As people go down the street, they settle naturally into a comfortable walking pace. As they walk a little faster, nothing changes dramatically -their stride rate increases somewhat and several leg muscles contract more strongly - but if they speed up a little further, they eventually shift into a running gait, with an abrupt decrease in the time their feet touch the ground and a large increase in the maximum forces. Mechanically, the two gaits are very different. But is the gait transition really as discrete for the brain as it appears to be visually? Maybe one `program', encoded in the brain and spinal cord, controls both, and the switch is due to some external cue, such as sensory feedback from the higher forces at high speeds. Or maybe running and walking are controlled by two entirely separate programs.

Answering this question requires some fancy math. G. Cappellini, Y. Ivanenko and their colleagues at the Scientific Institute Foundation Santa Lucia in Rome have been working for several years with a technique that lets them simplify the complex activation patterns of large numbers of muscles into a few relatively simple groups that act together. In their current study, they examined the activity of 32 muscles in the right legs, trunks, and right arms of walking and running humans. The researchers had their subjects run and walk at normal comfortable speeds, but also had them walk extra quickly and run very slowly, so that they had data for both gaits at the same speeds. Then,because the stride rate is faster at high speeds, they scaled all the times during a step when different muscles were active, relative to when the right foot hit the ground. Finally - this is the fancy bit - they ran the activity data through a statistical procedure called principal components analysis,which reduced the massive data set into activity patterns for five groups of muscles that acted together in consistent ways and found that the same five groups work together in about same way during both walking and running.

Each group of muscles showed an activity peak somewhere during the step cycle. For example, the hip and knee extensor muscles act together just after the foot hits the ground, and various muscles in the trunk turn on together as the leg begins to swing forward. Four of the five activity peaks showed up at the same time at the same speed, whether the subjects were walking or running. In both gaits, the peaks shifted earlier in the step cycle as the speed increased. But there was one rogue group which appeared to do different things in the two gaits at first sight; group number two, mostly muscles that point the toe, turned on late in the step during walking but early during running,suggesting that different programs were controlling that group.

Or so it appeared at first. Actually, it depends on how you define `early'and `late.' When the researchers looked relative to when the toe comes off the ground, then the anomalous group two patterns lined up for both gaits. The toe pointing muscles tend to come on at a consistent time just before lift off,unlike the other groups, which seem to be timed relative to heel strike.

Since all groups for both gaits are consistent relative to something, it appears that one program probably controls both gaits. The key difference between walking and running, once they get started, seems to be sensory information about the step cycle, not something centrally controlled.

Cappellini, G., Ivanenko, Y. P., Poppele, R. E. and Lacquaniti,F. (
). Motor patterns in human walking and running.
J. Neurophysiol.