Wednesday, April 6, 2011

A reply to Steve Magness.

In his Science of Running blog,  running guru Steve Magness recently posted a pair of articles called "How to Run: Running with proper biomechanics".

I posted a reply (a link to it will be here after it gets posted).  Please read it and the above article before continuing.

To summarize, Steve provides a very lucid description of the key elements of running, then provides a poor way to apply them.  His method requires you to run in order to learn to run.  A tautology at best, nonsense at worst.

I'd like to discuss what it means to 'create a runner', that is, to change a 'non-runner' into a 'beginner runner'.  I'll focus not on the science, but on what we should have a true newbie do to become a runner (backed by science, but not mentioned to the newbie).

So, what is a 'non-runner'?  Someone who can't run?  No, because if you can't run you probably need medical attention, not a run tutorial.  A non-runner is someone who doesn't run by choice.  Where adults are concerned, the vast majority of 'non-runners' are actually 'ex-runners'.  Most often, the decision not to run was forced by injury, pain, or sheer misery.  Non-runners are people who can run, but don't, because they don't know how to avoid repeating the same bad experiences.

I was such a runner.  In the mid-'80's (my late 20's) I joined the 10K craze, broke 45 minutes, then quit when I had pain not only during racing, then during training, then after running, and then eventually for days after running.  I stayed quit for 30 years, until almost 3 years ago, when I started triathlon.

I love triathlon!  It was the motivation I needed to do whatever it took to return to running.   I have not only reinvented myself as a runner, but I've had to do so three separate times, as age-related degeneration and injuries forced additional changes.  I own a dozen pair running shoes, all less than 3 years old.

Before turning someone into a good runner, you must first turn them into a safe and comfortable runner, one who is free of both pain and injury.  Speed and efficiency don't matter at this point: All that matters is getting the person to run in a way that is much more successful than whatever they tried before.

What can we do for the 'broken' runner who is always getting injured?  For the 'bad' runner who is always in pain, but keeps going until something breaks.  For the 'miserable' runner who finds no joy, but keeps grinding out the miles until they finally give up, or find pain and injury.

Do we give them 'cues'?  Or do we give them a whole new starting point?  Such runners must never be allowed to repeat prior damaging habits, as would be required by running while waiting for 'cues'.

Since I've had to relearn running 3 times in less than 3 years, I may have a rather unique perspective.  Being a new and enthusiastic triathlete, I had motivation to not give up, to persevere.  Being an engineer, I had the skills needed to approach the task analytically, in both an evidence-based and results-oriented manner.

I studied and applied everything I could find. Read the books, articles, blogs and scholarly papers.  Even got some textbooks to plow through some of the mumbo-jumbo, and skimmed some online college courses.

What did I learn?  The running community can't even agree on the words it uses, the terminology used to describe and define key aspects of running form, and how the body is best used to achieve this form.  Part of the lack of agreement is due to lack of knowledge and research.  Some is due to the building of isolated sub-communities (running cults) and ivory towers.

Bottom line, from the perspective of the true beginner runner, physiology is crap.  Kinematics is crap.  Biomechanics is crap.  Just about every knowledge domain related to running is crap.  It simply isn't what a beginner runner is interested in, or needs to know.  Telling them about it simply places yet another barrier, an intellectual barrier, between them and running.

So, what does the beginner runner really need?  Another way to look at it is by asking what it is not: When does a runner stop being a beginner, and become an intermediate?

Personally, I believe an intermediate runner is one who has achieved basic goals of speed and distance, and feels limited by their knowledge, rather than their bodies or their experience.  I believe Steve's article is aimed at just this type of runner, the intermediate who is done with being a beginner (or the coach helping such a runner).

So, what is the beginner runner?  Well, consider the minimal differences with the non-runner previously described.  Clearly, for a non-runner to run, they must avoid the circumstances that made them a non-runner in the first place, else to that domain they will quickly return.  This transition requires an approach to running that is certain to be free of injury, pain and misery.

That is my definition of the beginner runner: A non-runner who is learning an approach to running that is both safe and comfortable, possibly even enjoyable.  In this context, perhaps a more accurate term for 'beginner' runner would be 'remedial' runner, one who is genuinely starting all over again, rather than continuing on from their last point of failure.

Does Steve's article provide such an approach?  If it does, I didn't find it.  What would such an approach have to provide?  At the very least, it must provide a straightforward path to learning the simplest and safest running gait possible.  A gait that is simple enough for everyone, while simultaneously being general enough for everyone.

Does such a gait exist?  Clearly I must believe so, otherwise I wouldn't be writing all this.  But before I describe my version of it, let's pause to reflect on what we seek.

