Changes

The Science of Running Economy

2,353 bytes added, 11:58, 29 December 2017
no edit summary
[[File:Paula Radciffe NYC Marathon 2008 cropped.jpg|right|thumb|300px150px|Paula Radcliffe, the holder of the women's world record for the marathon (2:15:25) has gained much of her improvement through greater Running Economy.]]
[[Running Economy]] is how much energy it takes you to run. The better your economy, the faster and further you can run. Running economy is obviously determined to some extent by biomechanical efficiency, especially [[Cadence]]. There is some evidence that biochemical changes may have a significant impact on running economy. For instance, slow twitch [[Muscle]]s require less oxygen for the same level of work as fast which muscles do, and burning fat requires more oxygen than carbohydrates.
=Why Should You Care?=
Studies have consistently shown that heavier shoes reduce running economy<ref name="LussianaFabre2013"/><ref name="Burkett-1985"/><ref name="Sobhani-2014"/><ref name="Wierzbinski-2011"/>. Each 100g/3.5oz added to the weight of each shoe reduces running economy by about 1%<ref name="Franz-2012"/><ref name="Wierzbinski-2011"/><ref name="Frederick 1985"/><ref name="Frederick-1984"/>.
==Shoe Cushioning==
[[File:Reeves-2014-ShoeMass-RE.jpg|right|thumb|300px|Shoe mass against running economy<ref name="Reeves-2014"/>.Even on a cushioned treadmill, it appears that shoe cushioning can provide an improvement in running economy. The tests were performed at paces corresponding to various percentages of [[VO2max|V̇O<sub>2</sub>max]] (the shapes) and the vertical axis is the change in RE compared with barefoot. ]]
The results of studies that have compared barefoot and shod running (or running in un-cushioned and cushioned shoes) have provided conflicting information. Some studies show that running economy is worse when wearing shoes, with the bulk of the detrimental effect explained by the shoe weight. However, other studies have shown that the cushioning provided by shoes can compensate for the reduction in running economy due to shoe weight, and sometimes the shoes can actually be provide better economy than barefoot. I believe that the key to understanding this conflict lies in the type of running a surface that is used by the study. Not surprisingly, if a study uses a cushioned treadmill, the cushioning provided by the shoe does not confer any additional advantage over the barefoot condition. Of the 11 studies I found, 3 of the 4 studies using cushioned treadmills showed no improvement in running economy for shoe cushioning, while the 4 studies using un-cushioned treadmills and one using a cushioned treadmill showed an improvement for shoe cushioning. There were three studies that did not give sufficient information to determine the type of treadmill, and two showed no improvement while the remainder did. '''I conclude that a well cushioned running shoe can improve Running Economy by an estimated 2-3.5% compared with a weight matched un-cushioned shoe'''. In practice, most shoes will have a beneficial impact on Running Economy due to cushioning and a negative impact due to their weight.
* '''===Studies on cushioned treadmills'''. ===Most studies tend to use cushioned treadmills, as these are the most commonly available.** A thesis study showed that barefoot is more economic than shoed (New Balance M1500, 344-368g/shoe), but the same as shod when compensating for shoe mass<ref name="Flaherty-1994"/>. The barefoot condition had a greater [[Cadence]] than shod. The treadmill (Marquette model 1900) was cushioned.** Habitually FFS runners are 2.4% more efficient in minimally shod (FiveFingers with) than cushioned (GEL-Cumulus 10) shoes at 9 min/mile pace and with weights to ensure equal mass<ref name="Perl-2012"/>. This study used a Vision T9250 treadmill that is cushioned, with a reputable reviewer rating as 6.8/10 for cushioning<ref name="treadmilldoctor.com"/>.** A study compared barefoot (BFT), minimally shod (MS), and shod (SH) showed that BFT was 2.6-5.1% more efficient than SH<ref name="MooreJones2014"/>. The study also compared MS at the [[Cadence]] of BFT (165) and the cadence of SH (160), with an improvement only seen at the lower cadence. The participants had no prior barefoot or minimal running experience, and no adjustment was made for shoe mass. However, this study used a Woodway PPS treadmill with the manufacture claim that "Our patented Slat Belt running surface helps absorbs more of the impact than any other treadmill(<ref name="woodway.com"/>)".