Content uploaded by Joel M Stager
Author content
All content in this area was uploaded by Joel M Stager on Mar 10, 2015
Content may be subject to copyright.
76
www.IJSPP-Journal.com
ORIGINAL INVESTIGATION
Lower-Leg Compression, Running Mechanics,
and Economy in Trained Distance Runners
Abigail S.L. Stickford, Robert F. Chapman, Jeanne D. Johnston, and Joel M. Stager
The efcacy of and mechanisms behind the widespread use of lower-leg compression as an ergogenic aid to improve running
performance are unknown. The purpose of this study was to examine whether wearing graduated lower-leg compression sleeves
during exercise evokes changes in running economy (RE), perhaps due to altered gait mechanics. Sixteen highly trained male
distance runners completed 2 separate RE tests during a single laboratory session, including a randomized-treatment trial of
graduated calf-compression sleeves (CS; 15–20 mm Hg) and a control trial (CON) without compression sleeves. RE was deter-
mined by measuring oxygen consumption at 3 constant submaximal speeds of 233, 268, and 300 m/min on a treadmill. Running
mechanics were measured during the last 30 s of each 4-min stage of the RE test via wireless triaxial 10-g accelerometer devices
attached to the top of each shoe. Ground-contact time, swing time, step frequency, and step length were determined from accel-
erometric output corresponding to foot-strike and toe-off events. Gait variability was calculated as the standard deviation of a
given gait variable for an individual during the last 30 s of each stage. There were no differences in VO2 or kinematic variables
between CON and CS trials at any of the speeds. Wearing lower-leg compression does not alter the energetics of running at
submaximal speeds through changes in running mechanics or other means. However, it appears that the individual response to
wearing lower-leg compression varies greatly and warrants further examination.
Keywords: endurance athletes, energy cost, gait, ground-contact time, performance clothing
Endurance runners are using compression garments presumably
as a means to improve training, performance, and recovery, yet there
is little consistent, conclusive research exploring the efcacy of and/
or mechanisms underlying the use of lower-leg compression as an
ergogenic aid specically during running. Improved venous return
and clearance of metabolites,1 proprioception,2,3 force production,2–4
thermoregulation,3 and subjective measures2,5 have been evaluated
as potential mechanisms to improve athletic performance using
compression, with varying results. With regard to distance running
in particular, studies evaluating the use of compression stockings
to improve performance have been equivocal.5–8
A relatively constant nding among studies evaluating com-
pression garments is a decrease in active hip-joint range of motion,
suggesting the likelihood of an “ergogenic interplay” between
compression and biomechanical factors.2,3,7 In addition, there is
evidence to suggest that compression garments reduce longitu-
dinal and anterior–posterior muscle vibrations on landing during
repetitive jumping.3,4 This nding has been attributed to enhanced
muscle activation, as soft-tissue vibrations with ground impact can
be attenuated by increased muscle activity.9 Similarly, compression
garments appear to allow greater maintenance and consistency of
repeated-jump performances, although whether this would trans-
late to any kinematic measure during other modes of exercise is
unknown. A potentially comparable measure during running is
stride-to-stride variability, which has been shown to relate to the
energy cost of locomotion and increase with fatigue.10,11 Only
recently have kinematics been examined with subjects wearing
lower-leg compression garments during endurance running, and
no differences were observed when the results were compared with
a noncompressive control visit.8 However, the runners used in the
study can be considered moderately trained (mean 10-km best: 38
min), and no indication of gait stability was measured.
It is well known that individual differences in running mechan-
ics inuence the oxygen cost of running at submaximal workloads
(ie, running economy [RE])10–14 and that a strong relationship exists
between RE and distance-running performance.15–17 Small differ-
ences in RE often inuence performance outcomes, particularly at
the elite level, where all athletes possess a high maximal oxygen
uptake (VO2max) and can sustain efforts at a high percentage of
their VO2max for prolonged periods of time.17 Thus, any change in
running mechanics that inuences RE could ultimately affect an
athlete’s performance. It is currently unknown whether lower-leg
compression sleeves alter gait mechanics and/or stride-to-stride
variability in young, highly trained runners and, if so, what impact
this alteration will have on RE. We hypothesized that wearing
compression sleeves would reduce stride-to-stride variability and
improve economy during treadmill running.
Methods
Subjects
Subjects (N = 16) were highly trained men, all either current col-
lege or professional distance runners. Subject anthropometrics and
personal-best race performances are displayed in Table 1. Subjects
reported running 6.5 ± 0.9 d/wk and 100 ± 32 km/wk during the 6
months before the study. All runners were determined to be t to
participate in the study, as assessed by physical activity and training
The authors are with the Dept of Kinesiology, Indiana University,
Bloomington, IN. Address author correspondence to Abigail Stickford at
AbigailStickford@texashealth.org.
