Running efficiency: boosting speed and power by reducing the energy cost of your movement


To increase your max running speed you need to reduce the energy cost of your activity. Here’s how: If you use running in your sport – whether it be football, basketball, tennis or even running itself – one of your key training goals should be to reduce the energy cost of your running. If you do so, any given speed will represent a lower fraction of your maximal cost of movement, and thus will be easier for you to establish and sustain.

So how do you decrease your energy cost? To understand the fundamentals of cost-cutting while running, we should first take a look at the research of the late, great, Professor Dick Taylor of Harvard University. Taylor and his colleagues worked with a variety of different animals, training them to run on treadmill belts while wearing oxygen-collecting masks. By calculating how much oxygen the animals were using in the course of this activity, Taylor and co could determine their energy cost. As a general rule, as long as an animal is not moving too quickly, each cubic centimetre of oxygen it consumes is associated with the production of about 20 joules of energy; it matters not whether the animal is a mouse, a human, an eland, or an elephant (all of which Taylor studied, although the elephants broke the laboratory treadmills and had to be studied from golf buggies as they trotted around the zoo!). Calculating animals’ oxygen usage during exercise is actually a simple matter. Just two pieces of information are really required: the amount of air expired into the mask or hood, and the oxygen content of the expired air. Let’s postulate, for example, that an animal is expelling 12 litres of air per minute into his/her mask or hood, and the expired air is 19.95% oxygen by volume (according to the analysing equipment). Since normal air (the stuff the animal is breathing in) is 20.95% oxygen by volume, he must be utilising about 1% of 12 litres, i.e. 0.12 litres (120 cc) of oxygen per minute. Thus, the animal would be expending 120 x 20 = 2400 joules of energy per minute. Although Taylor worked with a variety of animals, some of his best work was carried out on the treadmill with small ponies. These creatures are especially interesting because they can employ three quite different gaits, walking when they are moving slowly, trotting at medium speeds, and break into a gallop when they want to move very quickly. Each of these styles of movement has a unique footfall pattern, and Taylor and his student, Dan Hoyt, were able to determine the energy costs (in joules per minute) associated with each (1). One of the most surprising things about the Taylor-Hoyt findings was that the metabolic cost of movement at a specific speed was entirely dependent on the chosen gait. (One might naively expect that cost and speed would be fairly tightly related, but they were not.) For example, when the ponies moved at velocities slower than about 1.7m per second (roughly 16 minutes per mile), walking was by far the cheapest way to travel and trying to trot at such speeds boosted cost by as much as 67%!

Learning from the donkeys

At speeds ranging from about 1.7m-4.6m per second, trotting was by far the best alternative method of transportation. For example, at a pace of around 3.5m per second (7:40 per mile), trotting cost almost 25% less than galloping. (3.5m per second was too fast for walking, so the walking cost could not be determined.) At all speeds above 4.6m per second (5:48 per mile), galloping was the cost-effective way to go; at 5.25m per second, for example, galloping consumed energy at a rate roughly 16% lower than trotting. We’ll come on to how these findings can improve your running in a moment. Before we do that, though, it’s important to note that when Taylor and Hoyt plotted energy cost as a function of speed (with speed on the x-axis), each gait (walking, trotting, or running) produced a ‘U-shaped’ graph (a horseshoe pattern with a low point and two curving arms rising on either side of it). In other words, each gait had a speed associated with a minimum energy cost per metre; moving faster than that speed raised the cost of movement, but so did moving slower. For each gait style, there was one speed which was most economical. This may help to explain why some athletes have trouble slowing down during their running workouts. They may have a particular running speed which is most efficient in that it is associated with a low energy and oxygen cost along with a modest heart rate. Attempting to run slower than this high-efficiency pace may paradoxically raise the energy cost and feel tougher! The natural tendency in such cases would be for the athlete to speed up in order to feel more comfortable during the workout; but such accelerations could, in turn, raise the risk of overtraining (remembering that the impact forces experienced by the legs increase as a function of running speed).

