Causation and Correlation

A new study has been published in JAMA showing that the supposed correlation between salt consumption and heart disease actually works in reverse. In other words, the more salt you ingest, the lower your risk of heart disease. It's been getting a bit of coverage (NYtimes, USA Today, CNN).

The study raises two important points. The first is that the authors only looked at a group with relatively low risk of high blood pressure. This leads to the plausible recommendation that if you're a person with high blood pressure or high risk of high blood pressure, then consuming high amounts of salt might still be considered risky, but if you're a person with normal to low pressure, then consuming more salt might be beneficial. Additionally, they note that people of African descent are more sensitive to sodium, and thus should also curb their salt intake. Too often health advice is framed in terms of "all people should do x," but there needs to be more emphasis on the different dietary needs among people. This one study is insufficient to prove that people with low blood pressure should maintain high salt consumption, whereas those with high blood pressure should curb their salt intake, but it raises the possibility.

The second important point is that it's not at all clear how strong the causal connection between sodium and heart disease is. Increased salt intake increases blood pressure. This is uncontroversial. People with higher blood pressure are at a higher risk of heart disease. This also is uncontroversial. But, the question is: is there a causal relationship between high blood pressure and heart disease. Namely, if your blood pressure rises for whatever reason, is your chance of getting a heart attack increased? Or is it perhaps the case that some underlying factor or set of factors increases your risk of heart disease and increases your blood pressure and that increases in blood pressure by some means (for example, consuming more salt) have no effect on heard disease risk? It's instructive that Ralph Sacco, who disagrees with this new study and thinks we should restrict salt intake, responds by saying "There are good randomized, controlled studies the gold standard of scientific studies that show a lower sodium diet has a meaningful effect on blood pressure." He doesn't say there are randomized controlled trials showing that increasing your salt intake increases your risk of heart disease, because there are aren't. The NY times article points out that the research on this question is mixed, and that such a study would be necessary to resolve the debate. In fact, Dr Alderman in his own study, according to the NY Times article, showed that even people with high blood pressure had a higher risk of mortality from all causes on a low salt diet.

At the end of the day we have to admit we just don't have strong evidence. Many groups will continue to advocate low salt diets and, though you should take their recommendation with a grain of salt (pun intended), you might want to follow their advice nonetheless.

A new study has been published in JAMA showing that the supposed correlation between salt consumption and heart disease actually works in reverse. In other words, the more salt you ingest, the lower your risk of heart disease. It's been getting a bit of coverage (NYtimes, USA Today, CNN).

The study raises two important points. The first is that the authors only looked at a group with relatively low risk of high blood pressure. This leads to the plausible recommendation that if you're a person with high blood pressure or high risk of high blood pressure, then consuming high amounts of salt might still be considered risky, but if you're a person with normal to low pressure, then consuming more salt might be beneficial. Additionally, they note that people of African descent are more sensitive to sodium, and thus should also curb their salt intake. Too often health advice is framed in terms of "all people should do x," but there needs to be more emphasis on the different dietary needs among people. This one study is insufficient to prove that people with low blood pressure should maintain high salt consumption, whereas those with high blood pressure should curb their salt intake, but it raises the possibility.

The second important point is that it's not at all clear how strong the causal connection between sodium and heart disease is. Increased salt intake increases blood pressure. This is uncontroversial. People with higher blood pressure are at a higher risk of heart disease. This also is uncontroversial. But, the question is: is there a causal relationship between high blood pressure and heart disease. Namely, if your blood pressure rises for whatever reason, is your chance of getting a heart attack increased? Or is it perhaps the case that some underlying factor or set of factors increases your risk of heart disease and increases your blood pressure and that increases in blood pressure by some means (for example, consuming more salt) have no effect on heard disease risk? It's instructive that Ralph Sacco, who disagrees with this new study and thinks we should restrict salt intake, responds by saying "There are good randomized, controlled studies the gold standard of scientific studies that show a lower sodium diet has a meaningful effect on blood pressure." He doesn't say there are randomized controlled trials showing that increasing your salt intake increases your risk of heart disease, because there are aren't. The NY times article points out that the research on this question is mixed, and that such a study would be necessary to resolve the debate. In fact, Dr Alderman in his own study, according to the NY Times article, showed that even people with high blood pressure had a higher risk of mortality from all causes on a low salt diet.

At the end of the day we have to admit we just don't have strong evidence. Many groups will continue to advocate low salt diets and, though you should take their recommendation with a grain of salt (pun intended), you might want to follow their advice nonetheless.

Causation and Correlation