Wednesday, 30 May 2012

Shoe weight and running speed - the scientific literature

Shoe weight and running speed - the scientific literature

Whilst I have done a few tests of my own on the effect of different running shoes on physiological effort associated with running, I thought it might be interesting to draw together the literature on the subject.

There have been quite a few studies, but many of them are inaccessible to the general public because they were published in Scientific Journals that are not open access. Where possible I have tried to get hold of these articles and I have explained what the authors did and what they concluded. This is very much my opinion of the research - you may well have a different view point (please feel free to comment). For my starting point I used an article by Greg Crowther, published in Northwest Runner back in 2001.

The early years

The first relevant article I can find seems to be Hettinger & Muller (1953). They published an article on the effect of shoe weight on the energy consumption during walking and carrying (sorry for my poor translation). I have managed to access it, but it does stretch my German to its limit. They investigated a range of conditions (in 265 trials) from barefoot, plimsolls (UK terminology), half shoes, normal laced shoes, ski-shoes, riding boots, military boots and a variety of other test shoes. I only mention them since it is the most complete test list of shoes I have come across - very Germanic! Whilst there is much good data to consider - most of it is hard to connect with more modern studies. However, the conclusion that; when walking and carrying a weight of 30 kg, energy consumption increases more rapidly in shoes than barefoot seems interesting. The next relevant study was by Catlin & Dressendorfer (1979) where the authors, apparently, looked at the effect of shoe weight on the energy cost of running. Since I am unable to get hold of the article and PubMed does not have an abstract for it I am limited to recounting what Jones et al. (1986) state about it. The study looked at marathon runners, on a treadmill wearing lightweight racing flats (520g total weight). They recorded a 0.9% increase in energy cost for each 100g increase in shoe weight (per pair).

Myers & Steudel (1985) reported the results of four people running at a stately 10.3 km per hour (5:50 min per km) on a treadmill (paced using a metronome) with weights (3.6 kg in total in canvas belts filled with lead shot) attached to them in four locations - waist, upper thigh, upper shank and ankle. They made measurements of oxygen consumption at steady-state (after 8-15 mins of running). The authors also modelled the physical work required to accelerate the additional weights using a few simple equations. They found that 3.6 kg around the waist had the smallest effect causing a 3.7% increase in energy (oxygen) consumption. Moving the weights (split equally across the limbs) caused a greater increase in energy consumption with the size of the increase related to how close the weights were to the foot. Thus, the ankle weights produced a 24.3% increase in energy consumption compared to running without the added weights and 20.7% compared to running with the weights around the waist. Thus, they demonstrated that weight around the feet has a much bigger effect (on the flat and on a treadmill) than weight on the waist. Interestingly they state the kinetic energy changes of the legs account for about a third of the energy used when running. They also noted that this kinetic energy component increases with the square of the running speed whilst the mechanical energy in moving the trunk rises linearly resulting in high energy costs for limb movements at higher speeds.

In the same year (Martin, 1985) the results of weight loading (thighs or feet) of 15 men running at 12 km per hour (5 min per km) was published. I only have access to the abstract, so I can make little comment other than to report what was summarized within it. The increase in oxygen consumption was 7.2% per kg of load (on the foot) which, when extrapolated to 3.6 kg, produces a value of ~26% and is in agreement with Myers & Steudel (1985). They also go on to state that the energy increase can be attributed to the additional weight rather than a change in gait.
Jones et al. (1986) looked at the energy cost of women walking and running in athletics shoes (514g) and leather military boots (1,370g)! They report an 8.3% increase in energy cost associated with the weight increase of 857g giving rise to the memorable number of 1.0% increase in energy cost per 100g increase in footwear weight.

Miller & Stamford (1987) went a little bit further, investigating men and women, ankle and arm weights during both walking and running! Again the experiments were on a treadmill but, at a range of speeds. The maximum running speed was halfway between Martin (1985) and Myers & Steudel (1985) at 7 mph (11.2 km per hour). The bottom line is that men and women had the same energy consumption and it rose by 8% per kg on the ankle (i.e. similar to Martin's study). Weights on the hand produced a greater increase (13%) in oxygen consumption - but, subjects ran with a 90 degree bend in the elbow with arms swinging with each step. Interestingly they also report that the increase in oxygen consumption rose linearly with weight added to the ankle all the way up to 4.5 kg.

Barefoot - the ultimate weight reduction

Warburton  (2001) published a brief review of barefoot running. In it he states that Laboratory studies show a 4% reduction in energy cost when running barefoot. But he also states; "Competitive running performance should therefore improve by a similar amount, but there has been no published research comparing the effect of barefoot and shod running on simulated or real competitive running performance." This short coming may have been solved by Buchholz (2007). He wrote a Masters Thesis on performance differences between bare foot and shod running, however, an electronic copy is not available. Divert et al. (2008) attempted to compare barefoot with shod running in order to test whether shoe weight or a change in foot strike was responsible for the differences in oxygen consumption reported. They used 12 subjects running at 13 km per hour on a treadmill with a range of shoe and additional weights. They found that additional mass increased oxygen consumption and that efficiency was reduced when wearing shoes.  They hypothesized that the damping effect of shoes reduced the storage of elastic energy lowering the efficiency of running in shoes.
Hanson et al. (2011) compared the cost of running barefoot and shod both on the treadmill and overground using both men and women at 70% of VO2max (10.7 km per hour). They found that shoes increased oxygen consumption and heart rate on both the treadmill and overground. Interestingly, the effects were more pronounced overground where using shoes increased oxygen consumption by 5.7% compared to on a treadmill where shoes increased oxygen consumption by 2%. An electronic copy is available at the time of writing, and care needs to be taken with reading it. Fig. 1, for instance, appears to show no significant oxygen uptake differences between the groups - which is compatible with their data since the SEM will be dominated by the variation between subjects rather than the different test conditions. It is hard to see why this figure is included and the statistically significant data (i.e. mean relative change in oxygen uptake for each condition) is not plotted - hey ho! But, there are studies which report the opposite. Jack Daniels, who worked for Nike in the early 1980s, states that as shoe weight is reduced there is a point at which further weight reductions began to increase the cost of running. This may well be the same observation reported by Franz et al. (2012). Again this is a treadmill comparison of barefoot versus shod. However, in this case running (12 km per hour) in shoes produced a lower oxygen uptake and the authors concluded that running barefoot offers no metabolic advantage over lightweight cushioned shoes. The paper got a lot of publicity across the web, but without being able to see it I cannot comment further.

