Content uploaded by Stephen R Stannard
Author content
All content in this area was uploaded by Stephen R Stannard on Oct 01, 2017
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=tejs20
Download by: [Massey University Library Te Putanga ki te Ao Mātauranga] Date: 30 May 2017, At: 19:32
European Journal of Sport Science
ISSN: 1746-1391 (Print) 1536-7290 (Online) Journal homepage: http://www.tandfonline.com/loi/tejs20
Validity and reliability of the Moxy oxygen monitor
during incremental cycling exercise
E. M. Crum , W. J. O’Connor, L. Van Loo, M. Valckx & S. R. Stannard
To cite this article: E. M. Crum , W. J. O’Connor, L. Van Loo, M. Valckx & S. R. Stannard (2017):
Validity and reliability of the Moxy oxygen monitor during incremental cycling exercise, European
Journal of Sport Science, DOI: 10.1080/17461391.2017.1330899
To link to this article: http://dx.doi.org/10.1080/17461391.2017.1330899
Published online: 30 May 2017.
Submit your article to this journal
View related articles
View Crossmark data
ORIGINAL ARTICLE
Validity and reliability of the Moxy oxygen monitor during incremental
cycling exercise
E. M. CRUM
1†
,W.J.O’CONNOR
1
, L. VAN LOO
2
, M. VALCKX
2
, & S. R. STANNARD
1
1
School of Sport and Exercise, Massey University, Palmerston North, New Zealand &
2
Haagsche Hogeschool, Den Haag,
Zuid-Holland, Netherlands
Abstract
Introduction: The Moxy is a novel, cutaneously placed muscle oxygen monitor which claims to measure local oxygen
saturation (SmO
2
) and total haemoglobin (THb) using near-infrared spectroscopy. If shown to be reliable, its data storage
and telemetric capability will be useful for assessing localised O
2
usage during field-based exercise. This study investigated
the reliability of the Moxy during cycling and assessed the correlations between its measurements, whole-body O
2
consumption (VO
2
) and heart rate (HR). Methods: Ten highly trained cyclists performed an incremental, step-wise
cycling protocol on two occasions while wearing the Moxy. SmO
2
, THb, VO
2
and HR were recorded in the final minute
of each five-minute stage. Data were analysed using Spearman’s Order-Rank Coefficient (SROC), Intraclass Correlation
(ICC), and Coefficient of Variance (COV). Significance was set at p≤.05. Results: SmO
2
showed a ‘strong’or ‘very
large’correlation between trials (SROC: r= 0.842–0.993, ICC: r= 0.773–0.992, p≤.01) and was moderately correlated
with VO
2
and HR (r=−0.71–0.73, p≤.01). SmO
2
showed a moderate to high reliability at low intensities, but this
decreased as relative exercise intensity increased. THb showed poor correlations between tests and with the other
measured variables, but was highly reliable at all power outputs. Conclusions: The Moxy is a reliable device to measure
SmO
2
at low to moderate intensities, but at higher intensities, greater variation in measurements occurs, likely due to
tissue ischaemia or increased movement artefacts due to more frequent muscular contractions. THb has low variation
during exercise, and does not appear to be a valid indicator of muscle oxygenation.
Keywords: Exercise, training, technology, physiology
Highlights
.The Moxy monitor is a small, lightweight device with data capture capacity, which claims to measure local muscle oxygen
saturation (SmO
2
) and haemoglobin content (THb) using near-infrared spectroscopy.
.The Moxy showed strong correlations between SmO
2
values during two repeated cycling exercise protocols, conducted
one week apart, however, the reliability of values decreased as the exercise intensity increased.
.THb values showed low variation during exercise of varying intensity and did not appear to be a valid indicator of muscle
oxygenation.
