Original Article
Industrial Health 2000, 38, 366–371
Thermal Perception Threshold Testing for the
Evaluation of Small Sensory Nerve Fiber Injury
in Patients with Hand-Arm Vibration Syndrome
Norikuni TOIBANA1, Hisataka SAKAKIBARA2*, Mamoru HIRATA3,
Takaaki KONDO4 and Hideaki TOYOSHIMA4
1
Tokushima Kensei Hospital, 4–9–1, Shimosuketou-machi, Tokushima 770-0963, Japan
Nagoya University School of Health Sciences, 1–1–20, Daiko-minami, Higashi-ku, Nagoya 461-8673, Japan
3
Osaka Prefectural Institute of Public Health, 1–3–69, Nakamichi, Higashinari-ku, Osaka 537-0025, Japan
4
Nagoya University School of Medicine, 65 Tsurumai-cho, Showa-ku, Nagoya 466-8550, Japan
2
Received May 8, 2000 and accepted July 17, 2000
Abstract: The aim of the present study was to investigate whether thermal perception threshold
testing is a useful method that could replace pain threshold testing in the evaluation of small sensory
nerve fiber injury in vibration-induced neuropathy. Vibration, pain, and thermal (warm and cold)
perception thresholds were examined on both middle fingers of 50 patients with hand-arm vibration
syndrome and 29 healthy controls of similar age. The patients were divided into three subgroups
according to the Stockholm Workshop sensorineural scale. Thermal (warm and cold) thresholds as
well as vibration and pain thresholds were significantly more deteriorated among the patients than
in the controls. Among the patients, warm thresholds elevated and cold thresholds lowered according
to the Stockholm Workshop scale. Thermal thresholds were significantly correlated with pain
thresholds, and the sensitivity of the thermal threshold testing tended to be greater than that of the
pain threshold testing. The present findings indicate that thermal threshold testing for warm and
cold perception can be a useful substitute for pain threshold testing to examine small nerve fiber
injury in vibration-induced neuropathy.
Key words: Thermal threshold, Vibration-induced neuropathy, Pain threshold, Sensory nerve
Introduction
Peripheral nerve impairments in the hand are often
encountered among workers exposed to hand-transmitted
vibration. These neurological abnormalities include
numbness and tingling of the fingers, damaged skin sensory
perception, reduced grip force, and occasionally impaired
manipulative dexterity. Histopathological studies have
demonstrated demyelinating neuropathy with severe
destruction and loss of the myelin sheath, and remarkable
loss of nerve fibers in finger skin specimens from vibrationexposed subjects1, 2). Nerve conduction measurements have
*To whom correspondence should be addressed.
shown reduced sensory and motor nerve conduction velocities
in vibration-exposed hands3, 4), which are associated with
impaired vibration perception thresholds and reduced grip
force5, 6).
Vibration perception threshold testing is widely accepted
as a useful measure for screening or diagnosing vibrationinduced sensory nerve impairments7). Such testing is thought
to evaluate the function of large sensory nerve fibers such
as myelinated A-β fibers. On the other hand, some studies
indicate that thermal perception is also damaged among
vibration-exposed workers, implying that small nerve fibers,
including unmyelinated C fibers and myelinated A-δ fibers,
are impaired in vibration-induced neuropathy8, 9). An animal
experiment with rats has demonstrated that vibration exposure
THERMAL PERCEPTION IN HAND-ARM VIBRATION SYNDROME
produced transient changes in non-myelinated fibers of the
plantar nerve as well10). Recently, Stockholm Workshop 94
has recommended thermal threshold testing as a suitable
measure of the function of unmyelinated C fibers and
myelinated A-δ fibers7).
In Japan, vibration and pain perception thresholds are
examined in order to evaluate vibration-induced sensory
nerve injury. However, there are some questions about pain
threshold testing from the ethical point of view7), so that it
would be better to find a substitute for it. Thermal threshold
testing is one candidate for such a substitute, because it is
supposed to examine the function of small sensory nerve
fibers, just as in pain threshold testing. Hence, the present
study investigated whether thermal threshold testing could
be a useful substitute for pain threshold testing in the
evaluation of vibration-induced neuropathy. In this study,
the authors examined thermal (warm and cold) thresholds
as well as vibration and pain thresholds of patients with handarm vibration syndrome and healthy controls, and compared
the effectiveness of these sensorineural tests.
