Criterion Measures of Energy Expenditure
Doubly Labeled Water (DLW)
Doubly-labeled water (DLW) is typically considered as the “gold standard” for the assessment of
energy expenditure in free-living individuals.  DLW requires that an individual consume a known
volume, based on body weight, and concentration of stable isotopes 180 and hydrogen 2H.  Urine
is measured over a 7 to 14 day period to determine the elimination rates of these two isotopes,
from which carbon dioxide and the respiratory quotient can be estimated and energy expenditure
determined (Montoye et al., 1996).
The ability of DLW to assess energy expenditure was initially examined for use in laboratory
animals in the 1950’s (Lifson et al., 1995), and has since been applied to humans.  DLW has
been shown to be accurate to within ± 8% of known values when used in laboratory animals
(Roberts 1989, Nagy 1980), whereas DLW has been shown to be accurate to within ± 5% in
humans when compared to a respiratory chamber (Schoeller et al., 1986, Seale et al., 1993,
Westerterp et al., 1988) or other continuous methods of measuring respiratory gas exchange
(Schoeller and van Santen 1982).  When applied to free-living environments, it has been
suggested that there is most likely a slight increase in the error of measurement with DLW
(Montoye et al., 1996).  Thus, DLW is typically considered as the most accurate technique for
assessing energy expenditure in free-living individuals.
Despite the potential accuracy of DLW for assessing energy expenditure in free-living
environments, there are disadvantages that limit the wide use of this technique in research
and clinical situations.  DLW requires the use of water containing stable isotopes (18O and 2H),
and this can be expensive with cost ranging from $500 to $1500 per subject for each measurement
period, with additional costs for laboratory equipment and well-trained technicians.  DLW also
requires that subjects collect urine over a 7 to 14 day period and transport this urine to a laboratory
for analysis, which may create a significant burden and barrier for subjects.  These factors may
limit the practical utility of assessing energy expenditure using DLW in many research and clinical
settings (Starling et al., 1999, Macfarlane 2001).
While DLW may provide an accurate representation of energy expenditure in free-living individuals,
this method provides little information about the pattern of physical activity behavior that influences
energy expenditure.  Because of the need to collect excreted urine over a 7 to 14 day period, DLW
can only provide a representation of the average total energy expenditure per day during this period
of time, rather than information with regard to more acute periods of physical activity.  Thus, this
may limit the utility of DLW if the desire is to obtain information about acute periods of physical
activity or how patterns of activity contribute to total energy expenditure and health-related outcomes.
Indirect Calorimetry (IC)
Open-circuit indirect calorimetry (IC) is commonly used as a criterion measure when assessing
energy expenditure.  Montoye et al. (1996) reported that indirect calorimetry is accurate to within
2% of energy expenditure measurements of doubly-labeled water (DLW).  Despite the potential
accuracy of IC, there are numerous disadvantages of this method for the assessment of energy
expenditure including expense, the need for trained technicians, and due to limited mobility this
method may not feasible for use outside of controlled laboratory conditions (Macfarlane 2001).
While there are portable IC systems commercially available that can be used in field settings,
these systems can be expensive (i.e. $20,000-$30,000), may only be used continuously for a
few hours before recharging is necessary, and may not be practical to wear in many settings
(i.e., work, home, social setting, etc.).  These limitations of IC negatively impact the utility of this
method of assessing energy expenditure in free-living adults.

Methods of Estimating Energy Expenditure
Self-Report Methods
A commonly used method to estimate energy expenditure is self-report of physical activity,
which can include the use of questionnaires, interviews, or physical activity diaries.  These
techniques typically involve the assignment of a score or value to a reported physical activity,
which are then summed over the measurement period and converted to energy expenditure
(Keim et al., 2004).  Self-report measures have advantages of being low cost and require relatively
minimal participant burden.  However, the accuracy of self report techniques for estimating energy
expenditure has been questioned.  This may be a result of these techniques being prone to
misinterpretation of instructions by respondents, inaccurate recall of activity behaviors, deliberate
misreporting of information, or the inability of these techniques to accurately capture all forms and
components of physical activity (Pereira et al., 1997, Montoye 1996).
The 7-day Physical Activity Recall (PAR) is commonly used in intervention research to assess
physical activity and estimate energy expenditure.  Leenders et al. (2001) compared energy
expenditure estimated from the PAR to DLW and reported no significant difference in the group
mean when represented using either of these techniques.  However, further examination of the
data reveal rather larger individual differences between the PAR and DLW for individuals with
relatively low or high levels of energy expenditure.  Individuals with the lowest levels of energy
expenditure tended to overestimate energy expenditure by 137 kcal/d, with individuals with the
highest energy expenditure tended to underestimate by 287 kcal/d.  Moreover, Irwin et al. (2001)
reported that the PAR differed from DLW by 30.6 ± 9.9%.  These discrepancies in energy
expenditure estimated from the PAR may indicate that this method is unable to accurately
capture individual differences in physical activity, which may limit the utility of this questionnaire.