Introduction
The Moxy muscle oxygen monitor is a small (dimen-
sions: 61 × 44 × 21 mm) lightweight (42 g)
cutaneously placed device which claims to measure
local oxygen saturation (SmO
2
) and total haemo-
globin (THb) in the muscle capillaries below its
point of position using near-infrared spectroscopy
(NIRS). NIRS models were developed as a non-inva-
sive method for measuring localised blood flow and
oxygenation, as opposed to the global measurements
of flow provided by earlier Doppler and plesthsmo-
graphy devices. The process is based on the differen-
tial proportion of wavelengths which are absorbed by
oxygenated and deoxygenated Hb (De Blasi, Cope,
Elwell, Safoue, & Ferrari, 1993; Jobsis, 1977).
However, traditional models are bulky and often
require multiple wire attachments to an external
device. Though suitable for research in a laboratory,
© 2017 European College of Sport Science
Correspondence: E. M. Crum, School of Sport and Exercise, Massey University, Corner University Ave and Albany Drive, Palmerston
North, New Zealand. E-mail: e.crum@massey.ac.nz
†
Data were collected by LV and MV with the help of EC and WO. Data analysis was performed by LV and MV. The study was written up by
EC with the help of WO and SS. All authors read and approved the final manuscript.
European Journal of Sport Science, 2017
https://doi.org/10.1080/17461391.2017.1330899
such products are limited in their use in field-based
studies. Thus, devices such as the Moxy, which
have data storage and telemetric capability, have the
potential to expand current knowledge of muscle
O
2
usage in ‘real-life’sporting situations and field-
based research.
In addition, the Moxy manufacturers claim that
their product will make local O
2
monitoring more
accessible to the general athletic population, and as
such provide a new method for measuring training
intensity ‘zones’. These zones can be used to guide
training prescription via the effect of a particular
mechanical workload on muscle O
2
requirements,
as opposed to using pace, power or heart rate (HR),
which are influenced by environmental conditions,
fatigue or mental stress. Further, the return of oxy-
genation values to a pre-determined plateau may
help determine the length of recovery between inter-
vals, or the number of intervals that should be per-
formed in a session (Fortiori Design LLC, 2015).
However, although the validity of the NIRS
method has been demonstrated through its highly
significant correlations with venous O
2
saturation
at rest and during exercise (r= 0.92) (Mancini
et al., 1994), and from the reproducibility of its
measurements during repeated incremental
running (r= 0.87–0.88) and cycling (r= 0.94–0.99)
tests to exhaustion (Austin et al., 2005), there is
little published information regarding the reliability
of the Moxy. This raises the question of whether it
is reliable enough to be used in scientific research
or in the monitoring of training. The group of
Born, Stöggl, Swarén, and Björklund (2016)
demonstrated that the Moxy’sSmO
2
measurement
inversely correlates with VO
2
(r=−0.7) during trail
running on hilly terrain. In addition, Cornachione,
McLaren, and Heil (2014) reported that SmO
2
measurements during various skiing styles changed
in response to intensity, location of the sensor
(arm versus leg placement) and variation of skiing
movement (arm versus leg-dominant). However,
neither study used replicated trials, so no evaluation
of the reproducibility of results obtained from the
Moxy could occur.
The current study thus aims to determine whether
data obtained from the Moxy during cycling testing
protocols can be replicated in subsequent identical
trials performed one week later by the same partici-
pants. In addition, we investigated whether the
Moxy can successfully monitor effort by determining
whether its measures correlate with other physiologi-
cal indicators of exercise intensity (HR, VO
2
).
Because it uses the previously validated NIRS
method, it is hypothesised that the Moxy measure-
ments will show strong correlations and reliability
between trials, and with HR and VO
2
.
Methods
Participants
Ten highly trained cyclists, including nine males and
one female, were recruited from the local cycling and
triathlon communities. The participants were all
regular competitors in local cycling events and had
an average age, height and weight of 23 years
(range: 16–30), 180 ± 9 cm and 71 ± 10 kg. All par-
ticipants provided written consent after being
informed about the nature of the experiment and
completed a health questionnaire to determine their
suitability for the study. The study was approved by
the Massey University Human Ethics Committee
(Southern A: 16/21) in accordance with the Declara-
tion of Helsinki.
