Monday, August 03, 2015

Salt: killer or scapegoat?

"New Scientist" is fighting a rearguard action to reduce salt in our diet, despite a lot of evidence that salt does no harm.  I reproduce the key passages below.  I will add my comments following that

In 2009, cardiologist Francesco Cappuccio of the University of Warwick, UK, pooled all the data and found a strong relationship between a salty diet and cardiovascular disease (BMJ, vol 339, p b4567).

 Another way is to intervene directly in people's diets - take two groups of people, get one of them to eat less salt for a while and see what the outcome is. These trials take more work than observational studies but several have been done. The biggest managed to get thousands of people to cut down on salt by about 2 grams a day for up to four years and saw a 25 per cent fall in cardiovascular disease (BMJ, vol 334, p 885).

 Or you can look at whole countries, taking the before-and-after approach. Fifty years ago northern Japan had one of the world's biggest appetites for salt - an average of 18 grams a day per person - and shockingly high numbers of strokes. The government implemented a salt reduction programme and by the late 1960s average salt consumption had fallen by 4 grams a day and stroke deaths were down by 80 per cent.  Finland, another salt-guzzling nation, achieved similar gains in the 1970s.

However, the evidence is not always so clear. In July the Salt Institute was presented with its biggest PR coup for years when the Cochrane Collaboration, an internationally renowned body dedicated to assessing medical evidence, published a long-awaited study on salt and cardiovascular disease. As is usual for Cochrane, the study was a "meta-analysis", pooling the results of all the best-designed randomised controlled trials that have been done, the highest standard of proof in medicine.

Seven trials met the quality criteria, with over 6000 subjects in total. The analysis did show that people who cut back on salt have slightly lower blood pressure and are less likely to die from heart attacks and strokes. But, crucially, the effect on deaths wasn't big enough to be statistically significant. The Cochrane team could not rule out the possibility that the reductions had happened by chance.

The research was published simultaneously by Cochrane and the American Journal of Hypertension (vol 24, p 843), whose editor-in-chief Michael Alderman is a long-time critic of salt reduction. In an accompanying editorial (vol 24, p 854), Alderman, who was once a paid consultant for the Salt Institute, repeated his oft-stated claims that there is not enough evidence for salt reduction. Sensing a story, many newspapers ran with his line.

 Is Alderman correct? Not surprisingly, MacGregor thinks not. For one thing, he claims the Cochrane study is flawed. When he reanalysed the same data in a slightly different way, he found a reduction that was statistically significant (The Lancet, vol 378, p 380). Alderman criticises this as "salami epidemiology", but even in the original analysis the link between salt and death rates only just slipped below statistical significance. Far from casting doubt on salt reduction, some argued that the findings supported it.

The Cochrane report wasn't the end of it. Last month Alderman's journal published a further meta-analysis purporting to show that salt reduction could actually be harmful (doi:10.1038/ajh.2011.210). It concluded that while cutting salt lowered blood pressure, blood levels of certain hormones and lipids were increased, which could theoretically raise cardiovascular risk.

But many of the studies included in the analysis lasted just a few days and involved big salt reductions. MacGregor accepts that sudden and steep salt reduction can lead to counterproductive hormonal changes, but says that modest reductions, say from 8 to 6 grams, do not. "There's no evidence whatsoever that a modest reduction does any harm," he says.

 Even the chief author of the Cochrane study, statistician Rod Taylor at the Peninsula Medical School in Exeter, UK, agrees with MacGregor that the findings lend further support to salt reduction. "Our results do not mean that asking people to reduce their intake of salt is not a good thing," he says. "We have much stronger evidence for salt than we do for fat, for the benefits of eating fruit and vegetables or losing weight," argues MacGregor.


Most of the studies  that found harm from salt were epidemiological and what such studies show is always contestable.   Let me illustrate:  The authors above use Japan to argue for salt reduction.  I can use Japan in exactly the opposite way.  Even after the various anti-salt campaigns, the Japanese are still huge salt consumers.  If ever you have tasted Japan's favourite sauce -- soy sauce -- you will know why.  It's almost solid salt.  Yet Japan is renowned for it multitude of centenarians and long life generally.  So, on that evidence, heavy salt consumption is clearly not harmful and may be beneficial.

So in that context the Cochrane study is all-important.   It filtered out the most contestable findings and zeroed in on the findings that are least contestable.   And that study showed no statistically sigificant harm from salt ingestion.

"New Scientist" acknowledges that but argues that a different analysis of the Cochrane data DOES produce statistical significance.  But to argue that way fails to understand what statistical significance does.  It compares a given result with what would happen by chance alone.  And if a result is on the borderline of conventional significance -- whether a bit above or a bit below -- hardly matters for policy decisions.  In any case, the effect is tiny.

If there were any kind of robust effect going on, statistical significance would hardly be worth calculating.   So the key thing that Cochrane showed was not the statistical  significance or otherwise of the result but rather that any effect from salt consumption was TINY -- and hence not worth bothering with.

And since anecdotes tend to be more persuasive than statistics, let me report that I have always put PLENTY of salt on my food  -- and yet my blood pressure is within the accepted safe range.  My blood pressure was up a bit once but I started doing a walk around the block most evenings and that fixed my blood pressure.  If you are worried about your blood pressure do some light exercise.  Exercise matters far more to your blood pressure than any trivial effect of salt. UPDATE: There's an article here which confirms the significant benefit of light exercise.

A final comment:  I liked the last sentence in the excerpts above:

"We have much stronger evidence for salt than we do for fat, for the benefits of eating fruit and vegetables or losing weight," argues MacGregor."

That rightly shows that all the other food fads are even more poorly founded.  LOL.

Further reading:  A big European study showed that LOW salt in your blood is most likely to lead to heart attacks.  See JAMA. 2011;305(17):1777-1785.  More here and here and here for similar findings.  Salt is harmless but a deficiency of it is not.  We need it.  See also here

Posted by John J. Ray (M.A.; Ph.D.).

No comments:


eXTReMe Tracker