Content uploaded by Javier Courel Ibáñez
Author content
All content in this area was uploaded by Javier Courel Ibáñez on Jun 26, 2019
Content may be subject to copyright.
Technical Note on the Reliability of the PowerLift App for Velocity-Based
Resistance Training Purposes: Response
JAVIER COUREL-IBA
´N
˜EZ ,
1
ALEJANDRO MARTI
´NEZ-CAVA,
1
ALEJANDRO HERNA
´NDEZ-BELMONTE,
1
JUAN JOSE
´GONZA
´LEZ-BADILLO,
2
and JESU
´SG. PALLARE
´S
1
1
Human Performance and Sports Science Laboratory, Faculty of Sport Sciences, University of Murcia, C/Argentina S/n.,
Santiago de la Ribera, Murcia, Spain; and
2
Faculty of Sport, Pablo de Olavide University, Seville, Spain
This comment refers to the article available at https://doi.org/10.1007/s10439-019-02265-6.
(Received 31 May 2019; accepted 7 June 2019)
Associate Editor Jane Grande-Allen oversaw the review of this article.
Dear Editor,
Thank you for the opportunity to respond to the
points raised in this letter from Mr. Balsalobre
regarding our recent article.
3
We appreciate the inter-
est shown in our study. However, it seems that this
author has misunderstood some points about the reli-
ability analyses that we conducted and, therefore, this
needs to be clarified. Our study
3
aimed to analyze the
agreement between five bar velocity monitoring de-
vices, currently used in resistance training, to deter-
mine which was the most reliable device based on the
results of reproducibility (between-device agreement
for a given trial) and repeatability (between-trial vari-
ation for each device) analyses. Among the five dif-
ferent technologies assessed, only the data regarding
the PowerLift smartphone app have been questioned
by the author. It should be noted that this author is the
main developer of the PowerLift app and the main
author of the paper in which it was allegedly vali-
dated.
2
The arguments put forward by the author to ques-
tion our results are based on the assumption that ‘‘the
observers who performed the video analysis did not use
the app properly, because, following their data, the raw
difference in frames would be higher that 20–30
frames’’. As clearly detailed in our manuscript,
3
‘‘Intra-
device reproducibility was assessed by comparing the
velocity outcomes for trial 1 simultaneously obtained
by each pair of the two (same brand and model) de-
vices’’. Thus, the results we presented here were
obtained using two same version PowerLift apps (v 4.0
iOS) installed on two iPhone 6 units running iOS 11.3,
and analyzed by the same examiner. Notwithstanding
the above, in the letter, the author noted ‘‘Specifically,
in powerlifting exercises like the bench-press, observers
need to manually select the frames in which the bar
takes-off the chest (beginning of the lift) and stops its
vertical ascent (end of the lift). Because of that, pre-
vious research analyzing the validity and reliability of
VBS [video-based systems] compared to scores of two
observers analyzing the same videos in order to detect
the inter-observer variability’’. Thus, it seems that the
author has confused the concept of ‘‘intra-device
agreement’’ (i.e., comparing the outcomes simultane-
ously obtained by two devices for a given trial) with the
concept ‘‘inter-observers’ agreement’’ (i.e., comparing
the outcomes obtained by two observers for a given
video file in the same smartphone). It should be noted
that we were fully aware of the need of counting both
with experienced examiners and reproducible proce-
dures to ensure the quality of the measures, particu-
larly in manually operated systems such as the
PowerLift app. In our study,
3
the examiner showed an
excellent reliability when he analysed the same video
file, in the same smartphone, in two separate occasions
(ICC >0.997, 95% ICC 0.984–1.000; CV <3.02%),
which are values higher than those reported in previous
investigations where the smartphone was hold in the
hands.
1,2
We therefore agree that the PowerLift is very
easy to use and reliable in terms of intra-observers’
agreement. However, this was clearly not the hypoth-
esis tested in our investigation, which leads us to think
that the author has not read the study carefully. The
two main goals of our work
3
were: (1) to identify the
errors arising from current velocity monitoring tech-
Address correspondence to Jesu´ s G. Pallare
´s, Human Perfor-
mance and Sports Science Laboratory, Faculty of Sport Sciences,
University of Murcia, C/Argentina S/n., Santiago de la Ribera,
Murcia, Spain. Electronic mails: alejandro.mcava@gmail.com,
jgpallares@um.es
Annals of Biomedical Engineering (2019)
https://doi.org/10.1007/s10439-019-02305-1
BIOMEDICAL
ENGINEERING
SOCIETY
2019 Biomedical Engineering Society
nologies (intra-device reproducibility and repeatabil-
ity) to determine which was the most reliable (i.e.,
which, in turn, could be taken as the gold standard or
reference device), and (2) to objectively quantify the
agreement between measurements from each device
against that gold standard.