The Moxy device
A diagram of the Moxy monitor is shown in Figure 1.
As with earlier NIRS devices, the Moxy monitor
(Moxy, Fortiori Design LLC, Minnesota, USA)
functions by sequentially sending light waves (630–
850 nm) from four light emitting diodes into the
tissue beneath it and recording the amount of
returned scattered light at two detectors positioned
12.5 and 25 mm from the light source. The pen-
etration depth of the light received at each detector
is half the distance between the light source and the
detector. The scattered light is processed by an algor-
ithm, which combines a tissue light propagation
model and the Beer–Lambert law to determine the
amount of light absorbed at wavelengths pertaining
Figure 1. Placement of the Moxy monitor.
2E. M. Crum et al.
to oxygenated and deoxygenated Hb. This allows cal-
culation of THb present beneath the device, as well as
the percentage of Hb containing O
2
(SmO
2
). Because
light passing into micro-vessels with a diameter
greater than 1 mm is expected to be completely
absorbed, it is assumed that the majority of reflected
light comes from the capillaries, and thus the
measurements obtained reflect the relative supply of
O
2
to the muscle versus its uptake (Fortiori Design
LLC, 2015; McCully & Hamaoka, 2000).
Experimental protocol
Each participant completed a bout of cycling exercise
protocol on two occasions separated by seven days.
Both trials were completed at the same time of day
and in the same air-conditioned lab which was set
to 17°C to ensure consistency across all trials. Partici-
pants were asked to refrain from heavy exercise in the
24 hours prior to each trial, and any exercise that was
completed during this period prior to the first trial
was replicated before the second trial. All participants
had previously been involved with research in our lab-
oratory and were familiar with the equipment and
testing protocols employed, thus a familiarisation
session was not required.
Pre-exercise protocol
On arrival into the laboratory, body mass was
obtained and a HR monitor (Polar Electro,
Kempele, Finland) was applied. The Moxy was posi-
tioned on the participant’s dominant leg, on the
vastus lateralis, halfway between the greater trochan-
ter and lateral epicondyle of the femur. The vastus
lateralis was chosen as the location for the Moxy
because it is part of the knee extensor group which
is the primary contributor to force generation in the
crank of the bicycle during down stroke of the pedal
(Raasch, Zajac, Ma, & Levine, 1997). Prior to place-
ment, this area was trimmed with an electric razor
and cleaned with alcohol swabs. The device was
secured with a light shield and athletic tape to block
ambient near-infra red light from interfering with
the detectors, and its exact position was recorded
and replicated for the second trial. After resting in a
seated position for five minutes, resting measures
for HR, SmO
2
and THb were recorded.
Exercise protocol
The cycling protocol was completed on an electroni-
cally braked cycle ergometer (Lode Excalibur, Gro-
ningen, the Netherlands) which was set up as
closely as possible to the participant’s own bike.
The set-up dimensions in the first trial were recorded
and replicated for the second trial. Following, the
pre-exercise measures, the participant mounted the
ergometer and completed a five-minute warm-up at
50 Watts (W). Subsequently, they completed a
step-wise protocol, which began at 100 W and
increased by 50 W every five minutes until exhaus-
tion. A step-wise protocol was chosen to allow assess-
ment of the Moxy’s reliability across a range of levels
of metabolic demand. During all stages, cadence was
kept at 90 rpm to ensure consistency between trials,
and exhaustion was defined as the inability to main-
tain this cadence or voluntary cessation of exercise.
This occurred after 4–6 stages (final stage = 250–
350 W) depending on each individual’s fitness level.
HR and Moxy data were recorded constantly
throughout the trial, with the average values from
the last minute of each stage being used for analysis.
