which would require several assessments over time.

The potassium measurement was even more questionable, Appel said. "I was shocked that they use the potassium urine excretion as a surrogate of dietary intake because it is so variable," he said. The authors do concede that "our approach is probably less reliable for estimating potassium intake than for estimating sodium intake, because the proportion of consumed potassium that is excreted in the urine is lower than the proportion of consumed sodium that is excreted."

When there is a J-shaped curve between a variable and risk of a poor outcome, "you have to put your antennae up," Appel continued, because a there may be a "hidden" explanation; for example, people with a very low body-mass index (BMI) can have worse outcomes than obese individuals because they are very sick. "You have to be careful about drawing inferences, much less changing policy," he said. "The best available evidence is that sodium reduction has beneficial effects on blood pressure and should lower the risk of cardiovascular disease [as shown in a recent meta-analysis[6]], and studies with flawed methods should not derail sound policy."

Follow a Healthy Diet, Avoid Surplus Salt

The researchers and Appel do agree on some implications. "The overall message is the importance of a healthy diet pattern [including fruits and vegetables that are high in potassium] and within that, avoiding high salt intake," O'Donnell said. "For high sodium intake . . . most of the risk is in people who had hypertension at the time they came into the study . . . and that's the group of people where it is most important to reduce their sodium intake; it's likely to be prudent to reduce their intake to the lower end of the range at which we observe the lowest risk of CV events."

"Does this [study] definitively prove [that the lowest risk is consumption of salt in moderation]? It doesn't. That would require a large randomized controlled trial with clinical outcomes."

Appel says, "I'm actually an advocate of doing everything you can. . . . Let's try to keep weight under control . . . [and] have a better diet with more fruits and vegetables."

Findings Argue Against Public-Health Policy to Lower Sodium Intake

Dr Suzanne Oparil (University of Alabama at Birmingham), writing in an accompanying editorial[7], points to major design-related weaknesses of PURE, including the lack of multiple, direct 24-hour urine measurements of sodium excretion.

"Nevertheless, this large study does provide evidence that both high and low levels of sodium excretion may be associated with an increased risk of death and cardiovascular-disease outcomes and that increasing the urinary potassium excretion counterbalances the adverse effect of high sodium excretion," she writes.

She calls for a randomized controlled trial and until then holding off on a public-health recommendation to lower sodium intake. "These provocative findings beg for a randomized, controlled outcome trial to compare reduced sodium intake with usual diet. In the absence of such a trial, the results argue against reduction of dietary sodium as an isolated public-health recommendation."

Appel, however, is doubtful that such a randomized trial will ever be undertaken.

For years, the PURE trial been praised for offering global insight into social, environmental, behavioral, biological, and genetic factors that contribute to the development of CVD in low-, middle-, and high-income countries, but the morning-urine sodium tests used for the blood-pressure analyses have drawn mounting criticism. At the HYPERTENSION 2014 meeting earlier this year, Dr Graham MacGregor (Wolfson Institute of Preventive Medicine, London, UK) crisply advised an audience member to "please let Salim Yusuf know that he should stop using spot urine analysis."

The main PURE study and its components are supported by the Heart and Stroke Foundation of Ontario, the Population Health Research Institute, the Canadian Institutes of Health Research, unrestricted grants from several pharmaceutical companies (with major contributions from AstraZeneca [Canada], Sanofi [France and Canada], Boehringer Ingelheim [Germany and Canada], Servier, and GlaxoSmithKline and additional contributions from Novartis and King Pharma), and various national or local organizations in participating countries. O'Donnell reports receiving lecture fees from Boehringer Ingelheim, Bayer, Bristol-Myers Squibb, and Pfizer. Disclosures for the coauthors are listed in the article. Appel received a research grant from the McCormick Foundation. Oparil reports research support from AstraZeneca, Duke University, Merck, the National Heart, Lung, and Blood Institute, Novartis, Takeda, and Medtronic. She receives other research support from Daiichi-Sankyo, Medtronic, and Vivus and serves as a consultant/advisor to Backbeat, Bayer, Daiichi-Sankyo, Medtronic, Novartis, and Pfizer


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.