** A study comparing barefoot running with the participants using their normal running shoes showed that barefoot condition had a 4.4% improvement in running economy<ref name="Reeves-2014"/>. However, 80% of the change in running economy could be explained by shoe mass. In fact, the slope of the line between shoe mass and running economy predicts that a shoe weighing 9.2oz/260g would have the same running economy has barefoot. The shoes used weighed between 8.5-15oz (239-430g). Interestingly while running economy improved, [[VO2max|V̇O<sub>2</sub>max]] did not. The treadmill used was a WoodWay Pro-XL, a highly cushioned "slatted" treadmill similar to the one noted above.* [[File:Reeves-2014-ShoeMass-RE.jpg|center|thumb|300px|Shoe mass against running economy<ref name="Reeves-2014"/>.Even on a cushioned treadmill, it appears that shoe cushioning can provide an improvement in running economy. The tests were performed at paces corresponding to various percentages of [[VO2max|V̇O<sub>2</sub>max]] (the shapes) and the vertical axis is the change in RE compared with barefoot. ]]===Studies on un-cushioned treadmills. ===** Running in lightweight (150g/5.3oz, Nike Mayfly) shoes was 2% more efficient than barefoot<ref name="Franz-2012"/>. The study used runners with a midfoot [[Foot Strike]] and an un-cushioned treadmill (Quinton 1860). ** A thesis study showed that running in lightweight shoes (Nike Mayfly) was 3.4% more efficient than weight matched barefoot, and 2.1% more efficient than barefoot<ref name="Wierzbinski-2011"/>. The study used runners with a midfoot [[Foot Strike]] and an un-cushioned treadmill (Quinton 1860). When the weight of the Nike Mayfly shoes was doubled using weights, the running economy was still better than barefoot. The study also added EVA foam to the treadmill surface, but this did not change the running economy of barefoot running. However, the cushioning was added in slabs, and participants noted that the gaps felt "like running on a trail". ** Running barefoot and in Nike 3.0 shoes had no difference in running economy on an un-cushioned treadmill (Quinton 1860)<ref name="Tung-2014"/>. However, adding 10mm of cushioning to the treadmill improved barefoot running economy by 1.9%<ref name="Tung-2014"/>. Using 20mm cushioning made no further improvement over 10mm. The cushioning was provided by EVA foam used in running shoes. (The EVA foam was fairly soft, measuring 52-58 on the Asker Type C scale<ref name="Tung-2014-PC"/>. That's about 27-33 on the Shore A scale.)** A comparison of barefoot and shoes weighing 150g and 350g showed the same running economy in barefoot and 150g shoes, but 3.6% lower running economy in 350g shoes<ref name="Divert-2008"/>. The treadmill used (ADAL 3D, HEF-Tecmachine) appears to be un-cushioned. This study also showed that compared with barefoot, shoes reduced the total energy required to run (work) without changing V̇O<ref name="Divert-2008"/>, suggesting that barefoot is using more elastic energy, further suggesting that the barefoot condition in this study uses forefoot strike<ref name="Ardigò-1995"/>.* '''===Unknown Treadmill'''. ===Sadly, many studies don't explicitly state the type of treadmill used.** Barefoot was 2.0% more economic than shod on a treadmill and 5.7% more economic on an indoor track<ref name="Hanson-2011"/>. However, the type of treadmill and indoor track was not described and may have been cushioned. No indication was given of the type or weight of the shoes used. ** Barefoot was 1.3% more efficient than running in 340g shoes, while Vibram FiveFingers was 2.8% more efficient<ref name="Squadrone-2009"/>. The [[Cadence]] was 91.2 barefoot, 88.3 in FiveFingers, and 86.0 in shoes. While it seems strange that the FiveFingers were more efficient than barefoot, I suspect this is likely to be a statistical anomaly due to the small sample size (8 runners). The runners were all experienced in barefoot running. It is unclear if the Zebris FDM-T treadmill is cushioned or not. ** A study using Nike Air shoes showed an average 2.4% improvement in running economy compared with firmer shoes of a similar weight<ref name="Frederick 1986"/>. The improvement in running the economy varied by individual, between 0.5% and 6%, SD=1.8%. No details of the type of treadmill used were recorded.
==Heel Rise (drop)==
The typical running shoe has a sole that is around 10mm thicker at the heel than the forefoot, something that is typically called "drop" or "pitch". Currently the available research in this area is extremely limited.