There was also real-time monitoring of data using
PerfPRO computer software (Hartman Technol-
ogies, Rockware, Michigan, USA) which was dis-
played on a screen that was only visible to the
researchers. At 3:30 minutes into each stage, a
mouthpiece and nose-clip were applied to the partici-
pant and expired air was collected in a Douglas bag
for one minute. The expired gases in the Douglas
bags were immediately analysed for O
2
and carbon
dioxide (CO
2
) concentrations and volume using a
calibrated gas analysing system (AD Instruments,
Dunedin, New Zealand), which was calibrated
using gases of known concentration (15.01% O
2
,
5.01% CO
2
).
Data analyses
Minute ventilation (V
E
) and concentrations of O
2
and CO
2
values were used to calculate the volume
of inspired air (V
I
) using the Haldane transformation,
where V
E
was corrected for barometric pressure,
ambient temperature and atmospheric water satur-
ation. Subsequently, VO
2
and expired CO
2
(VCO
2
)
could be determined and are reported as standard
temperature and pressure dry (STPD).
Statistical analyses
Average values for the data collected constantly
during the exercise (HR, SmO
2
, THb) were calcu-
lated for each minute of each trial and used in the
analyses described hereafter. Statistical analyses
were performed using statistical computer software
(SPSS Statistics, Version 23, IBM Corporation,
New York). Linear correlations were performed to
explore associations between the two trials. The
level of correlation between trials for the Moxy
Reliability of the Moxy oxygen monitor 3
variables (SmO
2
and THb) was determined using
Spearman’s Rank-Order correlation (SROC; r=1
perfect correlation; >0.75 strong; >0.5 moderate;
<0.5 weak) and Intraclass Correlation (ICC; r=
>0.9 almost perfect correlation; 0.7–0.89 very large;
0.5–0.69 large; 0.3–0.49 moderate; 0.1–0.29 small;
<0.1 trivial). The reliability of the Moxy measure-
ments between trials was assessed using the coeffi-
cient of variation (COV) with the equation SD/
mean ∗100. The COV data from each participant
was pooled to determine an average COV for each
exercise stage. In accordance with Tew et al.
(2011), COVs of ≤10%, 10–25% and >25% were
considered good, moderate and poor. To test for cor-
relations between the Moxy measures and other vari-
ables (VO
2
, HR, velocity, power), SROC was
performed and used for analysis as previously
described. A negative correlation value indicates an
inverse relationship between variables. Significance
was set at p≤.05.
Results
Changes in physiological variables
Data detailing the changes in physiological variables
are presented in Table I. As exercise intensity
increased, there was a significant increase in HR
and VO
2
(p< .01), and a significant decline in
SmO
2
(p< .01). THb showed little variation
throughout each trial and was not significantly associ-
ated with changes in exercise intensity (p= .29).
Correlations and reliability between trials
A strong correlation for SmO
2
was observed in all
participants with SROC and ICC (SROC: 0.834–
0.980; ICC: 0.773–0.992). For the THb measure-
ments, six participants showed a weak correlation
(<0.5), two showed a moderate correlation (0.5–
0.75) and two showed a strong correlation with
SROC (>0.75). For ICC, THb showed a moderate
correlation in one participant (0.482), but all other
participants showed a small (0.1–0.29) or trivial
(<0.1) correlation. Data detailing the reliability
between trials for SmO
2
is presented in Table II.
While COV demonstrated a good to moderate
reliability at lower relative exercise intensities,
reliability lessened as intensity increased. For THb
there was a strong reliability of ≤1% between tests
for all participants at all exercise intensities (Table
II) which corresponded to a day-to-day variation in
SmO
2
of ≤0.1 g dL
−1
.
Correlations between Moxy measures and other
variables
Data detailing the correlations between SmO
2
and
THb with VO
2
, HR and power output are presented
in Table III. These values represent the average
SROC value across the two trials. SmO
2
showed a
moderate inverse correlation with all other variables
(r=0.71–0.73, p< .01). VO
2
also showed a strong
correlation with HR (r= 0.83, p< .01). However,
THb did not correlate with any of the other measured
variables.