International Journal of Sports Physiology and Performance, 2015, 10, 76-83
http://dx.doi.org/10.1123/ijspp.2014-0003
© 2015 Human Kinetics, Inc.
Effect of Compression on Gait and Economy 77
questionnaires. Their specialty racing events included the 1500-m,
3000-m steeplechase, 5000-m, and 10,000 m. Primary inclusion
criteria were active training, age 18 to 30 years, and a 5000-m time
of ≤16:30 within the past year. Runners who had not raced a 5000-m
in the past year were deemed highly trained by VO2max measures
(>65 mL · kg–1 · min–1) obtained within the past year. The study was
carried out from late July through early September. Subjects were
informed of the risks and benets of the study and gave written
informed consent before testing. All protocols and procedures were
approved by the institutional review board of Indiana University.
Design
Subjects completed a single experimental session. Since subjects
could not be blinded to wearing compression garments, they also
completed a questionnaire dening their a priori experiences with
and beliefs regarding compression garments. The 7 dichotomous
questions were summed to give an overall belief score, with high
positive values indicating substantial experience with and positive
opinions on the effectiveness of compression garments.
Two separate RE tests took place during the session. The
sequence of tests, a trial with calf-compression sleeves (CS) and a
control trial without compression sleeves (CON), was randomized
and counterbalanced, with approximately 10 to 15 minutes rest
between. Graduated calf-compression sleeves (Zensah, Miami, FL)
generating 15 to 20 mm Hg of compression (per manufacturer’s
statement) were worn by the subjects during the CS trial; manufac-
turer recommendations for sizing based on height and maximal calf
circumference were used. The elastic of the sleeve reached from ~2
cm above the ankle to ~4 cm below the knee. For both trials runners
wore the same pair of their own lightweight shoes and low-cut socks.
Methodology
RE was assessed by measuring VO2 during 4-minute stages at
each of 3 constant submaximal speeds of 233, 268, and 300 m/min
on a motorized treadmill (Quinton, model 18-72, Bothell, WA).
Treadmill speed was veried through the use of a laser tachometer
(Mastech, model DT-2234C, San Jose, CA). RE was calculated from
the VO2 measured over the nal 60 seconds of each 4-minute stage
at each speed and the slope of line relating VO2 to speed.
Ventilatory and metabolic variables were continuously mea-
sured during exercise using a computer-interfaced, open-circuit,
indirect calorimetry system. Minute ventilation (VE) was deter-
mined using a pneumotach (Hans Rudolph #3813, Kansas City,
MO) and amplier (Hans Rudolph #1110) on the inspired side.
Subjects breathed through a low-resistance 2-way valve (#2700,
Hans Rudolph), and a 5-L mixing chamber was used for collection
of expired gases. Fractional concentrations of O2 (FEO2) and CO2
(FEO2) were determined from dried expired gas sampled at a rate of
300 mL/min, using separate O2 and CO2 gas analyzers (SA-3 and
CD-3A, respectively, AEI Technologies, Pittsburg, PA). Analyzers
were calibrated before each test using commercially available gas
mixtures within the physiological range. VE, VO2, and VCO2 were
averaged over each minute of exercise, with VE corrected to BTPS
(body temperature and pressure, saturated) and VO2 and VCO2
corrected to STPD (standard temperature and pressure, dry). These
variables, as well as FEO2 and FECO2, were continuously measured
and monitored with a data-acquisition control system (DASYLab
10.0, National Instruments, Norton, MA) sampling at 50 Hz.
To measure select kinematic variables related to running gait,
accelerometric data were gathered during the last 30 seconds of
each 4-minute stage of the RE test. Separate wireless triaxial 10-g
accelerometers (G-link, Microstrain, Williston, VT) were attached
to the shoelaces on each foot using plastic ties. The accelerometers
sampled each axis at 1024 Hz, with data from each 30-second stage
being stored in separate les. Accelerometer data were analyzed
using a custom in-house program, following a technique described
previously.18 Briey, the unltered output of the accelerometer in
the vertical and horizontal planes was used to identify contact and
toe-off time points for each step, allowing quantication of (1) foot
ground-contact time (tc), dened as the time (s) from when the foot
contacts the ground to when the foot toes off (ie, breaks contact
with the ground); (2) swing time (tsw), dened as the time (s) from
toe-off to ground contact of consecutive footfalls of the same foot;
(3) step frequency (SF), dened as the number of ground-contact
events (ie, steps taken) per second; and (4) step length (SL), dened
as the length (m) that the treadmill belt moved from toe-off to
ground contact in successive steps (opposite feet), calculated from
SF (steps/min) and treadmill speed (m/min). Values of tc, tsw, SL,
and SF were determined from the average of accelerometric values
obtained from a minimum 20 consecutive steps. Gait variability was
calculated as the standard deviation of a given gait variable for an
individual during the last 30 seconds of each stage. Mechanical
parameters of the spring-mass model during running, including
vertical displacement of the center of mass at lowest point, peak
displacement of the leg spring, maximal ground-reaction force, ver-
tical stiffness, and leg spring stiffness, were calculated as described
by Morin et al19 using measures of body mass, height, ight time,
and ground-contact time.