The general beginner gait must certainly meet the requirements of good form so well described in Steve's article, sacrificing only some speed and efficiency in pursuit of safety and comfort.

Unfortunately, I found no explicit mention of either safety or comfort in Steve's article.  Perhaps safety is implicit, but comfort is only defined by its absence, not as a desirable factor in and of itself.

I believe comfort is the single most important factor that is relevant to turning a non-runner into a beginner runner.  From the beginner's perspective as an ex-runner, little else matters.

How can we get from 'non-running' to comfortable running without passing through any pain zones?  More importantly, how can we teach good form without getting into all the (excellent!) technical stuff in Steve's article?

Should we take it 'slow & easy'?  That approach works well in many areas of physical endeavor, such as swimming and bicycling.  Unfortunately, running ceases to be running when it is done 'slow'.  And as for 'easy', I haven't found an 'easy' way.  My way is 'hard': Not in the sense of grueling or arduous, but more in the sense of challenging and requiring focus and attention.  More like learning to juggle as compared to becoming a bodybuilder.

As Steve says in his article, we must not treat running as 'natural', which makes doubly good sense when applied to an ex-runner.  So we can't ask an ex-runner to just start running and wait for 'cues'.

What should we do instead?  How can we get an ex-runner from a 'bad gait that shall not be repeated' to a 'new & improved' gait?  And just what is this 'new & improved' gait built from?

Steve's article defines many of the characteristics of such a gait.  But the article doesn't say a single word about how to obtain such a gait (aside from giving 'cues').

For the ex-runner, the new gait must meet just one requirement:
  1. Running should be comfortable, both during, right after, and long after.
The only discomfort a healthy beginner runner experiences should come from just two sources: Muscle growth and muscle fatigue.  Nothing from the bones, joints, tendons or ligaments.

The gait must also be both minimal and sufficient to be considered running. Before we go any further, let's agree on a working definition for running:
  • Running is a means of human locomotion that is faster than 10:30 minutes per mile, where no more than a single foot is in contact with the ground at any time.
  • Jogging is running at speeds of 10:30 minutes per mile and slower.
I realize the above definitions may seem a bit arbitrary, but I use them to make it clear that we want to create runners, not joggers.

Where to start?  Let's start with the minimal motions every runner must perform:  Pushing the leg down and back, and lifting the leg up and forward.  How can we get ex-runners to perform these motions without running?  We start by running in-place, focusing first on the up-down aspects of dynamically supporting the body against gravity before considering locomotion.

Before we ask any ex-runners to run in-place, we first must ensure they are able to do so.  That means ensuring their legs are able to support their weight.  Remember, many non-runners could have trouble getting up a flight of stairs.  We must ensure their legs are ready for the stresses of running in-place.

The simplest way to ensure this is to start with squats.  Feet shoulder width apart, gradually lower the body so that the knees bend, extending over the feet, not out to the side.  Stop and pause briefly when the top of the thighs are level to the ground.  Then gradually straighten the legs until fully upright.

Basic leg strength is assured when a person becomes able to do 30 squats, taking 1 second down, 1 second pause, and 1 second up.  That is at least 90 seconds of squats (assuming no pause at the top).

Unfortunately, running uses only one leg at a time, so we also need to ensure each leg is individually strong enough.  And that means doing single-leg squats.  The important thing here is to never fall to the ground.  Start with a chair behind and support nearby.  Lift one foot clear of the ground (an inch or so is adequate).  Initially, slowly bend the knee to 45 degrees, then slowly extend, without pause.  Repeat until able to slowly sit in the chair and get out of it.  Continue until able to barely touch the chair and return to vertical 5 times in a row.  Then take away the chair, and repeat until the top of the thigh becomes horizontal.  When able to do 5 reps with each leg, the legs are ready for running in-place.

Now, this degree of leg strength is not needed by most runners.  But for safety, it is required for all beginners.  We want to ensure not only success, but also safety and comfort while running.  A little overkill in the strength department is both warranted and required.  As an example, consider the successful obese runner (I know several).

Before we start running in-place, we need to set a cadence.  This cadence should be compatible with fast distance running that is still safe and comfortable.  I've arbitrarily selected a cadence of 190 beats per minute, at the low end for fast distance runners.

Why emphasize such a high cadence at the very beginning?  All the evidence I've seen, from formal research and anecdotes, and particularly from barefoot runners, indicates that learning to maintain a high cadence is the single most difficult skill to acquire along the path to becoming a fast and injury-free runner.  So why not tackle the hardest aspect of running right at the very beginning?

If long-term success is the goal, to create a lifetime runner, then learning a fast cadence early can only smooth the path to future improvement.  But it does far more than just that:  When we get the beginners to actually run, we want each step to be as gentle as possible, with minimal stress, strain and impact.  A bunch of fast, tiny steps will be vastly better than fewer big ones.  So a fast cadence also helps us succeed in the near term.