Table I. Average values for physiological variables measured during the exercise protocol.
100 W 150 W 200 W 250 W 300 W 350 W
SmO
2
52 ± 6 45 ± 9∗36 ± 11∗26 ± 11∗20 ± 8∗15 ± 6∗
THb 12.7 ± 0.2 12.7 ± 0.3 12.7 ± 0.3 12.7 ± 0.3 12.8 ± 0.3 12.7 ± 0.3
HR 112 ± 12 130 ± 14∗146 ± 17∗162 ± 17∗171 ± 14∗175 ± 14∗
VO
2
27 ± 4 36 ± 5∗45 ± 6∗55 ± 7∗64 ± 8∗71 ± 9∗
Notes: Data are presented as mean ± SD. SmO
2
, muscle oxygen saturation expressed as %; THb, total haemoglobin content in g dL
−1
; HR,
heart rate in bpm, VO
2
, whole-body oxygen uptake in ml min kg
−1
.
∗p≤.05 compared to the previous stage.
Table II. Average coefficient of variation values between tests for
the Moxy measurements.
Power output (W) SmO
2
THb
100 4 1
150 5 1
200 12 1
250 20 1
300 31 1
350 28 <1
Note: COV data are presented as %. SmO
2
, muscle oxygen
saturation; THb, total haemoglobin.
Table III. Correlations between Moxy measurements and other
measured variables.
SmO
2
THb VO
2
HR
SmO
2
–0.05 −0.73∗−0.71∗
THb 0.05 –0.11 0.03
VO
2
−0.73∗0.11 –0.83∗
HR −0.71∗0.03 0.3∗–
Notes: SmO
2
, muscle oxygen saturation; THb, total haemoglobin;
VO
2
, whole-body oxygen consumption; HR, heart rate; PO, power
output.
∗p< .01.
4E. M. Crum et al.
Discussion
This study was designed to determine whether the
Moxy is a reliable device to measure muscle oxygen-
ation characteristics (SmO
2
, THb) in the vastus later-
alis during incremental cycling exercise. The data
support the validity of the Moxy to measure SmO
2
,
with statistical analyses showing a strong or excellent
correlation between trials for all participants (SROC:
r= 0.842–0.993, ICC: r= 0.773–0.992, p< .01).
Further, SmO
2
measurements showed a moderate
negative correlation with VO
2
and HR (r=−0.71–
0.73, p< .01), indicating that this variable is an
acceptable index of metabolic demand in the
working muscle. However, while the reliability of
SmO
2
measurement was good or moderate at lower
intensities, this decreased as the relative exercise
intensity increased for each participant. In contrast,
the THb measurements showed a poor or weak cor-
relation in the majority of participants for both stat-
istical analyses and were not correlated with any
of the other physiological variables measured.
However, the reliability of THb measurements was
strong in every participant at all intensities.
The reliability of the SmO
2
measurements shown
in the current study combined with the compact
nature of the device and its data capture capacity,
suggest that the Moxy may be useful for measuring
changes in local muscle oxygenation characteristics
resulting from an intervention, environmental influ-
ences or changes in training state. While portable
devices exist to measure whole-body O
2
usage (Over-
street, Bassett, Crouter, Rider, & Parr, 2016), there
has previously been no convenient and portable
method of recording local muscle O
2
status. The
Moxy device is small enough to be worn without
encumbering most types of exercise and possess
data capture capability, so it will have some utility
in the field-based setting. Accordingly, the Moxy
has already been used to determine local muscle oxy-
genation in exercise modes which cannot be easily
conducted in a laboratory, such as trail running,
rock climbing and skiing (Born et al., 2016; Corna-
chione et al., 2014; Kodejska, Michailov, & Balas,
2015). The low COV values shown in the current
study during low to moderate exercise intensities cor-
responded to day-to-day variation in SmO
2
values of
5 units, while a moderate COV corresponded to
changes in SmO
2
values of 5–10 units. Although
the Moxy shows lower reliability at higher exercise
intensities, this is similar to the reliability pattern
found in measurements from more traditional
NIRS devices (Thiel, Vogt, Himmelreich, Hubscher,
& Banzer, 2011). Thiel et al. (2011) attributed the
lower reliability of NIRS measurements during
cycling exercise at 75% VO
2max
compared to 25%
VO
2max
to the impairment of muscle perfusion by
stronger contractions producing more variable
tissue oxygenation. In addition, we suggest that
greater movement artefact which is likely to occur
with increased rates of muscular contraction, may
also contribute.
