Follow me on

The Evidence in Favour of Reducing Sodium Intake is Less Than Conclusive

sharma-obesity-salt-shaker1Regular readers may well be aware that I spend the first 10 years of my research career studying the effects of salt on blood pressure. Although I am no longer directly involved in sodium research, I continue to follow this issue with keen interest.

Recently, I commented on the  Institute of Medicine’s report that there is little hard evidence on which to base a recommendation to reduce sodium intake to less than 2.3 grams per day to reduce cardiovascular risk.

In today’s post I want to draw readers’ attention to a special issue of the American Journal of Hypertension dedicated to responses and thoughtful commentaries on the Institute of Medicine’s findings.

In one article, David McCarron reminds us that in healthy individuals body sodium is indeed tightly homeostatically controlled, or as he puts it,

“These well-established physiologic pathways provide the biological basis of the fact that actual human sodium intake is set by physiology and not public policy.”

As McCarron points out, any decrease in sodium intake is countered by an increase in hormones like renin, angiotensin, and sympathetic activity, that seek to defend the body’s sodium stores. Interestingly enough, it is exactly these hormones that we block when we treat patients with ACE inhibitors, angiotensin receptor blockers, renin inhibitors, or beta-blockers – all with proven cardiovascular benefits. Stimulating these hormones and systems by restricting salt intake makes no physiological sense – indeed, it is the very opposite of what you may wish to biologically accomplish to reduce cardiovascular risk.

In another article, Salim Yusuf and colleagues point out that despite some studies showing a small blood pressure lowering effect of  sodium restriction,

“…none of these studies addressed the crucial question: Does lowering Na intake reduce CVD? In the absence of reliable randomized controlled trials (RCTs), some clues can be obtained from cohort studies examining the association between Na intake and CVD. Recently, 4 cohort studies with urinary estimates of Na consumption (which are better than dietary estimates) raised concerns that low Na intake (<3g/day) was associated with higher (or no lower) rates of CVD/mortality compared with moderate intake.”

With regard to feasibility and actual consumption, Yusuf and colleagues argue that,

“The average American consumes about 3.5g of Na per day, and this has remained constant for the last 50 years, despite guidelines and population-based efforts to reduce Na intake further. In addition, data from >30 countries have reported a similar level of Na intake. In the United States, only 9% of adults consume <2.3g/day, and just 0.6% consume <1.5g/day. Therefore, the human experience for very low levels of Na consumption is extremely sparse.”


“…in countries such as the United States, there have been marked reductions in CVD rates by about 50% over the last 25 years although Na intake has remained constant. So, Na reduction does not appear to be essential to reducing CVD.”

Yusuf and colleagues also note that there is no known benefit of further lowering their blood pressure in people with normal blood pressure (the vast majority of the population) – if anything, there may be some evidence of harm.

Where I wholeheartedly agree with Yusuf and colleagues, is in their statement that,

“The zeal to recommend extreme reductions in Na intake that are difficult to achieve in the entire population in the United States and other countries with moderate Na intake (e.g., <5 g/day) is a case of ideology replacing good science….As shown repeatedly in the past, well-meaning interventions based on insufficient science can mislead (e.g., hormone therapy, margarines, and total fat intake).”

Or, as the IOM committee chair Brian Strom stated:

“It’s not a question of studies showing benefit being better than those showing harm; there are no studies showing benefit.”

I do not expect for a minute that this state of affairs will in any way deter the anti-salt lobby to ease in their efforts to call for public health campaigns, policies, and other measures to convince people to eat less salt.

Indeed, the same issue of the American Journal of Hypertension features several “pro” articles, essentially rehashing the “epidemiological” case for salt restriction based on the blood pressure surrogate.

Thus, Jenifer Clap and co-authors, use the same age-old blood pressure argument, to make the rather dramatic but unsubstantiated assertion that,

“reducing sodium intake may be second only to controlling tobacco as the most effective intervention to prevent deaths from chronic disease.”

Even more interestingly, also for the “pro” side,  Lawrence Appel and Paul Whelton, fully acknowledge the shortage of conclusive evidence, but then make the surprising statement that,

“Flawed evidence should not derail sound policy.”

Perhaps it is this sort of thinking that does more to muddy the waters than the lack of better evidence.

Personally, I will continue limiting my recommendations to reduce salt intake to those who actually have high blood pressure and, even for them, it may be far easier to simply add a few milligrams of a diuretic to their meds.

Edmonton, AB


  1. I understand the message that reducing salt has been found to have a lower impact on blood pressure then was previously thought. However other lifestyle changes continue to be well supported in helping manage blood pressure.

    Studies have shown the majority of our salt intake ~80% is found in processed foods we eat rather than the salt shaker. Depending on a person’s cardiac risk factors, some dietary changes such as the DASH diet with fewer processed foods and more fruits and vegetables would have a variety of benefits and would naturally be higher in potassium and lower in sodium which have blood pressure benefits.

    Of course adding a few miligrams of diuretic is always easier however patients deserve the opportunity to explore information about lifestyle change. Avoiding further weight gain, getting some extra exercise, and manging stress can all help lower blood pressure. Medication may not be immediately required unless the blood pressure is signficantly elevated.

    Post a Reply
  2. “Flawed evidence should not derail sound policy??”