The inability of questionnaires to accurately estimate energy expenditure is not limited to the
PAR. Startling et al. (1999) compared the Minnesota Leisure Time Activity Physical Activity
Questionnaire (LTA) and the Yale Physical Activity Questionnaire (YPAS) to DLW in older men
and women (45 to 84 years).  Results showed the LTA underestimated physical activity by
approximately 50% to 60% compared with DLW, with no significant difference reported between
YPAS and DLW.   Jacobs et al. (1993) examined 10 commonly used physical activity
questionnaires and reported that most questionnaires may not be suitable for accurately
estimating energy expenditure during moderate and light intensity activity, but may be more
accurate for estimating energy expenditure during periods of more vigorous intensity activity.
Moreover, Montoye et al. (1996) reviewed the reliability and validity of various physical activity
questionnaires and concluded that there is a wide range of validity and reliability that may be
questionnaire specific.  Thus, it appears that there is variability in the accuracy of questionnaires
for estimating energy expenditure, and this should be considered when selecting a questionnaire
to assess physical activity in free-living adults.
Pedometers
Pedometers, which assess number of steps of locomotion, have been used to measure physical
activity.  Advantages may include objective measuring of physical activity, low cost, minimal
burden, and the ability to provide feedback to the user (Schneider et al., 2004, Tudor-Locke et al.,
2001).  However, inherent disadvantages of pedometers may make these devices less viable in
the assessment of energy expenditure.  One major disadvantage of pedometers is that the
accuracy varies in different models.  A recent study by Schneider et al. (2004) compared the
step values of 13 models of pedometers over a 24 hour period, with the Yamax Digi-Walker
SW-200 (YX200) model used as the criterion measure.  Results showed five of the pedometers
(Freestyle Pacer Pro, Accusplit Alliance 1510, Yamax Skeleton EM 180, Colorado on the Move,
and Sportline 345) significantly underestimated steps (p< 0.05), while three pedometers
(Walk4 Life LS 2525, Omron HJ-105, and Oregon Scientific PE316CA) significantly overestimated
steps (p< 0.05).  Underestimations were as high as 25% while overestimations reached 45% in
some models of pedometers.  Thus, these results indicate that the accuracy of steps taken and
energy expenditure estimation may depend on the brand of pedometer.
Accuracy of pedometers can also be impacted by the intensity and rate of the activity.
When comparing pedometers to hand counted steps, during slow treadmill walking of 54 m/min
many pedometers (Sportline 345, Yamasa Skeletone, Sportline 330, and Freestyle Pacer Pro)
significantly (p< 0.05) underestimated steps, while during fast treadmill walking (107 m/min)
pedometers (Yamasa Skeleton, Omron, Kenz Lifecorder, New Lifestyles 2000, Oregon Scientific
and Walk4Life LS 2525) significantly overestimated steps (Crouter et al., 2003).  There is some
evidence that pedometers may be most accurate for assessing steps at the speed of 80 m/min,
with some pedometers (Yamax, Omron, New Lifestyles, Yamasa Skeletone, Kenz Lifecorder,
Walk4Life LS 2525) measuring steps within ± 1% of actual steps when walking at this pace
(Crouter et al., 2003; Le Masurier et al., 2004).
Although some pedometers are accurate in assessing steps they are less accurate in assessing
distance and kilocalories.  Crouter et al. (2003) found most pedometers estimated distance within
10% at 80 m/min, but overestimated distance at slower speeds (54 m/min) and underestimated
distance at faster speeds (107 m/min).  When the investigators compared energy expenditure of
pedometers to indirect calorimetry, net kilocalories were overestimated at every speed (54, 67,
80, 94, and 107 m/min), while gross kilocalories were within 30% accuracy for all speeds.  This
study found that at slower speeds, the accuracy of the pedometers was compromised for step
counting, kilocalorie estimates, and distance traveled.  Thus, these results indicate pedometers
may not be suitable for use in populations with a slow gait, such as the elderly or obese, and
may be more accurate for counting steps rather than estimating energy expenditure.
Studies of pedometers have shown that these devices may not provide a comparable estimate
of physical activity when compared to questionnaires or accelerometers.  When comparing
pedometers to the 7 day PAR, low (r=0.34) to moderate correlations (r=0.49) have been reported
between steps counts and average energy expenditure (Welk et al., 2000).  Moreover, when
compared to a CSA accelerometer in laboratory or field settings, the Yamax pedometer detected
significantly lower steps than the CSA during treadmill walking at 54 m/min (75.4% versus
98.9%, p< 0.05) (Le Masurier & Tudor-Locke, 2003), whereas the Sportline 330 detected fewer
steps than the CSA (p< 0.05) (Le Masurier et al., 2004).  Thus, these results show some
pedometers may be less accurate than others for assessing energy expenditure and lower
intensity activities.
