Skip to main content

Salt, health and nutritional dogmatism

Salt may be a health issue now but the two most famous salt wars involved the City of Perugia versus Pope Paul III over a salt tax introduced by the Vatican in 1540 and then the Mexican-US salt wars or El Paso salt wars of 1877 over ownership of salt rich lakes in Texas. Today’s salt war is fought on the plains and hills of public health nutrition and is put under the microscope in a series of papers in the International Journal of Epidemiology. The central paper, from researchers at Boston and Columbia Universities, examines the objectivity of both camps in this public health feud[1].

It began with a search of the scientific literature for all reports relating salt to health. A total of 269 such reports were found of which two thirds were simply comments or letters in learned journals, a quarter were primary studies in which the hypothesis of a salt health link was directly tested in humans and the remaining 9% were either guidelines or reviews of primary data  (systematic reviews). Of this body of evidence, 54% supported a link between salt and health, 33% were contradictory of such a relationship and 135 were inconclusive.

The authors used this large data set to answer a number of questions. The first was to analyse the data to ascertain if a bias in citation of scientific paper existed. By that is meant the following: Is an author who supports one side or other of the hypothesis more like to cite other studies supporting their stance and less likely to pay attention to studies which are at variance with their stance. The answer to that question is emphatically, yes. Those supporting the salt health link were twice as likely to prefer to cite similar supporting articles. Among those who contradict the salt-health hypothesis, they were three times more likely to cite contradictory than supportive papers and among those who were inconclusive, the bias factor was all of 15 fold. What this means is that scientists in this field are very definitely not objective in their view of the problem and presumably its solution. A second question was the extent to which the literature is dominated by a small or large body of papers. In fact, the authors found that a small number of authors and a few reports do, in fact, dominate the debate. Thirdly the authors sought to see if there was consistency in the selection of primary research data. In fact there wasn’t. Primary data, which are the hub of objectivity, were selected according to the prevailing bias of each camp.

John Ioannidis from Stanford University was asked to not only write a comment but to seek similar comments from those in the field[2]. He immediately encountered the dogmatism that dominates public health discussion. One invitee who declined to add a commentary wrote thus: “ the paper ….is rubbish…there doesn’t seem to be any realization that the majority of those papers that are against salt reduction are funded by the food or salt industry, just like the tobacco industry did (or still does for that matter) for cigarettes….I wouldn’t have anything to do with it.” This belief that scientific bias is a function of industry funding is very naïve. Those who live in the EU will be familiar with the requirement in the majority of research funding, that industry involvement is a must. Without it, the grant doesn’t even get to review. The logic is that such research will generate knowledge, create jobs and lead to economic growth. “Blue skies” research is funded with no industrial requirement but this is a small fraction of the research spend. Society makes the rules and scientists play the game according to those rules.

The philosopher Thomas Kuhn understood the way science works. He pointed out that it exists as “normal” science or “revolutionary” science. Revolutionary science is a rare occurrence such as the discovery of the double helical nature of DNA and its transforming effect on our knowledge of cell division. Normal science is the antithesis of revolutionary science and it utterly dominates the scientific landscape. So scientists “defend” the paradigms of normal science and continue to do so until, from nowhere, emerges some new revolutionary finding, which, forever, changes the science landscape.

  • In the salt and health war, honest scientists have for one reason or another looked at the landscape and decided to support or contradict the thesis. Or did they? Is it possible that they first opted for one side or the other, and then sought the scientific basis to support these views? The Nobel Laureate in economics Daniel Kahneman would argue the latter. He points out that our brain operates the decision making process in two systems.
  • System 1: Fast, automatic, frequent, emotional, stereotypic, subconscious
  • System 2: Slow, effortful, infrequent, logical, calculating, conscious

System 2 decides first on which side of the salt (and general health) debate we reside. System 1 now bolsters this with analytical data.

Ioannidis argues that the scientific morass of salt and health cannot be resolved by reviewing the existing data no matter how fancy the statistical approach of that review might be. Primary data, which studies real human beings in real life settings, are what counts. As he points out, the design of real primary studies can be such as to more likely to lean in one direction than another.  The loser is Joe Soap who eventually does what the US electorate is presently doing. They will cry: ”A plague on both your houses” and go their own way. Scientific dogmatism is destroying science. Dissent is the oxygen of science and to belittle it and its proponents, as “industry hacks” is sad. These days the noisiest and most vociferous media friendly scientists win.

More on this in my new book: ”Ever seen a fat fox ~ Human obesity explored” due out in early May

[1]Trinquart L et al (2016). Why do we think we know what we know? A metaknowledge analysis of the salt controversy. Int J Epid, February 17th
[2]Ioannidis JPA (2016) Commentary: Salt and the assault of evidence. Int J Epid February 17th


Popular posts from this blog