The debate over the perils of salty diets may be one of the most polarized in all of science.
On one side, scientists warn ominously that most Americans are killing themselves with salt. On the other, scientists insist most Americans are fine.
The inability to resolve this question may seem puzzling. It is a question with deadly consequences, at least potentially. How much salt is healthy? Given the marvels of technology, it seems like that ought to be an easy one.
Now a review of hundreds of papers on the topic indicates that the inability to reach a consensus stems at least partially from the fact that the two groups of scientists operate, in essence, in parallel scientific universes.
In one, the scientists write papers about the dangers of our salt consumption, and typically cite other papers that point to the same conclusion. In the other, the scientists write papers dismissing or minimizing the danger, and typically cite papers agreeing with their position.
Each side, in other words, steers away from taking into account contrary results.
“We found that the published literature bears little imprint of an ongoing controversy, but rather contains two almost distinct and disparate lines of scholarship,” according to the paper from researchers at Columbia University and Boston University, and published by the International Journal of Epidemiology.
The researchers offer one particular image that shows how polarized the salt debate has become.
The blotches of red and blue here represent instances where scientists cited like-minded research; those in green show instances where scientists referred to research that challenges their results.
In a more perfect world, where scientists sought balance in the evidence they reviewed, you would see more green - signs that scientists were considering evidence that is contrary to their beliefs.
As you can see, the image is dominated by red and blue, a sign that scientists are more likely to cite the research that conforms to their outlook. Overall the papers they reviewed were 50 percent more likely to cite reports that drew a similar conclusion than to cite papers drawing a different conclusion.
"This shows the polarization within the scientific community," said Ludovic Trinquart, who completed the paper with colleagues David Merritt Johns and Sandro Galea.
Trinquart and his colleagues also turned up another factor that might pose even more profound problems in the salt research. It appears scientists could not even agree on what ought to be counted as evidence.
This finding arose from their review of ten "systematic reviews" of the evidence that have been conducted. In systematic reviews, scientists collect all of the primary research on a topic and, in effect, weight it on the whole. But there appears to have been widespread disagreement about what research papers ought to be included in a systematic review. If a research paper was selected for one systematic review, it was more than likely not selected for another, the researchers found.
"There is no agreement or very poor agreement on what should be counted as evidence," Trinquart said.
Exactly what governments and other public health organizations ought to tell people about salt has been the focus of fierce debate in recent years. The U.S. government, through the Dietary Guidelines for Americans, as well as the American Heart Association, have long warned that most Americans are consuming far too much salt, and that excess consumption raises risks for high blood pressure, strokes and heart disease. However, research in recent years by some prominent scientists have raised doubts about those warnings.
While advocates of salt restrictions often disparage their opponents by saying the food industry has funded their efforts, in truth, some milestones in the research were not funded by corporate interests. The new research did not assess which research efforts had industry funding.
But the researchers said their review found evidence of bias arising from scientists reluctant to change their view in light of new facts.
"Our findings support a strong bias towards the status quo and the absence of a genuine scientific conversation where each side engages the other."