Content uploaded by Maurice Mars
Author content
All content in this area was uploaded by Maurice Mars on Aug 16, 2023
Content may be subject to copyright.
CARDIOVASCULAR JOURNAL OF SOUTH AFRICA Vol 16, No. 4, July/August 2005 215
Summary
Aim: Compressed air massage is a new treatment modal-
ity that uses air under pressure to massage skin and
muscle. It is claimed to improve skin blood flow but
this has not been verified. Several pilot studies were
undertaken to determine the effects of compressed air
massage on skin blood flow and temperature.
Methods: Skin blood flow (SBF), measured using laser
Doppler fluxmetry and skin temperature was recorded
under several different situations: (i) treatment, at 1 Bar
pressure using a single-hole (5-mm) applicator head, for
1 min at each of several sites on the right and left
lower legs, with SBF measured on the dorsum of the
left foot; (ii) at the same treatment pressure, SBF was
measured over the left tibialis anterior when treatment
was performed at different distances from the probe;
(iii) SBF and skin temperature of the lower leg were
measured with treatment at 0 or 1 Bar for 45 min, using
two different applicator heads; (iv) SBF was measured
on the dorsum of the foot of 10 subjects with treatment
for 1 min at 0, 0.5, 1, 1.5 and 2 Bar using three different
applicator heads.
Results: (i) SBF of the left foot was not altered by
treatment of the right leg or chest, but was significantly
increased during treatment of the left sole and first
web, p < 0.0001. (ii) SBF over the tibialis anterior was
increased when treatment was 5 cm from the probe,
p < 0.0001, but not when 10 cm away. (iii) SBF was
significantly elevated throughout the 45-min treatments
at 1 Bar and returned to normal within 1 min of stopping
treatment. Skin temperature fell by 6.8°C and 4.3°C
after 45-min treatments at 1 Bar, and slight rewarming
occurred within 15 min. At 0 Bar, no change in SBF
or temperature was noted. (iv) A near-linear increase in
SBF was noted with increasing treatment pressure for
two of the three applicator heads.
Conclusion: Compressed air massage causes an immedi-
ate increase in SBF, and an immediate fall in SBF when
treatment is stopped. The effect appears to be locally
and not centrally mediated and is related to the pressure
used. Treatment cools the skin for at least 15 min after
a 45-min treatment.
Cardiovasc J South Afr 2005; 16: 215–219. www.cvjsa.co.za
Compressed air massage is a new treatment modality that
has recently been developed in South Africa. Its use has
been aggressively marketed nationally and internationally.
Anecdotal reports suggest that treatment may improve
peripheral circulation and neuropathic symptoms in diabetic
patients. There have also been claims of enhanced healing
of diabetic and other peripheral ulcers. It is not clear
whether the peripheral ulcers were venous, arterial or due
to pressure. Compressed air massage has also been used
for sports massage, where it provides an easy-to-administer,
cooling massage. Although it is used commercially, there
are no clinical data available on its safety and efficacy.
Compressed air massage can be likened to putting air
into the tyre of a car. Instead of the compressed air going
into the tyre, it is directed onto the skin and used to massage
skin and subcutaneous tissues. The apparatus consists of
an electrically driven air compressor, two reservoir tanks,
The effect of compressed air massage on skin
blood flow and temperature
MAURICE MARS, SUNIL S. MAHARAJ, MARK TUFTS
Medicine, University of KwaZulu-Natal
MAURICE MARS, M.B., Ch.B., M.D.
Department of Physiotherapy, Faculty of Health Sciences,
University of KwaZulu-Natal
SUNIL MAHARAJ, B.Paed.Sc., B.Ed., B.Sc. (Physio),
M.Med.Sc. (Sports Med.)
Department of Physiology, Faculty of Health Sciences,
University of KwaZulu-Natal
MARK TUFTS, M.Sc.
The authors have no financial or other interest in any of the
companies whose equipment was used in this study.