The moderate inverse correlation between SmO
2
and VO
2
(r=−0.73) is in accordance with Born
et al. (2016) who found that the changes in SmO
2
in the vastus lateralis of elite runners competing in
an undulating trail running race closely corre-
sponded to changes in VO
2
. This inverse relation-
ship reflects the increasing difference between the
mean muscle capillary O
2
content and the rate of
O
2
utilisation by the working muscle. That is, as
the contracting muscle is asked to produce more
aerobic-derived power, its rate of O
2
uptake from
the blood supply increases. Since SmO
2
is a
measure of the mean localised muscle capillary
content, if more is being taken up by the muscle,
then the mean O
2
content of the capillaries perfusing
the muscle will decrease, as this includes inflow of
O
2
as well as outflow of CO
2
(Bhambhani, 2004).
Indeed the greater O
2
concentration gradient
between the two compartments drives the movement
of O
2
into the muscles at an increased rate. In
addition, the greater rate of mechanical work requir-
ing the greater VO
2
results in an increased concen-
tration of intramuscular adenine diphosphate,
which signals a need for more oxidative ATP pro-
duction and further promotes O
2
movement into
the muscle (Bassett & Howley, 2000).
Although the current study showed a moderate
inverse correlation between HR and SmO
2
(r=
−0.71) and a strong correlation between HR and
VO
2
(r= 0.83) during exercise on a stationary bike,
others (Born et al., 2016) could find no correlation
for HR with either measure in trail runners. In that
study HR did not correspond to changes in terrain,
but instead remained steady throughout the race.
Stationary cycling allows a controlled exercise
environment where HR is predominantly determined
by the amount of O
2
required to perform a particular
power output and its value is well correlated with
other indicators of O
2
usage. However, during trail
running, HR is also influenced by the significant
psychological input and impact forces which occur
during downhill running on steep and technical
terrain (Born et al., 2016). It is likely that in field-
based situations which often involve a number of
influencing factors on HR, SmO
2
is a more precise
indicator of metabolic demands than HR. This
gives support to the use of training zones based on
muscle oxygenation, rather than the more tradition-
ally used HR, when exercising in a field-based
setting.
Reliability of the Moxy oxygen monitor 5
In contrast to the reliability indicated for the SmO
2
measurements, the THb data showed a poor corre-
lation between the two tests and was not associated
with any other physiological measures. However,
despite being poorly correlated, THb measurements
showed a strong reliability between tests for all par-
ticipants at every intensity, which corresponded to a
day-to-day variation in values for a given stage of
≤0.1 g dL
−1
. This may be explained by the very
small variation in THb values throughout each test
(average range across all stages: 0.2 g dL
−1
), despite
large changes in all other measured variables. It is
likely that our highly trained participants had a suffi-
cient supply of oxygenated Hb to the muscles
throughout the test, and were not limited by a lack
of O
2
availability, even at higher intensities. The dif-
fusion of O
2
into the mitochondria is largely driven by
the O
2
concentration gradient between the intra and
extra-muscular capillaries, as well as a build-up of the
by-products produced in the formation of ATP.
Thus, with increased muscular O
2
requirements,
the drive for O
2
diffusion into the muscles is
greater, decreasing the amount of O
2
in the capillaries
surrounding the muscles (SmO
2
) (Honig, Connett,
& Gayeski, 1992). In the current study, it appears
that this alteration in drive for O
2
diffusion was the
major determinant in O
2
usage, rather than an
increase in Hb transport. Consequently, THb does
not appear to be a valid way of indicating effort,
and SmO
2
may be a more relevant measure for moni-
toring changes in O
2
usage in highly trained athletes.
