ACCEL | PURE-Sodium: Are Efforts to Curb Salt Science or Faith-Based Medicine?

It's a matter of faith that Americans consume too much sodium and that's hurting American health. It's more "faith" than science because the data have not been convincing, despite advice from several professional organizations recommending a wholesale curbing of salt in the manufacture and preparation of food.

Currently, the average consumption of sodium (Na) in the United States is about 3.5 g/day. (See the TABLE for the amount of sodium in a given amount of table salt.) The 2010 Dietary Guidelines for Americans recommend reducing Na consumption to <2.3 g/day. The same guidelines suggest that people 51 years and older, African Americans, or those who have high blood pressure (BP), diabetes, or chronic kidney disease—about half the US population and the majority of adults—should further reduce sodium intake to 1.5 g/day. (For this higher-risk population, the guidelines add that these individuals should meet the potassium recommendation of 4,700 mg/day.) Is this 35-65% reduction in Na consumption for millions of Americans necessary, safe, and feasible? Is there evidence that such lowering of daily Na benefits health?

Table. Approximate Amount of Sodium for a Given Amount of Table Salt

Amount of Table Salt


Amount of Sodium

1/4 teaspoon


575 mg

1/2 teaspoon


1,150 mg

3/4 teaspoon


1,725 mg

1 teaspoon


2,300 mg

Mente et al. raised these questions and looked at whether nationwide efforts to reduce dietary sodium is science or zealotry.1 One of the most influential studies was INTERSALT, which reported a weak relationship between Na intake (as measured by 24-hour urine) and BP: there was a reduction of 0.94/0.03 mm Hg of blood pressure per gram of Na reduction.2 Andrew Mente, PhD, and his colleagues noted that an equally well-conducted study from Scotland, published side by side with INTERSALT, showed no significant association between Na excretion and BP, "yet received little attention, illustrating the biases with which papers are selectively emphasized."3 (There was an association between potassium excretion but not Na excretion.)

The primary basis for the current American Heart Association (AHA) guidelines as well as the aforementioned 2010 national dietary guidelines was the 2001 DASH study, which was a 30-day "proof of concept" study as to whether changes in multiple aspects of diet (including Na reduction) would lower BP under controlled situations (all meals were provided to the participants and their spouses).4 They wanted to reduce the level of sodium from the average intake in the United States at that time (approximately 150 mmol/day, which is equivalent to 3.5 g of sodium or 8.7 g of sodium chloride) to below the recommended upper limit of 100 mmol/day. They reported that large reductions in Na intake (1.8 g/day) lowered BP (by 4.9/2.6 mm Hg), but the effects were more modest (3.0/1.6 mm Hg) in those who consumed an otherwise healthy diet but with less of a reduction in daily Na intake.

Recently, investigators evaluated the incidence of all-cause mortality and cardiovascular disease events in populations exposed to dietary intakes of sodium:5

  • low sodium <115 mmol
  • usual sodium; evaluating both a "low" usual sodium of 115-165 mmol and "high" usual sodium of 166-215 mmol
  • high sodium >215 mmol

Data from 23 cohort studies and two follow-up studies of randomized controlled trials (N = 274,683) showed that both low sodium intakes and high sodium intakes are associated with increased mortality, consistent with a U-shaped association between sodium intake and health outcomes.

And Then There's PURE-Sodium
Measuring 24-hour urine is the ‘gold standard' for measuring Na and potassium (K) intake, but it's not feasible in large studies. Fasting morning urine has been used to estimate 24-hour urine excretion using a mathematical formula. So, Dr. Mente and colleagues used this mathematical model in their Prospective Urban Rural Epidemiological (PURE) Sodium study.

Urine was collected from almost 100,000 individuals in 17 countries and BP was measured using an automated sphygmomanometer. They estimated Na and K excretion using the Kawasaki formula, validated against 24-hour urine obtained on the same day in 649 individuals.

Sodium excretion ranged from a low of 3.8 g/day in Malaysia to 6.1 g/day in China. (North American and Europe were at the low end of daily intake, just slightly higher than that seen in Malaysia.) Each 1 g (43.5 mmol) increase in Na intake was associated with a 2.2 mm Hg increase in systolic BP and a 0.8 mm Hg increase in diastolic BP. In commenting on the study, Prof. Giuseppe Mancia, first author of the recent European Society of Cardiology (ESC) guidelines for managing patients with hypertension, said that one important finding of PURE-Sodium was confirmation of the positive relationship between sodium and blood pressure. Indeed, he said, the relationship is strengthened by the huge amount of data obtained in the study.

However, Dr. Mancia said, the relationship was not uniform across subgroups analyzed. The relationship between sodium and blood pressure is tight when BP is elevated and sodium intake is high, but strikingly less so when BP is normal and sodium intake is lower; the same was true for older versus younger subjects, with a much closer relationship between sodium and blood pressure among older individuals. For example, at low-to-moderate levels of sodium intake (<3 g/day [low] or between 3 to 5 g/day [moderate]), the relationship with blood pressure was flat, but systolic BP moved sharply higher once Na intake exceeded 5.0 g/day.

The clear-cut clinical implication, he said, is that as far as BP control is concerned, one gains little by pursuing a low-sodium diet strategy in the general population, while there can be important advantages in focusing this intervention on subgroups with specific demographic, dietary, and clinical characteristics.  Another critical element: benefits are greatest when a reduction in sodium intake can be matched with an increase in potassium intake, since the study showed a strong negative association between K intake and BP values.

Dr. Mancia did add this caveat: given its epidemiological nature, the study did not measure the effect of changing sodium and potassium intake within individuals. This is inferred by the BP differences between individuals.

According to Dr. Mente, PURE-Sodium suggests a targeted approach to Na reduction is more appropriate for populations with moderate intakes (such as North America and Europe), whereas in countries with higher Na intake (e,g., China) a population strategy may be more reasonable. Indeed, he said using a population-wide approach for sodium reduction in a moderate-intake country, such as the United States, would be impractical, having only a modest effect for the money and effort expended.

The study data indicate that despite all the effort in the United States to reduce population-wide levels of sodium intake, fewer than 5% of the US population have a daily intake <2.3 g/day, the threshold used in most guidelines for average healthy adults. Based on the PURE-Sodium study, at current levels of sodium intake in this country a population-wide effort would have minimal effect on BP.


  1. Mente A, O'Donnell MJ, Yusuf S. Am J Hypertens. 2013;26:1187-90.
  2. Intersalt Cooperative Research Group. BMJ. 1988;297:319-28.
  3. Smith WC, Crombie IK, Tavendale RT, et al. BMJ. 1988;297:329-30.
  4. Sacks FM, Svetkey LP, Vollmer WM, et al. N Engl J Med. 2001;344:3-10.
  5. Graudal N, Jürgens G, Baslund B, Alderman MH. Am J Hypertens. 2014 Apr 26. [Epub ahead of print]

To listen to an interview with Andrew Mente, PhD, on the results and implications of the PURE sodium study, scan the code. The interview was conducted by Spencer King, III, MD.

Keywords: Follow-Up Studies, Nutrition Policy, Blood Pressure, Sodium Chloride, Dietary, Sodium Chloride, Potassium, Diet, Sodium-Restricted, Sodium, Dietary, Sphygmomanometers, Models, Theoretical, Hypertension, United States, Diabetes Mellitus, Renal Insufficiency, Chronic

< Back to Listings