J Occup Health 2010; 52: 167–175
Journal of
Occupational Health
Technique for Continuously Monitoring Core Body Temperatures
to Prevent Heat Stress Disorders in Workers Engaged in Physical
Labor
Chikage NAGANO1, Takao TSUTSUI1, Koichi MONJI2, Yasuhiro SOGABE2, Nozomi IDOTA1and
Seichi HORIE1
1
Department of Health Policy and Management, Institute of Industrial Ecological Sciences, University of
Occupational and Environmental Health, Japan and 2Climatic Chamber Laboratory, Bio-information Research
Center, University of Occupational and Environmental Health, Japan
Abstract: Technique for Continuously Monitoring
Core Body Temperatures to Prevent Heat Stress
Disorders in Workers Engaged in Physical Labor:
Chikage NAGANO, et al. Department of Health Policy
and Management, Institute of Industrial Ecological
Sciences, University of Occupational and
Environmental Health, Japan—Objectives:
Measuring core body temperature is crucial for
preventing heat stress disorders in workers. We
developed a method for measuring auditory canal
temperatures based on a thermocouple inserted into a
sponge-type earplug. We verified that the tip of this
thermocouple is positioned safely, allowing the wearer
to engage in normal physical tasks; this position
averaged 6.6 mm from the tympanic membrane.
Methods: To assess this technique, we had six healthy
male students repeat three cycles of exercise and rest
(20 min of exercise and 15 min of rest) in a temperaturecontrolled chamber with temperatures set at 25, 30, or
35°C, while monitoring the auditory canal, esophageal,
rectal, and skin temperatures. Results: We observed
differences of a mere 0.30–0.45°C between rectal
temperatures and auditory canal temperatures
measured with the thermocouple, the smallest such
difference reported to date in studies involving auditory
canal temperature measurement. Conclusions: We
conclude that monitoring temperatures based on a
technique involving an auditory canal plug can be used
to estimate rectal temperatures accurately, and thereby
Received Dec 17, 2009; Accepted Mar 9, 2010
Published online in J-STAGE Apr 5, 2010
Correspondence to: Chikage Nagano, Department of Health Policy
and Management, Institute of Industrial Ecological Sciences,
University of Occupational and Environmental Health, Japan, 1–
1 Iseigaoka Yahatanisi-ku, Kitakyushu City, Fukuoka 807-8555,
Japan (e-mail: [email protected])
to avoid conditions leading to heat stress disorders.
(J Occup Health 2010; 52: 167–175)
Key words: Auditory canal, Body temperature, Heat
stress disorders, Occupational health, Skin temperature
Although controlling ambient temperatures is the ideal
approach to eliminating heat stress disorders, this is often
not possible. This makes it especially important to
optimize work schedules by reducing hours worked,
providing workers with frequent breaks, or having
workers work in alternating shifts. Ideal work schedules
should be based on evaluations of physiological response,
including perspiration rates and heat balance, as defined
in the ISO7933 Predicted Heat Strain model1). In practice,
however, measuring the parameters required for this
model poses major difficulties. The indices available for
workplace evaluations include various proxies for core
temperature (tcr) measurements, including measurements
of skin temperature (tsk), heart rate (HR), and body weight.
Monitoring t cr in vivo is essential to preventing heat
disorders.
According to ISO 98862), the term “core” refers to all
tissue located at depths sufficient to remain unaffected
by the temperature gradient existing on surface tissue.
Esophageal temperature (tes), rectal temperature (tre), intraabdominal temperature (t ab), oral temperature (t or),
tympanic temperature (tty), auditory canal temperature (tac)
and urine temperature (tur) have been proposed as core
temperature indices. The transducer of tes is placed in
the lower part of the oesophagus, which is in contact over
a length of 50 to 70 mm with the front of the left auricle
and with the rear surface of the descending aorta.
Consequently, tes reflects the temperature of the arterial
blood with a very short reaction time. tre is independent
of ambient conditions because the rectum is surrounded
168
by a large mass of abdominal tissues with low thermal
conductivity. A transducer is inserted in the rectum to
measure tre. Measurements of tes and tre are generally too
invasive for workplace implementation. A transducer is
swallowed to measure tab. The record of t ab will vary
according to whether it is located in an area close to large
arterial vessels or to organs with high local metabolism
or near the abdominal wall. While tab can be measured
using CorTempTM 2000 (a core temperature telemetry
monitoring system in the form of an ingestible thermistor
capsule), such devices also tend to be in constant motion
within the digestive tract, and the results sometimes
require expert interpretation. The transducer of tty is
positioned as close as possible to the tympanic membrane
whose vascularisation is provided in part by the internal
carotid artery which also irrigates the hypothalamus.