Reliability and validity of pedometers has improved as this technology has evolved.  Earlier
models of pedometers had poor reliability across models and errors in the estimation of steps
and distance walked (Washburn et al., 1980, Gayle et al., 1977); however, newer models have
shown improvements in reliability and validity with the Yamax Digi-walker measuring steps and
distance to within 1% of actual values (Bassett et al., 1996) and correlations of 0.76 between the
Tritrac accelerometer and the Yamax Digi-Walker (Differding et al., 1998).  However, pedometers
continue to have difficulty in accurately detecting changes in speed of walking and can not
accurately estimate the intensity or duration of an activity (Welk et al., 2000).
In summary, while pedometers may be appealing because of their low cost and ease of use,
the ability of the devices to accurately estimate energy expenditure across a variety of activities
is limited.  This may limit the use of pedometers in some populations and when performing certain
forms of activities.
Heart Rate Method
Heart rate monitors have been used to estimate energy expenditure, and is a result of these
devices providing an objective measurement, having the ability to measure different intensities
of physical activity, and because heart rate is significantly correlated with energy expenditure
during aerobic physical activity (Janz, 2002).  However, there are numerous disadvantages of this
method of estimating energy expenditure.  For example, heart rate response may be due to
non-related physical activity events such as emotions, room temperature, and training state
(Janz, 2002), which can typically result in an overestimation of energy expenditure.
Disadvantages of heart rate monitors include comfort level when being worn and some are
not useful for capturing energy expenditure of anaerobic activity (Janz, 2002).
To improve the accuracy of heart rate to estimate energy expenditure, a calibration test is
necessary to determine the relationship between heart rate and energy expenditure for each
individual.  Strath et al. (2000) examined the relationship between heart rate (beats/min) and
oxygen consumption (VO2 = ml/kg/min) during both laboratory and field-based moderate intensity
activities.  A moderate correlation was found between heart rate and VO2 (r = 0.68); however,
adjustments for age and fitness level increased the accuracy of predicted energy expenditure
to r = 0.87.
In a review of the literature, Montoye et al. (1996) reported similar correlations for energy
expenditure between heart rate and DLW (r = 0.73) or VO2 (r = 0.55), with the inaccuracy of
heart rate to estimate energy expenditure ranging from 2% to 22%.
Accelerometry
Accelerometry is a method of detecting body motion using either uniaxial (i.e., Caltrac or
CSA/MTI) or multi-axial (i.e., RT3 or TriTrac-R3D) devices. These devices use electronic
sensors to monitor body movement, and these movement counts can be used to estimate
energy expenditure.  Accelerometers are typically worn at the level of the waist, and there is
the ability to capture and store minute-by-minute data for periods of up to 4 weeks.  Thus,
accelerometry may have utility for monitoring energy expenditure in free-living individuals.
It appears that accelerometers may provide the most accurate estimate of energy expenditure
during periods of level walking.  When compared to a criterion measure of energy expenditure,
significant correlations have been shown for uniaxial (r = 0.94) (Pambianco et al., 1990) and
triaxial accelerometers (r = 0.99) (Levine et al., 2001) during periods of level walking, with
consistency during steady state walking ranging from 0.86 to 0.96 (Jakicic et al., 1999).
Despite these significant correlations, Pambianco et al. (1990) reported that accelerometry
may overestimate energy expenditure by an average of 9-13% compared to indirect calorimetry,
with significant differences of 13.5 kcal, 19 kcal, 25.5 kcal shown between accelerometry and
the criterion measure for speeds of  3.2, 4.8 and 6.4 km/h, respectively.  Haymes et al. (1993)
reported that accelerometry significantly overestimated energy expenditure at walking speeds
above 2mph (~3.6 kcal/min) and could not discriminate between running speeds of 5-8mph
(overestimated ~2.6 kcal/min), while Balogun et al. (1989) reported that accelerometry
significantly (p< 0.001) overestimated energy expenditure by 13.3 to 52.9% during level
walking at various walking speeds (54, 81, 104, 130 m/min).
A disadvantage of accelerometers is that these devices may not be sensitive to changes in
work rate during walking resulting from changes in grade or speed, which may result in the over-
or underestimation of energy expenditure.  Fehling et al. (1999) reported that the Caltrac
accelerometer significantly overestimated energy expenditure 10% during walking on a flat
surface; however, when the grade was increased the error in estimate increased to 52%.
Examination of the Tritrac accelerometer indicated that energy expenditure was significantly
underestimated by 19% during level walking and by 28% when walking grade was increased.
It has also been demonstrated that accelerometry may not be accurate for all forms of activity.
Montoye et al. (1983) found accelerometry compared with indirect calorimetry had a standard
error of estimate of 6.6 ml/kg/min for activities such as stepping, half knee-bends, flour touches,
as well as walking and running on flat and incline surfaces.  Jakicic et al. (1999) found
accelerometry significantly overestimated (p< 0.05) energy expenditure at the lowest walking
and running speeds by 1 kcal/min, however significant underestimations were found for all other
walking and running workloads, stepping, slideboard and cycling activities (29.8 to 50.0 kcal).
Thus, these studies indicate the accuracy of accelerometry is activity specific.
