ArticlePDF Available

Are we ready to measure running power? Repeatability and concurrent validity of five commercial technologies



Training prescription and load monitoring in running activities have benefited from power output (PW) data obtained by new technologies. Nevertheless, to date, the suitability of PW data provided by these tools is still uncertain. In order to clarify this aspect, the present study aimed to: i) analyze the repeatability of five commercially available technologies for running PW estimation, and ii) examine the concurrent validity through the relationship between each technology PW and oxygen uptake (VO2). On two occasions (test-retest), twelve endurance-trained male athletes performed on a treadmill (indoor) and an athletic track (outdoor) three submaximal running protocols with manipulations in speed, body weight and slope. PW was simultaneously registered by the commercial technologies StrydApp, StrydWatch, RunScribe, GarminRP and PolarV, while VO2 was monitored by a metabolic cart. Test-retest data from the environments (indoor and outdoor) and conditions (speed, body weight and slope) were used for repeatability analysis, which included the standard error of measurement (SEM), coefficient of variation (CV) and intraclass correlation coefficient (ICC). A linear regression analysis and the standard error of estimate (SEE) were used to examine the relationship between PW and VO2. Stryd device was found as the most repeatable technology for all environments and conditions (SEM≤12.5W, CV≤4.3%, ICC≥0.980), besides the best concurrent validity to the VO2 (r≥0.911, SEE≤7.3%). On the contrary, although the PolarV, GarminRP and RunScribe technologies maintain a certain relationship with VO2, their low repeatability questions their suitability. The Stryd can be considered as the most recommended tool, among the analyzed, for PW measurement.
Are we ready to measure running power? Repeatability and concurrent
validity of five commercial technologies
Human Performance and Sports Science Laboratory, Faculty of Sport Sciences, University of Murcia, Murcia, Spain &
Exercise Physiology Lab at Toledo, University of Castilla-La Mancha, Toledo, Spain
Training prescription and load monitoring in running activities have benefited from power output (P
) data obtained by new
technologies. Nevertheless, to date, the suitability of P
data provided by these tools is still uncertain. In order to clarify this
aspect, the present study aimed to: (i) analyze the repeatability of five commercially available technologies for running P
estimation, and (ii) examine the concurrent validity through the relationship between each technology P
and oxygen
uptake (VO
). On two occasions (test-retest), twelve endurance-trained male athletes performed on a treadmill (indoor)
and an athletic track (outdoor) three submaximal running protocols with manipulations in speed, body weight and slope.
was simultaneously registered by the commercial technologies Stryd
, Stryd
, RunScribe, Garmin
and Polar
while VO
was monitored by a metabolic cart. Test-retest data from the environments (indoor and outdoor) and
conditions (speed, body weight and slope) were used for repeatability analysis, which included the standard error of
measurement (SEM), coefficient of variation (CV) and intraclass correlation coefficient (ICC). A linear regression analysis
and the standard error of estimate (SEE) were used to examine the relationship between P
and VO
. Stryd device was
found as the most repeatable technology
for all environments and conditions (SEM 12.5 W, CV 4.3%, ICC 0.980),
besides the best concurrent validity to the VO
(r0.911, SEE 7.3%). On the contrary, although the Polar
, Garmin
and RunScribe technologies maintain a certain relationship with VO
, their low repeatability questions their suitability.
The Stryd can be considered as the most recommended tool, among the analyzed, for P
Keywords: Testing, biomechanics, exercise, kinetics, physiology
Defined as the product of force and velocity (Cross,
Brughelli, Samozino, & Morin, 2017), power
output (P
) is one of the variables most used to
monitor workload in endurance sports (Bourdon
et al., 2017). P
provides a more instantaneous,
reliable and sensitive measure of exercise intensity
than other internal and external parameters such as
the heart rate or speed (Allen & Coggan, 2010; Haa-
konssen, Martin, Burke, & Jenkins, 2013; Sanders,
Myers, & Akubat, 2017). As a consequence of the
extensive advantages of this parameter, a large
number of companies have developed measuring
devices that allow a real time assessment of P
the field. For instance, in cycling, P
can be
measured by registering directly the force applied in
different parts of the bicycle such as the pedal,
crank, bottom bracket, chain or rear wheel hub (Pass-
field, Hopker, Jobson, Friel, & Zabala, 2017). For
running activities, new emerging devices use infor-
mation from the global positioning system (outdoor
environments) and/or from inertial sensors, like
accelerometers and gyroscopes (indoor environ-
ments), to track velocities and accelerations and esti-
mate P
from there (Aubry, Power, & Burr, 2018;
Austin, Hokanson, McGinnis, & Patrick, 2018;
García-Pinillos, Latorre-Román, Roche-Seruendo,
& García-Ramos, 2019). Although these technol-
ogies providing measurements of running P
signified an important advance for training
© 2020 European College of Sport Science
Correspondence: Jesús García Pallarés Faculty of Sport Sciences, University of Murcia, C/ Argentina s/n. Santiago de la Ribera, Murcia,
Spain. E-mail:
European Journal of Sport Science, 2020
TEJS1748117 Techset Composition India (P) Ltd., Bangalore and Chennai, India 4/4/2020
prescription and load monitoring (Bourdon et al.,
2017), the feasibility of the P
data they provide is
still uncertain in running activities. In order to
clarify this aspect, the measuresrepeatability (i.e.
the variation observed in repeated trials or sessions
made on the same subject under identical conditions)
(Bartlett & Frost, 2008) is an essential aspect that
should be examined. Since of the main goal of P
monitoring is to determine the real effort being
incurred during training (Mujika, 2017), it is essen-
tial to identify whether the changes observed in P
between sessions are due to the actual changes in ath-
letesperformance or due to measurement error
(Hopkins, 2000).
On other hand, the use of the P
parameter for
training prescription and predicting athletesper-
formance could be expanded to calculations of
energy expenditure due to its tight relationship with
oxygen uptake (VO
). Whereas in cycling, both par-
ameters have been found to have a very close associ-
ation (Arts & Kuipers, 1994;Muniz-Pumares,
Pedlar, Godfrey, & Glaister, 2017;Pedersen,
Sørensen, Jensen, Johansen, & Levin, 2002), this
relationship is still unclear in running activities. To
the best of our knowledge, only one previous study
has analyzed this relationship finding a very weak
association between P
and VO
(Aubry et al.,
2018). These authors hypothesized that the mech-
anics through which P
is obtained in running activi-
ties (indirectly), as well as different factors (such as
the running surface, coefficient of drag, or wind)
could explain this weak association. However, the
findings of this study have been questioned due to
its major methodological flaws from protocol and
analysis techniques (Snyder, Mohrman, Williamson,
& Li, 2018). Therefore, this relationship should be
clarified by more evidence with a suitable
Taking into account the scarcity of scientific litera-
ture analyzing P
estimated by running devices, we
deemed it necessary to conduct this study with the
following purposes: (i) to analyze the test-retest
repeatability of different technologies for P
ing, and (ii) to examine the relationship between P
and oxygen uptake (VO
) (concurrent validity).
Twelve endurance-trained male athletes volunteered
to participate in this study (age 25.7 ± 7.9 years, body
mass 73.0 ± 8.7 kg, height 177.8 ± 4.9 cm, body fat
11.6 ± 1.4%, VO
61.3 ± 5.6 ml·kg
endurance training experience 6.7 ± 2.4 years). All
subjects were competing at regional and national
level of triathlon or running races and following a
regular training load of 46 days per week, 12
hours per day. Measurements were obtained during
the pre-competitive season. They underwent a com-
plete medical examination, including electrocardio-
gram (ECG), that showed all were in good health
for being part of the present study. No physical limit-
ations or musculoskeletal injuries that could affect
testing procedures were reported. None of the sub-
jects were taking drugs, medications or dietary sup-
plements known to influence physical performance.
The Bioethics Commission of the University of
Murcia approved the study, which was carried out
according to the declaration of Helsinki. Subjects
were verbally informed about the experimental pro-
cedures and possible risk and benefits. Written
informed consent was obtained from all subjects.