So, get a metronome (I use the Seiko DM-50: It's cheap, and has a convenient clip for attaching to clothing) and set it to 190 bpm.

To run in-place, the beginner will start by standing vertically, with knees slightly bent.  Then alternately lift each heel off the ground in time with the metronome, keeping the toes always in contact with the ground.  Have them experiment with the degree of knee bend, finding the minimum bend that also produces minimal bounce.  Continue for at least 60 seconds.

As comfort and endurance grow (possibly over days), start lifting the toes barely off the ground, still at a 190 cadence, still for at least 60 seconds.  Continue to raise the toes higher, until they are at least 2 inches from the ground.  The only caveat during this phase is to avoid ankle stress: Maintain proper alignment of the knee over the foot, and stop before entering a fatigue state.

The goal here is to get lots of time at a 190 cadence, and also to gradually transition from bulk leg strength to plyometric strength.  Even as other skills are added, running in-place should be used as a warm-up tool by beginner runners.  Being able to run on cold muscles is something most runners take for granted: For a true beginner, cold legs are an invitation to a stumble and fall.

While initially running in-place, the beginner will be using a 'natural' arm swing.  When the beginner is able to raise the toes, it is time to start adding variation to the arm swing, so the beginner can directly experience the effects before actually using them running.

Let's pause the beginner lesson for a moment to return to Steve's article:  I believe the arm swing is both more critical and less formulaic than Steve describes.  The underlying physics involves complex coupled pendula (the arms and legs coupled via the torso while the legs generate propulsion).

Arms, legs and torsos each have different relative masses and lengths in different runners.  Making all this balance will generally require differences in arm swing for each and every runner.  A further complication is pace and grade/terrain, which also require different arm swings.  Finally, various forms of fatigue also add their own complications to the arm swing.  And it is seldom the case that the elbow flexion mimics the knee:  Often there is no elbow flexing, but it is important to know when to add some!

How can we tell if we are using the best arm swing for a given situation?  I've looked for physical explanations, but the best way I've found is to go by feel: Experiment with a very wide array of arm swings and swing dynamics, and see which helps make the gait faster, smoother, easier, and more comfortable.

It is important for every runner to practice the full catalog of arm swings in a variety of situations.  This permits that 'feel' I mentioned above to be developed and calibrated, to become a useful tool.

To return to the beginner lesson, the first arm swing situation occurs while running in-place.  The beginners should be encouraged to try every arm motion they can think of that is compatible with their running in-place.  Wide swings and narrow swings.  Arms straight, and thumbs touching shoulders.  Elbow locked, and elbow bending during each swing.  The  only caveats being that the arms must stay in the plane of the leg motion (never cross in front of the body), and the wrist should not go above the chin or behind the ribs.

Get each beginner to find the arm swing that makes running in place 'easiest', which in this context means lasting longer before stopping due to fatigue.  The final arm swing should be fairly forceful.  The goal here is to recognize that proper use of shoulder muscles can make running easier.

The next step is to add forward motion to the running in-place.  Again, we first must consider what is needed to ensure this motion will be safe and comfortable.

When running, the leg must swing forward and backward.  And most non-runners (excluding yoga and pilates enthusiasts) will have tight hip flexors, so tight that they can't physically run without pushing their butt out to permit rearward leg swing.  This impediment must be corrected prior to permitting a person to start forward motion.

So, we need to add gentle stretches for the hip flexor.  The simplest and easiest stretch is to go down on one knee, torso vertical, with both knees bent at 90 degrees, and the up leg having a hip angle of 90 degrees, and the down leg having no hip angle at all.  In this position, slowly push the up knee forward until resistance is felt, then hold for 5 seconds before slowly returning to vertical.  Repeat, but stop before the down knee becomes uncomfortable.  Some will require a pad under the knee to perform this stretch.

This stretch can be modified to kill two birds with one stone, to strengthen the leg while simultaneously stretching the hip flexor:  After finishing one hip flexor stretch, slowly extend the legs to return to vertical, then slowly drop to do another hip flexor stretch (that is, adding a stationary lunge).  Alternate legs every 5-10 reps.

At this point, the beginner can start moving forward at a 190 cadence taking steps that are no longer than 6 inches (shoes always overlap).  This short step ensures that no stabilization issues will arise.  Run in this manner as often as needed until the beginner is able to comfortably cover at least 400 yards (one loop of a track).

Before attempting longer strides, we must again ensure the body is properly prepared for the stress.  The above short-stride exercises will, in time, build the calves to the needed strength and ensure ankle stability.  The other issues to consider include knee and hip stability, which includes lateral mobility and strength.