However, as the tympanic membrane is also vascularised
by the external carotid artery, tty is influenced by local
thermal exchanges existing in the area vascularised by
this artery. Measurements of tty with contact type sensors
are associated with risk of injury to the eardrum;
measurements with non-contact types may measure
temperatures only in the external auditory canal. tur varies
in tandem with tre, except the time constant is somewhat
greater and its actual value is systematically lower than
tre by 0.2 to 0.5°C. Measurements of tur are possible only
during urination and are dependent on the quantity of
urine available in the bladder. This leaves oral
temperature tor and auditory canal temperature tac as the
only feasible indices of core temperature under real
workplace conditions.
Various studies have been performed to elucidate the
characteristics of tor and tac and to determine the superior
index of core temperature, with comparisons generally
involving a rectal temperature reference. Cooper3) et al.
have concluded that tac more closely tracks changes in
the carotid artery than tor and that tac is unaffected by
changes in ambient temperatures, unlike t or. The
transducer of tor is placed underneath the tongue and is
therefore in close contact with deep arterial branches of
the lingual artery. tor is affected by external conditions,
being dependent on whether the mouth is open or not.
After measuring tac by various methods, IH Muir et al.
reported that a method involving a thermistor covered
with ordinary ear plugs was most suitable (tre being equal
to tac+0.6°C when measured by the method4)). Darowski
et al. measured tac 1–2 mm from the tympanic membrane
in external auditory canals closed with cotton balls,
assuming negligible temperature gradients inside the
external auditory canal5). Comparing measurements of
tac at various distances from the tympanic membrane, JE
Greenleaf et al. found that tac at 10 mm is least affected
by ambient temperature and most closely matches tre. On
this basis, they proposed the following equation to
calculate rectal temperatures at ambient temperatures
J Occup Health, Vol. 52, 2010
between 5–35°C6):
tre=0.8tac + 0.2tsk (only between ambient temperatures
of 5–35°C)
Measuring tac and tre as subjects exercised in hot and
cold environments, Morgans et al. report that tac is 0.5–
0.9°C lower than tre7). These studies suggest tre can be
estimated from tac. However, Edwards et al. report that
t ac measured at a point 10 mm from the tympanic
membrane when the external auditory meatus is sealed
with a cotton ball cannot be used to predict tre, due to the
effects of changes in temperature on the head skin
surface8). Hansen9) et al. report that tty measured with a
contactless infrared sensor does not adequately reflect
t re, due to the effects of perspiration and changes in
ambient temperature on the surface of the surrounding
skin. Fuller et al. report that tac does not adequately reflect
tre when measured during athletic exertion10). In short,
many studies support the validity of tac as a proxy for tre,
while some studies identify limitations, especially when
temperatures differ between tsk at the head and tac.
As noted in ISO 9886, the external auditory canal is
vascularised by the external carotid artery, and cutaneous
blood flows around the ear and in adjacent areas of the
head affect ear canal temperatures. Cooper3) et al. report
gradually widening differences between tty and tac as the
point of measurement moved farther from the tympanic
membrane (e.g., 0.5°C at 2 mm and 0.83°C at 17 mm).
Noting this temperature gradient, ISO 9886 emphasizes
the importance of adequately insulating the ear and
placing the transducer less than 10 mm from the
tympanum. ISO 9886 also states that tac can be interpreted
to reflect tcr only when tac is measured within environments
that differ in air temperature by less than 10°C.
In light of these preceding studies, we developed a
new method for measuring tac that involves the secure
insertion of a temperature sensor through a sponge-type
earplug that seals the external auditory canal. We
expected this method would be easy to use and would
not cause undue discomfort, since the sensor is buried in
the earplug and secured in place, even when actual work
tasks are performed. Applying this method, we measured
tac continuously and compared our measurements with tre
as our subjects performed various physical tasks in
various warm environments, seeking to determine if
changes in tac can be used to track changes tre.
Methods
Subjects
The subjects were six healthy male volunteers, all
Japanese, aged 20.5 ± 1.8 yr (mean ± SD) and measuring
171.8 ± 6.4 cm in height, 60.5 ± 4.4 kg in weight, with
BMI values in the range of 20.5 ± 1.3 kg/m2. The subjects
reported no history of heat stroke in responses to a
questionnaire. All experiments were performed during
autumn to spring when the subjects were not acclimatized
Chikage NAGANO, et al.: Technique for Monitoring Core Body Temperatures in Workers
169
to hot environments. All subjects signed a written consent
form. The study was reviewed and approved by the ethics
committee of the University of Occupational and
Environmental Health, Japan.
