I think this is important.

[Mark Rippetoe]

http://www.psychologytoday.com/blog/...disease-part-i

[Mark E. Hurling]

I almost hesitate to mention his name, but Art De Vany used to make very similar points about marathoners, death, and heart disease. I think he may have reviewed some of the same studies mentioned in the article.

[Jefferson]

It's only one study, and the results aren't statistically significant. He's just shoe horned his hypothesis onto the barely statistically visible differences in heart attack rates. If a similar study was done on weghtlifters, and the results were just as marginal, and he attributed it to the stress the heart undergoes when lifting very heavy weightsand the changes it makes in response to that stress, what would your reaction be?

[Mark Rippetoe]

I'd post the fucking link.

[tfarny]

@ Jefferson: Science is always more interesting when people find the opposite of what they set out to find, isn't it! Statistical significance is largely a distracting leftover from pre-computer days; it's far prefereable to just post P-values as Dr. Harris discusses. Almost any effect is "significant" at a large enough N size; almost no effect is significant at a small enough N size. I thought it was a great write up of a pretty great study. Decent effect size in the OPPOSITE of the expected (for many reasons!) direction, and the imaging findings were consistent with the follow up mortality rates. It's not exactly Kurt Harris' hypothesis either; the idea that high cortisol and too many carbs drives heart disease has had quite a few backers for a lot of years. Do you still believe that saturated fat clogs the arteries, or something?

[Jefferson]

@ Jefferson: Science is always more interesting when people find the opposite of what they set out to find, isn't it! Statistical significance is largely a distracting leftover from pre-computer days; it's far prefereable to just post P-values as Dr. Harris discusses. Almost any effect is "significant" at a large enough N size; almost no effect is significant at a small enough N size. I thought it was a great write up of a pretty great study. Decent effect size in the OPPOSITE of the expected (for many reasons!) direction, and the imaging findings were consistent with the follow up mortality rates. It's not exactly Kurt Harris' hypothesis either; the idea that high cortisol and too many carbs drives heart disease has had quite a few backers for a lot of years. Do you still believe that saturated fat clogs the arteries, or something?

I'm just saying that it isn't hard to believe that there was an 8% difference in anything between any two groups. If this study was repeated lots of times and there was a consistent difference he would be entitled to start getting excited. Is it possible he's right? Yes. Does 1 study with hardly an 8% difference prove anything? No.

[stef]

Statistical significance is to help assess risk of error, but tells you nothing about the biological significance. Both, of course, are up to judgement.

3x higher in the "healthy" group is rather striking, particularly for such a negative outcome.

[Mike C]

Statistical significance is largely a distracting leftover from pre-computer days

care to elaborate on this? All statistical significance is trying to explain is the likelihood that a particular result wasn't the product of chance or random variation.

[Scrofula]

care to elaborate on this? All statistical significance is trying to explain is the likelihood that a particular result wasn't the product of chance or random variation.

Yes, but it's a rather simplified, contrived way to do so. You pick a completely arbitrary significance threshold, and then pretend that anything that's above the threshold is unquestionable truth, and everything below the threshold is completely baseless random noise. Why not just directly report the probability rather than crudely thresholding it?

[Lon Kilgore]

This is just another data point in the at least decade old concept that competitive marathoners are damaging their hearts. About 10 years ago a few studies demonstrated lesions (aka scar tissue) present in their myocardium. A few years ago it was marathoners have valvular leaflet calcifications on par with those of sedentary individuals with diagnosed heart disease. Last year it was found that cyclist doing 100 miles in the Texas heat had blood chemistry changes indicative of myocardial damage. Now we have another example of how an extreme type of training can produce maladaptation. You have to also understand that in all of the previous studies it was demonstrated that although the marathoners (or cyclists) had these anomalies present, they still had a superior cardiac function than diseased or normal populations. In other words, they weren't at imminent risk of cardiovascular compromise.

To me, none of this suggests that endurance training or "cardio" will kill you. It suggests that there is a cost benefit ratio in endurance training that we need to be aware of and respect. Unfortunately, the way exercise is presented to the general population leads one to believe that more miles is better.