You're looking at Less Wrong's discussion board. This includes all posts, including those that haven't been promoted to the front page yet. For more information, see About Less Wrong.

G0W51 comments on Optimal Exercise - Less Wrong Discussion

50 Post author: RomeoStevens 10 March 2014 03:37AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Article Navigation

Comments (141)
Tags: habits exercise longevity

Comments (141)

You are viewing a single comment's thread.

Comment author: G0W51 03 August 2014 03:18:48PM *  0 points [-]

One review found the relationship between physical activity and mortality to be linear. Does anyone have any ideas as to why this review found such a relationship while others didn't?

Comment author: RomeoStevens 04 August 2014 05:04:48AM 1 point [-]

They found it to be linear up to at least 1000kcal/wk AFAICT. This is a pretty low bar. I expect it to be approximately linear up to 3500kcal/wk, but this is admittedly based on those handful of studies that actually test higher caloric expenditure. Unsurprisingly, most longevity studies are based on the elderly where relatively low activity levels are being measured.

Comment author: G0W51 04 August 2014 03:57:09PM 0 points [-]

Thanks for the response, I think you may be right. Can you give me the links to the studies that test higher caloric expenditure? Do you know if one can exercise too much?

Comment author: RomeoStevens 04 August 2014 09:00:55PM 0 points [-]

Actually I take back what I said, it looks like a levelling off is occurring even before 3500 is reached.

In our study, 1000 kcal/week were associated with a 7% (95% CI 4–9) reduction in men and 15% (95% CI 11–19) reduction in women.

Finally, our study was the first to combine data of studies that assessed total activity (daily living, active commuting, occupation, leisure time). The Eurobarometer study110 proposed 3000 MET-min/week accumulated over 7 days (~7 MET-h/day) as the cut-point for ‘sufficient total activity’. In our study, this level was associated with a reduction in mortality risk of 17% in men and 21% in women.

so 3.5x as much expenditure results in 2.4x the risk reduction. This review indicates that the studies on athletes (which is the only readily available population performing exercise at higher average expenditures than 3500kcal/wk) are terrible, with few or no controls for other factors.

Comment author: G0W51 05 August 2014 03:43:47PM 0 points [-]

"Sufficient total activity" doesn't necessarily mean optimal total activity or when diminishing return begins. The Eurobarometer study isn't free, so I can't tell exactly what it means by "sufficient total activity." What are your thoughts on this?

Comment author: RomeoStevens 05 August 2014 05:40:10PM 0 points [-]

Distinguishing optimal total activity isn't possible given the limitations of current studies. Sufficient total activity is pointing to the activity level where correlations between higher levels of activity and longevity become too noisy. The tendency in studies like this is for the error bars to grow at the top and bottom of the range due to multiple factors.

Comment author: G0W51 06 August 2014 03:14:22PM 0 points [-]

Do you know of any reviews about the "dose-response relationship" of exercise and mortality other than the 4 we've mentioned?

Comment author: RomeoStevens 06 August 2014 06:20:37PM 0 points [-]

I don't think so, or at least if I did find any they were poor enough to disregard.

Comment author: G0W51 07 August 2014 06:39:24PM 0 points [-]

Ok. Do you know of any evidence that mortality rates increase after any amount of exercise? I suppose reverse causality would make it very hard to tell, but intuitively an 80-year old trying to exercise 10 hours per day seems unhealthy.

Another question: Do you know if aerobic and anaerobic exercise have different effects on mortality? I recall hearing (from a potentially unreliable source) that aerobic exercise was healthier, but I haven't managed to find any scientific literature comparing them. Also, could one hit diminishing returns in anaerobic exercise without hitting hitting diminishing returns in aerobic? Again, I haven't found an literature on this.

Comment author: Lumifer 08 August 2014 03:23:44PM 2 points [-]

Do you know of any evidence that mortality rates increase after any amount of exercise?

I have vague memories that Olympic athletes aren't the most healthy people and don't live too long, but I don't have a link, sorry.

Comment author: RomeoStevens 07 August 2014 10:14:47PM *  1 point [-]

  1. Yes, long distance running has some evidence of harm (CVD Rates). I don't have the cite off hand.

  2. They certainly have different short term adaptations. I know of no study that seriously attempts to separate them other than by "class of athlete" e.g. soccer players vs olympic weightlifters. The issue is that it's hard to separate them. Soccer players (especially professionals) still do resistance training etc. But yeah, soccer players do live longer than pure power lifters. I do expect aerobic activity to have a larger effect ceteris paribus.

Comment author: gwern 03 August 2014 07:30:03PM 0 points [-]

Which others didn't?

Comment author: G0W51 04 August 2014 03:52:31PM *  0 points [-]

The review mentioned in the post didn't, although I've only looked at 2 studies on the dose-response relationship.

Edit: Actually, the review mentioned in the post didn't seem to really test the shape of the dose-response relationship, I think they just assumed mortality risk had an inverse linear relationship with the natural log with the amount of exercise.

Comment author: gwern 05 August 2014 07:50:33PM 0 points [-]

I think they just assumed mortality risk had an inverse linear relationship with the natural log with the amount of exercise.

Yeah, the table just gives linear estimates, but all of them do say more exercise is better, I think.

Powered by Reddit Powered by Reddit