Study design
Subjects performed 5 testing sessions on separate
days with 4872 hours rest in between: one familiar-
ization session, two in the laboratory (Indoor
) and two in the field (Outdoor
). In the first session, subjects were familiar-
ized with the testing procedures used in this investi-
gation and were subjected to a maximal graded
exercise test. On the second and third day (indoor
sessions), subjects performed three submaximal lab-
oratory running tests on a treadmill while VO
collected: (i) a submaximal multistage test with
speed increases (test 1), (ii) a submaximal speed mul-
tistage test with body weight increases (test 2) and
(iii) a submaximal speed multistage test with
changes in the slope (test 3) (described later in
detail). On the fourth and fifth day (outdoor ses-
sions), subjects completed the aforementioned tests
1 and 2 on an athletic track. Test-retest repeatability
analysis for the different technologies included was
made by comparing the first and the second session
in each environment (i.e. Indoor
vs. Indoor
vs. Outdoor
). Furthermore, data from
session 1 in each environment (Indoor
) were used to analyze the relationship
between PW and VO
(Concurrent validity).
Familiarization and graded exercise testing. During the
familiarization session, subjects were subjected to a
maximal graded exercise test under medical supervi-
sion to fulfil two objectives: (Cross et al., 2017)to
discard cardiovascular diseases, and (Bourdon
et al., 2017) to determine the athletes maximal VO
). SubjectsHR was monitored by standard
2V. Cerezuela-espejo et al.
12 lead ECG (Quark T12, Cosmed, Italy), venti-
latory performance (VO
, and ventilation)
was recorded on a breath-by-breath basis using a
metabolic cart (MetaLyzer 3B- R3, Cortex Biophysik
GmbH, Leipzig, Germany) averaging data every 5 s
and the rate of perceived exertion (RPE) was assessed
using the 620 Borg Scale every 2 min. VO
defined as the highest plateau (two successive
maximal within 150 ml·min
) when averaging data
every 5 s (Cerezuela-Espejo, Courel-Ibáñez, Morán-
Navarro, Martínez-Cava, & Pallarés, 2018).
Indoor sessions. Subjects underwent two similar lab-
oratory sessions (Indoor
and Indoor
) on a treadmill
with VO
continuous assessment, simulating changes
in three running conditions (Van Dijk & Van Megen,
2017): speed (test 1), body weight (test 2) and slope
(test 3):
Incremental speed test (test 1). The speed pro-
tocol carried out in this study was a modification of
the test used by Aubry et al. (2018). It was composed
of stages of 3 minutes of work and 4 minutes of rest
(3:4 ratio). Athletes started to run at 9 km·h
increments in each stage until finding a rel-
evant involvement of anaerobic metabolism
(described later in detail).
Incremental body weight test (test 2). Follow-
ing the same 3:4 ratio and at a fixed speed
(10 km·h
), athletes completed three stages in
which their body weight was modified through a
weighted vest: body weight +0, +2.5, and +5.
Incremental slope test (test 3). Following the
same 3:4 ratio and at a fixed speed (10 km·h
), ath-
letes completed five stages in which treadmill incli-
nation was modified: 6%, 3%, 1%, +3% and
The aforementioned tests were performed on the
same treadmill (HP Cosmos Pulsar; HP Cosmos
Sports & Medical GMBH, Nussdorf Traunstein,
Germany) in both sessions (Indoor
and Indoor
In order to simulate air resistance that commonly
exists in outdoor running, a fixed incline of 1.0%
was used in test 1 (incremental speed) and 2 (incre-
mental body weight) (Jones & Doust, 1996). Air ven-
tilation was controlled with a fan positioned 1.5 m
lateral to the subjects at a wind velocity of 2.55 m·s
Outdoor sessions. Outdoor
and Outdoor
were designed to simulate test 1 (incremental
speed) carried out in the laboratory, following the
same methodology. Subjectsrunning pace (speed)
was controlled using automated sound beeps in a
prerecorded file. Participant had to reach a pylon
placed every 25 m along the track, on each beep.
To ensure that mainly the aerobic energy system
was involved, the ventilatory parameters (VO
, and ventilation) were recorded on a
breath-by-breath basis using a metabolic cart and
lactate samples were collected at the end of each
stage using a lactate analyzer (Lactate Pro 2,
Arkray, Japan), in both environments (indoor and
outdoor) and conditions (speed, body weight and
slope) assessed. This main participation of the
aerobic metabolism was established while (i) respirat-
ory exchange ratio was lower than 1 or (ii) blood
lactate concentration did not exceed the resting
level by 1 mMol·L
(Bijker, De Groot, & Hollander,
2001;Cerezuela-Espejo et al., 2018;Coyle, Sidossis,
Horowitz, & Beltz, 1992;Pallarés, Morán-Navarro,
Ortega, Fernández-Elías, & Mora-Rodriguez,
2016). In addition, only the last 1.5 minutes of each
stage were used for the analysis in both environments
and conditions, in order to allow the stabilization of
(Whipp & Ward, 1990). The same
metabolic cart, under the same calibration pro-
cedures, was used in both environments (indoor
and outdoor sessions), made it possible to ensure
that VO
values resulted from the modification of
conditions (speed, body weight and slope), and not
from the inter-analyzers variation or the extra-
weight generated by the wearable metabolic analy-
zers. In outdoor sessions, this indirect calorimetry
was placed on a trailer and attached to a bike
ridden by a researcher in parallel to the subject.
In addition, taking into account the influence of the
cadence parameter in P
values, the athletes
cadence at 13 km·h
was recorded by a metronome,
prior to Indoor
session, and kept fixed during the
remaining environments and conditions (Austin
et al., 2018). Evaluations were performed under
similar climatological conditions (2124°C and 45
55% relative humidity) at the same time of day
(16:0019:00 hours) to minimize the circadian
rhythm effects (Pallarés et al., 2015).
Power output technologies
The measures of P
were obtained simultaneously
from 4 (indoor) and 5 (outdoor) different technol-
ogies during all tests carried out in the present study:
- Stryd (Stryd Summit Powermeter, firmware 1.2;
Stryd, Inc., Boulder, CO, USA). Because this
device can be connected with both a mobile
App and a sport watch, the Stryd was paired to
the sport watch Garmin Forerunner 235
(Garmin Ltd., Olathe, Kansas, USA)
) and to the manufacturer App
Are we ready to measure running power? 3
version 5.13 (Stryd
) installed on the smart-
phone Xiaomi A2 (Xiaomi, Haidian, Beijing,
Republic of China).
- RunScribe (RunScribe Plus V3, Scribe Labs, Inc.,
Half Moon Bay, CA, USA): System composed
by two pedometers, each one attached on the
right and left shoe of the subject. In this study,
RunScribe was paired to the sport watch
Garmin Forerunner 920 XT for data
- Garmin Running Power (v1.6, Olathe, Kansas,
USA) (Garmin
): This tool uses a combination
of data from the watch and an external accessory
in order to estimate the P
parameter. In this
regard, the present study used the sport watch
Garmin Forerunner 935 and the Tri
rate monitor.
- Polar Vantage V (firmware 3.1.7, Polar, OY,
Kempele, Finland) (Polar
): This sport watch
uses the GPS and barometer sensors to obtain
different parameters such as the maximum,
average or laps P
. Because the Polar
power by using the aforementioned sensors, it
only can be used in outdoor environments.
Each device was strictly assembled according to the
manufacturers specifications and placed on partici-
pantsfeet (inertial sensors) and wrists (sport
watches) in a counterbalanced way (Han, Kim,
Sun, Malaska, & Miller, 2020). A picture of each par-
ticipant wearing the sensors was taken to ensure a
constant location throughout the testing sessions.
Statistical analyses
Normality and homoscedasticity assumptions were
verified using the KolmogorovSmirnov test, the
BrownForsythe robust test, the QQ plots and scat-
tered plots of the residuals. Sphericity was checked
using the Mauchlys test. The analyses of repeatabil-
ity (i.e. the variation observed in repeated trials or ses-
sions made on the same subject under identical
conditions) and concurrent validity (between each
technology P
and VO
) included the calculation of:
- The standard error of measurement (SEM) was
calculated from the square root of the mean
square error term in a repeated-measures
ANOVA to determine the amount of variability
caused by measurement error. Results are pre-
sented both in absolute (W, Watts) and relative
terms as a coefficient of variation (CV = 100
SEM/mean). For most exercise performance
tests, the CV should be lower than 5%
(Hopkins, 2000).
