Shoulder Rotation May Improve Running Efficiency
The preliminary results below suggest thoracic (shoulder and trunk) rotation may be important for reducing the energetic demand while running, allowing us to run longer at a given speed. Thoracic rotation is likely important at higher speeds given the result showing people naturally increasing their rotation as they increased speed coupled with the finding that suggests rotation may reduce cost of transport (VO2) during running.
The preliminary results below suggest thoracic (shoulder and trunk) rotation may be important for reducing the energetic demand while running, allowing us to run longer at a given speed. Thoracic rotation is likely important at higher speeds given the result showing people naturally increasing their rotation as they increased speed coupled with the finding that suggests rotation may reduce cost of transport (VO2) during running.
How shoulder rotation affects our movement
Have you ever wondered why our shoulders swing back and forth while running? I was curious whether thoracic rotation - or rotation of the shoulders - helps us move more effectively. I directly tested whether thoracic rotation reduces our energetic demand to move by measuring people's oxygen consumption (VO2) in three different conditions: 1) preferred rotation, 2) no rotation, and 3) forced rotation (see gif to left). I also tested this in three different contexts, during walking, uphill walking and running to see if shoulder rotation plays a greater role in specific tasks. |
Shoulder Rotation Increases with Speed
I found that when I asked subjects to go from a walk to a run they significantly increased the amount of shoulder rotation as the speed increased. For reference most adults walk around 1 m/s and start to jog around 2 m/s, the maximum speed of 4 m/s here is just shy of 9 mph. The graph on the right shows us how much shoulder rotation I observed in subjects as they walked (circles) at really slow speeds all the way up through running (squares) at 4 m/s. Different colors represented different individuals. |
How Shoulder Rotation Affects Efficiency
In the above figure I am showing the amount of shoulder rotation (top row of graphs) and the energy consumption as a measure of relative VO2 (bottom row of graphs). Each box and whisker plot shows the distribution of the ten subjects with the whiskers representing the minimum and maximum values of the data (excluding outliers which are marked with a red asterisk), the box represents the middle 50% of the data and the red line near the center of each box represents the median or middle value for that condition.
In the case of walking (left-most graphs), we notice that subjects prefer to rotate only a little more than they do when asked to not rotate at all and they also don't seem to change the amount of preferred rotation after they have been asked to force their shoulder rotation or stop rotating. When we look at the VO2 for walking we notice that the distribution or box jumps upward for the forced rotation condition and the preferred rotational conditions are both relatively similar to the no rotation. This helps to explain why we don't strut everywhere we go, because it takes significantly more energy to force shoulder rotation while walking.
When we look to incline walking (middle graphs), we see a similar trend in the amount of rotation observed during the no, forced and preferred trials. This time keep in mind the two preferred conditions are preferred rotation while walking uphill (second distribution from the right) and preferred rotation walking on a level surface (right-most distribution), it appears people like to rotate a little more when walking uphill. What is interesting though is that when we look to the energetic cost of swinging our shoulders while walking uphill we see that no rotation, forced rotation and preferred rotation all have similar distributions for VO2. This suggests there is no advantage to swinging your shoulders going uphill and that we can walk up the hill just as efficiently with little to no rotation as we can forcing our rotation.
Finally when we look to running (right-most graphs), we see again that subjects prefer to have much higher amounts of rotation than during our walking trials, and it is about double what we see when subjects are asked not to rotate. What's fascinating is that when we take a look at the energetic demand to run under these rotational conditions I found that forced rotation actually required the least amount of energy.
In the above figure I am showing the amount of shoulder rotation (top row of graphs) and the energy consumption as a measure of relative VO2 (bottom row of graphs). Each box and whisker plot shows the distribution of the ten subjects with the whiskers representing the minimum and maximum values of the data (excluding outliers which are marked with a red asterisk), the box represents the middle 50% of the data and the red line near the center of each box represents the median or middle value for that condition.
In the case of walking (left-most graphs), we notice that subjects prefer to rotate only a little more than they do when asked to not rotate at all and they also don't seem to change the amount of preferred rotation after they have been asked to force their shoulder rotation or stop rotating. When we look at the VO2 for walking we notice that the distribution or box jumps upward for the forced rotation condition and the preferred rotational conditions are both relatively similar to the no rotation. This helps to explain why we don't strut everywhere we go, because it takes significantly more energy to force shoulder rotation while walking.
When we look to incline walking (middle graphs), we see a similar trend in the amount of rotation observed during the no, forced and preferred trials. This time keep in mind the two preferred conditions are preferred rotation while walking uphill (second distribution from the right) and preferred rotation walking on a level surface (right-most distribution), it appears people like to rotate a little more when walking uphill. What is interesting though is that when we look to the energetic cost of swinging our shoulders while walking uphill we see that no rotation, forced rotation and preferred rotation all have similar distributions for VO2. This suggests there is no advantage to swinging your shoulders going uphill and that we can walk up the hill just as efficiently with little to no rotation as we can forcing our rotation.
Finally when we look to running (right-most graphs), we see again that subjects prefer to have much higher amounts of rotation than during our walking trials, and it is about double what we see when subjects are asked not to rotate. What's fascinating is that when we take a look at the energetic demand to run under these rotational conditions I found that forced rotation actually required the least amount of energy.