Spontaneous selection of gaits

The phenomenon of slow paces feeling difficult might be particularly applicable to athletes with a relatively high percentage of fast-twitch fibres in their leg muscles; if they routinely train at fast speeds, making primary use of their fast-twitchers, slowing down might cause the fast-twitch fibres to take a break, leaving the athlete to rely on the less-used and therefore less-efficient slow-twitch cells. Returning to the ponies, their most efficient walking speed was about 1.25m per second, while their most economical speeds for trotting and galloping were just above 3m per second and just above 6m per second respectively. When Taylor and Hoyt filmed the ponies as they moved freely and spontaneously around the paddock, they found that the creatures spontaneously selected almost precisely those speeds for each of the three gaits. Were the ponies most efficient at certain walking, trotting, and galloping speeds because those were the speeds they repeatedly practised, or were they somehow gravitating toward the ‘locked-in’ paces which were naturally most economical for them, given their innate physiology and anatomy? We simply don’t know, but it is clear that humans tend to become most efficient at the speeds they most frequently practise. That is one reason why marathon runners often have great difficulty sustaining their goal pace for the full duration of a race: during their long training runs, they usually employ a speed which is slower than hoped-for race pace and thus build up considerable efficiency at this training pace. The efficiency at goal race speed, however, may remain untouched by these slower-than-race-pace amblings, and thus race day will frequently contain some unwelcome surprises. While goal speed might be maintained for 15 miles or so, the runner, when tired, will naturally gravitate toward the slower, more efficient speed with which he is so familiar.

Regardless of whether or not practice made perfect for the ponies, it was certainly good policy for them to move at their lowest-cost velocities, as R McNeill Alexander, of Leeds University, has pointed out (2). Such movements save energy and allow more resources to be funnelled into useful activities like recovery, growth, immune-system function, and so on. For athletes, it makes paramount sense to spend as much time as possible during training at competition-specific speeds in order to optimise efficiency in competitive situations.

However, there is another extremely interesting aspect to this story: since walking, trotting and galloping produce U-shaped curves, with the minimum cost of trotting occurring at a higher speed than walking and galloping a higher speed than trotting, the right arm of the ‘U’ for walking tends to cross over the left arm of the ‘U’ for trotting and ditto for trotting and galloping. Animals who want to go somewhere should avoid these crossover-point speeds, since they represent inefficient movement for each specific gait and may be associated with indecision about how to move. In Alexander’s brilliant analysis (p66), he points out that if a pony decided to go from point A to point B at about 1.7m per second, that would be a costly mistake because this is a crossover point between the cost of walking and running, with either gait strategy producing a rather expansive expenditure of 400 joules per metre. The pony could choose to walk or trot, but either pattern would increase metabolic cost. A far better strategy for the little horse would be to trot the distance at 3.2m per second, slowing down to walk at about 1.2m per second when tired. With this strategy, he would cover the distance in a faster time: even if he divided the gait patterns equally, the average speed of 2.2m per second would still be greater than the crossover speed of 1.7. He would also be performing at lower cost since trotting at 3.2m per second and walking at 1.2m per second each cost just 300 joules per metre! Heavens! Is it possible that US sports writer Jeff Galloway is right? For years, Galloway has been recommending the strategy of alternating between running and walking during marathons, arguing that this can reduce fatigue, enhance endurance, and even improve finishing time. Could it be that Galloway’s proteges are using their most-efficient running speeds and most-economical walking paces and thereby reducing the cost of running the race, much like our ponies? If they slowed down their running pace and attempted to cover the whole distance by running, would they augment energy costs and glycogen depletion and finish in slower times? Not likely! As it turns out, the cost crossover point in humans for walking and running occurs at about 2m per second (13:24 per mile). In other words, if you range above 2m per second, running is far more efficient than walking, while walking is more economical at slower speeds (3). Now, 13:24 per mile – the crossover point – produces a marathon finish time of about 5:51. Almost every marathoner finishes at a faster time than this – i.e. at a tempo where running is far more efficient than walking. Thus, the attractiveness of walking as a within-race strategy should be nil. Let’s say you did plan to run the marathon in 5:51, however. Would it be better to run half the race at 3m per second and walk the other half at 1m per second, close to the most-efficient points for each gait pattern; the finish time would remain 5:51, but if the energy cost were lower, more glycogen might be saved, lowering agony levels and maybe even permitting a burst of energy near the end of the competition; heck, you might even finish in 5:49! At 3m per second, you would be burning about 300 joules per metre, and at the walking pace of 1m per second, the energy cost would be 150, giving an average of 225. In other words, combining walking at 26 minutes per mile and running at 9 minutes per mile would be less expensive than running the whole thing at 13:24 pace and would lead to an identical finish time.