- The level of agreement between paired P
comes from Indoor
vs. Indoor
and Outdoor
vs. Outdoor
was assessed using BlandAltman
analysis and the calculation of systematic bias
and its 95% limits of agreement (LoA = bias ±
1.96 SD) (Bland & Altman, 1986).
- The intraclass correlation coefficient (ICC) was
calculated (1,k), one-way random-effects, absol-
ute agreement, multiple raters/measurements
model, was chosen due to the fact that each
incremental test was assessed by a different set
of technologies. ICC (1,k) and its 95% confi-
dence interval ranges (CI) were calculated
according to Koo and Li guidelines (Koo & Li,
2016). For the assessment of technological
equipment, cut-off values of 0.950.99 are con-
sidered good for research and clinical practice
(Martins & Nastri, 2014).
- Linear regression analysis and Pearsons corre-
lation coefficient (r) were used to assess the
extent of the linear relationship existing
between P
from each technology and VO
comes from the metabolic cart.
- The standard error of the estimate (SEE) was cal-
culated as the standard deviation of the residuals
as a measure of variation around the regression
line. The smaller the value, the closer the data
points are to the regression line and the better
the estimation is. Results are presented in rela-
tive terms (%) as SEE/mean100. Due the accu-
racy of the Pearsonsris very sensitive to the
number and distribution of data pairs included
(i.e. a low and grouped range of data could
mask the rvalue) (Aggarwal & Ranganathan,
2016;Armstrong, 2019;Vincent, 2004), the
SEE (%) was used instead to present the
results from the body weight and slope
Statistical calculations were performed using the
SPSS statistical software version 17 (SPSS Inc.,
Chicago, USA) and figures were designed using
GraphPad Prism 6.0 (GraphPad Software Inc., Cali-
fornia, USA).
Test-retest repeatability
Repeatability analyses in indoor environments
and Indoor
) found the Stryd
as the most repeatable technologies both
in speed, body weight and slope conditions
(SEM 7.4 W, CV 2.8%, ICC 0.980), followed
by RunScribe (SEM 30.1 W, CV 7.4%, ICC
0.709) and Garmin
(SEM 47.0 W, CV 9.4%,
4V. Cerezuela-espejo et al.
ICC 0.495). In outdoor sessions, the Stryd
showed again the best repeatability values
(SEM 12.5 W, CV 4.3%, ICC 0.989), followed
by Garmin
(SEM 24.5 W, CV 7.7%, ICC =
0.823). In contrast, Polar
(SEM 40.6 W, CV
14.5%, ICC = 0.487) and RunScribe (SEM
59.3 W, CV 14.8%, ICC = 0.563) showed the
greatest errors and the lowest repeatability (Tables I
and II).
Relationship between P
and VO
Figure 1 shows the scatter plots of the correlations
between the P
provided by each technology and
measured by the metabolic cart in the incremen-
tal speed test (test 1), indoor (left panels) and
outdoor (right panels). The best agreement between
both parameters during this condition was found
with Stryd device (r0.911, SEE 7.3%), whereas
this relationship was lower for Polar
(r= 0.841,
SEE = 9.7%), RunScribe (r0.582, SEE 13.7%)
and Garmin
(r0.539, SEE 17.5%) technol-
ogies, in both environments. The Stryd device,
especially the Stryd
, maintained this close
relationship when the body weight (SEE 6.31%)
and slope (SEE = 6.79%) were modified. On the con-
trary, the agreement between the VO
and P
mated by the Polar
(body weight, SEE = 9.7%),
(body weight, SEE 10.3%; slope, SEE
= 19.0%) and RunScribe (body weight, SEE
10.3%; slope, SEE = 18.5%) decreased during these
two conditions.
This study aimed: (i) to analyze the repeatability of
four (indoor) and five (outdoor) different technol-
ogies for estimating running P
(test-retest repeat-
ability), and (ii) to examine the relationship
between each technology Pw and VO
validity). We observed that the Stryd device is the
most repeatable technology, among the five analyzed,
for P
estimation. Furthermore, the concurrent val-
idity analysis indicated that P
estimated by the
Stryd device showed the closest relationship with
the VO
directly measured by the metabolic cart.
These superior values of repeatability (Tables I and
II) and agreement with VO
(Figure 1) were observed
for the two environments (indoor and outdoor) and
three conditions (speed, body weight and slope)
examined. Hence, the Stryd device can be considered
as the best tool, among the five analyzed, for P
mation in running activities. On the contrary, despite
the fact that the devices Garmin
, RunScribe and
maintain a certain relationship with VO
0.841), they showed low repeatability, especially
indoors and RunScribe and Polar
doors. To the best of our knowledge, this is the first
study that has simultaneously analyzed a series of
commercial technologies for P
estimation, in differ-
ent conditions and environments. An essential
requirement for sport technological devices is to
provide repeatable outcomes under identical con-
ditions (Hopkins, 2000). Therefore, taking into
account both the ICC cut-off values considered as
good for research and clinical practice (0.950.99)
(Martins & Nastri, 2014), as well as the maximal
CV for most exercise performance tests (<5%)
(Hopkins, 2000), only the Stryd device could be con-
sidered as an adequate tool for P
measurement in
both indoor (CV 2.8%, ICC 0.980) and outdoor
environments (CV 4.3%, ICC 0.989) (Tables I
and II). The second technology with the best values
of repeatability was the RunScribe (CV 7.4%,
ICC 0.709) indoor and Garmin
(CV 7.7%,
ICC = 0.823) outdoor. Although based on the
results of the current study, we cannot recommend
these technologies, future studies should analyze if
new updates could improve their repeatability values
and thus their feasibility for P
The second objective of this investigation demon-
strated an excellent relationship between the Stryd
device, especially the Stryd
, and the VO
tered by the indirect calorimetry, in the two environ-
ments and three conditions examined (r0.911,
SEE 7.3%). Unlike in cycling, where the associ-
ation between both parameters has been strongly
demonstrated (Arts & Kuipers, 1994;Muniz-
Pumares et al., 2017;Pedersen et al., 2002),a
linear association between VO
and running P
not been shown yet. To the best of our knowledge,
only one study has analyzed this relationship in
running by using wearable technologies just like the
present study (Aubry et al., 2018). After evaluating
the effect of 3 conditions (surface, speed and subjects
experience) on the relationship between P
mated by the Stryd device) and VO
, these authors
concluded that P
from this device was not suffi-
ciently accurate as a surrogate of VO
, particularly
in the elite running population. However, the meth-
odological flaws from protocol and analysis tech-
niques identified in this study question its findings
(Snyder et al., 2018). In order to solve these meth-
odological problems, clarify and extend this relation-
ship, the present study: (i) used an adequate and
equitable interval time (in both indoor and outdoor
environments) for measuring VO
after its stabiliz-
ation in each condition (Whipp & Ward, 1990),
and (ii) avoided the normalization of values by
expressing the results in absolute values (ml·min
Are we ready to measure running power? 5
Table I. Test-retest repeatability of P
estimated by each technology in indoor sessions.
(%) ICC
Bland Altman (W)
(%) ICC
Bland Altman (W)
Speed (km/h) 9 3.7 1.9 0.998 1.2 4.6 10.3 to 7.8 5.2 2.8 0.992 1.6 8.1 17.4 to 14.2
10 3.3 1.5 5.7 2.7
11 2.9 1.2 6.3 2.8
12 2.7 1.0 5.8 2.3
13 3.0 1.1 7.4 2.8
14 3.6 1.2 5.3 1.9
15 4.8 1.5 2.8 0.9
Body weight
0 3.3 1.5 0.991 0.8 4.2 9.1 to 7.5 5.7 2.7 0.980 2.2 6.5 14.9 to 10.6
+2.5 2.3 1.0 5.3 2.4
+5.0 3.3 1.4 3.0 1.3
Slope (%) 1.7 1.1 0.999 0.5 3.2 6.8 to 5.9 1.7 0.8 0.988 0.9 4.6 10.0 to 8.1
2.3 1.3 2.2 1.0
+1° 3.3 1.5 5.7 2.7
+3° 2.4 1.0 1.8 0.9
+6° 1.3 0.5 3.5 1.7
Speed (km/h) 9 57.6 15.7 0.290 7.0 96.7 182.6 to 196.5 68.8 16.3 0.709 19.1 87.8 191.1 to 152.9
10 57.0 14.6 64.5 14.1
11 66.5 16.2 54.4 11.0
12 72.8 17.6 50.3 9.5
13 73.0 16.6 45.1 8.15
14 70.5 15.5 73.6 12.6
15 82.7 18.0 86.9 13.7
Body weight
0 57.0 14.6 0.495 17.2 65.9 112.0 to 146.5 64.5 14.1 0.678 12.8 65.3 115.3 to 140.8
+2.5 36.5 9.4 33.5 7.4
+5.0 47.0 11.8 30.1 6.9
Slope (%) 52.1 13.0 0.121 5.4 79.1 149.5 to 160.4 75.7 18.1 0.589 16.9 86.8 153.3 to 187.0
53.5 13.2 69.6 16.0
+1° 57.0 14.6 64.5 14.1
+3° 50.5 12.3 49.9 10.5
+6° 63.7 15.5 42.2 8.5
W: watts; SEM: standard error of measurement; CV: SEM expressed as a coefficient of variation; ICC: intraclass correlation coefficient, model (1,k); LoA: limits of agreement.
