My Kung fu is stronger than your Kung fu: The Science of Training

Even well informed coaches and riders can argue frequently, and sometimes angrily, about training methods. Enough research has been published, one might think training would be reduced to a science: That there would be one proven and accepted training plan for all to follow, but there isn’t. The best training plans are consistent with the results of research, but have been developed based on the accumulated experience of thousands and coaches and riders, and not, for the most part by scientists themselves. Because experiences vary, training methods vary. Rather than validating precise sequences of exercises or even precise exercise details, the scientists have told us about the effects of different exercises and provided guidelines by which exercises and sequences of exercises are shaped. The shaping is still an art, and the source of arguments.

The effects of particular exercises may be understood, but that does not tell us how often they should be done, or in what combination with other exercises. To prove that one sequence of exercises is more effective than another, scientists would have to have groups of riders follow each sequence and then compare the results. Scientifically choosing a best plan would require comparison of all possible plans. The question of the optimal sequence of exercises is one that science cannot answer. Even if one could identify the best plan for the average rider, differences between individuals are great enough that that optimal plan would not be optimal for most individuals. Even if you could identify the best training plan for each rider, it would not be the best for long as someone else could manipulate a variable that your plan did not consider. A recent example of this sort of unexpected improvement is the use of altitude or altitude simulation as an adjunct to training. Trying to develop the permanent optimal training plan is as realistic as searching for the Fountain of Youth. You can have plenty of great adventures on the way, but don’t expect to reach the goal.

Even with heart-rate monitors and several kinds of power-meters available at somewhat affordable prices, even with multi-color graphing and data processing packages that include dozens of super-special functions, training today is as much based on superstition and personal bias as on science. Measurement does not make science. Science is a method of answering questions based on the testing of hypotheses. The changes that occur in the muscles, nerves and blood vessels during and as a result of exercise are partially understood, but the ways in which those changes differ depending on the power produced in training or the duration or the number of intervals have barely been studied. Despite this, coaches wearing the mantle of science are prescribing "scientific" training plans with intervals of certain lengths, rides at precise powers and so on. There is little research to support the use of intervals over steady paced efforts, much less distinguishing between numbers and lengths of intervals. There is no formal research whatsoever comparing the effects of training at slightly different power-outputs over the several years it takes to become a well-trained bike racer. I’m not arguing against doing intervals or training with a power-meter. Intervals work. Power meters provide motivation and feedback. I’m just pointing out that as soon as I, or any coach, prescribe intervals of a particular length or training at a particular wattage, we are spreading our own personal bias, rather than recommending a training method scientifically proven to be more effective than a similar plan with slightly different intervals or wattages.

Science Run Amok

The first problem we run into in trying to devise the ultimate scientific training plan is that we can’t really compare training plans. The next problem is that a study can only answer a well-posed question and the answer may only apply in the conditions of the study. Science can answer questions about measurable outcomes, such as, "what cadence produces the lowest oxygen consumption for a given power output?" but cannot answer more practical questions like, "what cadence gives me the best chance of winning a road race and how does that cadence change as the race progresses?" Many oxygen-uptake studies have been undertaken to identify an optimal cadence. They agree that optimal efficiency in less fit individuals pedaling at low power outputs occurs around 60 rpm. In studies of fitter riders and higher power-outputs, the optimal cadence is found to increase, up to about 80 rpm for the fittest individuals working the hardest in sustained efforts. Astute observers note that professionals ride above 90 rpm most of the time, and often over 100 rpm. Professionals don’t use the cadence that minimizes oxygen consumption. Why not? As I noted, scientific research only gives a firm answer to the question that is asked: in this instance, what cadence minimizes oxygen consumption for a given power-output? Oxygen consumption efficiency is important, but not the determining factor for racing success. Apparently something else is more important. Otherwise the pros would ride at 80 rpm when going hard and slightly lower cadence at lower powers. Incidentally, it turns out that the cadences actually adopted by competitive cyclists are consistent with the minimization of peak muscle tensions. Science only answers the question that is actually asked. A rider or coach could do some pretty foolish things by over-generalizing a valid scientific result.

