This would involve a selection bias. For example, evidence is mounting that endurance athletes die earlier than non-endurance athletes.

UPDATE: I have provided a more detailed account, with some citations, below.

> evidence is mounting that endurance athletes die earlier than non-endurance athletes.

Please provide a citation for that extraordinary claim.

Not intended to support the parent's assertions, but it has been established that endurance exercise increases the propensity for Atrial Fibrillation, a potentially dangerous arrhythmia of the heart. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2638655/.

Sure, but it also decreases mortality by reducing the risk for hundreds of diseases, including heart disease, the leading killer for over-65s. You have to balance pros/cons, obviously!

"It"? In addition to a specific type of exercise, we're talking about a continuous variable here. Are you so sure it's a linear relationship, with the heaviest exercisers experiencing the greatest benefits? Because that seems like a tenuous position.

The point under discussion was the effect of endurance exercise, not exercise generally.

I recall seeing this as well... Maybe even posted on HN.

A quick Google on marathon life expectancy:

http://www.webmd.com/fitness-exercise/news/20140401/too-much...

I believe there have been studies linking higher death rates to some endurance exercisers, and others showing the opposite (esp. among 'elite athletes', who may be especially adapted to this behavior).[0] Sorry I don't have more links... please apply skepticism and look deeper as you like. This is what I can provide.

First, it seems reasonable that, as a stressor, there is going to be a hormetic level of exercise, and then above that, a level that isn't healthy. That's in line with the linked article's advice, and also this summary of early death research from NYT:

'In the largest such study to date, involving more than 50,000 adults and presented at the 2012 annual meeting of the American College of Sports Medicine, participants who ran between 1 and 20 miles per week had almost 20 percent less risk of dying prematurely than people who didn’t exercise. But those who ran more than 20 miles per week enjoyed no such benefit. They had about the same risk of premature death as those who were sedentary.' [1]

If the above establishes the basic principle that too much exercise is counterproductive, there is almost certainly a level that actually starts subtracting lifespan (that study did not single out the most extreme endurance exercises AFAIK). Ergo, why there are so many deaths during Marathons (running even one M. can cause measurable heart damage).

I haven't checked out this angle, but it's also possible that lower muscle mass associated with endurance athletes is linked to (less) longevity, since, I believe, it is in general [2].

But again, one cannot approach this too simplistically. IIRC there is some evidence among pro bodybuilders of early death. More generally, exercise could have many effects, going in both directions, all with different curves.

This is a concept that seems particularly hard for folks who have learned to push their bodies well past the evolved resistance to high levels of exertion, and who take that unconditionally as a virtue.

EDIT: see also [3] below. I didn't have time to find the actual paper.

0 https://primalcycle.com/2011/09/17/life-expectancy-o-enduran...

1 https://well.blogs.nytimes.com/2013/12/05/ask-well-endurance...

2 https://www.scientificamerican.com/podcast/episode/muscle-ma...

3 http://www.webmd.com/fitness-exercise/news/20140401/too-much...

Find some 20 years old cars, one with 200,000 miles on the clock, one with 20. Which would you choose?

Humans aren't cars...? Don't use an analogy unless it has a logical basis.

Parts of the analogy are certainly correct...

Mechanical parts wear out from use, whether they are animal or machine. Animal parts will heal to an extent and may require a minimum level of use to stay healthy, but they still wear out.

You're right. It is much easier to repair a car.

Car's don't self repair.

If I go lift heavy weights, my bones self repair and get stronger.

If I sit on my butt all day on a computer, my bones get thinner and weaker.

No matter how much you drive a car, it is never going to get "stronger", only slowly wear out.

Since you are a human and not a car, I would suggest you maintain your body like it is alive and not made of metal.

In both cars and people I would ask the same thing: is there a 25-35000 miles option too ? Too low is worse, and 20 is a sure indicator of big problems. So is 200,000.

As to the general feeling here. I invite everyone here who thinks ridiculous exercise is good to volunteer at an emergency room for a week. I guarantee this idea will get cured on the first day in any large city.

Humans are cars that run constantly. So the human analogy to a car with 20 miles would be a baby.