Monday, December 16, 2013

Cycling for Health

Modesto Roadmen, 1967

It is widely accepted that exercise is good for health. But how much exercise? What kind? Is it possible to exercise too much? One of the main reasons I restarted cycling five years ago was for the health benefits, so these kinds of questions are important to me. When I began writing this post, I worried that basic health considerations would be beneath the considerations of a randonneur1. They may be for most randonneurs, but I was surprised by how relevant these considerations turned out to be for me.


As I have posted before, I am a card-carrying member of the medical establishment, so my point of departure was to get online and search for the position of the medical community on these issues. During my career in medical research, I, like most other biomedical research scientists, derived most of the funding for my research from the National Institutes of Health, so that's where I started. The National Institutes of Health consist of 27 institutes, each specializing on a specific aspect of human health and disease. Because of my age, I started by looking at the offerings of the National Institute on Aging (NIA). To get a more general perspective, I also looked at the offerings of the Department of Health and Human Services (HHS), the parent organization of the National Institutes of Health. Finally, I did some general web searching. The problem with general web searching is separating the sense from the nonsense, so I carefully considered the sources I came across. One source I found was the American College of Sports Medicine (ACSM). This is a professional association of researchers, personal trainers, physicians, and educators who work in the area of sports medicine, and as best I can tell, is well respected. Two other associations for physicians associated with sports medicine are American Orthopaedic Society for Sports Medicine and the American Medical Society for Sports Medicine. I do not yet know how these organizations relate one to another, but I started with ACSM and purchased the book they developed for the general public, "Complete Guide to Fitness & Health." This book is extremely basic, has a strong focus on the part of the general public who has yet to develop any exercise program, but does provide a foundation of basic information about what is known by the biomedical research community about the fundamentals of exercise and health. In that context, it is comforting that what I read in this book agrees with material provided by NIA and HHS, and I will use this book as a foundation for this post.

Recommendations of the ACSM

I may be marginal as a randonneur, but by the standards of the ACSM, I am "established", at the top of their scale in terms of exercise experience. As such, their recommendation is that I engage in at least 200 to 300 minutes per week of moderate aerobic exercise or at least 100 to 150 minutes per week of vigorous aerobic exercise or any combination of the two. Moderate exercise should be done at least 5 days per week, vigorous at least 3 days per week, and again can be combined. The levels of exercise are defined as follows:
Exercise sufficiently intense to prevent singing but to allow talking.
Exercise sufficiently intense to create an inability to say more than a few words without pausing for a breath.
Using the above definitions, my MAF test bike rides2 are moderate level exercise. (I rechecked this again today, I can talk in complete sentences while riding a MAF test but cannot sing.) According to the ACSM, I should only count the moderate part of the exercise, not the warmup and cool-down which makes my MAF test 45 minutes long. If I ride MAF tests the recommended minimum of five days a week, this would mean I would need to ride a minimum of three in a row (e.g. three days of MAF tests, one day of rest, two days of MAF tests, one day of rest) and this would add up to 225 minutes per week, at the lower end of the minimum range. To reach the top of the minimum, I would either need to increase the length of all five MAF test rides to 60 minutes or increase the number of rides to six or seven per week.3

Readers of this blog have suggested more than once that I may be underestimating the intensity of my MAF test rides. I'm not sure why I am giving that impression, perhaps it is because my heart rates are unusually high for a person my age. (I reached a heart rate of 179 during some recent interval training, whereas using the formula provided by ACSM4, my maximum possible heart rate would be estimated to be 164.) Besides cycling, an exercise I do fairly often is a brisk, 3 mile walk. Brisk walking is generally recognized as "moderate", where brisk is defined as a pace of at least 3 miles per hour and typically 3.5 to 4.5 miles per hour. (I usually walk between 3.5 and 4 miles per hour.) Yesterday, I wore my heart rate monitor while walking to compare it to my MAF test ride. By the end of the walk, my heart rate was at 135 beats per minute, smack in the middle of the range I use for my MAF test rides. Personally, I am convinced that my MAF test rides rate as moderate intensity exercise. Thus, whatever I am doing right or wrong in my training, riding too many MAF tests is unlikely to be my problem.

Exercise News

A career as a research scientist have left me with two convictions:
  1. Science is much harder to do correctly than it looks.
  2. The media, even prestigious newspapers such as the New York Times, do an unreliable job of reporting science.
I am attracted to news stories on the latest scientific findings on exercise as much as anyone, but when I stop to think critically, and especially when I take the time to research the background of the story, I become very skeptical. A recent example, published in the New York Times was entitled "Why A Brisk Walk Is Better", reported the research of Paul Williams and Paul Thompson, published in the journal PLOS One. (PLOS stands for Public Library Of Science.) The problem with the New York Times Article is that it falls afoul of the cum hoc ergo propter hoc fallacy; they assume that because people who walk faster live longer, walking faster will make you live longer. The alternative explanation is that people with health conditions which will ultimately lead them to die early also don't feel well and as a result, walk more slowly. It is even possible that if these people forced themselves to walk faster, doing so would cause them to die sooner. When I followed up on the New York Times article by reading the PLOS One article, it was clear that Williams and Thompson had not made this mistake, that they understood the limitations of their data. It was the reporter for the New York Times who made the false conclusion, perhaps to make their article seem more exciting.

Even if you take the time to refer back to the original article, and only draw conclusions from what is there, not what a newspaper reporter might add, my experience as a research scientist makes me unwilling to draw too many conclusions from any single research publication. Science is not successful because scientists never make mistakes, we most certainly do. Rather, science is successful because it includes a mechanism for the correction of mistakes, but this mechanism is only partially effective at the time a single paper is published, many such papers contain flaws or are even completely wrong. Over time, the scientific community checks and rechecks conclusions until a firm, relatively reliable consensus is reached. Thus, it is much more reliable to form one's opinions from compilations of results, such as ACSM's "Complete Guide to Fitness & Health", rather than from single research papers.