Procedures
The subjects were asked to strip to the waist and to
wear only underwear and short pants. They were barred
from eating within 4 h prior to the experiments, but were
permitted to drink water during this time.
The experiments were performed in a climatic chamber
(Model TBL-15FW5CPX, Tabai Espec) installed at the
Bio-information Research Center at the University of
Occupational and Environmental Health, Japan, in
November 2003 and March 2004. The relative humidity
in this chamber was kept at 60%. The experiments were
repeated at three different dry-bulb temperature (t a)
settings: 25, 30, and 35°C.
The subjects were asked to remain at rest in the room
for at least 30 min before the experiments began to allow
tsk, tre and HR to stabilize. We assumed an actual workrest cycle occurring in a warm environment and asked
the subjects to sit still for 10 min, then to pedal a bicycle
ergometer (Corival 400, LODE) at a pace sufficient to
generate 75 W continuously for 20 min, then to rest in
the saddle for 15 min. This 20 min exercise session was
repeated twice with intervening 15 min rest periods,
followed by a 20 min rest period after the last exercise
session.
Measurements
We configured thermal sensors to measure t ac as
follows: First, we formed a small canal in an ear plug
(3M 1110, 3M Health Care) along the long axis; then we
inserted a copper-constantan thermocouple probe into this
canal, and stuck out the thermocouple tip within 1mm
from the inner side of the ear plug. The probe tip was
covered with a cotton ball measuring about 2 mm in
diameter (Fig.1). After gaining adequate experience with
inserting these ear plugs, we placed this probe in the
external auditory canals of our subjects and set it at the
deepest location. We verified the actual locations of the
tip of the probe through CT scans of the sphenoid bone
area in three subjects.
Following zero adjustments and calibration, we used
an analog digital converter (Remote Scanner DE1200,
NEC) to convert voltage fluctuations from the
thermocouple to digital signals and saved this data on a
personal computer (Mate NX, NEC) at 5-sec intervals.
We also measured and monitored t re, t es and t sk; and
temperatures at the right frontal plane of the forehead
(tforehead), right chest (tchest), right forearm (tforearm), the dorsal
surface of the right hand (thand), right thigh (tthigh), right
calf (tcalf), and the dorsal surface of the right foot (tfoot)
using copper-constantan thermocouple probes. We
Fig. 1. Thermocouple probe inserted in ear plug (above
figure) and CT image of its site in the external auditory
canal (below figure).
measured tre with a polyethylene-sealed thermocouple
probe measuring 4 mm in diameter and covered with a
YSI thermocouple probe cover (Nikkiso YSI) made of
latex rubber, which was inserted 12–15 cm beyond the
anal sphincter. We asked the subjects to swallow a
polyethylene-sealed tes probe measuring 2 mm in diameter
for positioning at the heart level and monitored maximum
values for tes. The mean skin temperature (tsk-mean) was
calculated from skin temperatures measured at seven
different points using the weighting factors described by
Hardy & Dubois et al.5) An electrocardiogram with a
telemetric system (Bioview1000, NEC) was used to
measure heart rate (HR).
Statistical analysis
One-factor ANOVA was used for the evaluation of the
effect and a stepwise regression model was performed to
estimate tre from tes or tac. All the statistical tests were carried
out with Stat view 5.0 for Windows (SAS Institute Inc.).
Results
Location of tac probe
The distance between the tympanum and the
170
J Occup Health, Vol. 52, 2010
Fig. 2. Trends in tsk and tsk-mean at various ta (N=6). tforehead=temperature of forehead; tchest=temperature of chest; tforearm=temperature
of forearm thand=temperature of hand; tthigh=temperature of thigh; tcalf=temperature of calf; tfoot=temperature of foot; tskmean=mean skin temperature calculated from the 7 skin temperatures using the weighting factors proposed by Hardy and
Dubois11); ta=ambient temperature ( : 25°C, : 30°C, and : 35°C).
Chikage NAGANO, et al.: Technique for Monitoring Core Body Temperatures in Workers
171
Fig. 3. HR trends at various t a (N=6). HR=heart rate (beats/min);
ta=ambient temperature ( : 25°C, : 30°C, and : 35°C).
thermocouple of the tac probe, as detected by CT, was
3.0, 10.0, and 6.8 mm, and its average was 6.6 ± 3.5 mm
(average ± SD).