Materials and Methods
The subjects were 50 male patients with hand-arm vibration
syndrome and 29 healthy male controls. The patients had
been under treatment for hand-arm vibration syndrome, and
the controls were healthy volunteers who had not been
occupationally exposed to hand-arm vibration. These subjects
were selected from among those without complications causing
peripheral neuropathies such as diabetic, chemical and druginduced neuropathies, cerebral vascular diseases, other
neurological diseases, or severe injuries to the hand. Subjects
were also limited to those who consumed less than 80 ml a
day of alcohol equal to ethanol.
There was no significant difference in age between the
patients and the controls (mean ± standard deviation: 60.3 ±
3.5 and 60.3 ± 3.3, respectively). The patients had VWF (n=39,
78%), finger numbness or tingling (n=41, 82%), coldness in
the hands (n=48, 96%), and joint pain in the hand and arm
(n=40, 80%). The controls did not have VWF, finger numbness
or tingling, but some complained of coldness in the hand (n=4,
14%) and joint pain in the hand and arm (n=6, 21%).
Thermal (warm and cold) perception thresholds were
measured on the palmar surface of the distal phalanx of both
middle fingers using a thermo-esthesiometer11). The plate
temperature of the thermo-esthesiometer was initially set
to the finger skin temperature of each subject, which had
been measured just before using an infrared thermometer
(IT340S, Horiba Manufacturing, Japan), and then it
367
automatically increased or decreased at the rate of 0.2°C/s.
Warm or cold perception thresholds were determined as the
temperature at which each subject started to feel warm or
cold in the finger tested. The measurements were performed
twice in a random order. The value closer to the initial skin
temperature was adopted as the subject’s threshold. The
values of warm and cold thresholds and differences between
warm threshold and finger temperature, between cold
threshold and finger temperature, and the neutral zone
between warm threshold and cold threshold were evaluated
as indices of thermal perception testing.
Vibration perception thresholds at 125 Hz and pain
perception thresholds were also measured on the same finger.
The former was done with a vibrometer (AU-02B, Rion,
Japan). The level of vibration generated by the vibrometer
can be changed from − 10 dB to 40 dB in increments of
2.5 dB. The 0 dB is set at 0.218 m/s2 r.m.s for 125 Hz. The
latter was measured using a syringe needle (gauge 23) with
weights of 1, 2, 3, 5, 7, 10 or 15 grams in each syringe.
These measurements were carried out in summer under
room temperatures between 26°C and 28°C to keep fingers
warm. About four subjects, including both patients and
controls, were examined in the morning and another four
subjects were tested in the afternoon. Subjects rested in the
test room for about 20 or 30 min before they were examined.
All subjects were given full explanations about the present
study and consented to participation.
Examinations were made in the following order. First,
the palmar surface skin temperature of both middle fingers
was measured using an infrared thermometer (IT340S, Horiba
Manufacturing, Japan). As the finger skin temperature of a
few patients was lower then 30°C, their hands were warmed
with a heater to above 30°C. Second, vibration perception
thresholds at 125 Hz were measured. Third, pain perception
thresholds were examined, and thermal perception thresholds
were tested last. The same physician examined all subjects.
The differences in test values between the patients and
the controls were compared with Student’s t-test, χ2 test, or
Fisher’s exact test. Comparisons among subgroups of the
patients classified according to the Stockholm Workshop
sensorineural scale12) were made with multiple comparison
by Scheffe’s method.
Results
Comparison of examination results of patients and controls
As shown in Table 1, both warm and cold thresholds in
the patients were significantly more deteriorated than in the
controls (p<0.01). The differences between warm threshold
368
N TOIBANA et al.
