Humans and other animals often walk economically so that they don't waste energy, but when learning new movement patterns, does the nervous system select movements that reduce energy use? Natalia Sánchez and Surabhi Simha, as part of a collaboration between labs at University of Southern California, USA, and Simon Fraser University, Canada, wondered how humans learn a new stepping pattern when they walk on a split-belt treadmill that has two belts, so that one leg steps at a slow speed and the other leg steps with a fast speed. The researchers focused on how humans learn to adjust the difference in the length of the left and right legs’ steps. From a clever analysis of how the treadmill can transfer energy to the person walking on it, the team predicted that as steps on the fast treadmill belt became longer relative to the steps on the slow treadmill belt, the treadmill would assist the walker by providing them with additional energy, allowing the walker's muscles to decrease the work that they were doing, so that their metabolic energy consumption would decrease.
The team recruited 16 healthy young volunteers to walk on the split-belt treadmill while the left belt ran at 0.5 m s–1 and the right belt ran at 1.5 m s–1 over a range of step lengths across values where the step on the slow belt was longest to where the step on the fast belt was longest. In order to test how treadmill work, muscle work and metabolic energy changed with step lengths, the team placed a screen in front of the treadmill that showed the position of the volunteers’ feet as they walked and instructed the volunteers to step at specific locations, thereby enforcing different values of step lengths. Following the guided stepping trials, the volunteers then freely adapted their step lengths so that the team could determine their self-selected step lengths. From the forces that the volunteers exerted on the treadmill belts, the researchers calculated the mechanical work that the treadmill performed on the person and the mechanical work that the person's muscles produced. Last, the team calculated the volunteers’ metabolic energy use from the amount of oxygen that the volunteers inhaled through a mask.
Consistent with the team's predictions, the amount of work that the treadmill performed on the person increased by 28% as the step on the fast belt became longer than the step on the slow belt, while the amount of work that the person's muscles produced decreased by 13%. This suggests that the person walking on the treadmill used the additional energy provided by the treadmill to reduce the work that their own muscles must do. To understand whether the flow of energy from treadmill to person resulted in reductions in metabolic energy consumption, the team compared the amount of muscle work with the volunteers’ metabolic energy and found that metabolic energy decreased with decreases in muscle work. This suggest that the treadmill's work conferred an energetic benefit to the volunteers.
When the team looked at how metabolic energy changed with differences in the step lengths and how the volunteers adjusted their step lengths throughout the learning period, they found that the step length adjustments reduced the energy cost by 14%. This suggests that humans can learn to take advantage of the assisting work that a split-belt treadmill can provide by learning a new movement pattern to capitalise on the energy savings. Although the adapted step lengths that the volunteers learned were not necessarily energetically optimal, they can learn to adapt their step length to reduce energy costs. These results could help patients that are recovering from neurological damage by using the split-belt treadmill to improve their walking symmetry and energy economy.