Cardiovascular Topics
medical air filters, a pressure regulator valve, pressure
tubing and several different applicator heads. Compressed
air at pressures of 4 to 7 Bar passes from the compressor
into the first reservoir tank, where the air is filtered to
remove any particulate matter. A regulator valve controls
the pressure at which compressed air is released from
the first reservoir tank into the second tank, and sub-
sequently, through compression hosing to an applicator
head. Treatment involves gently massaging the skin with
a metal applicator head and the stream of compressed
air that passes through the applicator head supplements
the massage. The applicator heads differ in the number,
size and position of the holes through which the air passes.
Air pressure for treatment ranges from 1 to 2 Bar, and
additional pressure may be applied to the skin by the
therapist pushing the applicator head down on the skin
during the massage.
Possible proposed mechanisms of action of compressed
air massage include: (i) direct stimulation of local skin blood
flow; (ii) an intermittent positive pressure treatment effect
reducing superficial oedema, with subsequent improved
cellular oxygenation due to reduced oxygen diffusion
distance; (iii) skin cooling during treatment with reflex
vasodilatation and improved blood flow on rewarming;
(iv) localised tissue damage inducing an inflammatory
response; and (v) a localised hyperbaric oxygen effect in
the dermal cells.
To date, the only publications on the use of compressed
air massage report on the effect of massage on skeletal
muscle fibre and capillary morphology, and morphometry
in animal models.1,2 There have been no studies on the
potential risks associated with this treatment and its possible
mode of action has not been investigated. The aim of this
series of pilot studies was to investigate the effect of com-
pressed air massage on skin blood flow and skin temperature
in healthy subjects using laser Doppler fluxmetry.
Methods
The work was undertaken with the approval of the Ethics
Committee of the University of Natal Medical School, and
written informed consent was obtained from the subjects. All
subjects studied were non-diabetic, non-smokers, who were
asymptomatic of peripheral vascular disease. Compressed
air massage was performed with the subject lying supine
and all subjects were acclimatised to the ambient room
temperature of 20 to 21°C for 15 min before treatment.
Study 1. Acute effects of treatment on skin blood
flow
Four male volunteers aged 28 to 47 years were studied. An
unheated laser Doppler fluxmetry probe (LDF) (Vasomedics)
and a thermocouple temperature probe (Yellow Springs
Instrument Co. Inc Precision 4400 and a YSI 4000A
thermometer) were attached to the mid-dorsum of the
left foot of the subjects. Compressed air massage was
applied for 1 min at 1 Bar pressure, using a single-hole
(5-mm) applicator head at each of five sites on the left leg.
Treatment sites were: (i) the posterior aspect of the calf,
10 cm distal to the joint line; (ii) the anterior compart-
ment of the leg, 10 cm distal to the joint line; (iii) the
mid-sole of the foot; and (iv) the first web space on the
dorsum of the foot, 5 cm from the probes. The applicator
head was placed lightly on the skin and the head was
moved around in a circular motion over a diameter area of
approximately 3 to 4 cm, except at the level of the first
web space of the foot. There it was applied with the long
axis of the applicator head transverse to the long axis
of the foot, so as to maintain a distance of approximately
5 cm from the probes on the dorsum of the foot. Measure-
ments of skin blood flow, in arbitrary units (au), and skin
temperature were made after 60 seconds of treatment at
each site. The LDF was set to average readings over 10
seconds.
The process was repeated with the probes on the left
foot, but treating the right leg and an area on the anterior
chest wall 5 cm below the left clavicle in the mid-clavicular
line.
Study 2. Acute effects of treatment on skin blood
flow, probe placed over a large muscle
In study 1, the LDF and temperature probes were placed
on skin over tendons and bone on the mid-dorsum of the
foot. In this study of another four volunteers, the effect
of treatment on skin blood flow was investigated with the
probes positioned over a large muscle. The rationale was
that the applied air pressure would alter the pressure in
the muscle compartment and this might elicit a different
response. The probes were placed on the skin over the
tibialis anterior muscle at a site 10 cm distal to the left knee
joint line, and 2 cm lateral to the tibial margin. Treatment
was applied to the left leg at 1 Bar for 1 min, with the 5-mm
single-hole applicator head at each of the following sites:
(i) the sole of the foot; (ii) the mid-dorsum of the foot; (iii)
the posterior calf at the level of the probes; (iv) 10 cm,
and (v) 5 cm distal to the probes over the lateral lower leg.