* An undergraduate thesis showed no difference in running economy with subjects wearing shoes with 0mm drop, 4mm drop, and their usual running shoes (12-14mm)<ref name="Brown2013"/>.
* In one of the more bizarre bits of research, a comparison of running shoes and stiletto high heeled dress shoes (4.5cm and 7cm heels) showed that the running shoes were more economic<ref name="GuLi2013"/>.
==Shoe Flexibility==
There is some evidence that less flexible shoe may improve running economy<ref name="OhPark2017"/><ref name="Roy-2006"/>. However, there appears to be a "Goldilocks effect" where a little bit of additional flexibility improves running economy, but too much will actually have the opposite effect. It also appears that the amount of stiffness that is "just right" varies considerably from runner to runner, though this may be related to the runner's weight with heavier runners benefiting more. Converting the stiffness used in the studies to ordinary running shoes is not clear, and the two studies use different methods for measuring the stiffness. More importantly, the stiffer shoes were not because of thicker foam or more rubber outsole, which are the normal causes of inflexibility, but the addition of carbon fiber plates to highly flexible shoes. These carbon fiber plates are likely to be far more springy than typical inflexible running shoes. The first study used the Adidas Adistar Comp and compared it with modifications that increased the stiffness by 2.0x or 2.5x. In the graphs below you can see that the lighter runners have reduced running economy with stiffer shoes, while heavier runners do much better. However, even for a given weight, there is still wide variation.
[[File:RE and Stiffness2006.jpg|center|thumb|200px|The change in running economy for two different shoe stiffness's plotted against runner's weight.]]
=Running Biomechanics=
There are a number of biomechanical features of running that are relatively easily measured that may impact Running Economy.
* '''==Cadence'''. ==A review of the scientific studies showed consistently that an increased [[Cadence]] reduces shock at the hip, knee, and ankle, vertical oscillation, and ground contact time<ref name="SchubertKempf2013"/>. One study calculated that Cadence is 28% of the variation due to biomechanical differences<ref name="TartarugaBrisswalter2012"/>.* '''==Balance and step width'''. ==Maintaining side to side balance is estimated to cost 2% of the energy of running<ref name="ArellanoKram2011"/>. This may be due to step width, as increasing step width can reduce running economy by 11%<ref name="Arellano-2011"/>. [[File:Step Width.jpg|rightcenter|thumb|200px150px|The blue line is the center line of the body and the red line indicates the center of the foot placement. The distance between the two lines is the step width.]]* '''==Arm Swing'''. ==While it obviously costs energy to swing the arms while running, this arm swing actually improves running economy<ref name="ArellanoKram2011"/><ref name="Arellano-2011"/>, probably by improving balance. * '''==Ground Contact Time==It'''. It's s a generally believed that the less time a runner spends in contact with the ground, the more efficient they are. However, the scientific evidence is a little patchy. ** Two studies found that a shorter Ground Contact Time was indeed correlated with better Running Economy<ref name="PaavolainenNummela1999"/><ref name="Santos-ConcejeroGranados2013"/>** Plyometric training for 9 weeks improved 5K times, improved running economy and reduced Ground Contact Time<ref name="PaavolainenHakkinen2003"/>. It's unclear if the reduced Ground Contact Time was part of the improvement in either 5K time or running economy, but it is suggestive.** Two studies found no relationship between Ground Contact Time and Running Economy<ref name="HeiseMartin2001"/><ref name="TartarugaBrisswalter2012"/>.** One study found that a longer Ground Contact Time is correlated to a better Running Economy, the opposite of what is generally expected<ref name="Nummela-2007"/>. However, that study measured Running Economy at 3.9 m/s (6:50 min/mile), and the correlation with longer Ground Contact Time was only seen at speeds between 5.8-6.6 m/s (4:37-4:04 min/mile), rather radically faster.** Running to exhaustion may increase Ground Contact Time<ref name="FourchetGirard2015"/>. This may not tell us much about Running Economy, but it's possible that Ground Contact Time might be an indirect measure of fatigue.** In a wide range of animals, from a 32g kangaroo rat to a 140Kg pony, the energy cost of running has been shown to be proportional to the time each foot spends in contact with the ground<ref name="KramTaylor1990"/>. This suggests that the energy taken to run is mostly taken with supporting the weight of the body.* '''==Vertical Oscillation'''. ==It seems obvious that greater Vertical Oscillation is going to be less efficient, but the research is not there to support this. One study noted "The intuitive perception is that oscillation is adversely related to economy" but found no supporting research<ref name="Anderson1996"/>.