One final point I would like to make is to observe that the reason that the medical establishment has so little to offer us athletes is the result of how difficult research in this area is. The reason the study of Williams and Thompson cannot be used to determine if I should walk faster is because it is a retrospective, observational study. The authors went to walking events and ask walkers to voluntarily provide information for the study. Seven years later, the Social Security database was used to determine which of these walkers had died, and that was compared to the data about each of the walkers to ask questions such as "is there a greater percentage of deaths among slow walkers as compared to fast walkers?"5 As noted above, the problem with this approach is that you don't know if people who are more likely to die walk slowly, or if people who walk slowly are more likely to die. To avoid this problem, medical research, e.g. to determine the value of a new drug, uses randomization. The people who agree to be part of the study agree to be randomly assigned to the arms of the study, they don't know if they will be given the new drug or (e.g.) sugar pills. Thus, if a difference is observed, it cannot be alternatively explained by the difference between people who might choose a new drug as opposed to those who do not. This is essential for making a believable conclusion but is very hard on the participants. Imagine you have incurable cancer and sign up to try a new, experimental treatment. The reason you sign up, more likely than not, is that you hope the new treatment might save your life. However, to be allowed in the study, you have to agree to take a 50% chance you will not be given the treatment at all. How many of you would agree to participate in a study of interval training where, depending on the toss of a coin (figuratively speaking,) there was a 50% chance you would have to agree to sit on the sidelines while the other half of the group rides intervals? Randomization, essential for a firm conclusion, can only be justified where the expected health benefits are substantial. The medical community cannot justify the effort to help me ride a 200K brevet1, nor would I be likely to accept the constraints (randomization) their help would require.

Is it possible to exercise too much?

It is certainly possible to overtrain as I have discussed repeatedly in this blog, but that is not the question I am asking here. Suppose one trains effectively from a performance point of view, do athletes who choose to participate in particularly demanding events (e.g. 1200 kilometer bicycle rides1) harm rather than help their health? Again, my attention was caught by an article in the New York Times, in this case a regular column, "Ask Well", which solicits health questions from readers. The question being answered was "Is life expectancy affected if you do 'extreme' exercise like ultra­marathons, iron­man races, etc?" The expert The New York Times tapped to answer the question was Dr. James O'Keefe, who used appropriate caution by starting with "The short answer is, we don’t know" but finished by saying "my professional opinion is that if you are exercising to improve life expectancy and long ­term cardiovascular health, be moderate about it. More is not better." Dr. O'Keefe has published extensively on this topic and, to me, his position seems to be that training and competing in very long endurance events can be harmful to your health.

In this case, unlike the previous case, my examination of the original research articles behind the New York Times article did not reveal a difference. Also, the conclusion is supported by not one but several scientific publications. However, Dr. O'Keefe is an author on all of them which in general all come from the same research group. This brings me to another caution: multiple publications do not in and of themselves to confirm a conclusion. Scientific consensus comes from the debate between different groups of scientists. It is both common and appropriate for a single group to have a particular point of view which they defend vigorously.

What is my opinion, do I think it is possible to exercise too much? I don't have one, I am going to go with Dr. O'Keefe's first statement, "we don't know", which unfortunately is the consensus of the scientific community on so many issues of importance to us athletes. Dr. Phil Maffetone, author of "The Big Book of Endurance Training and Racing", the book that got me started on MAF test rides, has weighed in on this issue on the "Natural Running Center" blog, and his response makes a lot of sense to me. To summarize, he argues that everyone is different, it doesn't make sense to come up with one recommendation to fit everyone, that the best thing each of us can do is to listen to our bodies. And that's all I have to say about that.


1. For an explanation of randonneuring in general and what a brevet is specifically, see the Randonneurs USA website.

2. My MAF test rides consist of a warmup of about 20 minutes ridden at a heart rate less than 130 beats per minute, a 45 minute ride maintaining my heart rate between 130 and 140 beats per minute, and a cool-down of about 20 minutes designed to get my heart rate back down to no higher than 110 beats per minute, typically ridden between 9 and 10 miles per hour. My speed during a MAF test ride varies between 13.5 and 16.5 miles per hour, depending on the state of my training.

3. Of course, there are many ways to train and most people would probably argue that a training regimen consisting only of MAF tests would not be optimal. For example, if I have any hope of riding a brevet, I will need to include one long ride each week into my training regimen which would cause me to easily exceed my recommended weekly minimum of 300 minutes of moderate exercise.

4. Your maximum heart rate is defined as the highest heart rate it is possible for you to reach. It is difficult (and perhaps unsafe) to reach that heart rate and in particular, it is very difficult if not impossible to reach this heart rate while bicycling; most athletes can reach a higher heart rate running than they can cycling. Maximum heart rate is known to decline with age and so there are a variety of formulae for estimating your maximum heart rate that all have the form of some standard maximum with a value subtracted that depends on age. The simplest, and perhaps least accurate of these is (220 - Age). For a 64 year old, this formula would estimate a maximum heart rate of 156 beats per minute. ACSM uses the formula (206.9 - (0.67 x Age)). The problem with all these formulae is that every person is different; two people of the same age can have maximum heart rates differing by as much as 60 beats per minute.

5. This study is way more complicated than my simple statement would imply, of course. Throughout this paragraph, I have dramatically simplified the issues in pursuit of clarity, without, I hope, distorting the main point.

No comments:

Post a Comment