Skin temperature tsk
Figure 2 shows the trends for seven different t sk and
t sk-mean when the ambient t a was 25, 30, or 35°C. The
figure was plotted by drawing straight lines between the
mean values of t sk and t sk-mean measured at 5-minute
intervals. Among all tsk, tforehead demonstrated the smallest
range of temperature distributions throughout the
experiment. In contrast, average tchest tended to decrease
as the exercise was repeated. This trend became more
pronounced when ta was elevated. The effects of exercise
were more evident in tsk of the lower regions of the body
ta had clear effects on tsk at the extremities: thand, tforearm,
tthigh, tcalf, and tfoot. Higher ta values were associated with
higher values for these parameters. Calculating overall
mean ± SD values for tsk-mean during the experiment, we
obtained 32.9 ± 0.14, 33.9 ± 0.16, and 36.0 ± 0.19°C,
respectively, at ta of 25, 30, and 35°C. The increase in
overall mean values was pronounced when ta rose from
30 to 35°C.
Heart rate
Figure 3 shows the trends in mean HR at 5-minute
intervals. HR rose and decreased rapidly following the
start and completion of exercise. The overall HR mean
± SD values calculated during the experiment were 93.9
± 19.4, 98.8 ± 18.2, and 109.0 ± 18.6 beats/min,
respectively, at t a of 25, 30, and 35°C, significantly
different by one-factor ANOVA (p<0.0001). Overall
mean values clearly rose as ta increased from 30 to 35°C.
The mean ± SD values calculated for HR during the three
exercise periods were 112.7 ± 5.4, 116.0 ± 6.6, and 126.0
± 8.0 beats/min, respectively, at ta of 25, 30, and 35°C.
The corresponding values during the four rest periods
were 76.5 ± 8.1, 82.8 ± 8.1, and 93.3 ± 10.0 beats/min.
The upward tendency of these values was more
pronounced when ta rose from 30 to 35°C.
Core temperature tcr
The overall mean ± SD values of tac throughout the
experiment were calculated to be 36.9 ± 0.12, 37.1 ± 0.15,
and 37.4 ± 0.15°C, respectively, at ta of 25, 30, and 35°C.
Elevation of mean t ac in accordance with t a was
pronounced, with a steeper gradient between 30°C and
35°C. The overall mean ± SD values for tes were 37.4 ±
0.20, 37.4 ± 0.21 and 37.6 ± 0.20°C; the corresponding
values for tre were 37.4 ± 0.15, 37.5 ± 0.21 and 37.7 ±
0.23°C, respectively, at ta of 25, 30, and 35°C. The
increase in mean tes and tre following the increase in ta
was less remarkable than with tac. Figure 4 illustrates the
trends in mean ± SD values for tac, tes, and tre at 5-minute
intervals during the experiment.
We calculated the differences between tre and tac and
compared the values for t re -t ac at various t a . The
corresponding mean ± SD values were 0.45 ± 0.08, 0.36
± 0.11, and 0.30 ± 0.12°C, respectively at ta of 25, 30,
and 35°C.
Table 1 provides a tabular summary of the mean ± SD
values and minimum–maximum values for tsk-mean, tac, tes,
tre, and tre-tac during each period of rest and exercise at
various values of ta.
Multivariate analysis shows that the average of tac, tes,
and tre tended to rise over time and with repeated exercise.
In contrast, mean -tsk over the same timeframe tended to
decline. Mean tre-tac grew larger over time and with
repeated exercise. As ta rose from 25 to 30°C and from
30 to 35°C, the differences between the two measures
narrowed.
172
J Occup Health, Vol. 52, 2010
Estimating tre from tes or tac
We applied a multiple regression model to estimate tre
at each ta using the two independent variables of time
from the initiation of each exercise period (T) and tes or
t ac, using all data obtained during the three exercise
periods. Then, we also calculated an overall equation to
estimate tre using three independent variables of ta, T, and
tes or tac. Table 2 gives the equations, the values assigned
to the various terms, and the squares of the correlation
coefficient (R2).
Higher ta corresponded to decreasing R2 between tre
and tac. However, this tendency was not observed with
the R2 between tre and tes. R2 between tre and tac was always
higher than R2 between tre and tes under all ta conditions.
We applied a stepwise regression model to estimate tre
with t a. Table 2 also provides the equations used to
calculate these results and the R2 values. It shows that
the correlation between tre and t ac was higher than that
between tre and tes.
Discussion
Simply measuring ambient temperatures is a relatively
ineffective safeguard against heat stress disorders.
Individual factors such as underlying health, lifestyle, and
genetic background can also affect the occurrence of heat
stress disorders. A more reliable method for monitoring
core body temperatures in the workplace would allow
detection of individuals at risk before the symptoms of
heat stress disorder actually became visible.