Table 1. Examination results of patients with hand-arm vibration
syndrome and healthy controls
Right middle finger
Skin temperature (°C)
Warm threshold (°C)
Cold threshold (°C)
WPT − FST (°C)
CPT − FST (°C)
WPT − CPT (°C)
Vibration threshold (≥10 dB)
Pain threshold (≥5 g)
Left middle finger
Skin temperature (°C)
Warm threshold (°C)
Cold threshold (°C)
WPT − FST (°C)
CPT − FST (°C)
WPT − CPT (°C)
Vibration threshold (≥10 dB)
Pain threshold (≥5 g)
Vibration syndrome
patients (n=50)
Controls
(n=29)
32.0 ± 1.1
43.7 ± 3.6**
18.7 ± 6.8**
11.6 ± 3.5**
13.3 ± 6.6**
25.0 ± 9.0**
49 (98%)**
43 (86%)**
32.5 ± 1.3
35.6 ± 2.6
29.4 ± 3.3
3.2 ± 2.3
3.0 ± 2.9
6.2 ± 4.1
3 (10%)
4 (14%)
32.1 ± 1.1
43.4 ± 3.2**
18.7 ± 6.6**
11.3 ± 3.2**
13.6 ± 6.9**
25.0 ± 9.0**
50 (100%)**
41 (82%)**
32.4 ± 1.2
35.3 ± 1.9
29.4 ± 2.4
2.9 ± 1.7
3.0 ± 2.3
5.9 ± 3.1
3 (10%)
4 (14%)
**p<0.01: significant difference between patients and controls. WPT,
CPT and FST stand for warm perception threshold, cold perception
threshold, and finger skin temperature, respectively.
and finger temperature and between cold threshold and finger
temperature, and the neutral zone between warm threshold
and cold threshold were also significantly greater in the
patients (p<0.01). Finger skin temperatures did not differ
significantly between the patients and the controls.
Significantly increased vibration and pain thresholds were
also encountered in the patients (p<0.01). The cut-off value
for vibration threshold in the present study was set at 10 dB
or more, and that for pain threshold was set at 5 g and over.
These were determined as the point where specificity was
around 90% in the control subjects under study.
Examination results according to Stockholm Workshop
sensorineural scale
The patient subjects were divided into three subgroups
according to the Stockholm Workshop sensorineural (SN)
scale. In this study, the scale was defined as follows: 0 SN
(without numbness), 1SN (with numbness but without
reduced sensory perception as defined in the 2+3 SN
category), and 2+3 SN (with numbness and impaired sensory
perception of 15 dB or more in vibration threshold and 10 g
and over in pain threshold).
Age and finger temperature did not differ among the three
subgroups (Table 2). With higher SN scale, warm thresholds
Table 2. Examination results of patients with hand-arm vibration syndrome according to
Stockholm Workshop sensorineural (SN) scale
0 SN
1 SN
2+3 SN
Right middle finger
Age (years)
Skin temperature (°C)
Warm threshold (°C)
Cold threshold (°C)
WPT − FST (°C)
CPT − FST (°C)
WPT − CPT (°C)
Vibration threshold (≥10 dB)
Pain threshold (≥5 g)
(n=13)
60.6 ± 2.6
32.1 ± 1.1
41.1 ± 4.1
22.3 ± 5.4
9.0 ± 3.9
9.8 ± 5.2
18.7 ± 8.5
13 (100%)
9 (69%)
(n=18)
61.0 ± 3.3
32.1 ± 1.3
43.7 ± 2.7
19.4 ± 6.7
11.7 ± 2.6
12.7 ± 6.8
24.4 ± 7.7
17 (94%)
15 (83%)
(n=19)
59.5 ± 3.6
32.0 ± 1.1
45.4 ± 3.0*
15.6 ± 6.5*
13.4 ± 3.1*
16.7 ± 6.1*
29.8 ± 7.9*
19 (100%)
19 (100%)*
Left middle finger
Age (years)
Skin temperature (°C)
Warm threshold (°C)
Cold threshold (°C)
WPT − FST (°C)
CPT − FST (°C)
WPT − CPT (°C)
Vibration threshold (≥10 dB)
Pain threshold (≥5 g)
(n=16)
61.3 ± 2.8
32.4 ± 1.1
42.4 ± 3.4
21.0 ± 6.0
10.0 ± 3.3
11.4 ± 5.9
21.4 ± 8.3
16 (100%)
11 (69%)
(n=18)
60.7 ± 3.3
32.1 ± 1.1
43.2 ± 3.3
20.3 ± 6.4
11.1 ± 3.1
11.9 ± 6.5
22.9 ± 8.4
18 (100%)
14 (78%)
(n=16)
58.9 ± 3.6
31.7 ± 1.2
44.7 ± 2.4
14.7 ± 5.9*
13.0 ± 2.7*
17.9 ± 6.7*
30.8 ± 8.5*
19 (100%)
16 (100%)
*p<0.05; statistical difference from 0 SN. WPT, CPT and FST stand for warm perception threshold,
cold perception threshold, and finger skin temperature, respectively.