The most important concerns in this area are the ITB and the knee ligaments.  Rather than detail the simple and effective stretches and limbering motions available for these areas, I'll try to bring this long post to some kind of a conclusion.

The above path to running is both gradual and gentle.  If at any point along the above path the beginner experiences pain, a medical referral is highly advised.

My own path started with a return to heel-striking, this time getting shoes that would keep my heel as far from the road as possible: The Mizuno Wave Creation did this job admirably well, while also providing significant energy return.

Soon after getting some of my speed back, I developed massive low back pain, which I initially attributed to learning to ride my bike in a deep aero position.  After months wasted tweaking my training, I finally went to a Sports MD and learned that my L5-S1 disc had evaporated.  While 90% of folks with this condition also have a subluxated L5 and require spinal fusion, my L5 had managed to stay in place, so I was able to use muscle relaxants and PT to regain full function.  The PT gave me the leg strength I had previously lacked.  But I was no longer able to tolerate any heel impact.

At this point, I got the Chi Running and POSE books, started to switch to a mid-sole strike, and got new shoes (Asics).  Much better, but the pain soon returned.  I then switched to a forefoot strike and more new shoes (Sacuony Kilkenny racing flats), and thanks to the extra shock absorption provided by my calves, I was soon running without back pain.

Then, last October, just a month after completing my first Olympic distance triathlon, I got a stress fracture in a second metatarsal.  While that healed over the winter, I studied everything I could find, to learn that I needed to greatly increase my cadence.  And get new road shoes with a bit more cushioning than racing flats (Adidas Chill M).

As I returned to running, I also decided to try barefoot running, but my tender soles were not up to it.  So I got another pair of shoes (Vibram Bikila).

At this point, after three stride changes and many pairs of shoes, I'm still rebuilding my speed:  At age 54, taking the winter off really killed my conditioning, something that takes longer and longer to rebuild every time.

But I'm having more fun running than ever, and I'm confident I'll be a runner (and triathlete) for the rest of my life.

And I also think I've found a straight forward way to get non-runners and ex-runners into running.  This method will not produce an ideal gait for any runner:   But it will get any runner to the point where they will be ready to listen to 'cues'.


  1. Bob,
    Heel striking pain is a similar problem that I’ve struggled with. I have spent so much money on 4 pairs of custom orthotics, PT, A.R.T, Trigger Point Therapy and deep tissue massage. Every time my heel strikes the ground on my left side I get a shooting pain in my ischial tuberosity area and the only way I can avoid it is by not hitting my heel when I am running. Doc mentioned something about the L-5 and S-I joint being locked but I still have no clue what that exactly means. I have switched to much more minimalist running shoes and that has helped so far but as you know old habits are so hard to break.

    Thanks for your insightful post and I’m so happy you are having more fun running then ever!

  2. @Angela: See a Sports MD and request PT from an athlete-centered practice (most PT focuses on getting you walking, not running).

    Such injuries and pain are certainly not uncommon. Right after I started running again I got severe knee pain under one kneecap. My Sports MD had X-Rays taken, and the radiologist diagnosed chondromalacia (which was very easy to see on the X-Rays). My Sports MD didn't recommend any treatment for the knee: Instead, he diagnosed the proximate cause as ITBS, and gave me some simple stretches to gently loosen the ITB and permit my knees to return to proper alignment, after which the chondromalacia quickly faded away. (Soon I'll do a post describing those stretches.)

    I especially recommend getting a Sports MD who is NOT also an orthopedist. Way too many orthopedists get stuck in the rut of "MRI -> Cortisone -> Surgery" as the cure for whatever ails you. A Sports MD who is not also a surgeon will prioritize non-surgical solutions, both pharmacological and PT.

    Plus, recent studies indicate that MRI is not a great diagnostic tool by itself. Things that appear to be present in a static MRI may not be a factor in pain felt while in motion. Athletes often fail to benefit from a diagnosis generated solely by static observation. If your MD doesn't watch you run before making (or confirming) a diagnosis, find one who does!

    After you get cleared to return to running by the MD and PT, be sure to get your cadence up and shorten your stride. Return to running on softer surfaces such as grass sports field or on a padded track. With my PT, I started gentle running during the last 2 weeks, so the therapist could watch my gait and tailor my therapy to compensate for observed weaknesses. That's when we got very serious about single-leg squats.

    The coolest side-effect of my current stride is that I'm able to run on a treadmill! As a heel striker I would get knee pain within minutes of getting on the treadmill. I now find the treadmill to be an ideal running surface for me: Softer than the road, and without any downhill segments! (I'm still working on my downhill stride.)

    If you have any discomfort running on a treadmill, be sure to tell your Sports MD if the discomfort differs from your other running.