Measurements were taken as before.
Study 3. Effect of 45-min treatment on skin blood
flow
Five male subjects were studied on three separate occasions.
The laser Doppler and temperature probes were attached
to the mid-dorsum of the foot and baseline readings were
obtained. Treatment consisted of 45 min of massage of the
foot with the single-hole or multiple small-hole applicator
head at an air pressure of 1 Bar on the first two visits, and
on the third visit, to serve as a control, with the multiple
small-hole applicator head with no air passing through the
applicator head. Measurements of skin blood flow and skin
temperature were made every 5 min during treatment and
every min for 15 min thereafter. The first measurement,
after 5 min of treatment, was made with the treatment on the
dorsum of the foot adjacent to the probes. All subsequent
measurements were made while the applicator head was on
the skin and at least 5 cm away from the probes at the
following sites, the sole, heel, medial arch and over the
dorsal web spaces.
216 CARDIOVASCULAR JOURNAL OF SOUTH AFRICA Vol 16, No. 4, July/August 2005
Study 4. The effect of different air pressures on
skin blood flow
Skin blood flow was measured in 10 volunteers with the
laser Doppler probe placed on the dorsum of the foot.
Treatment was undertaken at pressures of 0.5, 1.0, 1.5 and
2.0 Bar for 1 min, with each of the single-hole, three-hole
and multiple small-hole applicator heads. Treatment was
over an area 5 to 7 cm away from the LDF probe. To serve
as a control, and to determine the possible effect of massage
with the applicator head on skin blood flow, each subject
underwent a further treatment consisting of massage with
an applicator head with no compressed air passing through
the applicator. Steady-state resting values of skin blood
flow were obtained for at least 5 min before the next air
pressure was used.
Statistical methods
Results were expressed as the mean and one standard devia-
tion, with 95% confidence intervals. Repeated-measures
analysis of variance with post hoc testing using the Tukey-
Kramer test was performed when repeated tests were under-
taken at different sites or at different times on each subject.
Two-factor analysis of variance was used to compare the
mean values obtained with different applicator heads at
different times or application pressures. Alpha was set at
5%. Statistical analysis was performed using GraphPad
Prism version 4.00 for Windows, GraphPad Software, San
Diego, California, USA.
Results
Study 1. Acute effects of treatment on skin blood
flow
Blood flow and skin temperature measured on the dorsum of
the foot were not affected by treatment of the opposite leg or
the anterior chest wall. With the probes on the foot, resting
blood flow was 2.8 ± 1.3 (95% CI: 0.7–4.95) arbitrary
units (au) and treatment over the anterior and posterior
compartments of the leg did not affect foot blood flow.
Blood flow was increased with treatment of the sole, 18.0
± 7.2 (95% CI: 6.5–29.5) au, and the first web space, 27.1
± 8.7 (95% CI: 15.4–38.1) au (repeated-measures ANOVA,
p < 0.0001). Post hoc testing using the Tukey-Kramer
multiple comparisons test showed significant increases in
blood flow with treatment of the sole and the first web
space (p < 0.001) and between the sole and the first web
space (p < 0.05) (Fig. 1).
Study 2. Acute effects of treatment on skin blood
flow, probe placed over a large muscle
With the probes placed over the anterior compartment,
resting blood flow was 1.3 ± 1.1 (95% CI: −0.5–3.0) au.
Skin blood flow was only altered when treatment was over
the lateral aspect of the leg. With treatment 10 cm from the
probes, blood flow was increased to 3.2 ± 2.8 (95% CI:
−1.2–7.6) au. With treatment 5 cm from the probes, it rose
to 19.0 ± 4.2 (95% CI: 12.2–5.6) au. Repeated-measures
ANOVA showed a significant difference in blood flow (p
< 0.0001). Post hoc testing showed the blood flow with
treatment 5 cm from the probe to be significantly greater
than at rest or with treatment at any other site (p < 0.001)
(Fig. 2).
