Comparison of Harvard Study Results with Chi Running - Chi Running

Comparison of Harvard Study Results with Chi Running

Posted by Danny Dreyer on Sun Feb 14th, 2010, 6 comments

Since the release of Dr. Lieberman’s study I’ve had a few requests from folks wanting me to show my test results from a gait analysis I had done at the UVA human performance lab. They’re curious to see how my Ground Reaction Force (GFR) chart compares with the one shown in the Nature video showing a barefoot runner (I’m assuming Lieberman himself) landing on a strike plate with a GFR chart beneath.

In the tests I did at UVA, I was wearing three different pairs of shoes (I was never tested barefoot): tai chi shoes, NB-MF800′s and NB-790′s. I’ve pasted in three graphs below. The first two are of a barefoot runner running with a heel strike, a barefoot runner running with a forefoot landing and the third graph is of me with the graphs of all three of my tests overlaid onto the same chart in three different colors.

GRF for a Heel Striker

GRF for a Heel Striker


GRF for a Forefoot Striker

GRF for a Forefoot Striker


GRF for Danny

GRF for Danny

Notice the differences in GFR between the three runners shown as a multiple of Body Weight measured during the initial weight-bearing phase. The Barefoot runner had a multiple of 1.85 x Body Weight. To make the comparisons fair, let’s assume all three tests were done with a runner who weights 150 lbs., that would mean (with a GFR of 1.85 x Body weight) that his heels were absorbing a force of 277.5 lbs. upon impact. Let’s say that his heel has an area of 5 sq. inches. that would mean that his heels were experiencing an impact of about 55.5 lbs./sq. inch. Ouch!!!

In the second figure, the runner is landing on his forefoot and his GFR is 2.64 x Body Weight during his support phase which figures out to 2.64 x 150 lbs. = 396 lbs. Now let’s say that the area of his forefoot is approximately 16 sq. inches. That would mean that his foot would be absorbing about 24.75 lbs./sq. inch. … or less than half of what the heel striker feels.

In the third figure I’m landing in a midfoot (or fullfoot) strike which means that the entire bottom of my foot is landing as my foot hits the ground. My GFR was measured at 2.47 x Body Weight or 2.47 x 150 lbs. = 370.5 lbs. Now, if the surface area of the entire bottom of my foot is roughly 30 sq. inches. (I’m a size 9.5 shoe), that would mean that the impact felt by my feet would be approximately 12.35 lbs./sq. inch. or half as much as the forefoot striker.  That’s less than 25% of the impact per square inch the heel striker feels and half of what the forefoot striker feels. You can also see that my GFR was the same no matter which shoes I was wearing. I imagine that it would still have been the same had I been tested barefoot… possibly implying that there’s something to be said for working on improving your running technique. It’s about the runner, not the shoe.

I’ve used 150 lbs. as a standard weight for all three runners because I don’t know how much Dr. Leiberman weights. I also used the size of my feet in my calculations because I also don’t know his shoe size. BUT, the point I’m trying to make here is that whether or not you run with shoes on, your impact with the ground will be most if you land in a heel strike because there’s a lot of force going into a relatively small area of your foot. Subsequently, if you land in a forefoot strike, you land with more area of your foot touching the ground and it spreads the force of impact over a larger area thereby reducing the amount of impact per square inch. And lastly, if you land in a fullfoot strike, you spread the impact out over a much larger area and lower your impact per square inch even more still.

Since greater impact with the ground can most likely be directly related to greater incidence of impact injuries, this all boils down to one big question. How do you lower your impact with the ground so that you don’t hurt yourself when you’re running. My premise is that although some types of shoes can reduce shock and impact, it’s the runner who, in the end, is responsible for learning how to manage his or her impact with the ground in a way that consistently works. This is the need that is presented to every runner and it is one of the main reasons why Chi Running was developed.

Dr. Liebermans study was a good beginning into seeing the difference in impact forces between shod and barefoot runners. Next, I would like to see a study comparing the difference in impact between heel strikers, midfoot strikers and forefoot strikers; with bare feet, with minimal shoes, racing flats, and with your basic high-heeled running shoes.

There are times to run with a forefoot strike. There are times to run with a fullfoot strike. And, there are times to run (believe it or not) with a heel strike. I’ll cover these in a future blog.

Run well…be well,



  • minimal shoes,
  • barefoot running,
  • dr. lieberman,
  • forefoot strike,
  • ground reaction force,
  • harvard study of barefoot and shod runners,
  • mid-foot strike,
  • racing flats

6 CommentsLeave a comment below


From reading the graph, isn’t the barefoot runner’s force of 1.85 x body weight when heel striking occuring at the spike at the beginnning of the graph?  Where the red circle is?  That spike is absent from the forefoot strike graph and your gait analysis graphs entirely (one of the arguments against heel striking I believe), and even in the heel striking case, the maximum GRF is about the same.  It’s the initial impact when heel striking that’s the issue, when it goes from zero to 1.85 x body weight, right?  With the forefoot and midfoot strikes, there’s a smoother curve in which the GRF builds up over time rather than spiking right at the beginning and then falling into nearly the same curve.  Am I interpreting that correctly?