Inter-unit variability may impact the accuracy of accelerometry and inter-unit correlations may
be affected by change in work rate.  Jakicic et al. (1999) found there was a significant difference
(p < 0.05) between two accelerometry units during walking, stepping, and slideboard exercises,
with the difference between these two units being 0.5 to 0.8 kcal/min.  Nichols et al. (1999)
reported inter-unit correlations of 0.87 during walking (r = 0.87), however the inter-unit correlations
were 0.84 during jogging and 0.73 during fast running.  These results illustrate there may be
inter-unit variability among accelerometers, which may suggest the need to using the same unit
when assessing energy expenditure within an individual over a period of time.
The accuracy of accelerometry may be dependent on the unit type (i.e., Caltrac versus Tritrac
versus CSA, etc).  When comparing the accuracy of accelerometry units during laboratory
conditions, Welk et al. (2000) reported that the CSA provided accurate estimates of energy
expenditure, while the Tritrac and Biotrainer overestimated energy expenditure (101 to 136%).
However, during field activities the CSA, Tritrac and Biotrainer all underestimated energy
expenditure (42 to 67%).  These results demonstrate that variability may exist between different
models of accelerometers, and this should be considered when these devices are used in clinical
and research applications.
In summary, while accelerometers detect motion and provide minute-by-minute data these
devices exhibit large over- and under estimations of energy expenditure particularly during
activities of increased work rate due to changes in speed or grade.  Both inter-unit variability
and model of the accelerometer may play a role in the accuracy of estimated energy expenditure.
Thus, the use of accelerometry to estimate energy expenditure may not be applicable for many
forms of activities that occur in the free-living environment.
Intelligent Device for Energy Expenditure and Activity (IDEEA)
A newer portable device called the Intelligent Device for Energy Expenditure and Activity
(IDEEA, MiniSun, CA) has been developed to estimate energy expenditure of physical activity.
The IDEEA system estimates energy expenditure through body and limb motions, which are
collected through five sensors attached to the chest, thighs and feet.  Signals from the sensors
are recoded to the device and later downloaded to a computer for analysis.  Few studies have
been published examining the validity of this device.  A study by Zhang et al. (2004) was
performed to examine the validity of IDEEA, compared to estimated energy expenditure from a
non-portable mask calorimeter (Hans Rudolph, Kansas City, MO) and a respiratory chamber with
open air circuits.  One of the experimental protocols included performing activities such as sitting,
standing, lying down, level treadmill walking and running at different speeds for 50 minute durations
while wearing the mask calorimeter, while the other protocol consisted of subjects living in a
metabolic chamber for 23 hours during which time they completed three exercise sessions on a
motorized treadmill (walk for 15 minutes, run for 10 minutes or walk for 15 minutes, and walk for
15 minutes).  Analysis of data showed the overall accuracy for estimated energy expenditure of
IDEEA and the calorimeters was 95.1 ± 2.3%.  However, it was also found that IDEEA
underestimated energy expenditure for certain subjects and overestimated energy expenditure
for others up to 10%.  The errors in the estimation of energy expenditure using IDEEA may limit
the use of this device for estimating energy expenditure in free-living individuals.
IDEEA may have limitations related to wearability of this device in free-living individuals.
For example, IDEEA sensors are attached to the body using medical tape and must be
removed during bathing (Zhang et al., 2004).  Moreover, the sensors are taped on the chest,
the frontal part of the thigh and under each foot, and these placements may potentially make
the device uncomfortable or less appealing to some individuals.  IDEEA sensors are also
connected by thin flexible wires which may be cumbersome or limit the willingness of individuals
to wear this device.  Zhang et al. (2003) also noted that the anatomical positions or angle of the
sensors may be impacted by the shape of the body (i.e. lean versus obese, male versus female
shape), and the variability in site location may affect the accuracy of this device.  Zhang et al.
(2004) reported that the IDEEA may also have limitations when detecting arm movements and
the transition from one activity to another (i.e. from running to walking).  These factors appear to
impact the accuracy of IDEEA, which may limit the utility in research and clinical environments
for the estimation of energy expenditure.
SenseWear Pro Armband
The SenseWear Pro Armband (SWA) is a portable sensor that gathers information on
movement, heat flux, skin temperature, near-body temperature, and galvanic skin response,
which are used to estimate energy expenditure.  The SWA is worn on the right arm over the
belly of the bicep muscle, and has the capability of capturing and storing minute-by-minute data.
It appears the SWA exhibits errors in estimation of energy expenditure, which vary according to
exercise modality.  Fruin et al. (2004) found during cycling exercise the SWA underestimated
energy expenditure compared to indirect calorimetry, with the most pronounced difference during
early exercise (minute 1-10, % difference = 8%).  When examining walking, King et al. (2004)
found that the SWA underestimated total energy expenditure during various speeds of walking
and running compared to indirect calorimetry (p< 0.001), while Fruin et al. (2004) found the SWA
overestimated energy expenditure while walking on a flat surface (14-38%) and underestimated
energy expenditure during walking on an incline (22%).