6V. Cerezuela-espejo et al.
Table II. Test-retest repeatability of P
estimated by each technology in outdoor sessions.
(%) ICC
Bland Altman (W)
(%) ICC
Bland Altman (W)
9 6.5 3.4 0.989 1.6 10.2 18.4 to 21.6 5.1 2.7 0.990 0.7 9.6 19.6 to 18.2
10 5.7 2.7 6.0 2.8
11 6.3 2.8 5.4 2.4
12 5.0 2.0 3.2 1.3
13 4.9 1.8 4.0 1.5
14 12.5 4.3 9.5 3.3
15 6.8 2.3 10.5 3.5
9 24.5 8.6 0.823 26.7 46.1 117.0 to 63.7 59.3 14.8 0.563 99.3 84.8 265.5 to 66.9
10 33.9 10.7 80.9 18.4
11 27.9 8.8 87.2 19.1
12 33.9 9.0 87.7 17.1
13 49.8 12.7 98.9 18.2
14 33.1 7.7 122.9 21.1
15 49.9 11.0 94.1 15.9
9 42.4 18.0 0.487 61.8 51.4 38.9 to 162.4
10 40.6 15.5
11 40.9 14.7
12 58.4 18.4
13 50.4 14.5
14 81.3 22.6
15 68.5 17.8
W: watts; SEM: standard error of measurement; CV: SEM expressed as a coefficient of variation; ICC: intraclass correlation coefficient, model (1,k); LoA: limits of agreement.
Are we ready to measure running power? 7
Figure 1. Scatter plots of the correlations between the P
provided by each technology and VO
measured by the metabolic cart during the
incremental speed test, indoors (left panels) and outdoors (right panels). SEE: standard error of the estimate (%).
8V. Cerezuela-espejo et al.
In addition, it is worth noting that the present inves-
tigation used the same metabolic cart under the same
calibration, in both indoor and outdoor environ-
ments. This methodology ensures that VO
resulted from each condition modification (speed,
body weight and slope) and not to the inter-analyzers
variation or extra-weight generated by the wearable
metabolic analyzers.
Despite the fact that the current study expands the
knowledge about the use of these technologies for the
daily training, it presents some limitations which
should be addressed in the future. With the objective
of maintaining a RER < 1 and avoiding confounding
factors, the present study established a fixed speed
(10 km·h
) and cadence (corresponding to
13 km·h
). Although these methodological aspects
substantially improved the quality of the repeatability
(Austin et al., 2018) (objective 1) and concurrent val-
idity (Bijker et al., 2001;Cerezuela-Espejo et al.,
2018;Coyle et al., 1992;Pallarés et al., 2016)(objec-
tive 2) analyses, they could reduce the ecological val-
idity. Therefore, future investigations should confirm
the reliability of these devices with a self-selected
speed and cadence. Furthermore, the current
research used only three body weights (0, +2.5, and
+5 kg) and five slopes (6%, 3%, 1%, +3%, and
+6%) to examine the relationship between technol-
and the VO
. It would be a great practical
value that future studies extend the knowledge by
examining a wider range of weights and slopes.
The emergence of technologies which offer P
benefited exercise prescription and load monitoring
in running activities, due the advantages of this vari-
able in comparison with other internal and external
parameters, as well as its tight relationship with
. Nevertheless, the scarcity of scientific literature
analyzing P
estimated by these emerging technol-
ogies, together with their increasing commercializa-
tion, makes necessary the analysis of the suitability
of P
provided by these tools. Since of the main
goal of P
measurement is to determine the real
effort being incurred during an endurance training,
coaches should guarantee that the device used is
repeatable enough to ensure that changes in P
due to improvements/decrements in athletesper-
formance, instead of measurement errors of the tech-
nology. Therefore, based on the higher repeatability
and relationship with the VO
of the Stryd technol-
ogy, we encourage practitioners to use this device
for P
measurement in running activities.
The authors wish to thank the participants for their
invaluable contribution to the study.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Víctor Cerezuela-Espejo
Alejandro Hernández-Belmonte
Javier Courel-Ibáñez
Elena Conesa-Ros
Ricardo Mora-Rodríguez
Jesús G. Pallarés
References Q4
Aggarwal, R., & Ranganathan, P. (2016). Common pitfalls in stat-
istical analysis: The use of correlation techniques. Perspectives in
Clinical Research,7, 187190.
Allen, H., & Coggan, A. (2010). Training and racing with a power
meter. VeloPress. Q5
Armstrong, R. A. (2019). Should Pearsons correlation coefficient
be avoided? Ophthalmic and Physiological Optics,39, 316327.
Arts, F., & Kuipers, H. (1994). The relation between power
output, oxygen uptake and heart rate in male athletes.
International Journal of Sports Medicine,15, 228231.
Aubry, R. L., Power, G. A., & Burr, J. F. (2018). An assessment of
running power as a training metric for elite and recreational
runners. Journal of Strength and Conditioning Research,32,
Austin, C., Hokanson, J., McGinnis, P., & Patrick, S. (2018). The
relationship between running power and running economy in
well-trained distance runners. Sports,6, 142.
Bartlett, J. W., & Frost, C. (2008). Reliability, repeatability and
reproducibility: Analysis of measurement errors in continuous
variables. Ultrasound in Obstetrics and Gynecology,31, 466475.
Bijker, K. E., De Groot, G., & Hollander, A. P. (2001). Delta effi-
ciencies of running and cycling. Medicine & Science in Sports &
Exercise,33, 15461551.
Bland, J. M., & Altman, D. G. (1986). Statistical methods for
assessing agreement between two methods of clinical measure-
ment. The Lancet,327, 307310.
Bourdon, P. C., Cardinale, M., Murray, A., Gastin, P., Kellmann,
M., Varley, M. C., Cable, N. T. (2017). Monitoring athlete
training loads: Consensus statement. International Journal of
Sports Physiology and Performance,12(Suppl 2), S2-161S2-170.
Cerezuela-Espejo, V., Courel-Ibáñez, J., Morán-Navarro, R.,
Martínez-Cava, A., & Pallarés, J. G. (2018). The relationship
between lactate and ventilatory thresholds in runners: Validity
and reliability of exercise test performance parameters.
Frontiers in Physiology,9, 1320.
Coyle, E. F., Sidossis, L. S., Horowitz, J. F., & Beltz, J. D. (1992).
Cycling efficiency is related to the percentage of type I muscle
fibers. Medicine & Science in Sports & Exercise,24, 782788.
Cross, M. R., Brughelli, M., Samozino, P., & Morin, J.-B. (2017).
Methods of power-force-velocity profiling during sprint
running: A narrative review. Sports Medicine,47, 12551269.
Are we ready to measure running power? 9
García-Pinillos, F., Latorre-Román, PÁ, Roche-Seruendo, L. E.,
& García-Ramos, A. (2019). Prediction of power output at
different running velocities through the two-point method
with the Stryd
power meter. Gait & Posture,68, 238243.
Haakonssen, E. C., Martin, D. T., Burke, L. M., & Jenkins, D. G.
(2013). Energy expenditure of constant- and variable-intensity
cycling: Power meter estimates. Medicine & Science in Sports
& Exercise,45, 18331840.
Han, H., Kim, H., Sun, W., Malaska, M., & Miller, B. (2020).
Validation of wearable activity monitors for real-time cadence.
Journal of Sports Sciences,38, 383389.
Hopkins, W. G. (2000). Measures of reliability in sports medicine
and science. Sports Medicine,30,115.
Jones, A. M., & Doust, J. H. (1996). A 1% treadmill grade most
accurately reflects the energetic cost of outdoor running.
Journal of Sports Sciences,14, 321327.
Koo, T. K., & Li, M. Y. (2016). A guideline of selecting and
reporting intraclass correlation coefficients for reliability
research. Journal of Chiropractic Medicine,15, 155163.