Statistical Analysis
Descriptive statistics were used to describe the characteristics of the
group, and Pearson correlations were used to quantify relationships
between mechanical and metabolic variables (SPSS 18, Chicago,
IL). To assess differences in the outcome measures of tc, tsw, SL,
SF, and submaximal VO2 at the different running speeds during CS
and CON testing sessions, a 2 × 3 repeated-measures ANOVA was
conducted. When the F value was considered statistically signicant,
differences between CS and CON at each speed were determined
using simple main effects. A modied Bonferroni adjustment was
done to account for the multiple planned comparisons. Data are
reported as mean ± standard error of the mean (SE) unless otherwise
stated. Signicance was set at P < .05.
Table 1 Subject Characteristics and Reported
Personal-Best Race Performances During the 1 Year
Before the Study, Mean ± SD
Characteristic Value
Age (y) 22.4 ± 3.0
Height (cm) 180.6 ± 4.6
Mass (kg) 66.4 ± 5.2
Event average time (min:s)
1500 m (n = 8) 3:56.2 ± 0:12.6
5000 m (n = 12) 14:47 ± 1:02.2
10,000 m (n = 4) 29:22 ± 0:35.7
Note: Some subjects (n = 9) reported personal bests in more than 1 event.
78 Stickford et al
Results
Running Economy
There were no differences in VO2 between CON and CS trials at
any of the speeds (P = .70–1.00) (Figure 1). At 233 m/min, VO2
was 46.7 ± 1.6 mL · kg–1 · min–1 (CON) and 46.5 ± 1.5 mL · kg–1 ·
min–1 (CS). At 268 and 300 m/min, submaximal VO2 was 54.0 ± 1.6
mL · kg–1 · min–1 for CON and 54.0 ± 1.7 mL · kg–1 · min–1 for CS
and 62.1 ± 1.7 mL · kg–1 · min–1 for CON and 62.2 ± 1.8 mL · kg–1
· min–1 for CS, respectively. In addition, there was no difference in
the slope of the lines relating submaximal VO2 and running speed
between the 2 experimental conditions (slope = 0.230 and 0.233
for CON and CS, respectively; P = .54) (Figure 1).
Running Mechanics
There were no differences in tc, tsw, SF, and SL between CON and
CS trials at any of the running speeds (P = .28–.94) (Table 2A).
Gait variability was also not different from CON to CS conditions
at any of the speeds (P = .42–.73) (Table 2B). Spring-mass model
parameters were comparable to those previously reported19 and did
not differ between CON and CS.
As seen previously, tc was inversely correlated with submaxi-
mal VO2. Figure 2 shows the relationship between the inverse of tc
(tc–1) and VO2 for each individual during the 3 speeds of the control
condition. The mean coefcient of determination (R2) was .96 ± .02.
A similar relationship was seen between these variables during the
treatment condition.
Individual Response to Compression
The individual metabolic response to compression was quite variable
(ΔVO2 range –4.8% to 5.1%), with some runners consistently show-
ing a higher VO2 during the CON condition and others displaying
higher VO2 while wearing compression (Figure 3). We completed
a post hoc analysis using subjects with the greatest improvements
(n = 4) and largest decrements (n = 4) in RE (from CON to CS)
to determine if there were any distinguishing characteristics that
explained these individual responses to compression. There were no
differences between groups in spring-mass model parameters (Table
Figure 1 — Submaximal oxygen consumption (VO2) during control (black
circles) and compression (white circles) conditions. Slopes of the running
economy lines are 0.230 during control and 0.233 during compression
trials. Values are mean ± SE.
