Journal Logo

Clinical Sciences: Clinically Relevant

Failure of manual massage to alter limb blood flow: measures by Doppler ultrasound


Author Information
Medicine & Science in Sports & Exercise: May 1997 - Volume 29 - Issue 5 - p 610-614
  • Free


Proponents for the use of sports massage in the enhancement of performance and post-exercise recovery have suggested that this modality raises muscle blood flow and thereby enhances tissue healing and lactate removal (2,3,10). Some evidence supports the hypothesis that massage increases muscle blood flow(2,3,20). However, there is also an accumulation of evidence that massage has a minimal, if any, role in elevating blood flow either at rest or following fatiguing exercise(2,6,16).

There are several possible sources of this controversy. First, Callaghan(3) suggested that massage may produce transient increases in blood flow, which are evident only in the early moments of treatment. However, continuous measures of femoral artery mean blood velocity (MBV) have suggested that effleurage massage does not alter leg blood flow at any time during the treatment (16). Second, the size of the muscle mass being massaged may alter the effects on blood flow. Reports supporting the effect of massage on increasing blood flow extend largely from studies where smaller muscles masses such as the forearm and calf have been treated(7,9). In contrast, when larger muscle masses have been massaged, no effect on blood flow was observed(12,16,20). Further, the type of massage used may affect muscle blood flow. For example, the tapotement form of massage has been suggested to have the greatest effect on muscle blood flow owing to muscle trauma and impact (3,9).

The purpose of this study was to test our earlier hypothesis that the size of the muscle mass being massaged may influence the blood flow response to massage (16). Further, the effect of type of massage on limb blood flow was investigated. Effleurage, petrissage, and tapotement forms of massage were performed on both the forearm (small muscle mass) and the quadriceps (large muscle mass) while limb blood flows were monitored by pulsed and echo Doppler for measures of MBV and vessel diameter, respectively.


Subjects. Ten healthy subjects (7 male and 3 female) volunteered for this study. The subjects were 35.8 ± 3.4 yr of age, 171.5 ± 3.0 cm in height, and 70.6 ± 2.3 kg in weight with 17.8 ± 2.1% body fat. The volunteers indicated that they were free of any form of cardiovascular or musculoskeletal disease, and none was on any medication, as assessed by a medical history. Each subject provided signed consent to the testing procedures on a form approved by the Office of Human Research at the University.

Experimental design. Following assumption of the supine position, the subjects were instrumented for measures of heart rate and limb blood flow(described below). No measures were obtained before 20 min of supine rest had passed so that cardiac stroke volume (1) and limb blood flows had stabilized.

Effleurage (30 strokes per min), petrissage (50-60 strokes per min) and tapotement (360 percussions per min) massage were used in the study. Effleurage massage consisted of rhythmic pressure strokes along the longitudinal axis of the muscle group. Petrissage consisted of kneading and squeezing motions over the muscle mass. For tapotement, the massage therapist used percussive motions where the ulnar surface of the hands contacted the quadriceps in a rapidly repetitive hacking motion. For the forearm, the tips of rigid fingers were used to produce the percussive effect. All massage treatments were administered by a certified massage therapist. The therapist was unaware of the effect of the massage treatments on the blood flow responses during the collection of data.

With the exception of tapotement, these techniques did not result in movement artifact or impact noise in the blood velocity tracing when performed on the distal two-thirds of the muscle mass. The tapotement technique did induce some impact noise into the pulsed Doppler tracing of MBV, but this was effectively eliminated with a high pass filter. All massage techniques were performed on both the right forearm and quadriceps muscles, allowing investigation of the role of muscle mass in determining the effectiveness of massage to elevate muscle blood flow. The order in which the massage techniques were administered to a given limb and the order of limbs receiving the treatments were counterbalanced. For each condition, the massage lasted for 5 min with 5 min of rest between each treatment. To determine the magnitude and timing of limb blood flow changes with massage, beat-by-beat measures of MBV were obtained from the brachial and femoral arteries over 30 s of rest, as well as over the first min and final 30 s, of each massage treatment.

After the three massage techniques were administered on a given limb, a mild voluntary contraction was performed to measure the effect of muscle activation versus massage on the blood flow response. For forearm exercise, the subjects were required to produce a mild isometric handgrip contraction(35.8 ± 3.4% of maximal voluntary isometric contraction) over a 2-s period. For leg exercise, a pad 10 cm in height was placed under the right knee, and the subject produced a mild unweighted knee extension maneuver in which the lower leg was lifted and lowered dynamically over a 2-s period. For these exercise tests, arterial MBV was measured continuously for 30 s at rest, during the contraction, and for 30 s of post-contraction flow response.