Study 3. Effect of 45-min treatment on skin blood
flow
During the 45-min treatments with 1 Bar pressure, skin
blood flow increased immediately and was always highest
when measured nearest to the LDF probe. The skin blood
flow remained significantly elevated throughout the 45
CARDIOVASCULAR JOURNAL OF SOUTH AFRICA Vol 16, No. 4, July/August 2005 217
Fig. 1. Skin blood flow measured at the dorsum of the
foot at rest and after 60 seconds of treatment over the
posterior compartment of the leg, the anterior compart-
ment of the leg, the sole of the foot and the first web
space region, expressed as the mean and one standard
deviation. Blood flow measured was significantly greater
with treatment on the sole and the first web space
regions than at the other sites (*p < 0.001) and the blood
flow at the first web was significantly greater than at
the sole (**p < 0.05).
0
5
10
15
20
25
30
35
40
Rest Post Comp Ant Comp Sole 1st Web
Treatme nt site s
LDF (au
)
*
*
**
Fig. 2. Skin blood flow measured over the anterior
compartment of the leg at rest and after 60 seconds of
treatment over the sole of the foot, the mid dorsum
of the foot, the posterior compartment of the leg (PC),
the anterior compartment of the leg at distances of 10
cm (AC10 cm) and 5 cm from the probe (AC5 cm),
and adjacent the probe, expressed as the mean and
one standard deviation. Blood flow measured with treat-
ment at the probe on the anterior compartment was
significantly greater than the resting values and those
obtained with treatment at the other sites (*p < 0.001).
0
5
10
15
20
25
Rest Sole Dors um PC AC(10cm) AC(5cm)
Treatment sites
LDF (au)
*
218 CARDIOVASCULAR JOURNAL OF SOUTH AFRICA Vol 16, No. 4, July/August 2005
min of treatment with 1 Bar pressure (repeated-measures
ANOVA, p < 0.001). There was no difference between the
single-hole 5-mm head at 1 Bar and the multiple head at 1
Bar. No change in blood flow was noted when massage was
performed with no air. Within 1 min of stopping treatment,
the skin blood flow returned to levels non-significantly
lower than the skin blood flow before treatment. Blood
flow during the recovery period, although not increased,
was significantly greater when no air was used. During
the 15-min observation period after treatment, skin blood
flow did not increase above pre-treatment baseline for
compressed air massage (Fig. 3).
Skin temperature fell significantly within 5 min of
starting treatment and remained relatively constant thereaf-
ter for both applicator heads used with 1 Bar pressure
(p < 0.001). After 5 min of treatment, the temperature
drop with the multiple small-hole applicator head of 5.9°C
was greater than with the single-hole applicator (4.1°C).
After 45 min of treatment, skin temperature had dropped
by 6.8 ± 1.8 (95% CI: 5.2–12.2)°C and 4.3 ± 0.8 (95%
CI: 3.5–7.8)°C. The difference in fall in temperature was
significant (p < 0.01). Following treatment with the multi-
ple small-hole applicator slight re-warming was noticed.
This did not occur after treatment with the single-hole
applicator (Fig. 4).
Study 4. The effect of different air pressures on
skin blood flow
An almost linear increase in skin blood flow was noted
with increasing air pressure when using the single-hole
and three-hole applicators. With the multiple small-hole
applicator head, the changes in skin blood flow were less,
and no change was noted between 1.5 and 2 Bar (Fig. 5). For
each applicator head, repeated-measures ANOVA showed
significant increases in skin blood flow to have occurred at
air pressures of 0.5 Bar and above. Some subjects reported
that treatment with air above 1.5 Bar was uncomfortable
and caused pain.
Discussion
The key findings of these studies were that compressed air
massage is associated with a concurrent rise in skin blood
flow measured at distances up to 5 to 7 cm away from
the point of treatment, and that skin blood flow returned to
normal values when the treatment stopped. The increase in
skin blood flow occurred concomitant with skin cooling.
No increase in skin blood flow occurred during the first 15
min after treatment, possibly due to reflex vasodilation. The
changes in skin blood flow were local phenomena and did
not appear to be centrally mediated. It was also noted
that changes in skin blood flow were related to the air
pressure used during massage and that the size and number
of holes in the applicator head influenced the changes in
skin blood flow.