Jakicic et al. (2004) examined the ability of the SWA to estimate energy expenditure, which
incorporated the use of both exercise-specific and general algorithms.  Use of the exercise
specific algorithms resulted in non-significant differences between energy expenditure estimated
by the SWA compared with indirect calorimetry.  However, when the general algorithm was used
the SWA significantly (p? 0.001) underestimated energy expenditure during walking (6.9%),
cycling (28.9%), and stepping (17.7%) and overestimated energy expenditure during arm
ergometry (29.3%).  Thus, these results indicate the SWA may be less accurate when using
the manufacturer’s general algorithm and may require the use of exercise specific algorithms,
which would limit the use of this device in free-living individuals.
Although armbands such as the SWA appear to be inaccurate for some forms of acute periods
of activity, they may be more useful in capturing longer periods of activity.  Mignault et al. (2005)
found no significant differences in mean energy expenditure between the SWA (2,237 ± 568
kcal/day), which is marketed as the HealthWear armband (Roche Diagnostics, Indianapolis, IN),
compared with doubly-labeled water (2,315 ± 625 kcal/day) during a 10 day period.  Although no
significant differences were found, the range of under-and over- estimation of the armband versus
DLW was -243 to 176 kcal/day.  Thus, while there are limitations in the accuracy of the SWA for
estimating energy expenditure during acute periods of physical activity, the accuracy of this
device may be improved when energy expenditure is estimated over longer periods of time.
Heat Flux to Estimate Energy Expenditure
The first model of the calculation of heat balance of the human body dates back to 1932
(Buttner 1932) while the first peer reviewed journals to publish studies on heat flux appeared in
the early 1980’s.  Heat balance is defined as the balance between heat produced and the heat
lost (English et al., 1990).  Original studies that incorporated the use of heat flux transducers
were used to examine heat loss in populations including divers and surgical patients.
The validity of heat flux transducers to measure heat loss has shown positive findings.
A study by ayton et al. (1983) was performed to examine the validity of heat flux transducers
by comparison to a suit calorimeter, which was served as the criterion measure or direct
calorimetry.  Subjects underwent 2 days of testing, each consisting of a series of cooling
and warming cycles, with the entire testing period lasting approximately 6 hours.  Subjects
rested in a seated position with their legs and feet resting on a hassock.  Water in the suit
calorimeter was cooled and heated to allow for changes in body temperature, which included
heat loss.  Temperatures used for the testing cycle included 28, 23, 18, 10 and 5 degrees
Celsius, with 35 degrees Celsius used to warm subjects after the coolest conditions.  During
the testing cycles both the heat flux transducers and suit calorimeter were worn for all testing.
Fourteen heat flux transducers were worn to provide heat loss information for 6 different
segments of the body.  Data analysis showed a correlation between heat loss rates measured
using the heat flux transducers and a suit calorimeter.  While heat loss measured between both
measures was similar for the torso and legs, the transducers measured less heat from the head
and arms than the suit calorimeter.  Based on the results it appears heat flux transducers may
provide a reasonable measurement of relative regional and total heat in human subjects during
rest in a supine position.
A more recent study examined the ability of heat flux transducers to measure heat exchange
in subjects who were exposed to four different temperatures (30, 33, 37 band 40 degrees Celsius)
(English et al., 1990).  Each temperature remained constant for twenty minutes and heat flux data
was recorded every minute.  Heat exchange was measured using six heat flux transducers, with
three worn on the back and three worn on the chest.  Heat exchange values obtained from the
heat flux transducers were used to compute heat exchange coefficients (radiant, convection,
combined radiant and conduction, and conductance) from pre-existing formulas.  Coefficients
for radiation (6.4), convection (8.7), combined radiation and conduction (9.7), and conductance (41)
were with-in accepted ranges (Allan., R.J. 1987 and Kerslake D., 1972).  The results indicate the
direct measurement of heat exchange with heat flux transducers may improve the understanding
of the body’s thermal balance.
The ability of heat flux transducers to measure heat loss during varying conditions has led to the
use of heat flux transducers for the assessment of energy expenditure.  The KAL-X Sensor is
a wireless sensor that uses heat flux technology to measure conductive, radiant, convective and
evaporative heat loss, to estimate energy expenditure (EE).  There is limited published data on
the validity of the KAL-X Sensor for estimating energy expenditure.  However, two pilot studies
published as abstracts have been conducted to assess the accuracy of the KAL-X Sensor.
Jakicic et al. (1993) examined the validity of a KAL-X prototype to measure energy expenditure.
