The very first picture is anathema to the trained eye: some overstriding, uncouth runner romping his way though the countryside, slamming his heels into the ground, straight-kneed as his legs stretch several feet in front of him. Now, this is probably just a stock photo from the archives, but it raises a good point: regardless of what we talk about regarding surfaces, running form is a big factor in injury prevention, and ignoring that is unwise. No matter what surface or shoe you wear, if you run like the man in the stock photo, you're going to have problems.
Moving on, the facts in the first part are more or less correct: there have not been any controlled studies linking soft surfaces to lower injury rates. This is a bit misleading because there have not been any controlled studies that link much of anything to lower injury rates. In survey-studies, such as this famous one, no link has been found between injury and shoe type, foot type, running surface, overall fitness, or weight. The only reliable predictors of injury are mileage and previous injuries. Some factors (like hip muscle imbalances) are related to injury risk, but have not been proven (yet) to have a causal relationship. Still other factors have evidence linking them to injury risk, but have not (yet) been shown to be related. This is where running surfaces belongs.
The NYT article moves on, touching on the surface-stiffness paradox: surprisingly, the impact force does not differ from surface to surface. For a given runner at a given speed, the foot will hit the ground with the same force no matter the hardness of the surface. Therefore, the article concludes, soft surfaces are no better than hard ones since the impact force is the same. Furthermore, the author concludes that soft surfaces are worse because they tend to be uneven, risking turned ankles and the like.
However, the surface-stiffness paradox is not the end of the story. Concluding that the injury risk is no different because the impact risk is no different presumes that injury is related to impact, which is not at all clear. Understanding how the body modulates impact on different surfaces is essential to understand the implications on injury risk, and this is not covered in the NYT article.
During impact with the ground, the leg muscles act like springs. Before impact, the muscles tense up to absorb the shock. Unlike mechanical springs, however, the body's muscles can have their stiffness altered. When you are running, the brain does this automatically. In order to optimize performance, the brain tenses the muscles to minimize the vertical motion of your center of gravity upon impact. Using feedback from the skin and muscles, the brain tenses the muscles so that the stiffness of the overall system remains the same. What is the overall system? It's the combined stiffness of the surface you are running on, the stiffness of the shoes you are wearing (if any), and the stiffness of your leg muscles. If the overall system stiffness is to remain constant, the body must modify the stiffness of the leg muscles in order to change it, since surface stiffness and shoe stiffness are out of your body's immediate control. So, when running on a soft surface, the leg muscles are tighter, and when running on a hard surface, the leg muscles are looser. And unlike the New York Times article suggests, the adjustment is not slow or gradual--it is instant. In a brilliant 1999 study, Daniel Ferris showed that runners adjust muscle stiffness before the first step onto a different surface. This is pretty cool. When you move from pavement to grass, your eyes see that you will land on the grass, your brain interprets that information, and relays it to the legs in the form of increased muscle stiffness. All this happens on auto-pilot.
The muscles and bones of the leg can be modeled as a system of levers and springs. A realistic model is illustrated on the left, and a more simple one is illustrated on the right. Unlike real springs, the muscles can be changed dynamically to have a different stiffness--imagine the effect of swapping out a loose spring in the model on the right for a stiffer one. This is the effect that different muscle pre-tension has on impact.
But impact force is not the whole story when it comes to collision with the ground. Daniel Lieberman, an anthropologist at Harvard University, has come into the spotlight recently for his work on barefoot running and footstrike style. More relevant to the current topic, however, is this important illustration of the effect of cushioning: in the graph below, the force at various times in a footstrike are plotted. The black line represents a barefoot heelstrike, while the red line represents a shod heelstrike. This is effectively a test of the effect of surface stiffness on the impact parameters (the surface just happens to be attached to the foot).
Irene Davis, a highly respected researcher, claims they do (and has data to back it up). But Benno Nigg (who is also highly respected) claims they don't, and also has data to back it up.
Impact loading rate is the most obvious difference between hard and soft surfaces, and I'll probably deal with the nuances of it in a separate post. But there are other differences too--the pressure distribution on the foot is one, and the evenness of the surface is another.
Plantar pressure distribution has been overlooked for a long time, since biomechanics researchers assumed impact control was passive for a long time. We now know that it is active--it involves dynamic feedback between your legs and your brain. The body senses what type of surface you are in contact with, and modifies muscle activity accordingly. There are some very interesting results from this: for example, it is easier to balance on a textured surface than on a smooth one. In running we know that the body attempts to minimize pressure on the sole of the foot. Sensory organs called mechanoreceptors detect the local pressure on the sole of the foot. If it is too high, the body modifies gait in an attempt to reduce it. When running barefoot, especially on a hard surface, your body forces you to take choppy, quick steps. A quicker stride frequency means less impact force per footstrike, so the pressures drop.
