Thursday, August 4, 2011

The cumulative damage theory of injuries

I've had injuries on the brain lately.  Why do they happen? My high school's training room had a sign outside that said "Running injury? TOO MUCH, TOO FAR, TOO SOON."  Needless to say, the trainer wasn't very helpful.  But the medical/scientific consensus isn't much more helpful than that.  Overuse injuries are "tissue damage that results from repetitive demand over the course of time" according to emedicine, which is a slightly less harsh wording of the same idea my high school trainer had.  But things are more complicated than that (this is going to become a theme on this blog).  It seems like the rules are always changing--mileage or workouts that were okay last year are problematic this year, or the other way around.  The topic of "why injuries happen" is way too broad to over all at once, but today I'd like to look at one small piece: the process by which damage accumulates.

Each of the various tissues in your body has its own injury threshold.  This is the amount of stress it can take before becoming injured, and it varies from person to person and from time to time.  Healthy training will increase the injury threshold of most tissue.  So a runner who has been running 50 miles a week will be able to handle a given stress (say, a 10 mile easy run) better than a runner who has only been doing 25 miles a week.  However, the higher-mileage runner is also incurring a greater stress on a day-to-day basis.  Statistically, higher-mileage runners get injured more often.  The problem is that, in most studies, "high mileage" is typically defined as 20+ or 25+ miles a week.  Most competitive runners don't consider anything under 30 miles a week to even be "training."  I wish I had the resources to do a large-scale study on factors that can predict running injuries.  There have been some very interesting studies, though they tend to either use a large number of serious runners and find very few conclusive results (these studies are often done via surveys) or they use a small number of recreational runners and get good results, but their applicability to serious athletes is questionable.

Here's an example: Let's say a study finds that recreational runners who weigh over 200 pounds have a greater risk of injury.  We might conclude "lighter is better, since it puts less stress on the body."  But in competitive runners (particularly females), a low body mass index is associated with an increased risk of injury.

Injuries occur when internal factors coincide with external factors.  I might have a biomechanically defective elbow (an internal factor), but since I don't play tennis (an external factor), I don't get lateral epicondylitis--tennis elbow.  I've started to put some things together for a post on the various possible internal causes of injury (e.g. weak hip stabilizers, low bone density), but today I'm going to look at external factors. 

For a given runner, how does stress on the body scale? That's what I want to answer.  What is the relative increase in injury risk for a 4, 6, 8, 10, or 12 mile run? There are three possibilities: the injury risk has a linear increase, a compounding increase, or a compounding recovery.


Compounding Recovery
 According to this model, the damage incurred during a run is greatest at the beginning.

Under a compounding recovery model, the risk of injury would quickly plateau as the body got warmed up.  Additional mileage would incur a smaller marginal injury risk.



There is some logic to this idea; namely, that the muscles and tendons are not warmed up, so injuries are more likely.  This led to major establishments (running magazines, doctors, etc) encouraging athletes to always stretch before they began their run.  However, it turns out that stretching pre-run has no effect on injury risk.  Additionally, if injury risk vs. distance had a compounding recovery, we would expect mileage to have little or no effect on injury risk, at least past a certain point--once the body was warmed up, additional mileage would have little risk of incurring injury.  However, there is some epidemiological evidence to dismiss this.  In this review, one study is covered which found that runners averaging more than 40 miles per week were 2.88 times more likely to sustain an injury over a 1.5 year period. 40 miles a week is, at a minimum,  5 or 6 miles a day.  One would think that four or five miles is more than sufficient for the body to get up to speed.  But according to this study, running seven or eight miles a day incurs significantly more damage than three or four.  So out goes compounding recovery. 

Linear Increase
Under a linear increase model, every step is equally damaging as the last--the stress on the body depends only on the total distance covered, and at what pace.

The consequences of a linear increase in stress on the body is that it doesn't matter if or how you break up your runs; only the absolute distance and speed matter relative to stress on the body.

Compounding Increase
Under this model, the stress on the body and the risk of injury increase incrementally as the duration of a run increases.  This means that the last mile of a 10-mile run is more stressful on your body than the first or the fifth.


