Subjects took a 5 min break between bouts a handheld

USA, Waukesha, WI, USA). Subjects took a 5-min break
between bouts. A handheld digital tachometer (Shimpo DT-
107A; Electromatic Equipment Inc., Cedarhurst, NY, USA)
was used to verify all treadmill velocities. To allow famil-
iarization, subjects breathed through an expired-gas analysis
system (True One 2400; Parvo Medics, Salt Lake City, UT,
USA) during this visit. Within 1 min after the completion
of each 5-min trial, we obtained a ff nger-prick blood sample
for blood [La] determination in duplicate (YSI 2300 lactate
analyzer; YSI Incorporated, Yellow Springs, OH, USA).
2.2.3 Visits 2 and 3
Visits 2 and 3 were completed in random order, with at
least 72 h between visits. Prior to their warm-up, subjects
had their height, weight, and body composition reassessed.
Subjects began running with a 15-min warm-up at their
own selected speed ≤ 14 km·h
−1
, in their own shoes. Fol-
lowing the warm-up, all subjects completed 4 × 5-min bouts
at 14 km·h
−1
on a ffat Woodway ELG treadmill (Woodway
USA) while wearing each of the shoe conditions in a ran-
domized order. In-between bouts, subjects took an ~ 3-min
break while they changed shoes. Note that runners mechani-
cally adapt their biomechanics very quickly in response to
changes in surface stiffness [ 18 ]. Subsequently, subjects
completed another 4 × 5-min trial at 18 km·h
−1
for men and
16 km·h
−1
for women, as previously described. Throughout
the running test, heart rate (Polar H7; Polar Electro, Kem-
pele, Finland) was recorded at 5-s intervals and expired
gases were continuously measured to determine VO
2
, carbon
dioxide production (VCO
2
), minute ventilation ( V
E
), and res-
piratory exchange ratio (RER). The mean VO
2
determined
during the last 2 min of each running stage was expressed
as running economy (ml·kg
−1
·min
−1
), gross oxygen cost
of transport (O
2
COT; ml·kg
−1
·km
−1
), and energetic cost
(W·kg
−1
) using the Peronnet and Massicotte equation [ 19 ,
20 ]. In our laboratory, the typical error of measurement [ 21 ]
of submaximal VO
2
was 1.9%. Rating of perceived exertion
(RPE) was determined using a 1–10 Borg RPE scale upon
completion of each interval [ 22 ]. Blood [La] was assessed
at the completion of each running velocity interval series.
Biomechanical measures (stride rate, stride length, con-
tact time, and ffight time) were determined using high-speed
video analysis (240 frames·s
−1
) while running at each veloc-
ity and shoe condition [ 23 ]. The average of each biome-
chanical parameter during the ff nal 30-s of each running
speed and shoe condition was used for analysis. We also
used the video recordings to determine the foot-strike pat-
terns of the runners during all trials (rearfoot strike vs. mid/
forefoot strike). However, only 5 of the 24 subjects qualiff ed
as mid/forefoot strikers, therefore our sample did not permit
an evaluation of foot-strike pattern interaction on the ener-
getic cost differences between shoes.