Table 2A Gait Variables, Mean ± SE
Gait Variable
Speed (m/min) Condition Ground-contact time (s) Swing time (s) Step frequency (Hz) Step length (m)
233 Control 0.204 ± 0.003 0.507 ± 0.010 2.831 ± 0.038 1.38 ± 0.02
Compression sleeve 0.205 ± 0.003 0.506 ± 0.009 2.827 ± 0.035 1.38 ± 0.02
268 Control 0.188 ± 0.003 0.503 ± 0.010 2.907 ± 0.037 1.53 ± 0.02
Compression sleeve 0.189 ± 0.003 0.500 ± 0.009 2.910 ± 0.035 1.53 ± 0.02
300 Control 0.175 ± 0.002 0.495 ± 0.008 2.990 ± 0.037 1.68 ± 0.02
Compression sleeve 0.175 ± 0.002 0.495 ± 0.009 2.994 ± 0.036 1.67 ± 0.02
Table 2B Calculated Variability (SD) in Measured Gait Variables, Mean ± SE
Gait Variability (SD)
Speed (m/min) Condition Ground-contact time (s) Swing time (s) Step frequency (Hz) Step length (m)
233 Control 0.0059 ± 0.0012 0.0096 ± 0.0013 1.62 ± 0.12 0.013 ± 0.001
Compression sleeve 0.0066 ± 0.0014 0.0088 ± 0.0012 1.57 ± 0.17 0.013 ± 0.001
268 Control 0.0055 ± 0.0010 0.0086 ± 0.0012 1.71 ± 0.22 0.015 ± 0.002
Compression sleeve 0.0052 ± 0.0011 0.0080 ± 0.0009 1.72 ± 0.21 0.015 ± 0.002
300 Control 0.0054 ± 0.0008 0.0086 ± 0.0010 1.93 ± 0.21 0.018 ± 0.002
Compression sleeve 0.0051 ± 0.0014 0.0082 ± 0.0015 1.81 ± 0.19 0.017 ± 0.002
79
Figure 2 — Mass-specic rates of oxygen consumption (VO2) increase linearly with inverse ground-contact time (1/tc) for all subjects. Average coef-
cient of determination (R2) is .96.
Figure 3 — Individual percentage changes in oxygen consumption (VO2) from the control (CON) to compression (CS) condition across 3 submaximal
speeds.
80 Stickford et al
Table 3 Main Mechanical Parameters During Treadmill Running (Mean ± SD)
Speed Parameter Group CON Group CS Positive CON Positive CS Negative CON Negative CS
233 m/min Δyc (m) 0.05 ± 0.01 0.05 ± 0.01 0.05 ± 0.01 0.05 ± 0.01 0.05 ± 0.01 0.06 ± 0.01
ΔL (m) 0.14 ± 0.02 0.14 ± 0.02 0.14 ± 0.01 0.14 ± 0.01 0.15 ± 0.03 0.15 ± 0.03
Fmax (kN) 1.80 ± 0.27 1.79 ± 0.24 1.82 ± 0.09 1.84 ± 0.05 1.73 ± 0.44 1.69 ± 0.38
kvert (kN/m) 36.53 ± 8.47 35.84 ± 7.44 35.26 ± 4.98 36.00 ± 3.49 33.60 ± 13.66 32.38 ± 11.59
kleg (kN/m) 13.67 ± 3.28 13.41 ± 2.88 13.33 ± 1.71 13.62 ± 1.11 12.63 ± 5.37 12.15 ± 4.90
268 m/min Δyc (m) 0.04 ± 0.01 0.04 ± 0.01 0.05 ± 0.00 0.05 ± 0.00 0.05 ± 0.01 0.05 ± 0.01
ΔL (m) 0.14 ± 0.02 0.14 ± 0.02 0.14 ± 0.01 0.15 ± 0.01 0.15 ± 0.02 0.15 ± 0.03
Fmax (kN) 1.88 ± 0.26 1.86 ± 0.25 1.91 ± 0.05 1.89 ± 0.05 1.84 ± 0.46 1.84 ± 0.44
kvert (kN/m) 43.42 ± 9.38 42.97 ± 9.20 43.12 ± 4.73 42.04 ± 4.91 41.95 ± 14.94 42.70 ± 15.89
kleg (kN/m) 13.41 ± 3.01 13.27 ± 2.97 13.49 ± 1.30 13.14 ± 1.35 13.03 ± 4.90 13.27 ± 5.22
300 m/min Δyc (m) 0.04 ± 0.00 0.04 ± 0.00 0.04 ± 0.00 0.04 ± 0.00 0.04 ± 0.00 0.04 ± 0.00
ΔL (m) 0.15 ± 0.02 0.14 ± 0.02 0.15 ± 0.01 0.15 ± 0.01 0.15 ± 0.02 0.14 ± 0.02
Fmax (kN) 1.96 ± 0.25 1.97 ± 0.27 1.97 ± 0.05 1.96 ± 0.08 1.95 ± 0.43 1.96 ± 0.43
kvert (kN/m) 52.14 ± 10.49 53.30 ± 12.77 49.45 ± 1.44 48.79 ± 3.97 52.44 ± 16.50 53.20 ± 16.38
kleg (kN/m) 13.68 ± 2.92 14.00 ± 3.54 13.14 ± 1.35 12.97 ± 1.15 13.84 ± 4.66 14.05 ± 4.62
Note: Spring-mass model parameters based on a sine-wave modeling of the force–time curve during contact (calculations based on Morin et al19). Positive refers to
improved running economy with compression (n = 4); negative refers to worsened running economy with compression (n = 4). No signicant differences in any measures
between conditions or groups.