Subjects were seven healthy males (age = 21.57 ± 5.06 years, BMI = 22.37 ± 1.91 kg/m2)
recruited to participate in three exercise trails (walking, cycling and stepping).  The trials were
each five minutes in length with both the KAL-X Sensor and indirect calorimetry worn to
measure energy expenditure at rest, during exercise and post exercise.  Four KAL-X Sensors
were worn on the upper arm, chest, back and thigh during each trial.  Protocols for the exercise
trials included the following treadmill walking at 3.0 mph at 0% grade, stepping on an 8 inch
bench at 80 cycles per minute and cycling at 1 kg resistance at 50 rpm.  Comparison of energy
expenditure measured by indirect calorimetry and by the KAL-X Sensor showed no significant
differences (p< 0.05) for walking (44.42 ± 6.12 (IC) vs. 42.46 ± 16.89 kcal (KAL-X), stepping
(47.26 ± 5.61 (IC) vs. 43.23 ± 18.48 kcal (KAL-X), and cycling (43.06 ± 4.65 (IC) vs. 43.08 ± 25.85
kcal (KAL-X). Although the sample size was small and the exercise duration was short, it appears
that the initial tests of the KAL-X system provide valid estimates of energy expenditure of selected
moderate intensity activities.
Winters et al. (1998) examined the validity of a KAL-X prototype to measure energy expenditure
during walking, cycling, stepping and slideboard exercises.  Twenty subjects (age = 21.5 ± 3.38
years; BMI = 23.3 ± 3.55 kg/m2) were recruited to participate in four exercise trials lasting 20-30
minutes.  The treadmill walking protocol was 30 minutes in length and consisted of walking at
3.5 mph at 0, 5, and 10% grades (each grade was a 10 minute bout).  Cycling, stepping and
slideboard exercises were all 20 minutes in length with the rate increasing at 10 minutes from
50 to 65 rpm, 17 to 21 cycles and 20 to 30 cycles for the three exercises respectively.  The KAL-X
Sensors were worn for all trials and were placed on the chest, back, right upper arm, and calf.
Heat flux data was recorded by the KAL-X Sensor during each minute of the exercise session
and this data was downloaded to a computer for analysis.  Indirect calorimetry served as the
criterion measure for all trials.  No significant differences (p< 0.05) were found between energy
expenditure measured from indirect calorimetry and from the KAL-X Sensor for any of the
exercise trials except for level walking.  Results for energy expenditure estimates for the KAL-X,
although not significantly different from indirect calorimetry, were based on a proprietary non-linear
regression from the walking data.
Based on the review of literature it appears the use of heat flux to measure energy expenditure
during physical activity shows promising results, however these results were based on a prototype
instrument, subjects in the studies were young, lean individuals and the sample size of the studies
was small, with one of the studies limiting the testing to males.  These initial studies provide
support for use of heat flux transducers in the assessment of energy expenditure, but also show
the need for a formal validation of the KAL-X Sensor.  Therefore, studies are needed to
establish the validity of the KAL-X Sensor to measure energy expenditure.

References
Allan JR:  The thermal environment and human heat exchange. In: Ernsting J,
King P (eds).  Aviation Medicine, 2nd ed. London: Butterworths, 1987; 219-234.
Balogun JA, Martin DA, Clendenin MA: Calorimetric validation of the Caltrac
accelerometer during level walking. Phys Ther 69: 501-509, 1989.
Bassett DR, Ainsworth BE, Swartz AM, Strath SJ, O’Brien WL, King GA:
Accuracy of five electronic pedometers for measuring distance walked.
Med Sci Sports Exerc 28: 1071-1077, 1996.
Bassett DR, Ainsworth BE, Swartz AM, Strath SJ, O’Brien WL, King GA:
Validity of four motion sensors in measuring moderate intensity physical activity.
Med Sci Sports Exerc 32(9 Suppl): S471-S480, 2000.
Buttner K: Physikalisches zum Warmehaushalt des Menschen.
Klin Wochenschr 36: 1508-1509, 1932.
Crouter SE, Albright C, Bassett DR Jr: Accuracy of Polar S410 heart rate monitor to
estimate energy cost of exercise. Med Sci Sports Exerc 36: 1433-1439, 2004.
Crouter SE, Schneider PL, Karabulut M, Bassett DR Jr: Validity of 10
electronic pedometers for measuring steps, distance, and energy cost.
Med Sci Sports Exerc 35: 1455-1460, 2003.
Differding J, Welk GJ, Hart P, Abate J, Symington S: The use of the
Digi-Walker step counter to measure levels of physical activity (Abstract).
Med Sci Sports Exerc 30: S11, 1998.
Haymes EM, Byrnes WC:  Walking and running energy expenditure estimated by
Caltrac and indirect calorimetry. Med Sci Sports Exerc 25: 1365-1369, 1993.
Hedley AA, Ogden CL, Johnson CL, Carroll MD, Curtin LR, Flegal KM: Prevalence
of overweight and obesity among U.S. children, adolescents, and adults, 1999-2002.
JAMA 291: 2847-2850, 2004.
Eckel RH, York DA, Rossner S, Hubbard V, Caterson I, St Joer ST, Hayman LL,
Mullis RM, Blair AN:  American Heart Association Prevention Conference VII: Obesity,
worldwide epidemic related to heart disease and stroke: Executive summary.
Circulation 110: 2968-2975, 2004. 
English MJM, Farmer C, Scott WAC: Heat loss in exposed volunteers.
J Trauma 30:  422-425, 1990
Fehling PC, Smith DL, Warner SE, Dalsky GP:  Comparison of accelerometers
with oxygen consumption in older adults during exercise. Med Sci Sports
Exerc 31: 171-175, 1999.