    Isn’t that just another way of saying “reality should not negate our foregone conclusions?”

    Evidence is evidence. I have to say I’m a little tired of when the evidence shows that salt/eggs/you name it is not the demon it’s made out to be, they acknowledge it, and follow it with, “But still don’t eat it! You’ll die!”

    Post a Reply
  3. This is an interesting and thought provoking artilcle the lack of proof should not mean that something is a fact. I wonder how much other reccamended daily intake stats are out there that are flawed with faux science.

    Post a Reply
  4. This lack of evidence that cutting the salt shaker improves blood pressure management is another suggestive clue that the harm from consuming processed food may derive from some other characteristic(s), not its salt content.

    Post a Reply
  5. Thanks for this Arya,
    I find it interesting how the bandwagon shifts according to the whims of the media. I don’t really have strong feelings either way on the sodium debate. Population wide sodium reduction is mostly a pipe dream anyway given the huge disconnect between current intakes and recommended targets.

    But, it’s pretty clear that sodium reduction lowers BP (actually a low sodium diet – which is technically not the same as sodium reduction). Certain groups experience clinically important BP reduction and others do not. However, while small differences in individual BP might not have much of an effect, small differences in population BP might.

    So, to those that contend we should discard the BP data, I would ask –
    1. we tell obese people to lose weight – where are the hard endpoint RCT data?
    2. some of us advocate for antiobesity drugs – where are the hard endpoint RCT data?
    3. ‘we’ have been telling people to eat a high carb, low fat diet for decades – where are the hard endpoint RCT data?
    4. we tell people to exercise – where are the hard endpoint RCT data?

    Post a Reply
  6. The advice to reduce sodium intake for people with high BP seems especially problematic to me. As you mention, it does not seem to affect mortality rate. Then, while patients focus on cutting down on salt, they don’t focus on something that could actually help (exercise, smoking cessation, alcohol moderation, etc. Though now I wonder if that advice is unproven too.). It gets worse when you look at renin / angiotensin / aldosterone and sympathetic activity as you mention. From memory, between 2/3 and 4/5 of people with hypertension have elevated r / a / a and s.a. Only a minority (the rest) are considered salt sensitive. And even among those, I would be curious to see who is really sensitive to sodium vs the rest of the crap that comes with sodium-rich foods. Finally, people taking medication for hypertension could be acutely hurt by cutting down on salt. Many of those drugs promote potassium retention and sodium loss already. Plus, “No Salt”, the salt replacer, is made of potassium chloride. It all seems like a very bad idea to me.

    Post a Reply
  7. The sodium and blood pressure issue reminds me of the “prevention paradox”–

    The link between quantification of risk and the targeting of health promotion interventions has been described by Geoffrey Rose (1981) as the prevention paradox. Prevention Paradox arises because many interventions that aim to improve health have relatively small influence on the health of most people. Thus for one person to benefit, many people will have to change their behaviour and receive no benefit from these changes. It leads to the use of two distinct and complementary strategies, the ‘high risk approach’ and the ‘population approach’. The pros and cons of these are summarised in Table 2.4 below.


    I think that there is conflicting evidence and confusion in nutrition science because of the prevention paradox–that many people need to adopt a health behaviour (e.g., reducing dietary sodium) in order that relatively few benefit (reduced blood pressure); and some individuals may not benefit from that particular behaviour (e.g., some individuals are not salt sensitive). By reducing the salt intake of everyone in the population, then the bell curve (of population’s blood pressure) shifts downwards and there can be a great reduction for most people (in blood pressure and in disease).

    From the link above, it states that population recommendations
    “Aims to lower the mean level of risk in the population and shift the distribution of risk”. This is an abstract concept in statistics and risk in the population, which might be difficult to understand for individuals. Diet recommendations are not as black and white as other recommendations such as wearing helmets, not smoking (where the prevention paradox also applies)…

    A diet high in vegetables, fruits, whole unprocessed foods, and lean proteins cooked in flavourful ways with herbs, garlic, onions, lemon, etc. is naturally lower in sodium content. Replacing high-sodium foods with these naturally-low-sodium foods should be the focus, rather than on choosing low-sodium products. Many other benefits come from increasing the intake of high-nutrient and higher-fibre foods.

    Post a Reply
  8. 1. RAAS blockers reduce CV risk in a subset of people with certain chronic disease. We do not know their effect in an otherwise healthy population, and cannot extrapolate to them.

    2. CVD rates have declined because of improvement in pharmacologic management of cardiovascular risk factors. These medications have side effects and are costly. We should still be seeking to find nonpharmacologic, population-based interventions that reduce CVD. The question of whether population-based sodium reduction decreases CVD remains an important one.

    3. There is a body of evidence that suggests increased sodium intake is linked to consumption of sugar-sweetened beverages and obesity.

    My take: Humans are hard-wired to get enough water and sodium. For low-sodium consumers, population-based efforts to reduce sodium intake will likely have little impact. However, there are many many people who are outliers and consume 2 times or 3 times more than the 3.5 g North American average. Population-based efforts to reduce sodium intake would affect these consumers most. Debate around sodium intake should focus on the potential benefits for this group of high-intake consumers.

    Post a Reply

Submit a Comment

Your email address will not be published. Required fields are marked *