Abbreviations: CON, control; CS, compression trial. Δyc, vertical displacement of the center of mass when it reaches its lowest point; ΔL, peak displacement of the leg
spring; Fmax, maximal ground-reaction force during contact; kvert, vertical stiffness; kleg, stiffness of the leg spring.
Table 4 Comparisons in Stride-Frequency (SF) and Stride-Length (SL) Variability (SD) During the Compression
Trial Between Subjects Who Showed Improvements (n = 4) and Decrements (n = 4) in Running Economy
Speed (m/min) Improved running economy Worsened running economy
P
Effect size (Cohen
d
)
SF Variability 233 1.11 ± 0.18 1.77 ± 0.38 .02* 2.22
268 1.37 ± 0.20 2.24 ± 1.23 .14 1.00
300 1.56 ± 0.29 2.20 ± 1.02 .18 0.85
SL Variability 233 0.009 ± 0.001 0.014 ± 0.003 .04* 2.24
268 0.012 ± 0.002 0.020 ± 0.012 .17 0.93
300 0.015 ± 0.001 0.021 ± 0.011 .20 0.77
Note: P value reported from comparison between groups (1-tailed independent-groups t test). Values are mean ± SD.
*Signicantly different at P < .05.
3). However, there was a clear trend for runners who improved RE
with compression to have lower SF- and SL-variability measures
(Cohen d = 0.77–2.24; Table 4). At the slowest running speed,
subjects with improved RE tended to further reduce gait-variability
measures when wearing compression, while subjects with worsened
RE increased gait variability (d = 1.22, 1.17, P = .09, .10 for SF and
SL variability, respectively; Figure 4). Runners with improved RE
also showed greater consistency in absolute SF and SL measures
from CON to CS than subjects whose RE worsened (d = 1.02, 1.08,
and 0.67, P = .14, .13, .21 for SF across speeds; d = 0.66, 1.08,
and 0.67, P = .20, .13, .21 for SL across speeds; Figure 5 shows
averages). In addition, there was a signicant inverse correlation
between belief scores and changes in VO2 with compression (r =
–.52, P = .04), as runners who exhibited more positive feelings about
compression garments displayed larger decrements in submaximal
VO2 during CS (Figure 6).
Discussion
This study investigated the impact of wearing lower-leg compres-
sion sleeves on the RE and mechanics of highly trained distance
runners. Our conclusions can be divided into 3 major categories:
RE, running mechanics, and the individual variability in response
to compression. While the use of lower-leg compression sleeves did
not affect group measures of RE or mechanics, subjects who had
a positive or negative change in RE with compression treatment
displayed different biomechanical responses to, and a priori beliefs
regarding, compression.
Running Economy
In the current study we found that lower-leg compression sleeves
do not alter whole-body VO2 during submaximal running. The
Effect of Compression on Gait and Economy 81
absence of a group effect of compression on submaximal VO2 is
consistent with previous ndings. Kemmler et al7 saw no change
in the VO2 of moderately trained runners (VO2max = 52 mL · kg–1 ·
min–1; 10-km best = 40:36) at various workloads (established based
on blood lactate values; submaximal speeds lower than those in the
current study) when wearing compressive stockings as compared
with during control conditions. Bringard et al20 also showed no
effect of compression on submaximal running gross VO2 compared
with regular elastic tights in trained runners (speeds 167–268 m/
min). However, they did observe a signicant reduction in the slow
component of VO2 while compression tights were worn, presumably
indicating improved energetics. Similarly, in subjects with tness
levels more comparable to that of our subjects (VO2max = 70 mL ·
kg–1 · min–1; 10-km best ≈ 38:30), there was no effect of compres-
sion on VO2 during a 40-minute run (at a speed corresponding to
the slowest speed, 233 m/min, of our study).6 Our ndings add to
and support the literature regarding compression garments and RE
in healthy individuals, providing additional evidence for no gross
effect of compression on submaximal VO2, specically in a group
of elite runners.
Running Mechanics
Few studies have investigated the effect of compression on the
selected mechanical variables measured in our study.8 However,
previous ndings indicate that lower-body compression garments
may inuence lower-body mechanics by limiting range of motion
(ROM).2,3 Decreased ROM is associated with increased stiffness,17
which, in turn, can alter the selected gait variables measured in the
current study. Furthermore, compression garments appear to allow
greater consistency in mechanics/performance during repetitive
power movements.4 Thus, previous ndings would suggest that
lower-body compression may affect the mechanics of movement.