However, it is possible that due to their lower content
of Hb, the VO
2
of untrained individuals during pro-
gressive exercise may be influenced by THb to a
greater extent, leading to a greater correlation
between THb and other physiological variables
during exercise.
Despite the mostly positive correlations in SmO
2
data shown by the current research, it is possible
that limitations in the use of the device may exist
when applied to different individuals or environ-
mental conditions. Our participants were trained ath-
letes with very little adipose tissue. However, thicker
tissue may alter the scattering of light from the Moxy
and inhibit the return of light to the sensors. Accord-
ingly, previous research has indicated that the cut-off
for successful measurement with NIRS is a body
mass index of ∼32 (McCully, Landsberg, Suarez,
Hofmann, & Posner, 1997). Thus, the Moxy
cannot be recommended for use in such individuals
until reliability studies using individuals with higher
body fat percentages have been conducted. Further,
during prolonged exercise, skin blood flow increases
as the need for heat dissipation becomes greater. Pre-
viously, an increase in either skin blood flow
(+149 AU) or whole-body temperature (+0.8°C)
have been associated with a rise in tissue oxygenation
(r= 0.95) (Davis, Fadel, Cui, Thomas, & Crandall,
2006). In the current study, a rise in skin and total
body temperatures were limited because the exercise
was performed in an air-conditioned laboratory, with
constant cool air-flow to the skin of the participant via
a fan. However, it is acknowledged that in field-based
exercise, these factors could influence Moxy data.
Lastly, although this study was designed to measure
the reliability of the Moxy in a test–retest situation,
it did not measure its validity compared to a standard
NIRS device. While we have shown that the Moxy
gives reliable results, future research could investigate
whether these results correlate with other devices.
In conclusion, the current study demonstrates that
the Moxy produces reliable measurements during
exercise of low–moderate intensity and can be used
alongside VO
2
to gain a clearer idea of how the
amount of O
2
absorbed by the body relates to that
taken up by the muscles. However, at higher intensi-
ties, the reliability of the SmO
2
measurement
decreases, possibly due to greater movement artefacts
or tissue ischaemia. Based on these results, the Moxy
appears to have similar reliability to traditional NIRS
devices.
Acknowledgments
The authors would like to thank the cyclists who gave
up their time to participate in this study.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This work was supported by a Massey University Doctoral
Scholarship.
ORCID
E. M. CRUM http://orcid.org/0000-0003-1060-
5735
References
Austin, K. G., Daigle, K. A., Patterson, P., Cowman, J., Chelland,
S., & Haymes, E. M. (2005). Reliability of near-infrared spec-
troscopy for determining muscle oxygen saturation during exer-
cise. Research Quarterly for Exercise and Sport,76, 440–449.
doi:10.1080/02701367.2005.10599317
Bassett, D. R., & Howley, E. T. (2000). Limiting factors for
maximum oxygen uptake and determinants of endurance per-
formance. Medicine and Science in Sports and Exercise,32,70–
84. doi:10.1097/00005768-200001000-00012
Bhambhani, Y. N. (2004). Muscle oxygenation trends during
dynamic exercise measured by near infrared spectroscopy.
6E. M. Crum et al.
Canadian Journal of Applied Physiology,29, 504–523. doi:10.
1139/h04-033
Born, D., Stöggl, T., Swarén, M., & Björklund, G. (2016).
Running in hilly terrain: NIRS is more accurate to monitor
intensity than heart rate. International Journal of Sports
Physiology and Performance, Advance online publication.
doi:10.1123/ijspp.2016-0101
Cornachione, K., McLaren, J., & Heil, D. P. (2014). Use of a wire-
less NIRS monitor to track changes in muscle oxygenation for
laboratory-based Nordic skiing test proto-col. In E. Müller, J.