In a study published in 2008, Vitor Tessutti and coworkers measured the pressure at different points on the sole of the foot during running on natural grass and asphalt. Predictably, running on grass resulted in longer contact time with the ground and lower peak pressure.
So if the body can adjust to any surface/shoe stiffness by modifying muscle stiffness, and high plantar pressures should be avoided, why not run with pillows strapped to your feet? Well it turns out that the body can't adapt to any surface/shoe stiffness, and that some plantar pressure is a good thing.
One emerging idea regarding muscle and surface stiffness holds that the body has a Zone of Optimal Leg Stiffness. Impacts that occur when the muscles are forced to operate outside this zone are definitely uncomfortable and probably injurious. The zone of optimal leg stiffness is easy to demonstrate using extremes: First, imagine sprinting as fast as you can, barefoot on asphalt. Of course, this would be a painful and regrettable experiment. Second, imagine trying to sprint in a high-jump pit. Not as painful, but nearly as uncomfortable. In the first case, the leg muscles can't be loose enough, so the legs are forced to absorb too much impact too fast. In the second case, the leg muscles can't be tight enough and cannot act as effective springs to return the energy from the impact.
I suspect that the zone of optimal leg stiffness changes as muscles become more fatigued--so at the end of a 90 minute run, your body is less tolerant of various surface stiffnesses than it was at the beginning. This is purely speculative, but from good old-fashioned experience, I (and most other serious runners) can tell you that 13 miles on pavement feels less comfortable than 13 miles on dirt and grass. I also suspect that most runners encounter the lower end of their zone of optimal leg stiffness more often than the upper end--that is, I suspect that the surface/shoe combination is too hard more often than too soft. Again, only a suspicion supported by experience. But if I'm correct, a soft surface is a much safer bet, as you will be in less danger of being outside that zone of optimal stiffness. Fortunately, there are some very smart folks over at the podiatry arena who are on the same page as me. Too bad the New York Times didn't interview them.
With regard to plantar pressure, emerging research is showing that, although the body tries to avoid high peak pressures, it is also used for feedback. "Proprioception" is the $5 word for the dynamic feedback between the plantar mechanoreceptor cells and the brain. As I mentioned earlier about the ease of balance on a textured surface, better feedback about the surface from the soles of your feet is a good thing. There is some evidence (though not by any means GOOD evidence) that part of the benefit in taping sprained ankles is not from the support, but from the increased proprioceptive feedback. A 2003 study found that soccer players with a textured insole had much better sensory feedback versus a smooth insole. Unfortunately, thicker cushioning means less proprioceptive feedback. Alas, this is a topic for another day--shoes, insoles, bare feet, etc.
Getting back to the topic at hand, one final advantage soft surfaces have is their unevenness--this is why I brought up proprioception in the first place. An uneven surface puts different stresses on the body, and the proprioceptive feedback from the varied surface will result in muscles being tensed slightly differently. Uphill, downhill, bumpy, and smooth surfaces all stress the body slightly differently. One step might stress the medial side of the foot more, while the next stresses the lateral side more. Accordingly, the muscles of the leg are stressed slightly differently. Sports orthopedists often blame hip and knee injuries on running on the same side of a cambered road, inferring that the same stresses over and over are a bad thing. So why not switch up the stresses on your body? Repetitive stress is a bad thing in many fields--if I keep typing up this article much longer, I'll get a repetitive stress injury. To my knowledge there are no studies looking at how slightly uneven surfaces affect injury or running gait. Most biomechanical studies are done in a lab on a smooth concrete surface, so the dearth of studies is completely understandable. It's hard enough in a lab, imagine doing it on a trail. This last paragraph is by far the least supported in this whole article, so toss it out if you must--but plantar pressure, impact loading rates, and leg stiffness are the more meaty arguments anyways.
So in closing: why is the New York Times article wrong? Because all surfaces are NOT the same. Hard surfaces may can push the body outside of its zone of optimal leg stiffness. They also increase the peak pressure on the sole of the foot, which we know is a bad thing because the body does all it can to avoid high peak plantar pressures. Finally, uneven surfaces stress the body in a more varied manner, lessening the risk of repetitive stress/overuse injuries. Saying there is "no evidence" that soft surfaces are better is a gross mischaracterization.