Martins, W. P., & Nastri, C. O. (2014). Interpreting reproducibil-
ity results for ultrasound measurements. Ultrasound in Obstetrics
& Gynecology,43, 479480.
Mujika, I. (2017). Quantification of training and competition loads in
endurance sports: Methods and applications. International Journal
of Sports Physiology and Performance,12(Suppl 2), S2-9S2-17.
Muniz-Pumares, D., Pedlar, C., Godfrey, R., & Glaister, M.
(2017). The effect of the oxygen uptake-power output relation-
ship on the prediction of supramaximal oxygen demands.
Journal of Sports Medicine and Physical Fitness,57,17.
Pallarés, J. G., López-Samanes, Á, Fernández-Elías, V. E.,
Aguado-Jiménez, R., Ortega, J. F., Gómez, C., Segura, J.
(2015). Pseudoephedrine and circadian rhythm interaction on
neuromuscular performance. Scandinavian Journal of Medicine
& Science in Sports,25, e603e612.
Pallarés, J. G., Morán-Navarro, R., Ortega, J. F., Fernández-Elías,
V. E., & Mora-Rodriguez, R. (2016). Validity and reliability of
ventilatory and blood lactate thresholds in well-trained cyclists.
PLoS One,11, e0163389.
Passfield, L., Hopker, J. G., Jobson, S., Friel, D., & Zabala, M.
(2017). Knowledge is power: Issues of measuring training and
performance in cycling. Journal of Sports Sciences,35, 1426
Pedersen, P. K., Sørensen, J. B., Jensen, K., Johansen, L., & Levin,
K. (2002). Muscle fiber type distribution and nonlinear VO(2)-
power output relationship in cycling. Medicine & Science in
Sports & Exercise,34, 655661.
Sanders, D., Myers, T., & Akubat, I. (2017). Training-intensity
distribution in road cyclists: Objective versus subjective
measures. International Journal of Sports Physiology and
Performance,12, 12321237.
Snyder, K. L., Mohrman, W. P., Williamson, J. A., & Li, K.
(2018). Methodological flaws in Aubry, RL, Power, GA, and
Burr, JF. An assessment of running power as a training metric
for elite and recreational runners. Journal of Strength and
Conditioning Research,32, e61.
Van Dijk, H., & Van Megen, R. (2017). The secret of running:
Maximum performance gains through effective power metering and
training analysis. Meyer. Q6
Vincent, W. J. (2004). Statistics in kinesiology. Human Kinetics. Q7
Whipp, B. J., & Ward, S. A. (1990). Physiological determinants of
pulmonary gas exchange kinetics during exercise. Medicine &
Science in Sports & Exercise,22,6271.
10 V. Cerezuela-espejo et al.
... Particularly, the running power meter (Stryd Summit Power Meter) is getting much attention from the running community due to its capacity to reflect external work and its repeatability across different running conditions compared with different commercially available devices. [7][8][9] To determine CP, the Stryd TM group proposed a test based on 2 maximum efforts of 9 and 3 minutes, respectively, with a 30minute active recovery break between the 2 efforts. 10 Then, based on a CP percentage, 5 training zones were established (ie, zone 1 easy: 65%-80%; zone 2 moderate: 80%-90%; zone 3 threshold: 90%-100%; zone 4 interval: 100%-115%; zone 5 repetition: 115%-300%). ...
... 21 Since its market launch, knowing which type of power Stryd reports has been of great interest among the running community. Cerezuela-Espejo et al 7,22 determined the relationship between the power output reported from 5 commercial power meters and 2 theoretical power models varying in speed, weight, and slope. The Stryd power meter showed the greatest sensitivity to these factors among the other meters (r ≥ .947), ...
... respectively). 7,23 Relative to the CP location with respect VTs, few studies have determined a proximity to RCP in cycling. 24 However, due to the sensitivity of CP estimation and RCP location to the protocols used, there is a need to highlight the dependent essence of these results according to the predictive trials and GXT test used. ...
Purpose: The critical power (CP) concept has been extended from cycling to the running field with the development of wearable monitoring tools. Particularly, the Stryd running power meter and its 9/3-minute CP test is very popular in the running community. Locating this mechanical threshold according to the physiological landmarks would help to define each boundary and intensity domain in the running field. Thus, this study aimed to determine the CP location concerning anaerobic threshold, respiratory compensation point (RCP), and maximum oxygen uptake (VO2max). Method: A group of 15 high-caliber athletes performed the 9/3-minute Stryd CP test and a graded exercise test in 2 different testing sessions. Results: Anaerobic threshold, RCP, and CP were located at 73% (5.41%), 86.82% (3.85%), and 88.71% (5.84%) of VO2max, respectively, with a VO2max of 66.3 (7.20) mL/kg/min. No significant differences were obtained between CP and RCP in any of its units (ie, in watts per kilogram and milliliters per kilogram per minute; P ≥ .184). Conclusions: CP and RCP represent the same boundary in high-caliber athletes. These results suggest that coaches and athletes can determine the metabolic perturbance threshold that CP and RCP represent in an easy and accessible way.
... Knowledge of the reliability and validity of these IMU devices is of paramount importance to collect and interpret data accurately. Some researchers have analyzed the Stryd's reliability and validity during running [2,5,[8][9][10][11]. However, the reliability and validity have been less investigated during walking [12,13], and never during walking on positive slopes using different backpack loads. ...
... In this sense, Stryd power is the estimation of the necessary work production to redirect the three-dimensional trajectory of the center of mass during walking or running. The scarce availability of gold standard devices to measure external mechanical power [8,11], together with the fact that the algorithms used to derive power from the Stryd sensor are proprietary [11], hinders appropriate validation of external mechanical power of the Stryd. Despite this limitation, results of this study could be of particular interest to military, fire, and rescue personnel, as well as their coaches and/or superiors to implement strategies to improve rescuers' performance. ...
... The main limitation might be that validation of power output using the gold standard force platform instrumentation was not performed. To date, only two studies attempted to validate power estimates of the Stryd during running [8,11]. In one study, power estimates of the Stryd were compared against respiratory measures based on the linear relationship between oxygen uptake and power output [8]. ...
Background: The Styrd Power Meter is gaining special interest for on-field gait analyses due to its low-cost and general availability. However, the reliability and validity of the Stryd during walking on positive slopes using different backpack loads have never been investigated. Research Question: Is the Stryd Power Meter reliable and valid to quantify gait mechanics during walking on positive inclines and during level walking incorporating load carriage? Methods: Seventeen participants from a police force rescue team performed 8 submaximal walking trials for 5-min at 3.6 km•h-1 during different positive slope (1, 10 and 20%) and backpack load (0, 10, 20, 30 and 40% of body mass) conditions. Two Stryd devices were utilized for reliability analyses. Validity of cadence and ground contact time (GCT) were analyzed against a gold standard device (Optojump). Results: The Stryd demonstrated acceptable reliability [mean bias: <2.5%; effect size (ES): <0.25; standard error of the mean: <1.7%; r: >0.76] for power, cadence, and GCT. Validity measures (mean bias: <0.8%; ES: <0.07; r: >0.96; Lin’s Concordance Coefficient: 0.96; Mean Absolute Percent Error: <1%) for cadence were also found to be acceptable. The Stryd overestimated (P < 0.001; ES: >5.1) GCT in all the walking conditions. A significant systematic positive bias (P < 0.022; r = 0.56 to 0.76) was found in 7 conditions. Significance: The Stryd Power Meter appears to produce reliable measurements for power output, cadence and GCT. The Stryd produced valid measurements for cadence during walking on positive slopes and during level walking with a loaded backpack. However, the Stryd is not valid for measuring GCT during these walking conditions. This study adds novel data regarding the reliability and validity of this device and might be of particular interest for scientists, practitioners, and first responders seeking reliable devices to quantify gait mechanics during walking.
... In recent years, this approach was extended by the introduction of running power. Running power is a measure that was introduced to analyze the metabolic workload during running in an instantaneous, reliable, and sensitive manner [116]. Different commercial enterprises have introduced ways to measure running power in real-time using either smartwatches (e.g. ...
... Nevertheless, both options require additional sensors like barometers, air resistance, or temperature sensors to increase the accuracy of the metric. In scientific studies, the IMU-based method has proven to determine running power reliably [94,95] and more accurately then the GPS-based approach [116]. ...