3
This approach is inspired
by the need of comprehensive information to guaran-
tee the suitability of emerging devices (based on new
technologies) in order to identify whether the changes
observed in velocity against certain workloads are due
to changes in athletes’ performance or are due to
measurement error.
5
We believe this is invaluable
information to both coaches and practitioners in order
to be able to determine the real effort being incurred
during resistance exercise.
We would like to take this opportunity to clarify
that what our results show is a very poor reliability of
the PowerLift measurements in terms of intra-device
(i.e., between two smartphones for a given trial or
execution) and inter-device (i.e., between the T-Force
System linear velocity transducer and one smartphone
running the PowerLift app for a given trial) agreement
when monitoring the full load-velocity relationship
(i.e., broad spectrum of mean bar velocity values, from
<0.10 to >1.50 m s
21
). In this regard, the author
provided a list of studies examining other physical
skills such as the running mechanics or vertical jump
by using specific apps for these purposes. However, we
are not questioning the use of these apps but the
PowerLift as a monitoring tool for velocity-based
resistance training purposes. In a more appropriate
manner, the author cited three studies which used the
PowerLift
1,2,8
; in two of them, the main author was
himself, the developer of the app, Mr. Balsalobre.
1,2
While none of these studies provided data about the
intra-device agreement, the three of them examined the
inter-device agreement of the PowerLift against a
supposedly gold standard. There are two main points
to clarify here. Firstly, these three studies considered as
their gold standard a different device (Smartcoach
encoder
1,2
and a 3D motion analysis system
8
) than that
taken by us (T-Force System) to assess the concurrent
validity of the PowerLift app. To the best of our
knowledge, there is no available information about the
inherent technical errors, standard error of the mea-
surement or minimal detectable changes, associated to
any of those devices for bar velocity measurement
during resistance exercise, unlike the evidence reported
by us for the T-Force System device.
3
Consequently,
comparisons of our results with those reported earlier
must be done with extreme caution. Secondly, these
three studies
1,2,8
only included data from resistance
exercises performed at slow
2
(<0.50 m s
21
) or mod-
erate (<1.00 m s
21
) velocities,
1,8
when compared to a
much broader load-velocity spectrum analyzed in our
study
3
(from 0.10 to >1.50 m s
21
). With the above-
FIGURE 1. Bland–Altman plots showing between-device agreement (T-Force vs. PowerLift) in mean velocity (MV) for trial 1 in the
bench press exercise. Magnitude of errors correspond to MV £0.5 m s
21
(green lines), 0.5 m s
21
<MV £1.0 m s
21
(yellow line)
and >1.0 m s
21
(red line). LoA limits of agreement, SDC smallest detectable change. Area shaded in grey indicates an
acceptable level of agreement between devices (Data pertaining to the study by Courel-Iba
´n
˜ez et al.
3
).
BIOMEDICAL
ENGINEERING
SOCIETY
COUREL-IBA
´N
˜EZ et al.
mentioned caution, if we divide the full load-velocity
relationship into three segments or parts of 0.50 m s
21
(Fig. 1, example for the bench press exercise; data from
Courel-Iba
´n
˜ez et al.
3
), it can be observed that our
findings partially agree with those earlier reported
suggesting the use of the PowerLift for measuring high
loads at slow velocities (<0.50 m s
21
), with a modest
margin of error (e.g., bench press: SDC = 0.15 m s
21
,
~9% 1RM, Fig. 1; full squat: SDC = 0.12 m s
21
,
~8% 1RM). The main problem here is that mean
velocities <0.50 m s
21
are associated with extremely
high loads for the most common resistance training
exercises (>80% in bench press,
5,9
90% in full squat
10
or even >100% in prone bench pull
9
), which limits its
practical application. More importantly, we found that
the reliability of the app drastically decreased as mean
lifting velocity increased, with the highest errors
occurring for lifts performed at veloci-
ties >1.00 m s
21
(SDC 0.34 m s
21
, 21% 1RM;
Fig. 1). We consider this fact a limitation for velocity-
based resistance-training purposes as the range of
velocities that can be monitored with sufficient sensi-
tivity is greatly reduced. More importantly, monitoring
high velocities (‡1.00 m s
21
) should be preferable to
assess changes in neuromuscular and functional per-
formance due to the higher specificity in relationship
with most sporting movements.
4,6,7
Finally, the author of the letter mentioned our lack
of knowledge to use the PowerLift app because we did
not evaluate the prone bench pull exercise. We
respectfully disagree, and we must clarify that the pre-
customized mode and the specific algorithm for the
prone bench pull exercise was not available when the
data were collected (PowerLift v. 4.0 iOS).