Don’t mimic the ponies

What about the four-hour marathon finisher who completes the 26.2-mile distance at an average pace of 2.9m per second? Could he/she run half the race at 3.9 and walk the other half at 1.9? Again, there might be some cost savings: 3.9 would cost about 300 joules per metre, while 1.9 would cost 260, and the average – 280 – represents a small decrease in cost, without sacrificing finishing time. There’s just one problem, though: 3.9m per second translates into a 6:52 per mile running pace. Could four-hour marathoners really handle this pace for half the race? Could three-hour marathoners tolerate even faster speeds in order to build in some efficiency via walking? If they could, wouldn’t they be better off – from a finishing-time standpoint – running the other half of the race at a speed faster than the walking tempo of 1.9m per second? The cost of doing so would not be prohibitive. But if the runners used the commonly recommended ratio of 4:1 or 5:1 (four-five minutes of running for every minute of walking), the cost saving would be so small that it would make little sense to incorporate walking. An athlete running the race at 4m per second (6:42 per mile pace), for example, would – if the need arose to slow down – be far better off jogging at 2.5m per second than walking at the same velocity. Walking would cost much more, and dipping down to a tempo at which walking saved energy (below 2m per second) would slow finishing time unreasonably. So perhaps humans shouldn’t mimic ponies during marathon running. Nevertheless, Alexander’s research has led to some important conclusions about training. That’s because, in addition to his fascinating work on the metabolic cost of exercise, Alexander has also determined the actual amount of work that muscles perform during the act of running. By pursuing the problem of muscular work, Alexander has uncovered a key paradox: when humans run at moderate speeds, the metabolic cost of running is often about four joules per kilogram of body weight per metre; however, the mechanical cost (computed from the forces exerted against the force platform) associated with such running is only two joules per kg weight per minute. In other words, humans run with an efficiency of about 50%, two of the four joules expended per minute being used for exerting muscular force and the other two lost as ‘waste’ heat.