Bringard et al7 and Kemmler et al20 seem to support this hypothesis
by proposing that compression treatment may act through increased
biomechanical support of the muscle–tendon unit, ultimately result-
ing in greater mechanical efciency. We were unable to conrm
this, however, as we found no overall effect of wearing lower-limb
compression on leg stiffness or running mechanics. The inability of
compression to exert an effect on running mechanics in the current
study could be due to a number of factors including the targeted gait
variables, the level and/or location of compression exerted by the
calf sleeve used, and, perhaps more likely, the subject population.21
Nonetheless, as will be discussed further, our ndings suggest that
there may be distinct individual metabolic and mechanical responses
Figure 4 — Percentage changes in average stride-frequency (SF) and
stride-length (SL) variability from control to compression in individuals
who showed an improvement in running economy (black bars) and those
whose running economy worsened (gray bars) at 233 m/min (d = 1.22,
1.17; P = .09, .10, for SF and SL variability, respectively). No differences
between groups at other speeds (d = 0.05–0.23, P = .38–.49). Values are
mean ± SE.
Figure 5 — Percent changes in stride frequency and stride length from
control to compression at a speed of 233 m/min in individuals who showed
an improvement in running economy (black bars) and those whose running
economy worsened (gray bars). Values are mean ± SE.
Figure 6 — Correlation between change in oxygen consumption (ΔVO2)
from control (CON) to compression (CS) and belief score (r = –.517, P
= .041). Score was determined from answers to pretesting questionnaire
about experience with and opinions on compression garments—positive
values assigned to experience/positive responses; negative values assigned
to no experience/negative responses.
82 Stickford et al
to compression, perhaps partially explaining the lack of a group
response in our, and previous, studies.
During submaximal running, the majority of stored energy
comes from muscles and tendons supporting the ankle and knee;
for the lower leg, this refers specically to the triceps surae and the
Achilles tendon.22 The lower band of the compression sleeve used
in our study sat immediately superior to the lateral malleolus. Thus,
although it covered most of the triceps surae and a portion of the
tendon, the compression did not cover the entire Achilles tendon,
nor did it surround the ankle joint, itself. In previous investiga-
tions nding decreased ROM, the garment completely covered the
joint about which ROM was reduced.2,3 On the other hand, jump-
performance consistency improves with a garment covering just
the hip joint, despite multiple muscles and joints being involved
in the movement,4 indicating that complete compression about all
involved muscles and joints is not necessary to see changes in global
mechanical variables. As running, too, involves the coordination
of various muscles and joints, it is unlikely that the lack of gross
mechanical effect in the current study was due to the garment not
covering a joint. Also of note, as is the case in the majority of studies
on compression garments,21 the precise level of pressure exerted by
the sleeves on the calf in our study is unknown; we relied fully on
the manufacturer’s statement. As such, the magnitude of compres-
sion used in our study may be different from that of garments used
in previous studies. Direct measurements of compression-garment
pressures have been shown to be comparable to manufacturer rat-
ings,6 so based on available reports, the compression appears to be
similar between our study (15–20 mm Hg) and others (range = 8–26
mm Hg).21 We chose this particular compression-sleeve garment as
it is relatively inexpensive, widely available, and commonly used;
in other words, ndings regarding the garment are quite applicable
to the target athlete population.
The subjects in our study were highly trained distance runners,
whereas many previous studies used less-t subjects and athletes
specializing in other events (eg, sprinters, skiers).2–7,20 Attempts to
alter running mechanics have been successful in clinical popula-
tions and in novice or moderately trained runners,23,24 but it appears
that altering habitual gait patterns in highly trained athletes is more
difcult. It has been demonstrated that highly trained runners con-
sistently select the most economically optimal locomotion style;
deviating from the preferred gait mechanics results in signicant
decrements in economy of locomotion.11,13 Furthermore, trained
runners are able to maintain similar gait patterns by adjusting, for
example, leg stiffness or muscle activity despite external interfer-
ences (eg, changes in surface properties).25 Therefore, our subject
cohort, being highly trained, may have been able to continue to select
the most economical running pattern despite the “interference” of
lower-leg compression.
Individual Response to Compression
Although the group mean metabolic response to compression was
not different from CON, there was substantial interindividual vari-
ability. A change of 1% in submaximal VO2 is reported to have
signicant performance implications, demonstrating a tight link
between RE and running performance.26,27 Therefore, we completed
a post hoc analysis to determine if there were any distinguishing
characteristics that explained these individual responses to com-
pression. When subjects were grouped according to changes in
RE, it became evident that those who improved RE while wearing
compression had lower measures of gait variability, particularly
at the slowest speed during CS (P = .04 [SL], 0.02 [SF]; Table 4).
Variability in intraindividual stride length, already lower in highly
trained runners than in nonrunners,28 decreases with faster running
speeds,29 so it is reasonable that we did not see signicant differences
in gait variability between groups at the faster speeds. In trained
runners, a large proportion of the increased energy cost of running
with fatigue is due to increased step variability,10 so decreased gait
variability may contribute to the decrease in submaximal VO2 with
compression seen in some subjects in our investigation.
