In my last post (Blog 51) I explained how the Institute of Medicine (IOM) has been re-evaluating the evidence on the relationship between salt and health. This was initiated because of growing doubts about the current recommendations to reduce the intake of salt and emerging concerns that a low salt intake may actually be damaging to health. I have now been looking at some of the relevant scientific papers and here is some of the information I have uncovered.

Essentially the justification for the present advice to lower salt intake is that such a measure will result in a reduction in blood pressure (BP). Since it is accepted that BP is a risk factor for stroke and for cardiovascular diseases (CVD) generally the assumption is made that salt reduction will reduce the chances of developing these conditions. According to one review, a decrease of 3 gm salt/day would lower the systolic BP (SBP) by between 3.6 and 5.6 mm Hg and the diastolic BP (DBP) by between 1.9 and 3.2 mm Hg in hypertensive patients. For normotensive individuals the reductions were much smaller (1).

However these values are critically dependent on the selection of the research results used to make the assessment. There is certainly not universal agreement that the intake of salt is a factor which has a major effect on the BP. For example, in the Scottish Heart Health Study the salt intake was determined by analysis of urine samples which is very much more accurate and reliable than that obtained from food consumption surveys. A total of 7354 men and women participated and they were from 22 different districts throughout Scotland. It was found that the strongest correlations for SBP in both men and women were with age and Body Mass Index (BMI). For DBP age and BMI were much less important. There was a weak positive association between the sodium excretion and BP. However it was also found that there was a weakly negative association between potassium excretion and BP.

The authors concluded that:

“These findings, which in general agree with those from other studies within populations, show that the association of sodium and blood pressure is weak and does not have any real role in explaining bloods pressures. The study was large enough to detect an effect of any importance”


“The most likely explanation of our findings, however, is that the true association between sodium and blood pressure is extremely weak” (2).

A meta-analysis based on 58 trials of people with hypertension concluded that a reduction of 6.8 gm salt/day reduced the SBP/DP by 3.9/1.9 mm Hg. In 56 trials with normotensive individuals the reductions were 1.2/0.26 mm Hg. It was the opinion of the researchers that such a small reduction in the BP of those who were normotensive does not support a recommendation for the population as a whole. These analyses also showed that when the sodium intake is reduced there is an increase in the levels of renin and aldosterone. The fact that raised levels of these constituents have been linked to increased death rates is another reason to question the advice to lower sodium intake in those with normal BP (3).

In a study conducted in Finland the sodium excretion was monitored over a 4-year period in patients with Type 1 diabetes. It was found that there was a gradual increase in the all-cause death rate above an intake of about 10 gm salt/day. However it was also found that at low intakes of salt (less than about 3 gm/day) there was very large increase in the all-cause death rate (4).

The sodium excretion was measured in 28,880 patients with
CVD between November 2001 and May 2004. Volunteers were from 733 centres located in 40 different countries. During this time there were 3430 deaths of which 2057 were attributed to a CV cause. The results in Table 1 show that compared to the Reference (equivalent to intakes of 10-15 gm salt/day) there were increases in the death rates at low intakes and at high intakes. This is further evidence that a low intake of salt is associated with a raised death rate. In the view of these authors the raised death rates occurred when the intakes of salt were less than 7 gm and more than 17.5 gm/day (5).

TABLE 1 Relative death rates due to all-cause and to CV

Sodium excretion,gm salt/day <5.0 5-7.5 7.5-10.0 10.0-15.0 15.0-17.5 17.5-20.0 >20.0
All-cause 1.42 1.27 1.11 1.00 1.11 1.08 1.50
CV 1.75 1.39 1.13 1.00 1.10 1.58 1.87


This investigation is in line with earlier research which suggests that a low intake of sodium may not be conducive to good health. In particular it raises further doubts about the validity of the official advice. Here in the UK it is recommended that the maximum salt intake should be reduced from an average of about 9 gm/day to 6 gm/day. If this is achieved it will mean that many people will have an intake which is much lower than the 7 gm/day which according to the above study could well be dangerous. What is even more worrying is that the National Institute for Health and Care Excellence (NICE) has recommended that the intake should be lowered to 3 gm salt/day by 2025. It claims that this would result in 14-20,000 fewer deaths from CVD every year and savings of £350 M in healthcare costs. In the light of the evidence I have cited here and of course the report by Institute of Medicine in the USA (7), it is difficult to understand how NICE can persist with this stance. Either they do not bother to read the scientific literature or they conveniently disregard anything that does not agree with own pre-conceived views.

When recommendations are being made which carry the full authority of governments then it is essential that they are fully supported by sound scientific research. There certainly should be no room for any doubts. Unfortunately this is not the case and the effect is to undermine totally any possible confidence in the process!


  1. F J He and G A Mac Gregor (2003) Hypertension 42 (6) pp 1093-1099
  2. W C S Smith et al (1988) British Medical Journal 297 (6644) pp 329-330
  3. NB A Graudal et al (1998) Journal of the American Medical Association 279 (17) pp 1383-1391
  4. M C Thomas et al (2011) Diabetes Care 34 (4) pp 861-866
  5. M J O’Donnell et al (2011) Journal of the American Medical Association 306 (20) pp 222-2238
  7. Available at