There are several non-invasive and minimally invasive
methods available to measure skin blood flow. The more
commonly used methods include 133xenon radioisotope skin
clearance,3 transcutaneous oxygen pressure measurement,4
thermography5 and laser Doppler fluxmetry.6 Each of these
methods has advantages and methodological problems.
Laser Doppler fluxmetry was chosen because it is a safe,
simple and rapid test of skin microcirculatory blood flow,
which can be performed without having to heat the skin to
facilitate maximal vasodilation.6 The transcutaneous oxygen
Fig. 3. Laser Doppler-derived skin blood flow expressed
in arbitrary units as the mean and one standard devia-
tion during treatment with the single-hole and multiple
small-hole applicator heads and with massage without
air. Measurements were taken at 5-min intervals during
a 45-min treatment and at minute intervals for the 15 min
thereafter (R = recovery after treatment).
Fig. 4. Mean skin temperature and one standard devia-
tion during treatment with the single-hole and multiple
small-hole applicator heads. Measurements were taken
at 5-min intervals during a 45-min treatment and at
minute intervals for the 15 min thereafter (R = recovery).
Fig. 5. Mean changes in skin blood flow and one
standard deviation, using different air pressures with
three different applicator heads.
0
5
10
15
20
25
30
Res t 5
min
10
min
15
min
20
min
25
min
30
min
35
min
40
min
45
min
1
min
R
5
min
R
10
min
R
15
min
R
LDF (au )
Sing le 5mm hole Mult iple 1mm hole s No Air
0
5
10
15
20
25
30
35
Res t 5
min
10
min
15
min
20
min
25
min
30
min
35
min
40
min
45
min
5
min
R
10
min
R
15
min
R
Temp
o
C
Single hole Mu ltip le s ma ll hole s
0
10
20
30
40
50
60
70
80
90
100
Res t 0 5 Ba r 1 Ba r 1 5 Bar 2 Bar
LDF (au)
Sing le 5mm Hole Mult iple 1mm Holes 3 Hole
CARDIOVASCULAR JOURNAL OF SOUTH AFRICA Vol 16, No. 4, July/August 2005 219
pressure measurement method was not used because it
requires heating the skin to 45°C to achieve maximum
dermal capillary dilatation,7 and compressed air massage
cools the skin. Thermography measures skin temperature
and not blood flow per se and would reflect the cooling of
the skin caused by the treatment. The intra-dermal pres-
sures generated during compressed air massage would
most likely exacerbate the known problem of leakage of
injected 133Xe back up the injection track, resulting in false
measurements.8
Laser Doppler fluxmetry, while similar in principle
to the use of Doppler ultrasound to detect flow in a
single vessel, detects a Doppler shift signal produced
when monochromatic laser light is reflected back off red
blood cells in the vasculature of the skin. Unlike Doppler
ultrasonography, in which a single vessel is evaluated, the
laser light penetrating the skin encounters many capillaries
and strikes them at varying angles of incidence. To further
complicate matters, in the hairpin dermal capillary loops,
the light will be reflected from cells moving both towards
and away from the light source.