All previously reported methods for measuring auditory
canal temperatures various limitations with respect to
actual implementation in workplace settings. Greenleaf
et al. concluded that the distance of the thermocouple
from the tympanic membrane should be no more than 10
mm to ensure accurate estimates of core temperatures6).
The only article evaluating the use of auditory canal
temperatures in actual workplace settings concludes that
the method was a poor index of core temperature;
however, that report failed to report the precise positioning
of the thermocouple10). Muir et al. used infrared sensors
to measure auditory canal temperatures and observed
differences from rectal temperatures exceeding 0.6°C, but
failed to verify the location of the sensor4). Some studies
have placed the thermocouple within a few millimeters
of the tympanic membrane. Despite this proximity, with
the ear canal closed with cotton wool, Edwards concluded
Fig. 4. Changes over time in mean ± SD values of tac, tes, tre
and ∆t re =t re -t ac at various t a (N=6). t re =rectal
temperature ( ); t es =esophageal temperature ( );
t ac= a u d i t o r y c a n a l t e m p e r a t u r e ( ) ; ∆ t re–
t ac =differences between t re and t ac ; t a=ambient
:30°C, and
:35°C);
temperature ( :25°C,
SD=standard deviation.
35°C
30°C
25°C
35°C
30°C
25°C
35°C
30°C
25°C
35°C
30°C
25°C
35°C
0.385 ± 0.025
(0.34 – 0.44)
0.240 ± 0.012
(0.21 – 0.26)
0.174 ± 0.018
(0.14 – 0.22)
37.2 ± 0.020
(37.12 – 37.21)
37.1 ± 0.020
(37.02 – 37.10)
37.2 ± 0.008
(37.20 – 37.24)
37.1 ± 0.021
(37.05 – 37.13)
37.0 ± 0.015
(36.97 – 37.04)
37.1 ± 0.020
(37.06 – 37.16)
36.8 ± 0.009
(36.76 – 36.81)
36.8 ± 0.011
(36.81 – 36.85)
37.0 ± 0.020
(36.99 – 37.07)
32.6 ± 0.055
(32.54 – 32.72)
33.8 ± 0.022
(33.73 – 33.81)
35.8 ± 0.039
(35.76 – 35.92)
rest 0
0.339 ± 0.013
(0.31 – 0.37)
0.213 ± 0.029
(0.16 – 0.26)
0.120 ± 0.025
(0.08 – 0.17)
37.2 ± 0.104
(37.03 – 37.38)
37.2 ± 0.118
(37.01 – 37.37)
37.4 ± 0.102
(37.21 – 37.53)
37.3 ± 0.219
(36.93 – 37.66)
37.3 ± 0.191
(36.88 – 37.53)
37.5 ± 0.193
(37.06 – 37.70)
36.8 ± 0.110
(36.69 – 37.05)
37.0 ± 0.145
(36.77 – 37.20)
37.2 ± 0.123
(37.06 – 37.43)
32.8 ± 0.227
(32.46 – 33.20)
34.0 ± 0.129
(33.76 – 34.17)
36.1 ± 0.160
(35.86 – 36.33)
exercise 1
0.392 ± 0.040
(0.32 – 0.45)
0.282 ± 0.067
(0.17 – 0.38)
0.235 ± 0.072
(0.10 – 0.34)
37.4 ± 0.037
(37.27 – 37.42)
37.4 ± 0.010
(37.36 – 37.43)
37.6 ± 0.013
(37.53 – 37.61)
37.4 ± 0.116
(37.23 – 37.64)
37.3 ± 0.074
(37.23 – 37.53)
37.5 ± 0.086
(37.40 – 37.72)
37.0 ± 0.074
(36.84 – 37.09)
37.1 ± 0.071
(37.01 – 37.22)
37.3 ± 0.066
(37.24 – 37.45)
33.1 ± 0.100
(32.86 – 33.20)
34.0 ± 0.102
(33.86 – 34.20
36.0 ± 0.120
(37.72 – 36.34)
rest 1
0.420 ± 0.010
(0.39 – 0.44)
0.334 ± 0.022
(0.30 – 0.38)
0.248 ± 0.038
(0.20 – 0.34)
37.4 ± 0.096
(37.23 – 37.54)
37.5 ± 0.088
(37.33 – 37.60)
37.7 ± 0.066
(37.56 – 37.77)
37.5 ± 0.176
(37.16 – 37.75)
37.5 ± 0.145
(37.22 – 37.70)
37.7 ± 0.101
(37.40 – 37.82)
37.0 ± 0.099
(36.81 – 37.13)