Of mice and men

If that makes it seem like humans are energy wastrels during running, bear in mind that it could be much worse. For example, mice run with an efficiency of only 3%, which may help to explain why they seem always to be looking for food. Generally, larger animals (like humans) tend to be a little more efficient during running than small animals like mice, for reasons that are not exactly clear. The real paradox is this, however: when muscles are stripped from the legs of freshly killed mice and tested in the laboratory, they operate with an efficiency of 25%. That same experiment can obviously not be carried out with humans, but it is known that when humans ride bicycles, their efficiency is also about 25%. Why is running such an inefficient process for mice and so efficient for humans? When mice run marathons, their efficiency of movement drops from 25% to just 3%, while for humans efficiency during running soars from the 25% achieved during biking to 50%! The answer to the mouse question is unknown, but the key to the human running response is obvious. Each time a human foot hits the ground while running, energy is stored as ‘elastic strain energy’ by the key ‘springs’ in the human leg – mainly the connective-tissue strips which run along the bottom of the foot, the Achilles tendon and its associated muscles, the relevant muscles and tendons around the knee and the relevant muscles and tendons surrounding the hip. All of these structures are stretched when the foot hits the ground, and this stretching process stores energy – i.e. increases the potential energy of the leg. When the structures recoil elastically during toe-off, they manage to return about 90% of the work required to stretch them (with only 10% lost as heat). If the tendons and muscles of the leg were not able to store energy during impact with the ground, the muscles would have to increase their work output and energy expenditure dramatically. In fact, Alexander estimates that when humans run at middle-distance speeds, the spring-like properties of the Achilles tendon and the arch of the foot alone cut the work the leg muscles have to do by half. Here lies the answer to our paradox: human leg muscles are still working with only 25% efficiency during running; they do not really become more efficient just because running is the chosen sport. If the mechanical cost of movement is two joules per kg body weight per minute, half of this cost is furnished almost for free by the legs’ springs. Thus the muscles cough up four joules per kg per metre to provide the other joule per kg/metre of mechanical cost – with an efficiency of 1/4 = 25%.

The bottom line for you as an athlete who uses running in your chosen sport is that the best way to decrease your cost of running – and thus run faster and longer – is by enhancing the function of your leg springs. Since they are able to store energy more effectively when the foot hits the ground and then release this energy more fully and in a more timely fashion during push-off, your metabolic cost of running at a specific speed will drop, and you will be able to move up to higher speeds during training and racing. Fortunately, there are drills you can carry out to improve this ability to store and release elastic strain energy. Below are the key exercises for improving your springiness and thus your energy efficiency. Carry out these routines 2-3 times a week after a thorough warm-up, making sure you are rested and recovered from prior training. If your workout for the day contains a quality running component, perform the exercises after your warm-up and before the quality running and cool-down. Once you have completed the drills at least twice a week for six weeks or so, you may back down to once a week for up to 3-4 months as you work on other aspects of your overall fitness.

Exercises for enhancing springiness and energy efficiency

1. Spring-jogging

  • After a nice warm-up, jog along with very springy, relatively short steps, landing on the mid-foot area with each contact and springing upwards after impact. As you move along, your ankles should act like coiled springs, compressing slightly with each mid-foot landing, then recoiling quickly, causing you to bound upwards and forwards. Move along for one minute with quick, spring-like strides, alternating right and left feet as you would during normal running;
  • After one minute, jog normally for about 10 seconds, and then ‘spring-jog’ for about 20m, alternating three consecutive spring-like contacts with your right foot with three contacts with the left (i.e. three hops on your right foot, three hops on your left, three more on your right etc until you have travelled about 20m);
  • Jog normally for 10 seconds again, and then spring-hop for 20m on your right foot only, before shifting over to 20m on the left foot alone (making sure you land in the mid-foot area with each ground contact).

2. One-leg hops on-the-spot

  • Stand with your left foot forward and your right foot back, with your feet about one shin-length apart from front to back and hip-width apart from side to side. Place the toes of your right foot on a block or step 6-8 inches off the ground and direct most of your weight through the mid-point of your left foot;
  • Hop rapidly on your left foot at a cadence of 2.5-3 hops per second for 40 second. Your left knee should rise by about 4-6 inches with each upward hop, while your right leg and foot remain stationary. Your left foot should strike the ground in the area of the mid-foot and spring upwards rapidly, as if landing on hot coals. Your hips should remain as level and motionless as possible throughout the exercise, with very little vertical displacement;
  • After hopping for 40 seconds on your left foot, shift over to your right;
  • Take a short break, and repeat once more on each foot.