The light penetrates to a depth of approximately 1.5
mm and undergoes significant scattering before reaching the
capillary. Similarly, the light undergoes further scattering
after reflection on its way back to the light sensor. The
resulting signal produced is the product of the number of
red blood cells moving in the sample volume and the mean
velocity of the red blood cells. It is therefore a measure of
flux and not blood flow.9
How can the paradoxical increase in skin blood flow
in the presence of skin cooling be explained? A possible
mechanism is based on changes in pressure in subcutaneous
and muscle tissue. It is well documented that pressure
applied externally to a limb is transmitted directly to the
underlying soft tissue and muscle.10 While it was previously
argued that muscle is a gel and therefore the pressure
should be dispersed equally throughout the muscle, it is
now accepted that pressure can differ in various regions of
the same muscle.11
Pressure will be highest at the point of application of
the stream of air, and will be a function of the pressure at
which the compressed air leaves the applicator head. Even
at 0.5 Bar, the locally transmitted pressure is far higher
than arterial, venous or capillary pressures and will occlude
vessels in the immediate region. Theoretically, there will be
a gradient of concentric rings of decreasing pressure around
the application point. Blood will effectively be shunted
away from the region of occlusion and where vessels remain
patent, blood will tend to flow down the concentration
gradient surrounding the point of application. This will
increase the velocity of the blood cells in the surrounding
patent vessels and the volume flow will also increase. These
are the changes that are measured by laser Doppler flux-
metry as an increase in skin blood flow. When compressed
air massage treatment stops and the externally applied pres-
sure is removed, the distribution of blood corrects itself and
blood flow to the skin returns to normal. The effect of skin
cooling then comes into play with local vasoconstriction
resulting in the skin blood flow being slightly lower than
the pre-treatment baseline values. From the data in these
studies, the alterations in haemodynamics appear to be
significant up to a distance of 5 to 7 cm away from the site
of treatment, but not at 10 cm from treatment.
In a histological and morphometric study in an animal
model in which compressed air massage was used in a
manner similar to deep transverse friction, we have shown
that skeletal muscle capillaries dilate for up to 24 hours
after treatment. When compared with deep transverse fric-
tion, these changes lasted longer, with fewer ultrastructural
changes seen in muscle or capillaries.1,2 The prolonged
vasodilation may be due to the initial cooling associated with
compressed air massage, resulting in a delayed hyperaemic
response, or it may indicate prolonged inflammation follow-
ing massage-associated minor muscle damage.
These preliminary pilot studies suggest that compressed
air massage causes skin blood flow to increase. Further
safety studies are required to determine what muscle dam-
age, if any, occurs with prolonged and repeated treatment
and, if there is damage, is it sufficient to cause changes in
plasma potassium concentration or to damage blood cells.
Studies on the pressure generated in muscle are also required
before clinical trials on diabetic patients can commence on
this potentially useful new treatment.
Compressed air massage equipment for this study was provided
by Jet Therapy South Africa.
References
1. Gregory MA, Mars M. Compressed air massage causes capillary dila-
tion in untraumatised skeletal muscle: a morphometric and ultrastruc-
tural study. Physiotherapy 2005; 91: 131–137.
2. Gregory MA, Mars M. The effect of compressed air massage
on untraumatised rabbit skeletal muscle − a morphometric and
ultrastructural study. S Afr J Physiother 2004; 60: 19–27.
3. Malone JM, Leal JM, Moore WS, Henry RE, Daly MJ, Patton
DD, Childers, SJ. The ‘gold standard’ for amputation level selection
xenon-133 clearance. J Surg Res 1981; 30: 449–455.
4. Mars M, Mills RP, Robbs JV. The potential benefit of pre-operative
assessment of amputation wound healing potential in peripheral
vascular disease. S Afr Med J 1993; 83: 16–18.
5. Sarin S, Shami S, Shields DA, Scurr JH, Smith PD. Selection of
amputation level: a review. Eur J Vasc Surg 1991; 5: 611–620.
6. Mars M, McKune A, Robbs JV. A comparison of laser Doppler
fluxmetry and transcutaneous oxygen pressure measurement in the
dysvascular patient requiring amputation. Eur J Vasc Endovasc Surg
1998; 16: 53–58.
7. Mars M. Hands up? A preliminary study on the effect of post-operative
hand elevation. J Hand Surg 1988; 13(B): 430–434.
8. Moore WS, Henry RE, Malone JM, Daly MJ, Patton D, Childers
SJ. Prospective use of xenon Xe 133 clearance for amputation level
selection. Arch Surg 1981; 116: 86–88.
9. Fagrell B. Advances in microcirculation network evaluation: an
update. Int J Microcirc Clin Exper 1995; 15(Suppl 1): 34–40.
10. Mars M, Brock-Utne JG. The effect of tourniquet release on intra-
compartmental pressure in the bandaged and unbandaged limb. J
Hand Surg Br 1991; 16: 318–322.
11. Mars M, Hadley GP. Raised intracompartmental pressure and compart-
ment syndromes. Injury 1998; 29: 403–411.