37.1 ± 0.108
(36.98 – 37.30)
37.4 ± 0.098
(37.25 – 37.54)
33.0 ± 0.084
(32.83 – 33.15)
34.0 ± 0.135
(33.76 – 34.22)
36.1 ± 0.140
(35.72 – 36.21)
exercise 2
0.495 ± 0.037
(0.40 – 0.55)
0.403 ± 0.057
(0.31 – 0.50)
0.371 ± 0.068
(0.24 – 0.45)
37.5 ± 0.051
(37.40 – 37.57)
37.6 ± 0.041
(37.50 – 37.65)
37.8 ± 0.014
(37.76 – 37.81)
37.4 ± 0.133
(37.22 – 37.73)
37.4 ± 0.086
(37.26 – 37.66)
37.6 ± 0.083
(37.48 – 37.78)
37.0 ± 0.086
(36.86 – 37.13)
37.2 ± 0.096
(37.02 – 37.32)
37.4 ± 0.075
(37.31 – 37.54)
33.0 ± 0.093
(32.87 – 33.17)
33.8 ± 0.189
(33.61 – 34.23)
35.8 ± 0.158
(35.62 – 36.17)
rest 2
0.479 ± 0.023
(0.43 – 0.54)
0.421 ± 0.046
(0.36 – 0.51)
0.372 ± 0.037
(0.32 – 0.45)
37.5 ± 0.093
(37.32 – 37.61)
37.6 ± 0.079
(37.48 – 37.73)
37.8 ± 0.062
(37.74 – 37.96)
37.5 ± 0.172
(37.19 – 37.74)
37.6 ± 0.145
(37.27 – 37.79)
37.8 ± 0.115
(37.50 – 37.92)
37.0 ± 0.113
(36.83 – 37.15)
37.2 ± 0.123
(36.99 – 37.35)
37.5 ± 0.097
(37.32 – 37.61)
32.9 ± 0.060
(32.86 – 33.09)
33.9 ± 0.156
(33.60 – 34.11)
35.9 ± 0.162
(35.61 – 36.16)
exercise 3
0.561 ± 0.036
(0.47 – 0.62)
0.503 ± 0.044
(0.38 – 0.56)
0.450 ± 0.041
(0.34 – 0.50)
37.5 ± 0.104
(37.30 – 37.65)
37.6 ± 0.077
(37.52 – 37.77)
37.9 ± 0.053
(37.81 – 38.00)
37.4 ± 0.151
(37.16 – 37.73)
37.5 ± 0.118
(37.33 – 37.80)
37.6 ± 0.101
(37.54 – 38.00)
36.9 ± 0.127
(36.76 – 37.14)
37.1 ± 0.110
(37.01 – 37.35)
37.5 ± 0.083
(37.38 – 37.64)
32.9 ± 0.089
(32.70 – 33.09)
33.8 ± 0.111
(33.65 – 34.11)
35.8 ± 0.104
(35.69 – 36.18)
rest 3
tsk-mean=mean skin temperature calculated from 7 skin temperatures using the weighting factors proposed by Hardy and Dubois 11). t ac=auditory canal temperature;
tes=esophageal temperature; tre=rectal temperature; ∆tre-tac=differences between tre and tac; ta=ambient temperature (25, 30, and 35°C); SD=standard deviation.
∆tre–tac (mean ± SD)
(minimum–maximum)
tre (mean ± SD)
(minimum–maximum)
tes (mean ± SD)
(minimum–maximum)
tac (mean ± SD)
(minimum–maximum)
25°C
tsk (mean ± SD)
(minimum–maximum)
30°C
ta
Body temperature (°C)
Table 1. Mean tsk-mean, tac, tes, tre, and ∆tre–tac for periods of rest and exercise at various values of ta (N=6)
Chikage NAGANO, et al.: Technique for Monitoring Core Body Temperatures in Workers
173
174
J Occup Health, Vol. 52, 2010
Table 2. Estimates of tre based on tes or tac (N=6)
ta
25°C
30°C
35°C
ta
Calculations
p value
tre=1.096tac–3.108
tre=0.52tes+17.906
tre=1.089tac–2.968
tre=0.768tes+8.677
tre=1.227tac–8.221
tre=0.798tes+7.640
tre=4.984+0.879tac+0.003T–0.011ta
tre=24.212+0.33tes+0.022T–0.00007169ta
<0.0001***
<0.0001***
<0.0001***
<0.0001***
<0.0001***
<0.0001***
<0.0001***
<0.0001***
R2
0.811
0.488
0.746
0.720
0.684
0.517
0.967
0.890
tre=rectal temperature; tes=esophageal temperature; tac=auditory canal temperature; ta=ambient temperature
(25, 30, and 35°C); T=time length after the initiation of exercise (min); p value=calculated by a stepwise
regression model, R2=squares of the correlation coefficient of these equations.