Differences between groups in the degree/direction of change
in SL and SF may also have affected RE responses to compression.
Subjects whose economy worsened while wearing compression
tended to decrease SL and increase SF, whereas subjects with
improvements in RE had little to no change in SL and SF across
conditions. As mentioned previously, highly trained athletes typi-
cally select gait patterns that minimize metabolic cost, and devia-
tions from preferred kinematics result in increases in oxygen cost.
If the control condition represents a preferred movement pattern, it
appears that the compression caused some subjects to deviate from
their most economical running mechanics for a given speed. The
underlying cause as to why compression would affect the stride
mechanics and gait variability of some subjects and not others is
unknown. One possible explanation may be subjective percep-
tions of the garment, as runners with personal experience and/or
positive opinions of the garment may have felt more comfortable
(psychologically and/or physically) wearing compression. How-
ever, as there were 4 individuals who exhibited positive responses,
8 showing neutral responses, and 4 exhibiting negative responses
to compression, there is certainly a possibility that the variation in
response occurred by chance, and the sleeves truly have no effect
on metabolic or mechanical variables in highly trained runners.
Practical Applications
In general, previous research and the current investigation indicate
that wearing lower-leg compression sleeves has no substantial
effect on distance-running performance.5–8,21 However, if, in fact,
differences do exist in metabolic response to lower-leg compression
among individuals, the performance implications could be particu-
larly important for highly trained runners. In the current study, the
average percentage change in submaximal VO2 with compression
treatment ranged from –4.8% to +5.1%. A 5% improvement in
RE has been shown to induce ~3.8% improvement in distance-
running performance, and a 10% improvement in RE results in ~7%
improvement in 3000-m performance,27,28 indicating that there could
certainly be a range of performance implications for the subjects in
our study when choosing to wear compression garments.
Finally, it is not unreasonable to suspect that the different
responses to compression are due to psychological effect. A sig-
nicant correlation was found between overall viewpoints on the
efcacy of compression and metabolic changes while wearing
compression, and the 2 subjects in our study with the greatest
improvements in RE with compression treatment were the only sub-
jects who had worn compression sleeves themselves and believed
that compression aided in training, racing, and recovery. Thus, in
determining whether an athlete should use compression garments
for training, competitive, and/or recovery purposes, coaches would
be well-advised to consider the athlete’s expectations, whether
positive or negative, of the garment.
Effect of Compression on Gait and Economy 83
Conclusions
Wearing lower-leg compression sleeves does not alter running
mechanics or economy in highly trained distance runners during
submaximal running. However, while there is no group difference
with the sleeves, it appears that the individual response to wearing
lower-leg compression varies greatly; as such, future research should
examine underlying physiological, anatomical, and psychological
characteristics in relation to the metabolic and/or kinematic out-
comes of wearing leg-compression garments.
Acknowledgments
Supported by a grant from USA Track and Field. We would like to
acknowledge Dr Dave Tanner for his work on the accelerometric gait-
analysis program and Dr S. Lee Hong for input regarding study design
and theoretical implications.
None of the authors of this article has any conicts of interest or nan-
cial conicts to report. The results of the current study do not constitute
endorsement of the product by the authors or the journal.
References
1. Berry MJ, McMurray RG. Effects of graduated compression stockings
on blood lactate following an exhaustive bout of exercise. Am J Phys
Med. 1987;66(3):121–132. PubMed
2. Bernhardt T, Anderson GS. Inuence of moderate prophylactic com-
pression on sport performance. J Strength Cond Res. 2005;19(2):292–
297. PubMed
3. Doan BK, Kwon YH, Newton RU, et al. Evaluation of a lower-body
compression garment. J Sports Sci. 2003;21(8):601–610. PubMed
doi:10.1080/0264041031000101971
4. Kraemer WJ, Bush JA, Bauer JA, et al. Inuence of compression gar-
ments on vertical jump performance in NCAA Division I volleyball
players. J Strength Cond Res. 1996;10(3):180–183.
5. Ali A, Caine MP, Snow BG. Graduated compression stockings: physi-
ological and perceptual responses during and after exercise. J Sports
Sci. 2007;25(4):413–419. PubMed doi:10.1080/02640410600718376
6. Ali A, Creasy RH, Edge JA. The effect of graduated compression stock-
ings on running performance. J Strength Cond Res. 2011;25(5):1385–
1392. PubMed doi:10.1519/JSC.0b013e3181d6848e
7. Kemmler W, von Stengel S, Kockritz C, Mayhew J, Wassermann A,
Zapf J. Effect of compression stockings on running performance in
male runners. J Strength Cond Res. 2009;23(1):101–105. PubMed
doi:10.1519/JSC.0b013e31818eaef3
8. Varela-Sanz A, España J, Carr N, Boullosa DA, Esteve-Lanao J.
Effects of gradual-elastic compression stockings on running econ-
omy, kinematics, and performance in runners. J Strength Cond Res.