Brachial and femoral artery diameters were obtained by echo Doppler at the beginning and end of massage treatments for a given limb prior to the voluntary exercise tests.

Data collection and analysis. Heart rate (Cambridge Model VS4 electrocardiograph, Cambridge, MA) and brachial artery MBV (pulsed Doppler velocimetry, Multigon Model 500V, Mt. Vernon, NY) were collected continuously on a computer-based system at 100 Hz. Brachial artery MBV was determined from the spectra of the pulsed Doppler ultrasound signal. A flat probe with an operating frequency of 4 MHz was fixed to the skin over the brachial artery in the antecubital fossa region of the right elbow. The angle of the transducer crystal relative to the skin was 45° and echo ultrasound imaging confirmed that the brachial artery ran parallel to the skin in this location. The ultrasound gate was set to insonate the total width of the artery lumen. The Doppler shift frequency spectra were processed by a quadrature audio demodulator (15) which provided the instantaneous MBV in real time. Beat-by-beat MBV was calculated as the average of the instantaneous MBV values over each cardiac cycle using the QRS complex of the ECG tracing to signal the end of one heart beat and the beginning of the next. MBV values, obtained at 5, 10, and 20 s and 5 min following the onset of massage, were compared with a 30 s average MBV obtained prior to the massage treatment.

The brachial and femoral arteries were imaged by echo Doppler (Toshiba Model SSH-140A, Pochtochiai-hen, Japan) for measures of conduit artery diameter. A hand-held 7.5 MHz linear probe, operating in B-Mode, was positioned over the sites from which the MBV data were collected. For brachial artery measures, this site was between the biceps aponeurosis and muscle belly. For femoral artery measures, the probes were positioned over the vessel 2-3 cm distal to the inguinal ligament. The imaged data were stored on VHS tape for analysis. An estimate of arterial diameter was made from the average of three measurements prior to and following the massage treatments. All diameter measurements were made during cardiac diastole. The diameter measurement markers of the Toshiba imaging system moved in increments of 0.1 mm. We have previously observed that, with these methods, the day-to-day coefficient of variation for these data range from 2-4% (unpublished data). From the corresponding MBV and diameter values, the limb blood flows (LBF) were calculated as LBF = MBV · πr2 where r is the vessel radius. Blood flow values were calculated for rest and at the times for which MBV values were measured.

Statistics. Statistical analysis was performed by a repeated measures two-way ANOVA for each of the forearm and quadriceps flow measures with massage. A one-way ANOVA with a repeated measure was used to analyze the effect of voluntary exercise on peak blood velocity and blood flows. The Student-Newman Keuls post-hoc test was used to differentiate significant effects. The level of significance was set at P < 0.05. All values included in the text are mean ± SE. The tracings for the figures represent the convergence averaged mean of all 10 subjects.


Forearm. At rest, MBV in the brachial artery was 4.76 ± 0.5, 5.77 ± 0.4, and 5.34 ± 0.2 cm·s-1 prior to the effleurage, petrissage, and tapotement treatments, respectively(P > 0.05). MBV at the 5, 10, and 20 s and 5 min time points following the onset of massage was variable but was not different from these rest values for any of the massage treatments (P > 0.05)(Fig. 1). For example, at 20 s following the massage onset, MBV during effleurage, petrissage, and tapotement techniques was 4.39± 1.8, 4.80 ± 1.3, and 4.78 ± 0.5 cm·s-1, respectively.

Brachial artery diameters at rest were also unchanged by the collective effects of the three massage techniques (P > 0.05). Therefore, the average of the pre- and post-massage diameter values (4.1 ± 0.3 mm) were used in determining the forearm blood flow response. The lack of effect of massage on artery diameter resulted in arm blood flow responses which were proportional to the MBV results; that is, massage had no effect on forearm blood flow at any time during the treatment (P > 0.05).

Following the voluntary handgrip contraction, forearm perfusion increased immediately following release of the contraction and peaked at 15.2 ± 1.2 cm·s-1 for MBV and 126 ± 19 ml·min-1 for blood flow (Fig. 1). These were both significantly greater than rest (P < 0.05).

Quadriceps. The leg hemodynamic responses during quadriceps massage were similar to those in the forearm. Femoral artery MBV was 9.73± 0.7, 9.38 ± 1.0, and 9.73 ± 0.7 cm·s-1 measured at rest prior to the effleurage, petrissage, and tapotement treatments, respectively. At 20 s following the onset of massage, femoral artery MBV was 10.1 ± 1.0, 8.55 ± 0.9, and 6.95 ± 1.7 cm·s-1 for the effleurage, petrissage, and tapotement techniques, respectively. These massage values were not different from their resting counterparts (P > 0.05).

