So, I always cringe a little when I read medical conclusions in the mainstream media. And I'll be upfront. I'm probably biased on the side of medicine on this one. I'm a former nephrologist.
The one study this person mentions by name (and dismisses) is the DASH-Sodium Study.
"Like the previous study, it was based on a large sample (412 participants) and was a multi-center, randomized, outpatient feeding study where the subjects were given all their food."
"The DASH-Sodium study found that reductions in sodium intake produced significantly lower systolic and diastolic blood pressures in both the control and DASH diets. Study results indicate that the quantity of dietary sodium in the control diet was twice as powerful in its effect on blood pressure as it was in the DASH diet."
So, positive conclusions (lower salt -> lower blood pressure) in a seemingly well designed study. People were randomized, it was prospective, and people given all their food. So, change one thing and see what happens.
The author here seems to cite other studies to make the opposing case (low salt = bad), but doesn't give exact references. I did a quick google search and found this one: http://www.nytimes.com/2011/05/04/health/research/04salt.htm... But you can already tell it's an inferior study. "It involved only those without high blood pressure at the start, was observational, considered at best suggestive and not conclusive."
So, let's try to compare apples to apples.
He points out that the DASH study didn't conclude anything about death or cardiovascular disease. Yes, that's true. So sure, there's a chance that despite blood pressure reductions a low salt diet could cause problems. But it's not the obvious conclusion. Anyone have any solid studies on this?
"The associations between systolic pressure and sodium excretion did not translate into less morbidity or improved survival. On the contrary, low sodium excretion predicted higher cardiovascular mortality."
> He points out that the DASH study didn't conclude anything about death or cardiovascular disease. Yes, that's true. So sure, there's a chance that despite blood pressure reductions a low salt diet could cause problems. But it's not the obvious conclusion.
That's the whole point, you say it yourself. The idea that "salt is bad" has just as shaky underpinnings as "lower blood pressure is good". The only thing we care about is death or disease, not lowering blood pressure. So even if less salt = lower blood pressure, that doesn't tell us if this is good, or harmful.
The obvious conclusion is that there is no conclusion.
> The idea that "salt is bad" has just as shaky underpinnings as "lower blood pressure is good".
Wait, "lower blood pressure is good" has shaky underpinnings? Is there a debate on that? I don't think there's a debate that high blood pressure results in worse (death/disease) outcomes.
"Wait, "lower blood pressure is good" has shaky underpinnings? Is there a debate on that? I don't think there's a debate that high blood pressure results in worse (death/disease) outcomes."
There's no debate that high blood is associated with heart disease. There's always "a debate" if you try to go from correlation to cause. You always have to be careful this.
You have to show that a particular mechanism for lowering blood pressure results in better outcomes. You can't unambiguously say "we've lowered the blood pressure, we know the results will be better now".
"There's no debate that high blood is associated with heart disease. There's always "a debate" if you try to go from correlation to cause. You always have to be careful this."
I always thought that high blood pressure was caused by destruction of the lining of the arteries, and stiffening of the tissue, rather than water retention.
Control of blood pressure is dependent on this delicate balance of tissues against the pumping of the heart, as well as production of the substances that allow you to be sexually active.
Hence the fortuitous discovery of Viagra whilst searching for blood pressure control medications.
I acknowledge that jumping from a to b to c (low salt to low blood pressure to less death) has gaps in causality. But why do you not take objection to his uncited assertions that low salt intake and bad outcomes together imply causality?
So then it sounds like one might say that lower blood pressure isn't necessarily good unless you already have high blood pressure (or are on the cusp of having high blood pressure)?
Populations with low sodium intake, and other factors, have no hypertension, and low rates of death and disease from heart disease.
The randomized studies have found lower death and disease from lower sodium. So far the probability that this is just chance is about 10%. So the standard p number is not reached.
The most likely outcome is that a bigger better studies will produce significance.
Thank you for such precision concerning statistical semantics.
But I also find the non-experimental data compelling. I think more long term data will clear this up eventually.
It is urgent that we solve foundational debates in statistics. The health and economies of the world hang on this very point. So far Judea Pearl's work is the most compelling I've found. Do you have any references you'd suggest?
Because kidney function is damaged by pressure but its control mechanisms are governed by osmolarity (flow). A high pressure, low-flow kidney will send signals to the body to increase BP further, until the kidney essentially annihilates itself. You've got to unload the kidney down to a reasonable level, while keeping sufficient pressure to perfuse the heart and brain. In the setting of cardiovascular disease (stiff pipes with reduced cross-sectional area), the safe range becomes progressively narrower.