Industrial Health 2000, 38, 366–371
369
THERMAL PERCEPTION IN HAND-ARM VIBRATION SYNDROME
Table 3. Specificity and sensitivity of thermal, vibration and pain
perception threshold testing
Specificity
Sensitivity
89.7%
89.7%
89.7%
93.1%
93.1%
86.2%
89.7%
86.2%
89.7%
93.1%
94.0%
94.0%
94.0%
94.0%
92.0%
96.0%
90.0%
96.0%
90.0%
96.0%
Vibration perception threshold (≥10 dB)
Right finger
Left finger
89.7%
89.7%
98.0%
100.0%
Pain perception threshold (≥5 g)
Right finger
Left finger
86.2%
86.2%
86.0%
82.0%
Thermal perception threshold
Warm threshold (≥38.5°C)
Cold threshold (≤27°C)
WPT − FST (≥5.5°C)
CPT − FST (≥5.5°C)
WPT − CPT (≥12°C)
Right
Left
Right
Left
Right
Left
Right
Left
Right
Left
WPT, CPT and FST stand for warm perception threshold, cold perception
threshold, and finger skin temperature, respectively.
increased and cold thresholds decreased. The differences
between warm threshold and finger temperature and between
cold threshold and finger temperature, and the neutral zone
between warm threshold and cold threshold, tended to be
greater according to the SN scale. The warm and cold
thresholds and these differences were significantly different
between the 0 SN and the 2+3 SN (p<0.05).
Specificity and sensitivity of thermal threshold, vibration
threshold and pain threshold testing
As the cut-off value was set at the point where specificity
was around 90% in the control subjects, the sensitivity of
warm and cold threshold testing in the patients was 94%
(Table 3). On the other hand, the sensitivity of vibration
threshold testing was 98% and 100% for the right and left
hand, respectively, and that of pain threshold testing was
86% and 82%. Thus, the sensitivity of thermal threshold
testing tended to be greater than that of pain threshold testing.
Correlation of thermal thresholds with vibration threshold
and pain threshold
Table 4 shows Spearmann’s correlation coefficients of
warm and cold thresholds with vibration and pain thresholds.
Thermal thresholds (both warm and cold thresholds) were
more closely linked with pain thresholds than with vibration
thresholds.
Table 4. Spearman’s correlation coefficients among thermal,
vibration and pain perception thresholds
Right middle finger
Warm threshold (°C)
Cold threshold (°C)
WPT − FST (°C)
CPT − FST (°C)
WPT − CPT (°C)
Left middle finger
Warm threshold (°C)
Cold threshold (°C)
WPT − FST (°C)
CPT − FST (°C)
WPT − CPT (°C)
Vibration threshold
Pain threshold
0.439**
− 0.157
0.469**
0.204
0.350**
0.457**
− 0.215
0.429**
0.285*
0.390**
0.206
− 0.100
0.165
0.171
0.193
0.428**
− 0.349**
0.398**
0.456*
0.468**
**p<0.01, *p<0.05. WPT, CPT and FST stand for warm perception
threshold, cold perception threshold, and finger skin temperature,
respectively.
Discussion
The present study showed that thermal perception thresholds
for warmth and cold as well as vibration and pain perception
thresholds were significantly more deteriorated among patients
with hand-arm vibration syndrome than in healthy controls,
which agrees with previous studies8, 9, 13–16). These findings
suggest that both small and large sensory nerve fibers are
impaired in vibration-induced neuropathy. Animal experiments
have also demonstrated damage to small and large nerve fibers
in the vibration-exposed extremities10, 17, 18).