This is the model that makes the most intuitive sense to me, but as we've seen before, intuition and common sense can be misleading.  There is, however, some evidence that this model makes the most sense.

Biomechanically speaking, we should be able to see the effect of fatigue on the body.  If stress on the body does not increase as the duration of a run increases, kinematics (motion of the joints) and impact parameters shouldn't change over time, and therefore damage follows the linear increase model.  But if they do, and we can link these changes to increased risk of injury, we know that stress on the body increases in a compounding fashion.

After reviewing the literature, it is fairly clear that kinematics change while fatigued.  This occurs both under metabolic fatigue (the familiar "burning" from hard anaerobic efforts and muscular/CNS fatigue (the kind that occurs during cycles of heavy training, not necessarily anaerobic), though the differences between the two are hard to discern.  Additionally, individual variability is very high when it comes to what exactly changes in the gait cycle. Some studies find no overall differences among groups of individuals, but large differences within an individual.  This makes sense--a fatigued runner with weak hip abductors (a hip external rotator) may have increased internal hip rotation, while a different fatigued runner with a weak tibialis anterior (shin muscle) may have increased ankle plantarflexion upon impact.

This highlights a weakness of simple kinematics studies--they cannot examine the underlying causes of why a joint moves differently or abnormally.  Fortunately, there is emerging research that is using electromyography and 3-dimensional accelerometers (the same thing that's in a Wii remote) to study the way muscles act during a run. Benno Nigg's work on tissue vibration and wavelet analysis is as fascinating as it is complicated, but today, we're only concerned with his results: in a fatigued state (in this case, a 8-12km run to exhaustion), muscles oscillate for longer and at higher amplitudes following impact versus in an unfatigued state.

Are these muscle oscillations linked to injury? Probably.  From Benno Nigg's earlier work, we know that the body does all it can to minimize soft tissue vibration following impact.  Unfortunately, retrospective or prospective studies on how soft tissue oscillations relate to running injury are lacking (though I'd bet they are in the work).  For now, this is a jump in logic we'll have to accept.

So, if stress on the body indeed increases in a compounding way, what are the implications? It's far too easy to say "well we'd better stick to lower mileage!"  It dodges the issue of maximizing potential.  The fact is, you will never reach your full potential at long distances (5k/10k) without high volume training.  What's high volume? For a junior or senior in high school, it's working up to at least 60 minutes a day (50-60 miles per week depending on pace).  For an upperclassman college runner, likewise for 80-90 minutes a day.

Now that I've made a bold statement (literally), I'll back down from it.  It makes no sense to run 80 miles a week if you can't stay healthy at 70.  Much of being able to handle higher mileage is a function of internal factors, mentioned at the beginning of this post.  But today's topic is external--if you are going to run 80 miles a week, what is the least injurious way to do so?

If we accept that stress over the duration of a run is cumulative, we can draw some practical (and controversial) training recommendations:

