Biomechanical wear testing of running shoes
Jens Heidenfelder, Thorsten Sterzing, Thomas L. Milani
Department of Human Locomotion – Chemnitz University of Technology, Chemnitz, Germany
During running a deterioration of footwear occurs when shoes are exposed to many repetitive
impacts. However, only little information exists about the rate of the deterioration over increasing
mileage. A 45% loss of initial shock absorption was found by simulating 800 km of running during
mechanical testing (Cook et al., 1985). However, the authors reported that this mechanical test revealed
differences up to 25% in shock absorption for machine-simulated running compared to in vivo loaded
shoes. Shoes worn over a relatively short distance of 220 km showed reduced shock attenuation by 9%
and increased rearfoot motion of 18% (Hennig & Milani, 1995). Kong et al. (2008) reported an increase
of stance time of 2% for worn (320 km) compared to new running shoes.
No research was found about in vivo testing of running shoes during lifetime of these shoes.
Furthermore, no information was found whether deterioration in running shoe properties occurs linearly
or progressively with respect to mileage of wear. Therefore, the purpose of this study was to evaluate
systematically biomechanical variables at four stages of the shoe lifecycle. A second research aspect was
to examine whether different running shoes show different wear effects over mileage and time.
A subject pool of 24 injury-free recreational runners was involved in this study (age: 27.8 ± 3.8
years, height: 176.9 ± 4.5 cm, weight: 72.8 ± 5.8 kg). All runners were rearfoot strikers. Six
commercially available running shoes of different brands within the middle price segment (100-125 €)
were included in the study. Ten runners of the subject pool performed the wear runs with a length of
approximately 10 km. To achieve identical wear, each subject had to run in all shoe conditions along the
same self selected individual course. Shoes were systematically switched between subjects.
The experimental laboratory design included four stages of biomechanical measurements after a
running distance of 0 km, 100 km, 300 km, and 600 km to investigate the changes in running shoe
properties. At least 19 subjects of the pool had to run (3.5±0.1 m/s) in all shoe conditions for laboratory
testing at each stage. The experimental setup consisted of a force platform (KISTLER 9287BA;
60x90cm) that was integrated into a 13-m indoor running surface. Parameters of vertical ground reaction
force like maximum force rising rate (FRR), peak impact force (PVF1) and its corresponding time
(TPVF1) were calculated. A light weight electrogoniometer (Megatron MP10 1kΩ) was fixed to the
shoe heel counter to analyze kinematic rearfoot parameters in the sagittal plane like maximum pronation
velocity (MPV) and the range of total pronation movement (TPR). A miniature accelerometer (Analog
Devices ADXL78 ± 35g) was attached to the medial aspect of the tibia to calculate peak tibial
Mean values of 5 valid trials were used for statistical analysis. A repeated measures ANOVA
with Post-Hoc tests according to Fisher’s LSD (p<0.05) was used to analyze differences in each variable
of interest. The within-group factors were mileage and shoe conditions.
A continuous change of impact parameters was found within four stages of biomechanical
testing (Tab.1). After 600 km PTA was increased by 15%, whereas TPVF1 was decreased by 4%. This
systematic trend to higher load could be observed for all six shoe conditions. FRR was increased as well
by 6% at stage four but showed no continuous change across the four stages. The slightly unsystematic
0 300 600[km]
[%] S1 S2 S3
S4 S5 S6
behavior of FRR is also caused by differences between shoes. While one shoe showed no alteration of
FRR during 600 km, another shoe showed an increase of 14%.
For kinematic rearfoot parameters most changes occurred already at the second stage of
biomechanical testing. After 100 km MPV was increased by 12% and TPR by 7%. Between 100 km and
600 km MPV remained nearly unchanged in five of six shoes while TPR decreased in all shoe
Between shoes, differences up to 25% were found within MPV and MPA. For impact parameters
differences between shoes ranged from 5% to 14% after 600 km of running.
Increased FRR and PTA as well as decreased TPVF1 indicate reduced shock attenuation for all
shoe conditions after 600 km of running. It could be shown that shock attenuation declined
continuously. Assuming that shock attenuation is mainly influenced by the properties of midsole
materials, a continuous deterioration of midsole material was observed.
In contrast, progression of TPR indicates that pronation control of running shoes does not alter
continuously. The fact that all running shoes showed the same trend indicates that the investigated
changes are systematically. Differences between shoes up to 25% indicate that changes of MPV and
MPA are shoe specific.
It is concluded that deterioration in shoe properties occurs during 600 km of running. Cushioning
properties decreased consistently, whereas pronation control of tested shoes decreased mainly during the
first 100 km of running. In further studies the change of pronation behavior during the first 100 km
should be investigated with higher resolution. Thus, it might be possible to give evidence how many
kilometers new running shoes must be worn before performing meaningful biomechanical testing. Big
differences between shoes showed that for developing running shoes the constancy of shoe properties
during shoe lifetime needs to be improved.
This research was supported by Puma Inc., Germany.
Cook S.D., et al., 1985. Shock absorption characteristics of running shoes. Am j sp med, 4:248–253.
Hennig E.M., T.L. Milani, 1995. Biomechanical profiles of new against used running shoes.
Proceedings American Society of Biomechanics, 43–44.
Kong P.W., et al., 2008. Running in New and Worn Shoes – a comparison of three types of cushioning
footwear. Brit j sp med, online since October 9th.
0 km 35.4 81.4 5.9 468 5.6 12.8
100 km 35.0 86.7 6.2 522 5.8 13.7
300 km 34.7 88.1 - 531 5.6 13.3
600 km 33.9 86.1 6.8 514 5.4 12.9
Tab. 1: Alteration of biomechanical parameters, mean of six
running shoes for in vivo running
*p<.05 **p<.01 (PTA data not available for 300 km) Fig. 1: Change (%) of TPR for six running shoes
at four stages during 600 km wear test