Flegal KM, Carroll MD, Johnson CL: Overweight and obesity in the United States:
Prevalence and trends.  Int J Obes Relat Metab Disord 22: 39-47, 1998.
Fruin ML, Rankin JW: Validity of a multi-sensor armband in estimating rest and
energy expenditure.  Med Sci Sports Exerc 36: 1063-1069, 2004.
Gayle R, Montoye HJ, Philpot J: Accuracy of pedometers for measuring
distance walked. Res Q 48: 632-636, 1997.
Haymes EM, Byrnes WC: Walking and running energy expenditure estimated by
Caltrac and indirect calorimetry. Med Sci Sports Exerc 25: 1365-1369, 1993.
Irwin ML, Ainsworth BE, Conway JM: Estimation of energy expenditure from physical
activity measures: Determinants of accuracy. Obes Res 9: 517-525, 2001.
Jacobs DR Jr, Ainsworth BE, Hartman TJ, Leon AS: A simultaneous evaluation of 10
commonly used physical activity questionnaires. Med Sci Sports Exerc 25: 81-91, 1993.
Jakicic JM, Kang J, Robertson RJ, Wing RR, Goss FL: Validity of a portable heat sensing
system to measure energy expenditure. Med Sci Sports Exerc 25(6 Suppl): 1993.
Jakicic JM, Winters C, Lang W, Wing RR: Effects of intermittent exercise and use of
home exercise equipment on adherence, weight loss, and fitness in overweight women
a randomized trial. JAMA 282: 1554-1560, 1999.
Jakicic JM, Marcus M, Gallagher KI, Randall C, Thomas E, Goss FL, Robertson RJ:
Evaluation of the SenseWear Pro Armband to assess energy expenditure during
exercise.  Med Sci Sports Exerc 36: 897-904, 2004.
Janz KF: Use of heart rate monitors to assess physical activity.
In: Welk GJ (ed). Physical Activity Assessments for Health-Related Research.
Champaign, IL: Human Kinetics, 2002.
Keim NL, Blanton CA, Kretsch MJ: America’s obesity epidemic: Measuring physical
activity to promote an active lifestyle.  J Am Diet Assoc 104: 1398-1409, 2004.
Kerslake DMcK: The Stress of Hot Environments. Cambridge:
Cambridge University Press, 1972.
King GA, Torres N, Potter C, Brooks T, Coleman KJ: Comparison of activity monitors to
estimate energy cost of treadmill exercise.  Med Sci Sports Exerc 36: 1244-1251, 2004.
Klesges RC, Kiesges LM, Swenson AM, et al: (1985). A validation of two motion sensors
in the prediction of child and adult activity levels. Am J Epidemiol 122: 400-410, 1985.
Lawrence VJ, Kopelman PG: Medical consequences of obesity.
Clin Dermatol 4: 296-302, 2004.
Layton RP, Mints WH Jr, Annis JF, Rack MJ, Webb P:  Calorimetry with heat flux
transducers: Comparison with a suit calorimeter. J Appl Physiol 54: 1361-1367, 1993.
Leenders NYJM, Sherman WM, Nagaraja HN, Kien CL: Evaluation of methods to assess
physical activity in free-living conditions. Med Sci Sports Exerc 33: 1233-1240, 2001.
Le Masurier GC, Tudor-Locke C: Comparison of pedometer and accelerometer
accuracy under controlled conditions. Med Sci Sports Exerc 35: 867-871, 2003.
Le Masurier GC, Lee SM, Tudor-Locke C: Motion sensor accuracy under controlled
and free-living conditions. Med Sci Sports Exerc 36: 905-910, 2004.
Levine JA, Baukol PA, Westerterp KR: Validation of the Tracmor triaxial accelerometer
system for walking. Med Sci Sports Exerc 33: 1593-1597, 2001.
Levine JA: Non-exercise activity thermogenesis (NEAT).
Best Pract Res Clin Endocrinol Metab 16: 679-702, 2002.
Lifson N, Gordon GB, McClintock R: Measurement of total carbon dioxide
production by means of D2O18.  J Appl Physiol 7: 704-710, 1955.
Lohman TG, Roche AF, Martorell R:  Anthropometric Standardization
Reference Manual. Champaign, IL: Human Kinetics, 1991.
Macfarlane DJ: Automated metabolic gas analysis systems: A review.
Sports Med 31: 841-861, 2001.
Melanson EL Jr, Freedson PS: Validity of the Computer Science and
Applications, Inc. (CSA) activity monitor. Med Sci Sports Exerc 27: 934-940, 1995.
Mignault D, St-Onge M, Karelis AD, Allison DB, Rabasa-Lhoret R: Evaluation of
the Portable HealthWear Armband: a device to measure total daily energy expenditure
in free-living type 2 diabetic individuals.  Diabetes Care 28: 225-227, 2005.