that temperatures did not accurately track rectal
temperatures8). Morgan observed differences from rectal
temperatures exceeding 0.5°C7). Hansen also observed
differences exceeding 1.0°C even when using molded
medical-use earplugs 9) . We hypothesize that these
temperature discrepancies are due to a narrow open space
around the thermocouple inserted alongside the earplug,
which we eliminated by inserting the thermocouple into
a sponge-type earplug. To avoid injury, the thermocouple
cannot be positioned too close to the tympanic membrane.
Ideally, according to Greenleaf, measurements of auditory
canal temperatures should be performed within 10 mm
of the tympanic membrane6), to minimize divergence from
rectal temperatures while ensuring safety in actual
workplace settings. Our method for inserting the
thermocouple and placing it at the tip of the earplug meets
their recommendations and the criteria specified by ISO
9886 for measuring core temperatures via auditory canal
temperatures 2). Our present study is the only study
featuring verified thermocouple positioning to report
differences from rectal temperatures of less than 0.5°C.
The auditory canal temperatures measured in this
experiment successfully tracked rectal temperatures, with
average differences of 0.30–0.45°C, less than in any other
study. This new method minimizes the temperature
gradient inside the auditory canal by tightly sealing the
canal. We measured auditory canal temperatures for a
period of 120 min, longer than any other past study.
Idealistically, we should have asked the subjects to
continue exercise and follow up the trend of tac until their
tre reached 38.5°C, the maximal allowable limit of trained
workers. However, we curtailed the heat exposure at
120 min to avoid excessive health risks to the subject.
We encountered no problems or breakdowns in the
apparatus or discomfort reported by the subjects. We
measured the temperature at 5-sec intervals, the shortest
interval reported in the studies cited above, confirming
that this method enables monitoring of minimal changes
in body temperature as an early sign of heat stress
disorder.
When ambient temperatures rose, auditory canal
temperatures closely followed rectal temperatures,
indicating the method we used is an ideal early detector
of heat stress disorders in workers engaged in physical
labor in hot environments. We observed narrowing
divergence between rectal temperatures and auditory
canal temperatures as subjects repeated the exercise
sessions, indicating that auditory canal temperatures may
accurately track rectal temperatures during physical
exertion. In a unique feature of this study, subjects were
asked to perform exercises at 15-minute intervals, since
time-consuming physical tasks in hot working
environments are rare in actual workplace settings
(workers in such settings are generally allowed breaks).
We observed increasing differences between rectal
temperature and auditory canal temperature as the subjects
repeated the exercise. However, during the course of the
experiment, the effects of exercise and ambient
temperatures tended to diminish, enabling consistent
estimates of rectal temperature based on temperatures
measured with the ear plug sensors. Sufficient repetition
of this procedure should result in a value for the difference
between the two temperatures that approaches a constant.
Such conditions are often encountered in real workplace
settings, and auditory canal temperatures represent a
promising reliable proxy for measurements of rectal
temperature.
Based on the equation developed to estimate rectal
temperatures based on auditory canal temperatures and
taking ambient temperatures into account, we calculated
the permissible maximum working times to avoid core
temperatures of 38.5°C, the threshold temperature
associated with high risk of heat stress disorder. In this
experiment, we also monitored tes to evaluate whether it
might react as a more sensitive indicator for heat disorder
of the subjects. In general, it reflects the temperature of
Chikage NAGANO, et al.: Technique for Monitoring Core Body Temperatures in Workers
the arterial blood with a very short reaction time; however,
it is easily affected by salivary temperature from
swallowing, especially at lower t a and, moreover, it
fluctuates based on the rapid change of tsk, especially after
the exposure to ta lower than 24°C21). Therefore, we
decided to use tre rather than tes as an index of core body
temperature of the subjects in this experiment.
Despite the limited number of subjects in this
experiment, the relationship among the three core
temperatures during exercise in three different
environments was similar for all the individuals. We
believe measurements of auditory canal temperatures
would also prove valid for the general population.
Although we did not evaluate the effects of movement of
the upper limbs on auditory canal temperatures, work
described in a previous report12) indicates such movements
would actually reduce discrepancies between auditory
canal and rectal temperatures.
Acknowledgments: We thank the members of our
department whose enormous support and insightful
comments were invaluable during the course of this study.
We would also like to thank Suikokai General Hospital
which cooperated in taking the elaborate CT images.