2011;25(10):2902–2910.
9. Wakeling JM, Nigg BM. Modication of soft tissue vibrations in the
leg by muscular activity. J Appl Physiol. 2001;90(2):412–420. PubMed
10. Candau R, Belli A, Millet GY, Georges D, Barbier B, Rouillon JD.
Energy cost and running mechanics during a treadmill run to vol-
untary exhaustion in humans. Eur J Appl Physiol Occup Physiol.
1998;77(6):479–485. PubMed doi:10.1007/s004210050363
11. Hunter I, Smith GA. Preferred and optimal stride frequency, stiffness,
and economy: changes with fatigue during a 1-h high-intensity run.
Eur J Appl Physiol. 2007;100(6):653–661. PubMed doi:10.1007/
s00421-007-0456-1
12. Anderson T. Biomechanics and running economy. Sports Med.
1996;22(2):76–89. PubMed doi:10.2165/00007256-199622020-00003
13. Cavanagh PR, Williams KR. The effect of stride length variation
on oxygen uptake during distance running. Med Sci Sports Exerc.
1982;14(1):30–35. PubMed doi:10.1249/00005768-198201000-00006
14. Williams KR, Cavanagh P. Relationship between distance running
mechanics, running economy, and performance. J Appl Physiol.
1987;63(3):1236–1245. PubMed
15. Conley DL, Krahenbuhl GS. Running economy and distance run-
ning performance of highly trained athletes. Med Sci Sports Exerc.
1980;12(5):357–360. PubMed doi:10.1249/00005768-198025000-
00010
16. Costill DL, Thomason H, Roberts E. Fractional utilization of the aero-
bic capacity during distance running. Med Sci Sports. 1973;5(4):248–
252. PubMed doi:10.1249/00005768-197300540-00007
17. Foster C, Lucia A. Running economy: the forgotten factor in
elite performance. Sports Med. 2007;37(4-5):316–319. PubMed
doi:10.2165/00007256-200737040-00011
18. McMahon TA, Cheng G. The mechanics of running: how does stiff-
ness couple with speed? J Biomech. 1990;23(Suppl 1):65–78. PubMed
doi:10.1016/0021-9290(90)90042-2
19. Morin JB, Dalleau G, Kyrolainen H, Jeannin T, Belli A. A simple
method for measuring stiffness during running. J Appl Biomech.
2005;21(2):167–180. PubMed
20. Bringard A, Perrey S, Belluye N. Aerobic energy cost and sensation
responses during submaximal running exercise—positive effects of
wearing compression tights. Int J Sports Med. 2006;27:373–378.
PubMed doi:10.1055/s-2005-865718
21. MacRae BA, Cotter JD, Laing RM. Compression garments and
exercise: garment considerations, physiology and performance.
Sports Med. 2011;41(10):815–843. PubMed doi:10.2165/11591420-
000000000-00000
22. Alexander M. The spring in your step. New Sci. 1987;4:41.
23. Crowell H, Milner C, Davis IS, Hamill J. Short-term retention of gait
changes after real-time feedback to reduce tibial shock. Med Sci Sports
Exerc. 2005;37(5):S346.
24. Messier SP, Cirillo KJ. Effects of a verbal and visual feedback system
on running technique, perceived exertion and running economy in
female novice runners. J Sports Sci. 1989;7(2):113–126. PubMed
doi:10.1080/02640418908729830
25. Ferris DP, Louie M, Farley C. Running in the real world: adjusting
leg stiffness for different surfaces. Proc Biol Sci. 1998;265:989–994.
PubMed doi:10.1098/rspb.1998.0388
26. Di Prampero PE, Capelli C, Pagliaro P, et al. Energetics of best perfor-
mances in middle-distance running. J Appl Physiol. 1993;74(5):2318–
2324. PubMed
27. Falloweld J, Wilkinson D. Improving Sports Performance in Middle
and Long Distance Running: A Scientic Approach to Race Prepara-
tion. New York: Wiley & Sons; 1995.
28. Nakayama Y, Kudo K, Ohtsuki T. Variability and uctuation in run-
ning gait cycle of trained runners and non-runners. Gait Posture.
2010;31(3):331–335. PubMed doi:10.1016/j.gaitpost.2009.12.003
29. Jordan K, Challis J, Newell K. Long range correlations in the stride
interval of running. Gait Posture. 2006;24(1):120–125. PubMed
doi:10.1016/j.gaitpost.2005.08.003