Montoye HJ, Washburn R, Servais S, Ertl A, Webster JG, Nagle FJ: Estimation of
energy expenditure by a portable monitor.  Med Sci Sports Exerc 15: 403-407, 1983.
Montoye HJ: Measuring Physical Activity and Energy Expenditure.
Champaign, IL: Human Kinetics, 1996.
Nagy KA: CO2 production in animals: Analysis of potential errors in the doubly
labeled water method. Am J Physiol 238: R466-R473, 1980.
National Heart, Lung, and Blood Institute:  Clinical guidelines on the identification,
evaluation, and treatment of overweight and obesity in adults: the evidence report.
Obes Res 6(Suppl 2): 51S-209S, 1998.
Nichols JF, Morgan CG, Sarkin JA, Sallis JF, Calfas KJ: Validity, reliability,
and calibration of the Tritrac accelerometer as a measure of physical activity.
Med Sci Sports Exerc 31: 908-912, 1999.
Pambianco G, Wing RR, Robertson R: Accuracy and reliability of the Caltrac
accelerometer for estimating energy expenditure.  Med Sci Sports Exerc 22: 858-862, 1990.
Pereira MA, Fitzgerald SJ, Gregg EE, Joswiak ML, Ryan MJ, Suminski RR, Utter AC,
Zmuda JM: A collection of Physical Activity Questionnaires for health-related research.
Med Sci Sports Exerc 29(Suppl 6): S1-S205, 1997.
Pi-Sunyer FX: Comorbidities of overweight and obesity: current evidence and
research issues.  Med Sci Sports Exerc 31(11 Suppl): S602-S608, 1996.
Roberts SB: Use of doubly labeled water method for measurement of energy
expenditure, total body water, water intake, and metabolizable energy intake in
humans and small animals. C J Physiol Pharmacol 67: 1190-1198, 1989.
Schneider PL, Crouter SE, Bassett DR: Pedometer measures of free-living physical
activity: Comparison of 13 models. Med Sci Sports Exerc 36: 331-335, 2004.
Schoeller DA, Ravussin E, Schutz, Acheson KJ, Baertschi P, Jequier E:
Energy expenditure by doubly labeled water: Validation in humans and
proposed calculation. Am J Physiol 250: R823-R830, 1986.
Schoeller DA, van Santen E: Measurement of energy expenditure in humans by doubly
labeled water method. J Appl Physiol: Resp Environ Exer Physiol 53: 955-959, 1982.
Seale JL, Conway JM, Canary JJ: Seven day validation of doubly labeled water
method using indirect room calorimetry. J Appl Physiol 74: 402-409, 1993.
Starling RD, Matthews DE, Ades PA, Poehlman ET:  Assessment of physical activity in
older individuals: A doubly labeled water study. J Appl Physiol 86: 2090-2096, 1999.
Strath SJ, Swartz AM, Bassett DR, O'Brien WL, King GA, Ainsworth BE: Evaluation
of heart rate as a method for assessing moderate intensity physical activity.
Med Sci Sports Exerc 32(9 Suppl): S465-470, 2000.
Swan PD, Byrnes WC, Haymes EM: Energy expenditure estimates of the Caltrac
accelerometer for running, race walking, and stepping. Br J Sports Med 31: 235-239, 1997.
Tudor-Locke CE, Meyers AM: Methodological considerations for researchers
and practitioners using pedometers to measure physical (ambulatory) activity.
Res Q Exerc Sport 72: 1-12, 2001.
Washburn R, Chin MK, Montoye HJ:  Accuracy of pedometer in walking and running.
Res Q Exerc Sport 51: 695-702, 1980.
Welk GJ, Differding JA, Thompson RW, Blair SN, Dziura J, Hart P:
The utility of the Digi-walker step counter to assess daily physical activity patterns.
Med Sci Sports Exerc 32(9 Suppl): S481-S488, 2000.
Welk GJ, Blair SN, Woof K, Jones S, Thompson RW:
A comparative evaluation of three accelerometry-based physical activity monitors.
Med Sci Sports Exer, 32(9 Suppl): S489-S497, 2000.
Westerterp KR, Browns F, Saris WHM, ten Hoor F: Comparison of doubly labeled water
with respirometry at low- and high- activity levels. J Appl Physiol 65: 53-56, 1988.
Westerterp KR: Physical activity assessment with accelerometers.
Int J Obes Relat Metab Disorder 23(Suppl 3): S45-S49, 1999.
Winters C, Lagally K, Jakicic JM, Wing RR, Robertson RJ: Estimated energy
expenditure using KAL-X heat flux monitor during several modes of physical activity.
(Abstract).  Med Sci Sports Exerc 30(5 Suppl): S134, 1998.
Zhang K, Werner P, Ming Sun F, Pi-Sunyer X, Boozer CN: Measurement of
human daily physical activity. Obes Res 11: 33-38, 2003.
Zhang K, Pi-Sunyer FX, Boozer CN: Improving energy expenditure estimation
for physical activity. Med Sci Sports Exerc 36: 883-889, 2004.

Core-C MainLinksOnline ReviewsWorkshopsConsultation