How many of hours of aerobic exercise are you doing each week? How much should you be doing?
50+ year olds that run were less likely to die and had lower disability scores as they aged compared to non-runners. The new research paper was published by Chakravarty et al., 2008 a group from Standford University in Archives of Internal Medicine.
Now most of you would say of course - that is obvious, and I would agree with you, but now there is evidence that our common sense is backed up by scientific testing - which might provide further impetus for people to exercise.
However, despite me being a big proponent of exercise, including running (at the endurance and sprint level), there are potentially a number of problem with this study that we should look more closely at before we accept the results at face value.
Methods and problems:
In this study they took two groups at age 50+, runners (average age 58) versus non-runners (average age 62) in 1984, and what is most interesting about this study followed these same subjects for the next 21 years. The subjects are now in their late 70 early 80s.
The runners were chosen from a 50+ runners association and the controls were chosen from permanent staff and faculty from Standford university and the demographics matched between the two groups. They ended up with 538 runners and 423 controls. Each year the participants filled out extensive questionnaires (medical history, exercise habits, and the health assessment questionnaire disability index (HAQ-DI) - there main outcome measurement. In 2005, there was 284 runners and 156 control members that continued to participate (obviously there were a number of dropouts over the 21 years but also some had died). The attrition rate for the runner group was 3% per year, while it was 6% for the control group.
Now as the authors point out at baseline (1984) the runner’s group was younger (58 vs 62), leaner (BMI 22.9 vs 24.4), and less likely to smoke (1.9% vs 9.5%) compared to the community control group (the overall results are similar for the all subjects enrolled and completers - so I only report one of these). Additionally, a far greater percentage of the runners were male (84%) while the control population consisted of 56% males. Now I don’t think the BMI (leanest) is too much of a concern because both groups are fairly lean compared to the general population, and the larger proportion of females should actually help the control group regarding lifespan (women live on average longer then men). But females might be less mobile as measured by HAQ-DI as they age compared to males. To the authors credit they do later on in their paper break down the groups into females and males and in both sexes runners have a lower disability score than non-runners. So the differential proportion of sexes in the two groups can be eliminated. The large difference in smoking at the beginning of the study is still troublesome though. (other measurements were consistent between the two groups: education level, alcohol intake). But maybe most importantly, their main outcome score (other than lifespan) the HAQ-DI score (which from now on I will refer to as disability score) was already lower in the runners group (86.6% of runners scored 0 (no disability) while in the controls only 61% of this group scored 0).
Later on they further broke up the groups into ever runners and never runners (for obviously some of the staff and faculty from Stanford university could be runners). Subjects were considered ever runners if they said yes to: “Have you ever run for exercise for a period of greater than 1 month?” So the ever run groups includes members who currently run but are not members of the running association, but also runners who no longer run.
Now even after they break all subjects into ever-runners and never-runners many of the same problems exist. Differences in age (58.4 vs 63.1) and smoking prevalence (2.1% vs 12.9%). For the disability score 82% of the ever-runners had a 0 disability score, while only 58.6% of the never-runners scored a 0 disability score.
So when the authors give the results they openly admit the differential scores between the groups (no matter how they break them up) and no great surprise throughout the next 21 years the runner groups continue to have a lower disability score than the community control group or never-runners. The authors main point that seem to support their conclusions is that the rate of disability is different between groups, with the controls and never-runners having a faster rate of increasing disability. And I would agree with this part of the paper - but there could still be a threshold effect.
How much did they exercise:
In 1984 the completers in the runners group were running 4 hours per week, while the community control group averaged 25 minutes (not yet separated into ever-runners and never runners), but the total aerobic exercise for the two groups were: 5 hrs and 7 min. versus 1 hour and 40 minutes. If we break them into ever-runners and never runners: running 3 hrs 21 min. vs 0 hours 2 min. and for total aerobic exercise: 4 hrs 32 min. vs 1 hr 7 min.
Now if you look at 21 years later (2005) (so talking about average age of 78-81) the ever-runner group was down to running only 1 hr 1 min per week while the never-runners was at 0, and for total aerobic exercise the ever-runners were at 4 hrs 29 min, and the never-runners at 1 hr and 27 min.
Basically, the running groups performed 3-4 more times aerobic exercise per week compared to the non-runners. Simplifying things, the runner group was doing 4 hours of aerobic exercise while control were getting 1 hour per week. What appeared to happen was as the running group aged they switched from running to other forms of aerobic exercise, but despite being in their 70s still put in 4.5 hours of aerobic exercise each week (they stayed very active).
Now I wish the authors had included the amount of running and other aerobic exercise was performed though-out the different time periods (though this might have been covered in earlier papers when they reported earlier results of this same study). But also I wish they had broken the running groups into those that continued to exercise and those that didn’t.
Lifespan effect:
At the end of 2003 15% of the runners had died and 34% of the controls (but remember there were difference in starting age and smoking prevalence). But even after adjusting for other variables at baseline the authors report a significant survival advantage. Bottom line runners had a greater survival rate.
Runners not only had a reduction in death from cardiovascular disease, but also interestingly cancer and neurological disorders.
(as expected initial age, sex, and initial disability score was associated with survival rates: surprising to me smoking did not reach statistical significance).
Conclusion:
Despite some problems with this study that centers around different starting points for the two groups (which can be a problem with long term longitudinal studies) the rate of increase of disability as they aged was lower in runners compared to non-runners. Additionally, runners also had a higher survival rate. What more could a person ask for, better survival rate and more functional ability.
Now with these type of studies there is always a problem of self-selection. Are people that decide to run more concerned about their health, do they eat better, and do other things in their life that give them advantages over non-runners. An open question. But at least the two groups were comparably at the socioeconomic level (based on education levels).
Also I would have liked to know how much running did the runners do before they reached their 50s. Were the ones that gained the biggest advantage running for 40-50 years when they reached their late 70s? How many years do you need to run to see these effects? There is always more questions. I am sure we will here more in the future.
Take home message:
Hopefully this data either gives you further impetus to continue to exercise program, or for you to start your exercise program today. But also remember the running group was putting in 5 hours of aerobic exercise per week (4 hours of running) in their late 50s, and in their late 70s still put in 4.5 hours of aerobic exercise per week (1 hour of running). That is impressive!
How many hours of aerobic exercise are you doing each week? And I bet most of you are not in your late 70s - something to think about. Keep moving, and keep moving a lot.