References
1) IOS 7933. Hot environments-analytical determination
and interpretation of thermal stress using calculation
of required sweat rate. Geneva: International
Organizaton for Standardization (ISO), 1989
2) ISO 9886. Evaluation of thermal strain by physiological
measurements. Geneva: International Organization for
Standardization (ISO), 1992.
3) Cooper KE, Cranston WI, Snell ES. Temperature in
the external auditory meatus as an index central
temperature changes. J. Appl. Physiol 1964; 19: 1032–
5.
4) Muir IH, Bishop PA, Lomax RG, Green JM. Prediction
of rectal temperature from ear canal temperature.
Ergonomics 2001; 44: 962–72.
5) Darowski A, Weinberg JR, Guz A. Normal rectal,
auditory canal, sublingual and axillary temperatures
in elderly afebrile patients in a warm environment. Age
Ageing 1991; 20: 113–9.
6) Greenleaf JE, Castle BL. External auditory canal
temperature as an estimate of core temperature. J Appl
Physiol 1972; 32: 194–8.
7) Morgans LF, Nunneley SA, Stribley RF. Influence of
ambient and core temperatures on auditory canal
temperature. Aviat Space Environ Med 1981; 52: 291–
3.
8) Edwards RJ, Belyavin AJ, Harrison MH. Core
temperature measurement in man. Aviat Space Environ
Med 1978; 49: 1289–94.
9) Hansen RD, Amos D, Leake B. Infrared tympanic
temperature as a predictor of rectal temperature on
warm and hot conditions. Aviat Space Environ Med
175
1996; 67: 1048–52.
10) Fuller A, Oosthuyse T, Maloney SK, Mitchell D.
Evaluation of miniature data loggers for body
temperature measurement during sporting activities.
Eur J Appl Physiol Occup Physiol 1999; 79: 341–6.
11) Hardy JD, Dubois F. The technique of measuring
radiation and convection. J Nutrition 1938; 15: 461–
75.
12) Price MJ, Campbell IG. Thermoregulatory responses
during prolonged upper-body exercise in cool and
warm conditions. J Sports Sci 2002; 20: 519–27.
13) Regan JM, Macfarlane DJ, Taylor NA. An evaluation
of the role of skin temperature during heat adaptation.
Acta Physiol Scand 1996; 158: 365–75.
14) Chwalbinska-Moneta J, Hanninen O. Effect of active
warming-up on thermoregulatory, circulatory, and
metabolic responses to incremental exercise in
endurance-trained athletes. Int J Sports Med 1989; 10:
25–9.
15) Almirall H, Ferrer R, Sanchez-Turet M. External
auditory canal temperature and reaction time
relationship during long performances. Int J
Psychophysiol 1988; 6: 215–20.
16) Randle IP, Legg SJ. A comparison of the effects of
mixed static and dynamic work with mainly dynamic
work in hot conditions. Eur J Appl Physiol Occup
Physiol 1985; 54: 201–6.
17) Nunneley SA, Flick CF. Heat stress in the A-10 cockpit:
flights over desert. Aviat Space Environ Med 1981;
52: 513–6.
18) Kabayashi K, Horvath SM, Diaz FJ, Bransford DR,
Drinkwater BL. Thermoregulation during rest and
exercise in different postures in a hot humid
environment. J Appl Physiol 1980; 48: 999–1007.
19) Richardson JE, Capra MF. Physiological responses of
firefighters wearing level 3 chemical protective suits
while working in controlled hot environments. J Occup
Environ Med 2001; 43: 1064–72.
20) Ayling JH. Regional rates of sweat evaporation during
leg and arm cycling. Br J Sports Med 1986; 20: 35–7.
21) Horie S, Tsutsui T, Sakata S, Monji K, Sogabe Y.
Optimum room temperature during rest periods
between repetitive exercises under heat stress.
Environmental Ergonomics, Elsevier Ergonomics Book
Series Volume 3; 2005: 85–90.
22) Keatinge WR, Sloan RE. Deep body temperature from
aural canal with servo-controlled heating to outer ear.
J Appl Physiol 1975; 38: 919–21.
23) Moran DS, Mendal L. Core temperature measurement:
methods and current insights. Sports Med 2002; 32:
879–85. Review.
24) Lee SM, Williams WJ, Fortney Schneider SM. Core
temperature measurement during supine exercise:
esophageal, rectal, and intestinal temperatures. Aviat
Space Environ Med 2000; 71: 939–45.
25) Gass EM, Gass GC. Rectal and esophageal
temperatures during upper- and lower-body exercise.
Eur J Appl Physiol Occup Physiol 1998; 78: 38–42.