Full-text available
Body-worn sensors, so-called wearables, are getting more and more popular in the sports domain. Wearables offer real-time feedback to athletes on technique and performance, while researchers can generate insights into the biomechanics and sports physiology of the athletes in real-world sports environments outside of laboratories. One of the first sports disciplines, where many athletes have been using wearable devices, is endurance running. With the rising popularity of smartphones, smartwatches and inertial measurement units (IMUs), many runners started to track their performance and keep a digital training diary. Due to the high number of runners worldwide, which transferred their data of wearables to online fitness platforms, large databases were created, which enable Big Data analysis of running data. This kind of analysis offers the potential to conduct longitudinal sports science studies on a larger number of participants than ever before. In this dissertation, both studies showing how to extract endurance running-related parameters from raw data of foot-mounted IMUs as well as a Big Data study with running data from a fitness platform are presented.
... (1) Running economy = metabolic power (mlO2/kg) Velocity (km/h) Although the specific calculation algorithms still undisclosed by the companies and validity of the data obtained with these wearable devices has not been yet compared with the 'Gold Standard' (i.e., a forceplate-instrumented treadmill or a long force platform system), a recent study [15] assessed their concurrent validity with metabolic demands (i.e., VO2) showing consistent results for some of these novel devices (r > 0.9). ...
... Mean power output (normalized by body mass), step frequency, form power and running effectiveness were calculated using the Stryd ™ power meter (Stryd Power meter, Stryd Inc. Boulder CO, USA) attached on the upper part of the running shoes. This sensor provides accurate kinematic [31,32] and consistent power output metrics [15]. Data from Stryd ™ power meter were obtained into the fit file via the manufacturer's website (https:// www. ...
Full-text available
Background The advent of power meters for running has raised the interest of athletes and coaches in new ways of assessing changes in running performance. The aim of this study is to determine the changes in power-related variables during and after a strenuous endurance running time trial. Methods Twenty-one healthy male endurance runners, with a personal record of 37.2 ± 1.2 min in a 10-km race, completed a 1-h run on a motorized treadmill trying to cover as much distance as they could. Before and after the time trial the athletes were asked to perform a 3-min run at 12 km h ⁻¹ . Normalized mean power output, step frequency, form power and running effectiveness were calculated using the Stryd™ power meter. Heart rate (HR) and rating of perceived exertion (RPE) were monitored, and data averaged every 5 min. Results Despite high levels of exhaustion were reached during the time trial (HRpeak = 176.5 ± 9.8 bpm; RPE = 19.2 ± 0.8), the repeated measures ANOVA resulted in no significant differences ( p ≥ 0.05), between each pair of periods for any of the power-related variables. The pairwise comparison ( T test) between the non-fatigued and fatigued constant 3-min runs showed an increase in step frequency ( p = 0.012) and a decrease in form power ( p < 0.001) under fatigue conditions, with no meaningful changes in normalized mean power output and running effectiveness. Conclusions Trained athletes are able to maintain power output and running effectiveness during a high demanding extended run. However, they preferred to reduce the intensity of vertical impacts under fatigue conditions by increasing their step frequency.
... A recent study reported poor to moderate reliability of RunScribe for evaluating power during indoor and outdoor running (ICC = 0.5-0.7; Cerezuela-Espejo et al., 2021). However, a subsequent study reported good absolute (coefficient of variation (CV) = 1.68 ± 1.49 %) and relative (ICC = 0.86) reliability for power assessment (Cartón-Llorente et al., 2021). ...
... %). Therefore, other technologies should be used to assess power, with Stryd TM looking most promising for now (Cerezuela-Espejo et al., 2021). Interestingly, another recent study compared power obtained from Stryd TM and RunScribe, and found good agreement between the devices (ICC = 0.85; Cartón-Llorente et al., 2021). ...
The aim of this study was to investigate the reliability of running biomechanics assessment with a wearable commercial sensor (RunScribeTM). Participants performed multiple 200-m runs over sand, grass and asphalt ground at the estimated 5-km tempo, with an additional trial with 21-km tempo at the asphalt. Intra-session reliability was excellent for all variables at 5-km pace (intra-class coefficient correlation (ICC) asphalt: 0.90–0.99; macadam: 0.94–1.00; grass: 0.92–1.00), except for shock (good; ICC = 0.83), and contact time and total power output (moderate; ICC = 0.68–0.71). Coefficient of variation (CV) were mostly acceptable in all conditions, except for horizontal ground reaction force (GRF) rate in asphalt 5-km pace trial (CV = 24.5 %), power (CV = 14.3 %) and foot strike type (CV = 30.9 %) in 21-km pace trial, and horizontal GRF rate grass trial (CV = 15.7 %). Inter-session reliability was high or excellent for the majority of the outcomes (ICC≥0.85). Total power output (ICC = 0.56–0.65) and shock (ICC = 0.67–0.75) showed only moderate reliability across all conditions. Power (CV = 12.5–13.8 %), foot strike type (CV = 14.9–29.4 %) and horizontal ground reaction force rate (CV = 12.4–36.4 %) showed unacceptable CV.
... Stryd has been evaluated during treadmill running [2][3][4][5], track running [3] trail running [6] and walking [6, 7]. Stryd provides reliable measures for power [2-4, 7], though a minimum sampling time of 10 seconds is required at a constant speed [8]. ...
Stryd is a foot pod that reliably estimates running power. Our objectives were to examine the efficacy of the website-generated Stryd critical power (CPSTRYD) as a meaningful parameter for runners. 20 runners performed their regular training while wearing Stryd for a minimum of 6 weeks to generate CPSTRYD. Runners completed laboratory graded exercise testing, and outdoor 1500 m and 5000 m time trails. CPSTRYD was most similar to the second ventilatory threshold (VT2) or the onset of blood lactate accumulation (OBLA) and is highly predictive of running performance. Stryd ground contact time (GCT) was a predictor of performance when comparing runners at the same submaximal treadmill speed. CPSTRYD generated from outdoor running is equivalent to that calculated using an established CP model. However, variance between different methods of CP estimation must be a consideration for runners and coaches. Stryd offers meaningful data for runners including a realistic estimate of CP.
... While several studies have already analysed power output in running [18,19] and others have investigated the relation between VO2max and power production [16,20], to the best of the authors' knowledge, there are no studies assessing the difference in power output between shod and barefoot running. In order to bridge this gap, this study aims to identify the effect of footwear on power output in endurance runners. ...
Full-text available
Several studies have already analysed power output in running or the relation between VO2max and power production as factors related to running economy; however, there are no studies assessing the difference in power output between shod and barefoot running. This study aims to identify the effect of footwear on the power output endurance runner. Forty-one endurance runners (16 female) were evaluated at shod and barefoot running over a one-session running protocol at their preferred comfortable velocity (11.71 ± 1.07 km·h−1). The mean power output (MPO) and normalized MPO (MPOnorm), form power, vertical oscillation, leg stiffness, running effectiveness and spatiotemporal parameters were obtained using the Stryd™ foot pod system. Additionally, footstrike patterns were measured using high-speed video at 240 Hz. No differences were noted in MPO (p = 0.582) and MPOnorm (p = 0.568), whereas significant differences were found in form power, in both absolute (p = 0.001) and relative values (p < 0.001), running effectiveness (p = 0.006), stiffness (p = 0.002) and vertical oscillation (p < 0.001). By running barefoot, lower values for contact time (p < 0.001) and step length (p = 0.003) were obtained with greater step frequency (p < 0.001), compared to shod running. The prevalence of footstrike pattern significantly differs between conditions, with 19.5% of runners showing a rearfoot strike, whereas no runners showed a rearfoot strike during barefoot running. Running barefoot showed greater running effectiveness in comparison with shod running, and was consistent with lower values in form power and lower vertical oscillation. From a practical perspective, the long-term effect of barefoot running drills might lead to increased running efficiency and leg stiffness in endurance runners, affecting running economy.
... The Stryd sensor is used according to the manufacturer's instructions and clipped to the subjects' shoe laces to measure the athlete's running PO during all runs, mean power output (MPO) during 5 km time trial and peak power output (PPO) during ramp test. A recent study found that the Stryd device is the most repeatable technology, among five analyzed systems, for running PO estimation [41]. Data are collected in the Garmin Connect application, which will be reviewed daily during the intervention by researchers to ensure program adherence. ...