  • Split up mileage.  If, by our logic, the last five miles of a ten-mile run is much more damaging than the first five, an easy way to migitate the damage of those last five miles is to do them eight hours after the first! Doing doubles instead of one long, single run will be better from an injury standpoint.  Now, I can already hear the critics saying, "But John, the duration of the run is what gives it such a great training effect!" or, smugly, "Well why not do ten one-mile runs per day?"  I must agree with the first arguement, and I'm not arguing that two five-mile runs is always better than one ten-mile run.  There are no studies on whether the fitness benefit is higher from one run or two shorter runs totaling the same duration.  All I know is this: 1) All else equal, one ten mile run is better fitness-wise than one five mile run.  2) All else equal (ex. getting enough sleep), two five-mile runs are less injurious than one ten-mile run.  3) One ten mile run and one five mile run are better fitness-wise than one ten mile run.  Anything other than that is pure conjecture--would a group of people running 70 miles a week in 14 runs be more or less fit than a group of people running 70 miles a week in 7 runs? I don't know.  I'd love to do that study, though.  I do know (all else being equal) that the doublers would be less likely to get injured.  In response to the second criticism about doing three or four or ten runs in a day, I have to defer to experience garnered in the real world.  Nearly all world-caliber athletes run twice a day, and a very small subset of the most elite have been reported to run three times a day on occasion.  This subset, to my knowledge, consists of Kenyans in training camps (which are, might I add, often blamed for burning out promising Kenyan athletes), Gerry Lindgren (probably the most talented high school runner in the history of the US--sorry, Rupp--and famous for running prodigious volumes of 160+ miles a week), and Lasse Viren, though his triple-days were usually limited to 5/5/5 miles.  So practical experience tells us that twice a day the max for reasonable people.  If you live in the woods, run on pine needle trails, and receive a generous government stipend like Lasse Viren did, you can go ahead and give triples a shot.  Otherwise stick to two-a-day max.  One of York High School's staple sessions in the glory days of Joe Newton was an "interrupted run" of 45, 30, and 30 minutes, the first at an easy pace, the second at a moderate pace, and the third easy again.  Between the segments is a 5-10 minute break for water, stretching, etc.  I can't help but think that even this short break reduces some of the damage to the body that would otherwise occur during a continuous 105-minute run. 
  •  Rethink the weekly long run, especially if you are not training for a marathon.  Anathema to many, including Lydiard fans like myself.  The "long run is 25% of your weekly mileage" rule has been a mainstay of training for years--it even survived the onslaught of the low-mileage trend in the '80s and '90s.  The long run exists precisely because there are benefits to going for long, continuous runs, mostly in the department of increased capillary growth.  In 1962, Lydiard insisted on a 20+ mile long run every Sunday for aerobic fitness. But a surprising number of coaches have come to advocate biweekly (every two week) long runs, including Joe Vigil and John Kellogg, who may well be Arthur Lydiard reincarnated.  Additionally, the 20-25% of weekly mileage run starts to fall apart at high mileages, especially when doubling.  Running 90 miles a week on 11-12 runs with a long run of 18-22.5mi long run is questionable at best.  Downright irresponsible at worst.  When I was running 115-120 miles a week, I certainly wasn't doing a 30 mile long run on the weekends.  A better rule I've been using is making the weekly or biweekly long run 1.5x further than your typical main run of the day. So if a high school junior typically runs 7 miles a day, he could do a 10-11 mile long run.  If a college junior is doing a 10 mile run each morning and a 4-5 mile afternoon run 3-5 times a week, a 15 mile long run would be appropriate.  Doing a long run only every two weeks is advantageous because it allows you to do 14-day training cycles and avoid the race-long run-workout triple that is the bane of many college runners.  You should realize that your long run will likely be the most damaging workout you do and you should treat your recovery as such.  Younger runners would do well with an off-day after a true long run, and more experienced runners would do well to take an easy day. 
  • Avoid long cooldowns. After a hard workout or race, your legs have taken a significant beating.  As such, they're more susceptible to additional stress, so tacking on an eight-mile cooldown after an early-season cross country race to "get the miles in" is a bad idea.  Instead, do a six-mile double earlier or later in the day.
  • Front-end fast running.  Workouts that involve more or faster running late in the workout will be more injurious than workouts that put the longer, faster segments first. So, it's better to do 3km-2km-2km-1km-1km-1km than the other way around.  Sometimes, workouts are designed to back-end fast running, particularly for road race training.  A classic American marathon workout is 10 miles easy + 10 miles at marathon pace.  While this will be much more damaging than 10 miles at marathon pace + 10 miles easy, it also more accurately mirrors the demands of the race.  In cases like this, you'll simply have to acknowledge that it'll be a more damaging workout than usual and treat it as such.
 Several of these recommendations are contingent upon the speed of recovery from runs--how long does it take to adequately recover? That's a question we don't have an answer to, but could provide a counter argument for my recommendation of splitting up mileage. How much better is 24 hours of recovery than 12 or 8? Anecdotal, I notice a major difference between 5 hours and 7--I prefer 8 or 9 hours between runs if I'm doubling.  But hey, what do I know?  That's all for this time.  Sorry for the big gap between the last post and this one!

1 comment:

  1. Dear John,

    If you are ever willing to build models of running injuries, please contact me. I work on artificial intelligence, if data is available, we could use several methods to detect what variables are important in the process, and if so, how much.

    Juan Flores
    juanfie@gmail.com

    ReplyDelete