** Greater vertical impulse on a force plate related to lower RE but noted exceptions to this seem common<ref name="HeiseMartin2001"/>.** In 16 national level runners, greater Vertical Oscillation was associated with improved Running Economy, the opposite of what would be expected<ref name="TartarugaBrisswalter2012"/>. The study found that Vertical Oscillation accounted for 7% of the variation due to biomechanical differences. ** A 1997 study found elite runners have less Vertical Oscillation than sub-elite runners<ref name="CavanaghPollock1977"/>.** Running with exaggerated Vertical Oscillation reduces Running Economy<ref name=" Tseh2008"/>. This suggests that it's possible to use excessive bounce to intentionally compromise your running, and might indicate that some runners could have a problematic level of Vertical Oscillation, but there is no clear evidence to indicate that's the case.** A study of 31 runners found that there was a non-significant relationship between increased Vertical Oscillation and reduced Running Economy, with the most economic runners having a Vertical Oscillation of 9.1cm and the least economic having 9.6cm<ref name="Cavanagh1987"/>.** Running to exhaustion may increase Vertical Oscillation<ref name="FourchetGirard2015"/>, which might make Vertical Oscillation a metric for evaluating fatigue.** Vertical Oscillation tends to decrease with increased Cadence<ref name="HalvorsenEriksson2012"/>. This makes perfect sense from a simplistic model of running, as a higher cadence means that each stride is shorter, so less time is spent in the air where the runner is in a ballistic curve. This relationship can confuse the analysis of Running Economy, as it's easy to attribute changes in Cadence to Vertical Oscillation, or Ground Contact Time.** Feedback to reduce Vertical Oscillation resulted in impaired Running Economy<ref name="HalvorsenEriksson2012"/>. The study used real-time feedback to encourage runners to reduce their Vertical Oscillation, but the study lasted a relatively short time. It's possible that the impaired Running Economy was a result of the initial change in biomechanics rather than indicating that reduced Vertical Oscillation is actually a bad thing. Of course, the other interpretation is that there is an optimal level of Vertical Oscillation and attempting to reduce it is actually a bad thing.Vertical Oscillation and attempting to reduce it is actually a bad thing.==Braking==Another intuitively obvious factor in running economy is breaking. After all, anything that slows a runner seems likely to reduce their running economy. There are three obvious places to measure breaking; the foot, the ground, and the body. A runner's foot has to decelerate so that there is no horizontal movement when it's in contact with the ground, and then accelerate to be faster than the body to move ahead for the next footfall. If the foot has decelerated so that there is no horizontal movement prior to the footfall, then you'd expect no braking force to be detected in the ground, whereas if the foot is still moving forward, there would be a braking force detected. The third place to measure breaking is the body (or center of mass), which may also vary in horizontal velocity. It seems a reasonable that variations in horizontal velocity of the body would reduce running economy.* A study of 17 runners found that the braking force detected on the ground explained most of the variation in running economy<ref name=" KyröläinenBelli2001"/>.* A study of 11 elite runners found that there was a statistically significant correlation between the sum of the peak horizontal and vertical forces and running economy, with lower forces being more economical<ref name="StørenHelgerud2011"/>. However, the study did report on horizontal forces alone, or the total energy involved in the breaking.I've not found any research covering braking forces on the body, which would seem to be a natural subject of inquiry.==Foot strike==* '''Foot strike. '''It's often assumed that a forefoot landing is more efficient than a rear foot landing, but there's no clear evidence to support that.** A study found that while there was no difference in Running Economy between different foot strike types, there was a higher work produced with Forefoot Strike, suggesting more elastic storage<ref name="Ardigò-1995"/>.** A study looked at runners with different foot strike patterns and compared them running with their habitual stride and intentionally modifying to an alternative foot strike<ref name="GruberUmberger2013"/>. The study found no differences with the exception of rear foot strike runners changing to a forefoot strike.** Sub-elite rear foot strike runners are more economic than midfoot strike, and while a cadence was the same the rearfoot runners had a higher Ground Contact Time<ref name="Ogueta-Alday-2014"/>.
=Body weight and fat percentage=
The relationships between body weight, body fat, and running economy are not what you might expect.