Taken into account all of the above, and according
to what we detailed in the original paper,
3
we sustain
that the PowerLift smartphone app should be not
recommended as a monitoring tool for velocity-based
resistance-training, given the substantial errors in-
curred and uncertainty of the outcomes obtained.
There is no doubt that the intentions to develop
smartphone apps as cheaper methods to measure bar
velocity are laudable. In this same line, we have iden-
tified quite good alternatives to linear velocity trans-
ducers for ~550 e. But whatever the price, unless a
measurement device provides enough accurate and
reliable measurements, its use should be discouraged.
If a device lacks enough sensitivity and the errors in-
curred when using it exceed a certain magnitude (see
the concepts of smallest detectable change and maxi-
mum error in our original study
3
) the device renders
completely useless for its intended purpose. Now more
than ever, in this era where we are experiencing the
apps, gadgets and wearables booms, we must be very
critical when choosing our measurement instruments.
REFERENCES
1
Balsalobre-Ferna
´ndez, C., D. Marchante, E. Baz-Valle, I.
Alonso-Molero, S. L. Jime
´nez, and M. Mun
˜o
´z-Lo
´pez.
Analysis of wearable and smartphone-based technologies
for the measurement of barbell velocity in different resis-
tance training exercises. Front. Physiol. 8:649, 2017.
2
Balsalobre-Ferna
´ndez, C., D. Marchante, M. Mun
˜oz-Lo
´-
pez, and S. L. Jime
´nez. Validity and reliability of a novel
iPhone app for the measurement of barbell velocity and
1RM on the bench-press exercise. J. Sports Sci. 36:64–70,
2018.
3
Courel-Iba
´n
˜ez, J., A. Martinez-Cava, R. Moran-Navarro,
P. Escribano-Pen
˜as, J. Chavarren-Cabrero, and J. J.
Gonzalez-Badillo. Reproducibility and repeatability of five
different technologies for bar velocity measurement in
resistance training. Ann. Biomed. Eng. 47:1523–1538, 2019.
4
Franco-Ma
´rquez, F., D. Rodrı
´guez-Rosell, J. M. Gonza
´-
lez-Sua
´rez, F. Pareja-Blanco, R. Mora-Custodio, J. M.
Yan
˜ez-Garcı
´a, and J. J. Gonza
´lez-Badillo. Effects of
combined resistance training and plyometrics on physical
performance in young soccer players. Int. J. Sports Med.
36:906–914, 2015.
5
Gonza
´lez-Badillo, J. J., and L. Sa
´nchez-Medina. Move-
ment velocity as a measure of loading intensity in resistance
training. Int. J. Sports Med. 31:347–352, 2010.
6
Ortega-Becerra, M., F. Pareja-Blanco, P. Jime
´nez-Reyes,
V. Cuadrado-Pen
˜afiel, and J. J. Gonza
´lez-Badillo. Deter-
minant factors of physical performance and specific
throwing in handball players of different ages. J. Strength
Cond. Res. 32:1778–1786, 2018.
7
Pallare
´s, J. G., A
´.Lo
´pez-Samanes, V. E. Ferna
´ndez-Elı
´as,
R. Aguado-Jime
´nez, J. F. Ortega, C. Go
´mez, R. Ventura,
J. Segura, and R. Mora-Rodrı
´guez. Pseudoephedrine and
circadian rhythm interaction on neuromuscular perfor-
mance. Scand. J. Med. Sci. Sports 25:e603–e612, 2015.
8
Pe
´rez-Castilla, A., A. Piepoli, G. Delgado-Garcı
´a, G.
Garrido-Blanca, and A. Garcı
´a-Ramos. Reliability and
concurrent validity of seven commercially available devices
for the assessment of movement velocity at different
intensities during the bench press. J. Strength Cond. Res.
33:1258–1265, 2019.
9
Sa
´nchez-Medina, L., J. J. Gonza
´lez-Badillo, C. E. Pe
´rez,
and J. G. Pallare
´s. Velocity- and power-load relationships
of the bench pull vs bench press exercises. Int. J. Sports
Med. 35:209–216, 2014.
10
Sa
´nchez-Medina, L., R. Mora
´n-Navarro, C. Pe
´rez, J.
Gonza
´lez-Badillo, and J. G. Pallare
´s. Estimation of relative
load from bar velocity in the full back squat exercise. Sport.
Med. Int. Open 01:E80–E88, 2017.
Publisher’s Note Springer Nature remains neutral with re-
gard to jurisdictional claims in published maps and institu-
tional affiliations.
BIOMEDICAL
ENGINEERING
SOCIETY
Technical Note on the Reliability of the PowerLift App