Strength training for endurance athletes is a well-researched topic where the research results have been misapplied. For several decades strength training has been tested for its ability to change physiological variables known to be important to racing. Successful endurance athletes have relatively high VO2-max, so the influence of strength training on VO2-max has been measured. Strength training has a substantial positive effect in untrained individuals, but little if any effect in highly trained athletes. Recently it has been recognized that power at lactate threshold is more important than VO2-max as a determiner of racing success, so the effect of strength training on power at lactate threshold has been tested, again with a negative result in athletes. Finally it is known that increasing body mass decreases climbing speed, and that strength training can cause muscle hypertrophy. With these facts in hand (strength training makes you gain weight but does not improve VO2-max or power at lactate threshold) some otherwise intelligent individuals have concluded that strength training can’t be good for cyclists and is in fact bad. The benefit or detriment to racing brought about by strength training has never been tested by a direct comparison in a large group of subjects. One problem with the conclusion that strength training is bad for cyclists and should not have a role in the development or maintenance of bike racing fitness is that many top cyclists winning races at the international and world level do in fact making use of strength training as a fundamental part of their programs.

Apparently there are additional factors beyond VO2-max and power at lactate threshold that can determine racing success. Some research suggests that while strength training does not increase VO2-max, it does increase the time for which one can sustain the power-output that corresponds to VO2-max in the muscles trained. This is not a generally accepted conclusion, but it is consistent with the real world evidence: Many top pros are lifting and continuing to win. The ability to sustain a VO2-max level effort is clearly also important and may be the key variable. Until the question is conclusively answered in favor of or opposition to strength training for individual cyclists, scientists can’t even begin to approach the question of the optimal strength-training program. Until then, we coaches will have to continue to recommend what we know to work, or at least not to hurt, even though science has not yet told us why.

Training Principles, Physiological Principles and the Field of Battle In Between

Science is good at answering the question, "what is the effect of this particular mode of training on this particular physiological variable?" With enough time and enough resources, questions of this sort can be answered. Science is relatively ineffective at answering questions of the form, "what is the best combination of training exercises to yield racing success?" Even if we could test all possible plans on large numbers of people to pick the "best" one, we’d find that while certain plans had better results on the average than other plans, individuals would not all do best on the same plan. There would be a few common factors in the plans that worked best for the largest numbers of riders though: Plans with more volume would be more effective than plans with less volume, up to some point and so long as the individual was not maintained in a state of fatigue. Plans that increased volume and intensity gradually would be more effective than those that started with a large volume and tapered for an extended period of months, or those that randomly jumped up and down in volume or intensity. Those that had a balance of harder days, easier days and rest days would be more effective than those that consisted of only one sort of day year round. Plans with the majority of training done on the bike will beat plans with the majority of time spent doing strength or other cross training. These things coaches believe based on our own experience with thousands of athletes, though they have not been rigorously tested scientifically.

There are certain rules that effective training plans will follow. Exercise scientists give them names like: the Law of Specificity, the Law of Progressive Overload, and the Principle of Balance of Rest and Work. While studies on many of the finer distinctions like interval length or number, or exact boundaries of heart rate zones are lacking, quite a bit is known about how training affects specific tissues. Between very general training principles and very specific physiological knowledge about the effects of particular types of training, lies a large area of speculation. When there is really no information, decisions are arbitrary and subject to argument. Coaches have to make training plans far more detailed than can actually be justified based on scientific research. We make use of training principles, a variety of personal and vicarious experiences, common sense, handed down wisdom, superstition and individual bias. The real testing occurs in competition.

Testing training theories by having riders follow training plans and then having them race is not the sort of rigorous scientific protocol of which the editorial boards of research journals would approve. There are too many variables beyond our control and the race outcome is not clearly enough related to the variables that we can manipulate. I’m not going to get my clients’ results published in Science or Nature any time soon. So long as they keep getting published though, we won’t complain. I’ll keep prescribing training plans that are in agreement with scientifically proven fact, but directly based on accumulated experience, even if that means that I’ll be waiting for the scientists to catch up, and even if that means having fight with other coaches.

The situation that ensues when two coaches discuss the merits of their own training plans reminds me of an old Bruce Lee movie in which the protagonists meet and one says to the other "my Kung fu is more powerful than your Kung fu". You know how they are going to settle the question. In the case of race coaches, rather than fight ourselves we send our clients to do battle. If my clients win or improve over their previous performances, I win.