Hi Daniel,
Yes, you are absolutely interpreting this correctly. The force is spread out over a much longer period of time…AND over a much larger area of impact, thus reducing the chance of any impact injury to the foot. I will say that landing on the forefoot has caused some people to experience soreness in their metatarsal heads and sometimes pain on the dorsal side of the foot.

Landing with a midfoot (fullfoot) strike also allows the runner to avoid loading the calf muscles and achilles tendon upon impact, thus reducing the chance of any lower leg injuries.


Dan Connelly Feb 16th, 2010 09:05am

The foot pressure discussed here may be a factor in foot injury, but the knee sees the total force, independent of the area of the foot over which it is distributed.

The intriguing factor to me from Lieberman’s paper, and evident in the data here, is the heel impact results in a substantially (6.5 times) greater mean rate of loading.  This suggests to me the leg muscles may be less able to adapt to the changing load conditions, and thus more of the force may end up being born by tendons or ligaments, even for the same peak force.  I don’t know if this sort of analysis stands up to anatomical scrutiny, but to my simplistic perspective it seems reasonable.

Lieberman suggests mid-sole strike would fall between heel strike and toe strike, but the data here seem to show the mid-sole strike may be much closer to the toe strike in the rate of loading.

Unmentioned in the paper is the relationship between cadence and peak force.  Simple physics suggests the total impulse will be inversely proportional to cadence, so the relatively high cadence promoted by Chi running relative to the longer stride, lower cadence of a traditional heel strike should provide additional benefit.

Anyway, thanks for the great blog, and for your fantastic work with Chi Running!

Thanks for posting this, Danny!  I agree with the other comments that Lieberman emphasizes the rate of rise of the GRF curve, rather than the peak level.  The sudden impact of heel-strike can’t be a good thing.

Has anyone discovered what is actually plotted? It it the vertical component of the force or the magnitude of the force. I would like to see the vertical and horzontal components, the vertical is related to the stress as discussed above and the horzontal component should show the breaking effect of heel striking.

I agree that the rate of the vertical ground reaction force is critical.  Tendons and ligaments are viscoelastic.  That means they have different material properties when loaded at different rates, like Silly Putty.  Stretch Silly Putty slowly and it will stretch forever without breaking.  Stretch it quickly and it will break.  Our tendons and ligaments are similar.  The slow loading rate of mid and fore foot strikers allows their tendons and ligaments time to stretch and absorb the impact.

Excellent question about what is actually plotted.  I assumed it was the vertical component but it doesn’t specify.

More videos are available at Nature and NPR websites.



I like your blog, but your analysis of foot strike here misses the point entirely. As some of the comments have suggested, it’s the RATE of force application that matters in the Harvard study, not the total force applied. This is easy to see if you consider that I can hold a barbell of, say, 500 pounds on my shoulders, and then roll back on my heels without the slightest fear of injury. The total pressure [psi] on my foot in that case exceeds that of all your cases with significant margin. But I’m at no risk of injury because the RATE of loading is so slow.

Also, your analysis seems to suggest that someone with a larger, flatter foot, or someone who ran in great wide flippers (large impact surface area) would lessen his chance of injury, which is of course nonsense.

What’s significant about the Harvard study, and what proponents of barefoot running will all tell you, is that it indicates that the anatomical structures of the unshod foot (the arch, pronation and dorsiflexion) are highly efficient attenuators of the RATE of force loading. This is why the motion of the foot hitting the ground is so beautiful and complex—it’s doing a lot of fancy engineering… footwork…(excuse the pun) to slow down the rate at which that 300-400 pounds gets applied and distributed through the body.

No doubt it will occur to you that this precisely ChiRunning’s strength as well: your emphasis on alignment and posture throughout the spine, core, and lower extremities are all very useful ways of lowering the rate of loading throughout the body. And of course this is why the metaphor of “chi” or “softening” resonates so strongly. The strength you gain through softening (iron and cotton) is real and measurable.

However, it does not follow that because ChiRunning can effect this load rate attenuation that running unshod or in minimalist footwear does not ALSO have the same benefit—in a different way. They are two sides of the same coin. And the science coming out recently is very strong in support of the assertion that being able to “feel” the ground dramatically reduces impact transients and, hence, the overall torque loading of joints. Said less like a scientist ... running barefoot is very Chi!


BTW, I worked one-on-one with Kathy Griest a few times here in Pasadena and though could not afford a continuing arrangement I very much enjoyed our sessions. I am an avid study of running mechanics and appreciate the work you’ve done in the field. If you have any questions about what I’ve said here, feel free to contact me.

Hi Isaiah,
Thank you for your great insight on my interpretation of the rate of force and how my explanation was a bit incorrect. In the GFR graphs done at the UVA lab it shows the rate of impact on my footstrike as a much more gentle curve than the steep slope or spikes of other runners tested. This would bear out what you say is going on…that I’m able to dissipate the impact over a longer period of time and thus reduce it’s effects on my body, whether it’s taken at my heel or in my legs.

What are your thoughts?

Please enter the word you see in the image below:

Top 100 Runnning Blogs award

Over a million happy runners can't be wrong.

Watch our FREE video series and run easier and better than you ever imagined. Feel the difference in your next run!