Full-text available
Background Performing multiple high-intensity interval training (HIIT) sessions in a compressed period of time (approximately 7–14 days) is called a HIIT shock microcycle (SM) and promises a rapid increase in endurance performance. However, the efficacy of HIIT-SM, as well as knowledge about optimal training volumes during a SM in the endurance-trained population have not been adequately investigated. This study aims to examine the effects of two different types of HIIT-SM (with or without additional low-intensity training (LIT)) compared to a control group (CG) on key endurance performance variables. Moreover, participants are closely monitored for stress, fatigue, recovery, and sleep before, during and after the intervention using innovative biomarkers, questionnaires, and wearable devices. Methods This is a study protocol of a randomized controlled trial that includes the results of a pilot participant. Thirty-six endurance trained athletes will be recruited and randomly assigned to either a HIIT-SM (HSM) group, HIIT-SM with additional LIT (HSM + LIT) group or a CG. All participants will be monitored before (9 days), during (7 days), and after (14 days) a 7-day intervention, for a total of 30 days. Participants in both intervention groups will complete 10 HIIT sessions over 7 consecutive days, with an additional 30 min of LIT in the HSM + LIT group. HIIT sessions consist of aerobic HIIT, i.e., 5 × 4 min at 90–95% of maximal heart rate interspersed by recovery periods of 2.5 min. To determine the effects of the intervention, physiological exercise testing, and a 5 km time trial will be conducted before and after the intervention. Results The feasibility study indicates good adherence and performance improvement of the pilot participant. Load monitoring tools, i.e., biomarkers and questionnaires showed increased values during the intervention period, indicating sensitive variables. Conclusion This study will be the first to examine the effects of different total training volumes of HIIT-SM, especially the combination of LIT and HIIT in the HSM + LIT group. In addition, different assessments to monitor the athletes' load during such an exhaustive training period will allow the identification of load monitoring tools such as innovative biomarkers, questionnaires, and wearable technology. Trial Registration :, NCT05067426. Registered 05 October 2021—Retrospectively registered, . Protocol Version Issue date: 1 Dec 2021. Original protocol. Authors: TLS, NH.
... External loads were monitored using a Stryd Power Meter (Stryd, Inc., Boulder, CO, USA) placed on the laces of the right soccer boot with a plastic clip [29]. Stryd Power Meter was used to carry out this study because its repeatability has been demonstrated (SEM ≤ 12.5 W, CV ≥ 4.3 %, ICC ≤ 0.989) [30]. Offline data collection was activated in accordance with the manufacturer's recommendations. ...
Full-text available
The aim of this paper was to examine the differences in the external and internal load in amateur match officials between the 1st and 2nd half and among different 15 min periods. Twenty-three field referees (FRs) and 46 assistant referees (ARs) from the Spanish División de Honor participated in this study. Match external and internal loads were monitored showing that FRs recorded a lower Powermean, Speedmean, Cadencemean and Stiffnessmean (p < 0.05; d = 0.52 to 0.57) during the 2nd half and they also recorded a lower HRmean, and HRpeak, and spent less time in zone 5 (p < 0.05; d = 0.50 to 0.62). The FRs’ match load decreased during the match but they performed higher Powermean and covered more distance in the last 15 min of the match (p < 0.01; d = 0.87 to 4.28). The ARs external load did not show significant variations between halves, but ARs recorded a lower HRmean and spent less time in zone 5 (p < 0.01; d = 0.41 to 0.63), and the highest values of Powermean, Speedmean, Cadencemean and Vertical oscillationmean during the first 15 min of the match (p < 0.05; d = 0.45 to 0.75). The highest values of HRmean and distance covered were in the 0–15 min period. Results suggest that match load decreases as the match progresses because of the neuromuscular fatigue but increases in the last 15 min.
Runners are known to have a lower risk of death from cardiovascular disease and overall mortality compared to their non-training peers. This review analyzes the literature data on the effect of running on some indicators of a person’s biological age. The influence of running on markers of aging, such as the telomere length and the redox and inflammatory statuses of the body, is considered. The positive effects of running on mental health and cognitive performance are discussed. The problem of determining the optimal intensity of physical activity for a unique beneficial effect on health and longevity is analyzed.
Full-text available
The purpose of this study is to establish evidence of validity for wearable activity monitors providing real-time cadence against a criterion measure. Thirty-six healthy adults, aged 18–65 years, participated in the study. Four activity monitors including 2 watch-based monitors and 2 cadence sensors attaching to shoelaces were tested. Each participant completed the study protocol consisting of 2 distinct components: (1) treadmill protocol and (2) overground protocol. Lin’s concordance correlation and mean absolute percentage error (MAPE) were calculated for the comparisons between the criterion and measures of the monitors. Bland–Altman analysis was performed to determine the mean bias and 95% limits of agreement. All activity monitors showed high correlations with the criterion measures (p < .01). Lower correlations were observed at slow walking speeds in the watch-based monitors. In contrast, consistent and strong correlations were found with both cadence sensors regardless of walking speeds (p < .01). Similar patterns were observed in the MAPE scores. Greater than 90% of the participants were able to maintain prescribed walking intensity using real-time cadence. The results suggest that the wearable activity monitors are an acceptable measure of real-time cadence and provide the potential to improve intensity-based prescription of physical activity using the monitors.
Full-text available
Background: The force- and power-velocity (F-V and P-V, respectively) relationships have been extensively studied in recent years. However, its use and application in endurance running events is limited. Research question: This study aimed to determine if the P-V relationship in endurance runners fits a linear model when running at submaximal velocities, as well as to examine the feasibility of the "two-point method" for estimating power values at different running velocities. Methods: Eighteen endurance runners performed, on a motorized treadmill, an incremental running protocol to exhaustion. Power output was obtained at each stage with the Stryd™ power meter. The P-V relationship was determined from a multiple-point method (10, 12, 14, and 17 km·h-1) as well as from three two-point methods based on proximal (10 and 12 km·h-1), intermediate (10 and 14 km·h-1) and distal (10 and 17 km·h-1) velocities. Results: The P-V relationship was highly linear ( r = 0.999). The ANOVAs revealed significant, although generally trivial (effect size < 0.20), differences between measured and estimated power values at all the velocities tested. Very high correlations ( r = 0.92) were observed between measured and estimated power values from the 4 methods, while only the multiple-point method ( r2 = 0.091) and two-point method distal ( r2 = 0.092) did not show heteroscedasticity of the error. Significance: The two-point method based on distant velocities (i.e., 10 and 17 km·h-1) is able to provide power output with the same accuracy than the multiple-point method. Therefore, since the two-point method is quicker and less prone to fatigue, we recommend the assessment of power output under only two distant velocities to obtain an accurate estimation of power under a wide range of submaximal running velocities.
Full-text available
A novel running wearable called the Stryd Summit footpod fastens to a runner’s shoe and estimates running power. The footpod separates power output into two components, Stryd power and form power. The purpose of this study was to measure the correlations between running economy and power and form power at lactate threshold pace. Seventeen well-trained distance runners, 9 male and 8 female, completed a running protocol. Participants ran two four-minute trials: one with a self-selected cadence, and one with a target cadence lowered by 10%. The mean running economy expressed in terms of oxygen cost at self-selected cadence was 201.6 ± 12.8 mL·kg−1·km−1, and at lowered cadence was 204.5 ± 11.5 mL·kg−1·km−1. Ventilation rate and rating of perceived exertion (RPE) were not significantly different between cadence conditions with one-tailed paired t-test analysis (ventilation, p = 0.77, RPE, p = 0.07). Respiratory exchange ratio and caloric unit cost were significantly greater with lower cadence condition (respiratory exchange ratio, p = 0.03, caloric unit cost, p = 0.03). Mean power at self-selected cadence was 4.4 ± 0.5 W·kg−1, and at lowered cadence was 4.4 ± 0.5 W·kg−1. Mean form power at self-selected cadence was 1.1 ± 0.1 W·kg−1, and at lowered cadence was 1.1 ± 0.1 W·kg−1. There were positive, linear correlations between running economy and power (self-selected cadence and lower cadence, r = 0.6; the 90% confidence interval was 0.2 to 0.8); running economy and form power (self-selected cadence and lower cadence r = 0.5; the 90% confidence interval was 0.1 to 0.8). The findings suggest running economy is positively correlated with Stryd’s power and form power measures yet the footpod may not be sufficiently accurate to estimate differences in the running economy of competitive runners.