The present study also demonstrated that indices of thermal
thresholds tended to worsen among patients according to
the Stockholm Workshop sensorineural scale; a deterioration
of thermal thresholds in the 2+3 SN (moderate and severe)
group were significantly greater than in the 0 SN group.
This finding indicates that thermal threshold testing could
reflect the severity of the nerve damage. Hirosawa et al.8)
similarly reported that thermal thresholds were deteriorated
in vibration-exposed patients with advanced nerve damage.
The results by Strömberg et al.16) showed that temperature
sense in vibration-exposed workers did not have a clear
tendency to worsen with higher Stockholm Workshop scale,
but it was likely to be more severely impaired in vibrationexposed workers with numbness than in those without it.
In the present study, a significant correlation was found
between thermal and pain thresholds, while such relations
were not as close between thermal and vibration thresholds.
This is physiologically reasonable because both thermal and
pain sensations are mediated through unmyelinated C-fibers
370
and myelinated A-δ fibers, and heat above 45°C and cold
below 15°C are perceived as heat pain and cold pain via
pain nerve fibers as well19). It, therefore, seems that thermal
threshold testing could be a substitute for pain threshold
testing in examining the impairments of small sensory nerve
fibers. In fact, the greater sensitivity of thermal threshold
testing in the present study suggests that the thermal threshold
testing might even be superior to pain threshold testing.
There are some differences among studies concerning the
sensitive indices of thermal thresholds testing for the
evaluation of vibration-induced nerve impairments. Some
studies pointed out that the neutral zone between the warm
and cold threshold was a sensitive indicator of nerve
damage8, 9), while others reported that cold thresholds were
more useful than warm or heat thresholds in the detection
of vibration-induced neuropathy9, 13). One study indicated
that heat threshold testing did not detect any significant
impairment in vibration-exposed workers20). On the other
hand, others showed that warm perception in particular were
impaired in vibration-exposed workers8, 16). In the present
study, the patients were almost equally damaged in warm
and cold thresholds and the neutral zone, judging from the
sensitivity in the tests. These differences in the sensitive
indices among studies could result from differences in the
subjects examined. For example, it is supposed that patients
with severe nerve injury will have abnormalities in almost
all indices of thermal thresholds. It is possible that the present
patients were such severe cases because most of them had
increased vibration thresholds. Another possibility is that
the differences are due to differences in the methods of
measuring thermal thresholds, as some studies have indicated
that thermal thresholds are affected by the starting
temperature21, 22). Further studies are necessary to determine
a sensitive indicator of thermal threshold testing for
evaluating vibration-induced neuropathy.
As mentioned above, vibration and pain thresholds are
measured in Japan to evaluate vibration-exposed neuropathy,
but some controversy still surrounds the ethics of pain
threshold testing. The present study has demonstrated that
thermal threshold testing for warm and cold perception can
reflect the severity of vibration-induced neuropathy and
would be a useful substitute for pain threshold testing.
Measurements of vibration thresholds for detecting the
dysfunction of large nerve fibers, and thermal thresholds
for testing impairments in small nerve fibers would be suitable
for the evaluation of vibration-exposed nerve impairments.
N TOIBANA et al.
Acknowledgements
We are much obliged to Prof. Hideki Mito of Kansai
Welfare University for the use of his thermo-esthesiometer.
This study was partly supported by a grant-in-aid for scientific
research from the Japan Ministry of Education, Science, and
Culture.
References
1) Takeuchi T, Futatsuka M, Imanishi H, Yamada S (1986)
Pathological changes observed in the finger biopsy of
patients with vibration-induced white finger. Scand J
Work Environ Health 12, 280–3.
2) Takeuchi T, Takeya M, Imanishi H (1988)
Ultrastructural changes in peripheral nerves of the
fingers of three vibration-induced persons with
Raynaud’s phenomenon. Scand J Work Environ Health
14, 31–5.
3) Brammer AJ, Pyykkö I (1987) Vibration-induced
neuropathy: Detection by nerve conduction
measurements. Scand J Work Environ Health 13, 317–
22.
4) Sakakibara H, Kondo T, Miyao M, Yamada S (1994)
Digital nerve conduction velocity as a sensitive
indication of peripheral neuropathy in vibration
syndrome. Am J Ind Med 26, 359–66.