Is strike type important?

To finish off I want to just bring a recent article to your attention, whilst it is not directly related to shoe weight it does impact on shoe type. Perl et al. (2012) looked at running economy in minimal shoes versus standard running shoes using a range of foot strike patterns. Again I am hampered by not being able to access the paper, but here is the gist of what they report in the abstract. First, they controlled for shoe mass (I don't know how) and they found that regardless of whether you forefoot or rearfoot strike minimal shoes require less energy (2.4% less for forefoot, 3.3% less for rearfoot). That is an effect of the shoe type not weight. Again, like Divert et al. (2008) they suggest that shoe cushioning may get in the way of the more efficient elastic recoil in your musculature.


It is clear that adding weight to your feet will cause a rise in oxygen consumption. Treadmill data suggests about 1% for each 100g of total shoe weight. There is only one study of oxygen consumption that used a hard surface (an indoor track) and that was by Hanson et al (2011). However, they only tested one shoe type against barefoot running, and the shoe weight does not seem to be given. Others (e.g.  Buchholz, 2007) must have run tests on the effect of shoe weight overground, however, they are hard to get hold of - perhaps because they are done as student projects and without laboratory control are considered unsuitable for publication in the scientific literature.

It is also clear that the physics dictates that increasing the mass at the end of a lever system must result in an increased energy input. The counter argument is that some of the weight, in a shoe, improves efficiency by altering the gait or foot-strike or improves race performance by allowing the runner to get to the finish line without injury. Minimal running shoes may also present a benefit beyond weight reduction in allowing the musculature to return more elastic energy. However, barefoot running tests have produced mixed results. There are suggestions that this may depend upon gait or foot-strike, experience, adaptations etc. However, I think it is probably fair to say that there is not yet enough evidence to be at all sure.

Finally, there is the BIG question as to whether a reduction in energy consumption leads to better race performance. Certainly extrapolating from mathematical models like Rapoport's (which is a good read) then one would conclude that it can. But, races are about more than simple energetics......


Buchholz MP Performance differences between the conditions of running with the foot bare and running with the foot shod (2007) Thesis (MS) Springfield College Listing
Catlin MJ, Dressendorfer RH (1979) Effect of shoe weight on the energy-cost of running Medicine and Science in Sports and Exercise 11, 80.
Divert C, Mornieux G, Freychat P, Baly L, Mayer F, Belli A (2008) Barefoot-shod running differences:shoe or mass effect? Int J Sports Med, 29, 512-518. PubMed
Franz JR, Wierzbinski CM, Kram R (2012) Metabolic cost of running barefoot versus shod: is lighter better? Med Sci Sports Exerc Mar 2 [Epub ahead of print] PubMed
Hanson NJ, Berg K, Deka P, Meendering JR, Ryan C (2011) Oxygen cost of running barefoot vs. running shod. Int J Sports Med PubMed Full Text
Hettinger T, Muller EA (1953) Der Einfluss der Schuhgewichtes auf den Energieumsatz beim Gehen und Lastentragen. Arbeitsphysiologie 15, 33-40. Full Text
Jones BH, Knapik JJ, Daniels WL, Toner MM (1986) The energy cost of women walking and running in shoes and boots. Ergonomics 29, 439-443. Full Text
Martin PE (1985) Mechanical and physiological responses to lower extremity loading during running. Med Sci Sports Exerc 17, 427-433. PubMed 
Miller JF, Stamford BA (1987) Intensity and energy cost of weighted walking vs. running for men and women. J Appl Physiol 62, 1497-1501. PubMed Full Text
Myers MJ, Steudel K (1985) Effect of limb mass and its distribution on the energetic cost of running. J Exp Biol. 116, 363-373. PubMed Full Text
Perl DP, Daoud AI, Lieberman DE (2012) Effects of footware and strike type on running economy. Med Sci Sports Exerc [Epub ahead of print] PubMed
Warburton M (2001) Barefoot running. Sportscience 5.


  1. Thanks for the great information on shoe weights!

  2. This comment has been removed by the author.

  3. This comment has been removed by the author.

  4. Thanks Christof for putting all this together - great article.

    Could you clarify whether the 1% increase in oxygen consumption per 100g of shoe weight guideline refers to 100g per shoe or 100g combined?

    Thanks, John

    1. 1% increase for each 100g increase in total weight of the pair. Thus, in theory, if you could drop the weight of a single shoe by 50g, without changing its other characteristics, you might expect ~1% reduction in energy cost.