> I always cringe a little when I read medical conclusions in the mainstream media.
I usually do too, especially articles about nutrition which are notoriously unreliable due to the near complete lack of randomized, controlled, prospective studies and the abundance of observational studies, but this one doesn't seem too bad. In fact, the same author published a similar article in Science last millennium http://www.stat.berkeley.edu/users/rice/Stat2/salt.html so I don't think you can blame the mainstream media too much here if they're just printing an article similar to what would satisfy Science's editorial standards.
If you look at the numbers in the Wiki link you provided, the reduction in BP in DASH dieters was small in non-hypertensives and not exactly huge in hypertensives.
"Respectively, the DASH or “combination” diet lowered blood pressures by an average of 5.5 and 3.0 mm Hg for systolic and diastolic, compared with the control diet. The minority portion of the study sample and the hypertensive portion both showed the largest reductions in blood pressure from the combination diet against the control diet. The hypertensive subjects experienced a drop of 11.4 mm Hg in their systolic and 5.5 mm Hg in their diastolic phases."
I question whether the 5.5/3 mmHg is really worth eating bland low sodium food. Statistical significance does not necessarily imply clinical significance, and you're asking people to give up something pretty big. Willpower is a limited resource. We should have good reasons to ask people to use it up.
I'd agree. I have had times in my life where I have had to increase my salt consumption to avoid problems. These include everything from altitude sickness to dehydration caused by eating a diet too high in fruit.
However, at the same time I don't think that things are as clear-cut as in the article. One of the serious issues we find in current dietary trends is that we eat more prepared foods which tend to be high in salt, hydrogenated oils, and sugars.
So I personally side both with the idea that we have gone overboard in trying to reduce salt and also the idea that the modern American diet tends towards too much salt. I would also make the somewhat controversial argument that I think that anything we replace salt with will be even worse. The solution is to eat more fresh foods and fewer junk foods and prepared foods, not try to engineer some sort of low sodium equivalents.
there's a chance that despite blood pressure reductions a low salt diet could cause problems
For that matter, there's a chance that because of blood pressure reductions, a low salt diet could cause problems -- if the kidneys respond to low blood pressure by secreting renin, and renin causes more cardiovascular damage than high blood pressure.
I believe the contention is that studies like DASH show that changes in sodium consumption, rather than levels of sodium consumption, can effect blood pressure. That is, if you go onto a low sodium diet your blood pressure will fall, but then rise again to its previous level as your body adjusts.
Taubes has written a bit about salt before. Unfortunately, the only thing he's written that I can find that goes into more detail is this Science news focus article:
It talks about the DASH study and some others in a little more detail, but unfortunately it's 14 years old now, so I don't know what research has been done since.
arn, thanks for leading this one! Let me chime in that salt no doubt drives caloric intake. One need look no further than chefs and fast food to figure that out. So, in my mind, salt is a double whammy: BP and a secondary cause of weight gain, cholesterol, and everything else that comes with weight gain.
The one study this person mentions by name (and dismisses) is the DASH-Sodium Study.
http://en.wikipedia.org/wiki/DASH_diet#Study_Results
"Like the previous study, it was based on a large sample (412 participants) and was a multi-center, randomized, outpatient feeding study where the subjects were given all their food."
"The DASH-Sodium study found that reductions in sodium intake produced significantly lower systolic and diastolic blood pressures in both the control and DASH diets. Study results indicate that the quantity of dietary sodium in the control diet was twice as powerful in its effect on blood pressure as it was in the DASH diet."
So, positive conclusions (lower salt -> lower blood pressure) in a seemingly well designed study. People were randomized, it was prospective, and people given all their food. So, change one thing and see what happens.
The author here seems to cite other studies to make the opposing case (low salt = bad), but doesn't give exact references. I did a quick google search and found this one: http://www.nytimes.com/2011/05/04/health/research/04salt.htm... But you can already tell it's an inferior study. "It involved only those without high blood pressure at the start, was observational, considered at best suggestive and not conclusive."
So, let's try to compare apples to apples.
He points out that the DASH study didn't conclude anything about death or cardiovascular disease. Yes, that's true. So sure, there's a chance that despite blood pressure reductions a low salt diet could cause problems. But it's not the obvious conclusion. Anyone have any solid studies on this?