5) Färkkilä M, Pyykkö I, Jäntti V, Aatola S, Starck J,
Korhonen O (1988) Forest workers exposed to
vibration: A neurological study. Br J Ind Med 43, 826–
9.
6) Sakakibara H, Hirata M, Hashiguchi T, Toibana N,
Koshiyama H, Zhu SK, Kondo T, Miyao M, Yamada S
(1996) Digital sensory nerve conduction velocity and
vibration perception threshold in peripheral neurological
test for hand-arm vibration syndrome. Am J Ind Med
30, 219–24.
7) Anonymous (1995) Clinical and laboratory diagnostics
of neurological disturbances in workers using handheld vibrating tools. In: Stockholm Workshop 94, Handarm vibration syndrome: Diagnostics and quantitative
relationships to exposure, Proceedings. eds. by Gemne
G, Brammer AJ, Hagberg M, Lundström R, Nilsson T,
187–93, Sweden’s National Institute of Occupational
Health, Solna.
8) Hirosawa I, Watanabe S, Fukuchi Y, Nishiyama K,
Hosokawa M (1983) Availability of temperature sense
indices for diagnosis of vibration disease. Int Arch
Occup Environ Health 52, 215–22.
Industrial Health 2000, 38, 366–371
THERMAL PERCEPTION IN HAND-ARM VIBRATION SYNDROME
9) Ekenvall L, Nilsson BY, Gustavsson P (1986)
Temperature and vibration thresholds in vibration
syndrome. Br J Ind Med 43, 825–9.
10) Lundborg G, Dahlin LB, Hansson H-A, Kanje M,
Necking LE (1990) Vibration exposure and peripheral
nerve fiber damage. J Hand Surg 15A, 346–51.
11) M i t o H , S h i m i z u T ( 1 9 8 1 ) E x p e r i m e n t a l
thermoesthesiometer incorporating peltier thermal
modules —its rationale and instance of applications—.
Acta Med Kinki Univ 6, 125–30.
12) Brammer AJ, Taylor W, Lundborg G (1987)
Sensorineural stages of the hand-arm vibration
syndrome. Scand J Environ Health 13, 279–83.
13) Virokannas H, Virokannas A (1995) Temperature and
vibration perception thresholds in workers exposed to
hand-arm vibration. Cent Eur J Pub Health 3 (Suppl),
66–9.
14) Zamyslowska-Szmytke E (1998) Efficacy of vibration,
electric current and thermal perception tests in diagnosis
of hand-arm vibration syndrome. Int J Occup Med
Environ Health 11, 247–54.
15) Lindsell CJ, Griffin MJ (1999) Thermal thresholds,
vibrotactile thresholds and finger systolic blood
pressures in dockyard workers exposed to handtransmitted vibration. Int Arch Occup Environ Health
371
72, 377–86.
16) Strömberg T, Dahlin LB, Rosén I, Lundborg G (1999)
Neurophysiological findings in vibration-exposed male
workers. J Hand Surg 24B, 203–9.
17) Chang KY, Ho ST, Yu HS (1994) Vibration induced
neurophysiological and electron microscopical changes
in rat peripheral nerves. Occup Environ Med 51, 130–
5.
18) Ho ST, Yu HS (1989) Ultrastructural changes of the
peripheral nerve induced by vibration: An experimental
study. Br J Ind Med 46, 157–64.
19) Guyton AC (1981) Somatic sensations: I. the
mechanoreceptive sensations. In: Textbook of medical
physiology (sixth edition). ed. by Guyton AC. 597–
625, Saunders, Philadelphia.
20) Nilsson T, Lundström R, Burström L, Hagberg M (1995)
Assessment of heat pain perception in relation to
vibration. Cent Eur J Pub Health 3 (Suppl), 70–2.
21) Hilz MJ, Glorius S, Beric A (1995) Thermal perception
thresholds: Influence of determination paradigm and
reference temperature. J Neur Sci 129, 135–40.
22) Ruffell CM, Griffin MJ (1995) Effect of starting
temperature on the repeatability on thermotactile
thresholds. Centr Eur J Publ Hith 3 (Suppl), 81–5.