Full-text available
The aims of this study were (1) to establish the best fit between ventilatory and lactate exercise performance parameters in running and (2) to explore novel alternatives to estimate the maximal aerobic speed (MAS) in well-trained runners. Twenty-two trained male athletes ( V ˙ O2max 60.2 ± 4.3 ml·kg·min-1) completed three maximal graded exercise tests (GXT): (1) a preliminary GXT to determine individuals' MAS; (2) two experimental GXT individually adjusted by MAS to record the speed associated to the main aerobic-anaerobic transition events measured by indirect calorimetry and capillary blood lactate (CBL). Athletes also performed several 30 min constant running tests to determine the maximal lactate steady state (MLSS). Reliability analysis revealed low CV (<3.1%), low bias (<0.5 km·h-1), and high correlation (ICC > 0.91) for all determinations except V-Slope (ICC = 0.84). Validity analysis showed that LT, LT+1.0, and LT+3.0 mMol·L-1 were solid predictors of VT1 (-0.3 km·h-1; bias = 1.2; ICC = 0.90; p = 0.57), MLSS (-0.2 km·h-1; bias = 1.2; ICC = 0.84; p = 0.74), and VT2 (<0.1 km·h-1; bias = 1.3; ICC = 0.82; p = 0.9l9), respectively. MLSS was identified as a different physiological event and a midpoint between VT1 (bias = -2.0 km·h-1) and VT2 (bias = 2.3 km·h-1). MAS was accurately estimated (SEM ± 0.3 km·h-1) from peak velocity (Vpeak) attained during GXT with the equation: MASEST (km·h-1) = Vpeak (km·h-1) * 0.8348 + 2.308. Current individualized GXT protocol based on individuals' MAS was solid to determine both maximal and submaximal physiological parameters. Lactate threshold tests can be a valid and reliable alternative to VT and MLSS to identify the workloads at the transition from aerobic to anaerobic metabolism in well-trained runners. In contrast with traditional assumption, the MLSS constituted a midpoint physiological event between VT1 and VT2 in runners. The Vpeak stands out as a powerful predictor of MAS.
Full-text available
Aubry, RL, Power, GA, and Burr, JF. An assessment of running power as a training metric for elite and recreational runners. J Strength Cond Res XX(X): 000-000, 2018-Power, as a testing and training metric to quantify effort, is well accepted in cycling, but is not commonly used in running to quantify effort or performance. This study sought to investigate a novel training tool, the Stryd Running Power Meter, and the applicability of running power (and its individually calculated run mechanics) to be a useful surrogate of metabolic demand (V[Combining Dot Above]O2), across different running surfaces, within different caliber runners. Recreational (n = 13) and elite (n = 11) runners completed a test assessing V[Combining Dot Above]O2 at 3 different paces, while wearing a Stryd Power Meter on both an indoor treadmill and an outdoor track, to investigate relationships between estimated running power and metabolic demand. A weak but significant relationship was found between running power and V[Combining Dot Above]O2 considering all participants as a homogenous group (r = 0.29); however, when assessing each population individually, no significant relationship was found. Examination of the individual mechanical components of power revealed that a correlative decrease in V[Combining Dot Above]O2 representing improved efficiency was associated with decreased ground contact time (r = 0.56), vertical oscillation (r = 0.46), and cadence (r = 0.37) on the treadmill in the recreational group only. Although metabolic demand differed significantly between surfaces at most speeds, run power did not accurately reflect differences in metabolic cost between the 2 surfaces. Running power, calculated via the Stryd Power Meter, is not sufficiently accurate as a surrogate of metabolic demand, particularly in the elite population. However, in a recreational population, this training tool could be useful for feedback on several running dynamics known to influence running economy.
Full-text available
Monitoring the load placed on athletes in both training and competition has become a very hot topic in sport science. Both scientists and coaches routinely monitor training loads using multidisciplinary approaches, and the pursuit of the best method- ologies to capture and interpret data has produced an exponential increase in empirical and applied research. Indeed, the eld has developed with such speed in recent years that it has given rise to industries aimed at developing new and novel paradigms to allow us to precisely quantify the internal and external loads placed on athletes and to help protect them from injury and ill health. In February 2016, a conference on “Monitoring Athlete Training Loads—The Hows and the Whys” was convened in Doha, Qatar, which brought together experts from around the world to share their applied research and contemporary prac- tices in this rapidly growing eld and also to investigate where it may branch to in the future. This consensus statement brings together the key ndings and recommendations from this conference in a shared conceptual framework for use by coaches, sport-science and -medicine staff, and other related professionals who have an interest in monitoring athlete training loads and serves to provide an outline on what athlete-load monitoring is and how it is being applied in research and practice, why load monitoring is important and what the underlying rationale and prospective goals of monitoring are, and where athlete-load monitoring is heading in the future.
Full-text available
Purpose: This study aims to evaluate training intensity distribution using different intensity measures based on session rating of perceived exertion (sRPE), heart rate (HR) and power output (PO) in well-trained cyclists. Methods: Fifteen road cyclists participated in the study. Training data was collected during a 10-week training period. Training intensity distribution was quantified using HR, PO and sRPE categorized in a 3-zone training intensity model. Three zones for HR and PO were based around a first and second lactate threshold. The three sRPE zones were defined using a 10-point scale: zone 1, sRPE scores 1-4; zone 2, sRPE scores 5-6; zone 3, sRPE scores 7-10. Results: Training intensity distribution as percentages of time spent in zone 1, zone 2 and zone 3 was moderate to very largely different for sRPE (44.9%, 29.9%, 25.2%) compared to HR (86.8%, 8.8%, 4.4%) and PO (79.5%, 9.0%, 11.5%). Time in zone 1 quantified using sRPE was large to very largely lower for sRPE compared to PO (P < 0.001) and HR (P < 0.001). Time in zone 2 and zone 3 was moderate to very largely higher when quantified using sRPE compared to intensity quantified using HR (P < 0.001) and PO (P < 0.001). Conclusions: Training intensity distribution quantified using sRPE demonstrates moderate to very large differences compared to intensity distributions quantified based on HR and PO. The choice of intensity measure impacts on the intensity distribution and has implications for training load quantification, training prescription and the evaluation of training characteristics.
Purpose: To survey the use of Pearson's correlation coefficient (r) and related statistical methods in the ophthalmic literature, to consider the limitations of r, and to suggest suitable alternative methods of analysis. Recent findings: Searching Ophthalmic and Physiological Optics (OPO), Optometry and Vision Science (OVS), and Clinical and Experimental Optometry (CXO) online archives using correlation and Pearson's r as search terms resulted in 4057 and 281 hits respectively. Coefficient of determination, r square, or r squared received fewer hits (65, 8, and 22 hits respectively). The assumption that r follows a bivariate normal distribution was rarely encountered (3 hits) although several studies applied Spearman's rank correlation (70 hits). The intra-class correlation coefficient (ICC) was widely used (178 hits), but fewer hits were recorded for partial correlation (43 hits) and multiple correlation (13) hits. There was little evidence that the problem of sample size was addressed in correlation studies. Summary: Investigators should be alert to whether: (1) the relationship between two variables could be non-linear, (2) the data are bivariate normal, (3) r accounts for a significant proportion of the variance in Y, (4) outliers are present, the data are clustered, or have a restricted range, (5) the sample size is appropriate, and (6) a significant correlation indicates causality. In addition, the number of significant digits used to express r and the problems of multiple testing should be addressed. The problems and limitations of r suggest a more cautious approach regarding its use and the application of alternative methods where appropriate.
Training quantification is basic to evaluate an endurance athlete's responses to the training loads, ensure adequate stress/recovery balance and determine the relationship between training and performance. Quantifying both external and internal workload is important, because the external workload does not measure the biological stress imposed by the exercise sessions. Generally used quantification methods include retrospective questionnaires, diaries, direct observation and physiological monitoring, often based on the measurement of oxygen uptake, heart rate and blood lactate concentration. Other methods in use in endurance sports include speed measurement and the measurement of power output, made possible by recent technological advances, such as power meters in cycling and triathlon. Among subjective methods of quantification the RPE stands out because of its wide use. Concurrent assessments of the various quantification methods allow researchers and practitioners to evaluate stress/recovery balance, adjust individual training programmes and determine the relationships between external load, internal load and athletes' performance. This brief review summarizes the most relevant external and internal workload quantification methods in endurance sports, and provides practical examples of their implementation to adjust the training programmes of elite athletes in accordance to their individualized stress/recovery balance.