25 November 2009

Why Nutrition Researchers Are Rightly Considered Bad Scientists

And as I've harped on before, awful journalism doesn't help either.

A recent headline screams, "High Salt Intake Directly Linked to Stroke and Cardiovascular Disease". Well, that's pretty attention-grabbing. Seems like it should have "!!!" at the end. I'm intrigued by the "directly" statement. As far as I knew, observed dietary salt intake links with various maladies were observational and ambiguous.

So what does the article say? Here are some excerpts.

The link between high salt intake and high blood pressure is well established,
um... it is?

and it has been suggested that a population-wide reduction in dietary salt intake has the potential to substantially reduce the levels of cardiovascular disease.
that's true -- it certainly has been suggested.

[....]
Collaborative research conducted by Professor Pasquale Strazzullo at the University of Naples, Italy and Professor Francesco Cappuccio at the University of Warwick, UK analysed the results of 13 published studies involving over 170,000 people that directly assessed the relationship between levels of habitual salt intake and rates of stroke and cardiovascular disease.
Again with the "directly". These are just observational studies. "Directly" is meaningless editorializing here.

Differences in study design and quality were taken into account to minimise bias.
But of course.

Their analysis shows unequivocally
"unequivocally"!! what, a 100% CI?

that a difference of 5 g a day in habitual salt intake is associated with a 23% difference in the rate of stroke and a 17% difference in the rate of total cardiovascular disease.
Is that supposed to sound scary? Those hazard ratios are pretty low. But I'm sure there are no confounding factors.

Based on these results, the authors estimate that reducing daily salt intake by 5 g at the population level could avert one and a quarter million deaths from stroke and almost three million deaths from cardiovascular disease each year.
Aha! In other words: the authors of the study unequivocally stated, "we are not credible scientists, please do not take us seriously". The authors then went on to assess a direct link between shoe size and math ability in over 170,000 children and their analysis showed unequivocally that a 5-size larger shoe is associated with 100% improved math ability. Based on these results, the authors estimate that providing all children with shoes 5 sizes larger than their current ones will double the math aptitude of the school-age population.


Furthermore, because of imprecision in measurement of salt intake, these effect sizes are likely to be underestimated, say the authors.
Heh. It's funny that imprecision gives them even more confidence in their conclusion.

6 comments:

JustJoeP said...

Eviscerated! You've shattered all their hocus pocus with factual perspective, critical review of their methods, and careful analysis of their word selections. Nicely done.

I am surprised that the "British Medical Journal" would publish this. I'd expect such crap to be published in "Dietitian Today" or some other barely-peer-reviewed journal.

I am particularly incensed with popular theory diet advice masquerading as legitimate, factual, science based medicine, after speaking with my father last weekend. The dietitian he goes to on Medicare for his advanced Type II Mellitus was upset with his last 2 weeks of itemized (logged) food log. See, 2 weeks ago, I'd convinced him to cut DOWN on carbs (have a slice of cheese, or lunch meat, instead of a banana if your sugar is not too low, for example), and increase his protein & fiber & cocoa content in a once-a-day small chocolate snack, citing the successes of my friends Ron, Mike S, Matt, Steve Z, and myself. And what does the damn dietitian tell him? Oh.... with all that protein, he's headed toward liver, kidney, and other multiple organ failures.

I thought my head was going to EXPLODE. Now, certainly, I am not a primary care physician or endocrinologist, but to claim that a reasonably balanced moderate diet that is low in carbs and higher in protein will lead to organ failure - based upon some bogus straw-grabbing studies such as you have sited here... 5 grams of salt! OMG! ....it's ridiculous, disingenuous, misleading, and just plain wrong.

Not only does increase in shoe size lead to better math ability Ron, but getting a degree as a dietitian and advising them to avoid protein leads to repeat office visits, repeat Medicare billing, and increased revenues as well! Grrrrrrrr.

pyker said...

I don't understand why anyone would recommend a low-fat, high-carb diet to someone with type II diabetes. Would your dad read Bernstein's book if you got it for him?

pyker said...

The thing about salt is: if you're not eating food from a factory (extruded onto a conveyor belt, mixed in industrial vats, etc.), you're probably doing fine on your salt intake. Industrial foods tend to have high amounts of sodium. This is a big confounding factor. I think if the only variable you changed in peoples diets was sodium, you wouldn't see anything dramatic in health outcomes. Avoiding industrial foods is a good thing in and of itself, a side effect of which would be lower salt intake.

The best thing I've read on salt remains this: http://www.nasw.org/awards/1999/99Taubesarticle1.htm

pyker said...

Joe, if he won't tackle Bernstein's book, maybe warm him up with this article from Men's Health

JustJoeP said...

I shall recommend it.

He is a disciple of my mother... so it has to get past that filter first, and sadly, she is a strong proponent of "doing as the dietitian says". Ugh.

Ame said...

Joe, your parents are just like most end-users (in this case of health) in that they trust the "experts". If you can give lots of documentation that states otherwise, from "real sources" such as books or well known magazines, they might become confused enough to listen. :)

I know some of it has worked for me with my mum. But, it is hard work.