Kröll, S. Kröll, S. Lindiger, J. Pfusterschmeid, & T. Stöggl
(Eds.), Science and Skiing VI, (pp. 369–376). Maidenhead:
Meyer & Meyer Sport (UK) Ltd.
Davis, S. L., Fadel, P. J., Cui, J., Thomas, G. D., & Crandall, C. G.
(2006). Skin blood flow influences near-infrared spectroscopy-
derived measurements of tissue oxygenation during heat stress.
Journal of Applied Physiology,100, 221–224. doi:10.1152/
japplphysiol.00867.2005
De Blasi, R. A., Cope, M., Elwell, C., Safoue, F. & Ferrari, M.
(1993). Noninvasive measurement of human forearm oxygen
consumption by near infrared spectroscopy. European Journal
of Applied Physiology and Occupational Physiology 67, 20–25.
doi:10.1007/BF00377698
Fortiori Design LLC. (2015). Introduction to muscle oxygen monitor-
ing with Moxy. Retrieved from http://cdn2.hubspot.net/hub/
188620/file-433442739-pdf/docs/moxy-ebook-intro-to-muscle-
oxygen.pdf?t=1488816603832
Honig, C. R., Connett, R. J., & Gayeski, T. (1992). O
2
transport
and its interaction with metabolism: A systems view of aerobic
capacity. Medicine and Science in Sports and Exercise,24,47–53.
Europe PMC ID:1548995.
Jobsis, F. F. (1977). Noninvasive, infrared monitoring of cerebral
and myocardial oxygen sufficiency and circulatory parameters.
Science,198, 1264–1267. doi:10.1126/science.929199
Kodejska, J., Michailov, M. L., & Balas, J. (2015). Forearm
muscle oxygenation during sustained isometric contractions in
rock climbers. Acta Universitatis Carolinae: Kinanthropologica,
51,48–55. doi:10.14712/23366052.2015.31
Mancini, D. M., Bolinger, L., Li, H., Kendrick, K., Chance, B., &
Wilson, J. R. (1994). Validation of near-infrared spectroscopy
in humans. Journal of Applied Physiology,77, 2740–2747.
PMID: 7896615.
McCully, K. K., & Hamaoka, T. (2000). Near-infrared spec-
troscopy: What can it tell us about oxygen saturation in skeletal
muscle? Exercise and Sport Sciences Reviews,28, 123–127. PMID:
10916704.
McCully, K. K., Landsberg, L., Suarez, M., Hofmann, M., &
Posner, J. D. (1997). Identification of peripheral vascular
disease in elderly subjects using optical spectroscopy. The
Journals of Gerontology Series A: Biological Sciences and Medical
Sciences,52A, B159–B165. doi:10.1093/gerona/52A.3.B159
Overstreet, B., Bassett, D., Crouter, S., Rider, B., & Parr, B.
(2016). Portable open-circuit spirometry systems: A review.
The Journal of Sports Medicine and Physical Fitness,57, 227–
237. doi:10.23736/S0022-4707.16.06049-7
Raasch, C. C., Zajac, F. E., Ma, B., & Levine, W. S. (1997).
Muscle coordination of maximum-speed pedaling. Journal
of Biomechanics,30, 595–602. doi:10.1016/S0021-9290
(96)00188-1
Tew, G. A., Klonizakis, M., Moss, J., Ruddock, A. D., Saxton, A.
D., & Hodges, G. J. (2011). Reproducibility of cutaneous
thermal hyperaemia assessed by laser Doppler flowmetry in
young and older adults. Microvascular Research,81,177–182.
doi:10.1016/j.mvr.2010.12.001
Thiel, C., Vogt, L., Himmelreich, H., Hubscher, M., & Banzer, W.
(2011). Reproducibility of muscle oxygen saturation.
International Journal of Sports Medicine,32, 277–280. doi:10.
1055/s-0030-1269922
Reliability of the Moxy oxygen monitor 7