3. Box-hop with ‘sticks’

  • Hop quickly up to the box on your right foot, then onto the platform and immediately off the opposite side;
  • When your right foot hits the ground, react explosively – hopping forward as quickly as possible, resisting the temptation to let your ankle, knee, and hip flex dramatically and also the natural tendency to spend a lot of time on the ground before hopping forward. The movements should be smooth and quick at all times. Continue hopping for three more hops, and then ‘stick’ the final landing on your right foot, holding your body position, as a gymnast would do at the end of a routine;
  • After a second or two of holding, run back to your original position, and repeat the series of hops;
  • After 60 seconds, shift over to the same routine on your left foot;
  • Rest for a moment, then repeat.
  • Key points: maintain relaxed, upright body posture at all times and avoid looking at your feet or the ground as you hop along. Be sure to begin this exercise on a very forgiving surface – beach sand, soft dirt, soft grass or a relatively springy basketball court floor. Once you have built up considerable hopping strength, you may use harder surfaces.

4. Indian-hopping

  • Jog for a few strides, and then take a jogging stride diagonally to the right with your right foot;
  • When your right foot makes contact with the ground, hop once, then explosively hop diagonally to the left, landing on your left foot;
  • When your left foot strikes terra firma, hop once, then explode diagonally to the right;
  • Carry on alternating left and right in this fashion for 45 seconds. Rest for 15 seconds, then repeat. Stay relaxed at all times, moving in a coordinated and rhythmic manner and keeping the hops very springy and quick. Avoid the tendency to look at your feet.

5. One-leg squats with lateral hops

  • Assume exactly the same starting position as in 2 above;
  • Now, bend your left leg and lower your body until your left knee is bent at an angle of 90°;
  • Then hop about 6-10 inches to the left on your left foot, keeping the right foot in place;
  • Hop back to ‘centre’, then repeat on the other leg and return to the starting position, maintaining upright posture with your trunk and holding your hands at your sides;
  • Perform 2 sets of 12 squats on each leg, with a one-minute break between sets.

6. High-knee explosions

  • Stand erect but relaxed, with your feet directly below your shoulders;
  • Begin by jumping very lightly on the spot, but then suddenly – while maintaining fairly erect posture – jump vertically while swinging both knees up toward your chest;
  • Land back on your feet in a relaxed and resilient manner, then explode upwards again, aiming for as little time on the ground as possible while trying to maximise vertical jumping height;
  • Complete 15 high-knee explosions, rest for a few seconds and repeat.

7. Shane’s in-place accelerations.

  • Perform 3 x 20 seconds of Shane’s In-Place Accelerations. To carry these out:
  • Stand as in 6 above;
  • Begin by simply jogging on the spot, but then – when you feel ready – begin to dramatically increase your ‘stride’ rate, building up fairly quickly to as rapid a speed as you can sustain on the spot. Keep your feet close to the ground: you’re not going for high knee lift but rather for dramatically minimised foot-contact times – and an ability to get your feet to spring off the ground as soon as they make contact. Maintain erect but relaxed posture at all times. As you accelerate up to ‘top speed’, it sometimes helps to turn your legs slightly outwards at the hips until you become more accustomed to the exercise.
  • Perform 3 sets of 20 seconds.

8. Hop sprints.

  • At a park or on a decent track (resilient surfaces work best for this drill, especially initially), mark off a distance of 30 metres. Go to one end of the marked distance, then hop the 30m as fast as you can on your right foot, staying relaxed and trying to avoid excess vertical displacement of your centre of mass. Focus on two principles:
    1. push off as hard as you can each time the right foot hits the ground, but forwards rather than upwards
    2. make each stance phase (the amount of time your right foot is in contact with the ground) as short as possible. You are looking for very powerful, productive hops.

Once again, avoid the tendency to look downwards, keeping your gaze fixed ahead, as you would do during normal running.

After about four weeks of carrying out these exercises twice a week, you’ll begin to notice a real difference in your springiness and explosiveness, and after 6-8 weeks, the effects will be even more dramatic. You’ll begin to notice an increased springiness in your running strides, and you will also notice that high-quality running paces are beginning to feel easier.

Owen Anderson

Share this

Follow us