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?
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.
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?
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.
“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?
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.
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.
Yes, long distance running has some evidence of harm (CVD Rates). I don’t have the cite off hand.
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.
The site seems sketchy, as the US Food and Drug Administration warned the site to stop making illegal claims, and many claims on the website go against mainstream medicine.
I would ignore the site, it’s just a handy collation of those eight studies so you can actually check them. Most health websites that attempt to synthesize research are pretty bad.
Edit: not sure why comment above this was downvoted? Checking sources is a good habit.
Mayo Clinic recently published proceedings00638-7/pdf) suggesting that 5 hrs/week of vigorous exercise was the upper limit for safe exercise. They didn’t state their methodology for finding studies on the topic, but I don’t see any reason for Mayo Clinic to be biased about it.
They also discussed a meta-analysis00519-9/pdf) that suggested that elite athletes (who presumably exercise a ton) were much healthier than the general population. The proceedings explained that the meta-analysis had many methodological limitations, such as elite athletes being physically gifted, practicing other healthy habits, and having high socioeconomic status.
Oh cool! The estimate of an actual MET level for best longevity is great! It seems reasonable too, 10 is pretty hard to sustain. I would have been suspicious of a lower number. They also note some limited evidence that intermittent vigorous exercise with full days off seems to be better than daily exercise. This matches the current model of vigorous exercise as a eustress, where the recovery is what is important.
WRT the elite athletes, the discussions I’ve seen of reverse causality seem fairly convincing. Those people were going to live longer regardless of their chosen profession, so it’s hard to tease out what specifically the additional benefit of exercise was. (IIRC there was a twin study that looked at pairs with one becoming a pro athlete and the other not.)
Sorry about the broken links. Anyways, IIRC, The Mayo Clinic proceedings only recommend limiting vigorous exercise. Do you think one could still get more health benefits by exercising non-vigorously? i.e. They recommend limiting vigorous exercise to ⇐ ~50 MET/wk (assuming the exercise burns 10 MET/hr). Do you think one would get additional health benefits for exercising moderately for, say, 100 MET/wk?
At some point the stress effects cross the exercise effects in size. That crossing might be slightly different for different people, but for most even 5 hours/wk is a big ask. An additional 100 MET per week seems like it would be pretty disruptive to trying to lead a normal life, hold down a job, and socialize. I think some people become addicted to exercise and do it to a fault.
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.
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?
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.
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?
Actually I take back what I said, it looks like a levelling off is occurring even before 3500 is reached.
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.
“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?
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.
Do you know of any reviews about the “dose-response relationship” of exercise and mortality other than the 4 we’ve mentioned?
I don’t think so, or at least if I did find any they were poor enough to disregard.
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.
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.
Yes, long distance running has some evidence of harm (CVD Rates). I don’t have the cite off hand.
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.
I’ll try to find the study evidencing increased CVD in long distance runners. I appreciate the response.
This article cites 8 studies on endurance training http://fitness.mercola.com/sites/fitness/archive/2013/08/23/extreme-endurance-exercise.aspx
No direct link to longevity has been shown IIRC, so take it with a grain of salt.
The site seems sketchy, as the US Food and Drug Administration warned the site to stop making illegal claims, and many claims on the website go against mainstream medicine.
I would ignore the site, it’s just a handy collation of those eight studies so you can actually check them. Most health websites that attempt to synthesize research are pretty bad.
Edit: not sure why comment above this was downvoted? Checking sources is a good habit.
Yeah, and as we all know the FDA is completely infallible.
Mayo Clinic recently published proceedings00638-7/pdf) suggesting that 5 hrs/week of vigorous exercise was the upper limit for safe exercise. They didn’t state their methodology for finding studies on the topic, but I don’t see any reason for Mayo Clinic to be biased about it.
They also discussed a meta-analysis00519-9/pdf) that suggested that elite athletes (who presumably exercise a ton) were much healthier than the general population. The proceedings explained that the meta-analysis had many methodological limitations, such as elite athletes being physically gifted, practicing other healthy habits, and having high socioeconomic status.
What do you think of this all?
Oh cool! The estimate of an actual MET level for best longevity is great! It seems reasonable too, 10 is pretty hard to sustain. I would have been suspicious of a lower number. They also note some limited evidence that intermittent vigorous exercise with full days off seems to be better than daily exercise. This matches the current model of vigorous exercise as a eustress, where the recovery is what is important.
WRT the elite athletes, the discussions I’ve seen of reverse causality seem fairly convincing. Those people were going to live longer regardless of their chosen profession, so it’s hard to tease out what specifically the additional benefit of exercise was. (IIRC there was a twin study that looked at pairs with one becoming a pro athlete and the other not.)
I’m surprised there are that many matching twins floating around.
4/1000 means they’re not that rare.
No, but we aren’t exactly brimming with pro athletes either.
Sorry about the broken links. Anyways, IIRC, The Mayo Clinic proceedings only recommend limiting vigorous exercise. Do you think one could still get more health benefits by exercising non-vigorously? i.e. They recommend limiting vigorous exercise to ⇐ ~50 MET/wk (assuming the exercise burns 10 MET/hr). Do you think one would get additional health benefits for exercising moderately for, say, 100 MET/wk?
At some point the stress effects cross the exercise effects in size. That crossing might be slightly different for different people, but for most even 5 hours/wk is a big ask. An additional 100 MET per week seems like it would be pretty disruptive to trying to lead a normal life, hold down a job, and socialize. I think some people become addicted to exercise and do it to a fault.
Which others didn’t?
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.
Yeah, the table just gives linear estimates, but all of them do say more exercise is better, I think.