[[Delayed Onset Muscle Soreness]] reduces running economy<ref name="BraunDutto2003"/><ref name="Smith-1992"/>. This matches most runners experience, as it's tough to run with DOMS. My personal suspicion is that DOMS damages the muscle fibers that are used first, and these fibers are the most trained and the most efficient.
=Flexibility=
[[File:Standing External Hip Rotation.jpg|right|thumb|200px|Greater Standing External Hip Rotation reduces running economy.]]
Greater flexibility reduces running economy. Runners with greater standing external hip rotation and dorsiflexion of the foot have poorer running economy<ref name="Craib-1996"/>. (Dorsiflexion of the foot is bending the ankle so the toes move towards the shin, which is what happens when you do a calf stretch. Standing External Hip Rotation is shown above.) Another study showed that greater overall flexibility is associated with poorer economy, with the most flexible third of the studied group using 9% more energy than the least flexible third<ref name="Gleim-1990"/>.
* '''Stretching Programs'''. Most studies show that [[Stretching]] programs do not reduce running economy<ref name="Nelson-2001"/><ref name="Godges-1993"/>.
* '''Stretching before running'''. One study<ref name="Wilson-2010"/> has shown that stretching directly before running reduces performance and running economy, but most studies indicate no impact<ref name="AllisonBailey2008"/><ref name="Hayes-2007"/>, even though the stretching impairs other muscular functioning. (One study on subjects with limited hip range of motion showed that stretching before running improved running economy<ref name="Godges-1989"/>, but the subjects were only tested with 4 minutes of running, and a steady state requires 4-15 minutes<ref name="Morgan-1989"/>).
[[File:Standing External Hip Rotation.jpg|center|thumb|200px|Greater Standing External Hip Rotation reduces running economy.]]
=Gender =
Women may improve their running economy in response to training more effectively than men<ref name="Bourdin-1993"/>.
=Height=
In elite runners, being taller was correlated with being less economic<ref name="StørenHelgerud2011"/>.
=Muscle Fiber Types=
Fast twitch [[Muscle]] fibers require more energy to provide contraction than slow twitch muscles<ref name="Huxley-1974"/><ref name="Horowitz-1994"/><ref name="Coyle-1992"/>, so a runner with a higher proportion of slow twitch fibers will have better running economy.
<ref name="Santos-ConcejeroGranados2013">Jordan Santos-Concejero, Cristina Granados, Jon Irazusta, Iraia Bidaurrazaga-Letona, Jon Zabala-Lili, Nicholas Tam, Susana Gil, DIFFERENCES IN GROUND CONTACT TIME EXPLAIN THE LESS EFFICIENT RUNNING ECONOMY IN NORTH AFRICAN RUNNERS, Biology of Sport, volume 30, issue 3, 2013, pages 181–187, ISSN [http://www.worldcat.org/issn/0860-021X 0860-021X], doi [http://dx.doi.org/10.5604/20831862.1059170 10.5604/20831862.1059170]</ref>
<ref name="Moore2016">Isabel S. Moore, Is There an Economical Running Technique? A Review of Modifiable Biomechanical Factors Affecting Running Economy, Sports Medicine, volume 46, issue 6, 2016, pages 793–807, ISSN [http://www.worldcat.org/issn/0112-1642 0112-1642], doi [http://dx.doi.org/10.1007/s40279-016-0474-4 10.1007/s40279-016-0474-4]</ref>
<ref name=" KyröläinenBelli2001">Heikki Kyröläinen, Alain Belli, Paavo V. Komi, Biomechanical factors affecting running economy, Medicine and Science in Sports and Exercise, volume 33, issue 8, 2001, pages 1330–1337, ISSN [http://www.worldcat.org/issn/0195-9131 0195-9131], doi [http://dx.doi.org/10.1097/00005768-200108000-00014 10.1097/00005768-200108000-00014]</ref>
<ref name="StørenHelgerud2011">Øyvind Støren, Jan Helgerud, Jan Hoff, Running Stride Peak Forces Inversely Determine Running Economy in Elite Runners, Journal of Strength and Conditioning Research, volume 25, issue 1, 2011, pages 117–123, ISSN [http://www.worldcat.org/issn/1064-8011 1064-8011], doi [http://dx.doi.org/10.1519/JSC.0b013e3181b62c8a 10.1519/JSC.0b013e3181b62c8a]</ref>
</references>