Resting femoral artery diameters also were unchanged by the massage(P > 0.05) so that the average of pre- and post-massage diameters(9.0 ± 0.3 mm) were used in the leg blood flow calculations.

Overall, massage did not alter the leg MBV or blood flow response at any time during the treatments (P > 0.05). There was a tendency for the petrissage and tapotement forms of massage to reduce femoral artery MBV(Fig. 2), but this failed to reach statistical significance. The light quadriceps contraction was followed by a rapid increase in femoral artery MBV and blood flow which peaked at 28.1 ± 3.1 cm·s-1 and 1087 ± 144 ml·min-1, respectively (Fig. 2). These responses were significantly greater than rest (P < 0.05).


Earlier investigations have both supported(2,3,20) and refuted(12,16,20) the hypothesis that manual massage elevates muscle blood flow. We have discussed the possible reasons for this discrepancy as including differences in blood flow measurement methodology, timing of the blood flow measures, and the muscle mass being massaged(16). This is the first study to measure beat-by-beat changes in limb blood flow with different massage techniques and muscle masses. The results do not support the hypothesis of a muscle mass dependency on the blood flow responses to manual massage techniques. Rather, these data clearly show that manual massage did not elevate total limb blood flow irrespective of massage technique on a large (quadriceps) or small (forearm flexors) muscle group.

It has been hypothesized that the tapotement form of massage could be more effective than other manual massage techniques in enhancing muscle blood flow(3,9). However, the results of the current study suggest that, even with tapotement, total limb blood flow was unchanged from rest. This was observed for both the large and small muscle masses investigated.

For the effleurage and petrissage techniques, in particular, the massage did not change mean blood flow response, but the variability about that mean was altered compared with rest. This was likely a result of the effects of muscular compression and release on altering the direction of flow through the conduit artery. Compression of the muscle resulted in a brief retrograde arterial flow, whereas release of the pressure allowed a surge of orthodromic flow through the artery. This effect is also observed with muscle contractions and relaxation (19) and may partially explain the variable effect of petrissage on muscle blood flow observed previously(9).

The continuous nature of the MBV measures indicate that the effect of massage on limb blood flow was not altered over the duration of the treatment period. In contrast, light contractions of the quadriceps or forearm flexors resulted in significant elevations in limb MBV. With minimal changes in the dimensions of the conduit artery following the single contraction, the elevated MBV equates to a significant increase in limb blood flow. Furthermore, it is doubtful that such a brief exercise would alter core temperature and skin perfusion so that the bulk, if not all, of the increase in limb blood flow with the exercise represented elevated muscle blood flow. Therefore, the results of the current study support previous suggestions that, if enhanced muscle blood flow does in fact promote muscle healing, light exercise may be more beneficial than massage(3,7,16).

Critique of methods. The primary methodology used in this study was pulsed and echo Doppler ultrasonography to determine mean blood velocity(MBV) and diameter, respectively, in the brachial and femoral arteries at rest and during massage. Pulsed Doppler technology provides the opportunity to monitor continuously red blood cell velocity so that the mean velocity over a single cardiac cycle can be calculated. Importantly, ultrasonic methods of blood flow determination have been shown to be highly correlated with other techniques(4,5,11,13,17,18,21). Furthermore, we have demonstrated a strong linear relationship (r = 0.99) between the measured MBV and actual blood velocities through plastic tubes(Doppler MBV = 0.9 velocity + 2.4 cm·s-1)(17). Our measures of conduit artery diameter were not continuous with the MBV measures, but they do indicate that the combined effects of the three massage techniques did not alter vessel dimensions. Therefore, the blood flow response with massage could be determined.

One possible disadvantage of Doppler ultrasound methodology, as well as other muscle blood flow measurement technology, is that intramuscular distribution of blood flow with massage cannot be discriminated. However, this technique is noninvasive and does offer the advantage of providing beat-by-beat MBV and diameter measures so that transient and steady-state levels of total flow to the limb can be detected. In contrast, venous occlusion plethysmographic blood flow measures cannot be obtained during massage because of the sensitivity of this technique to motion artifact. It is expected that most of the effect of massage on blood flow would occur during the treatment. Only minimal residual effects may be present immediately after the cessation of massage when plethysmographic measures of massage-related blood flow have usually been taken. Also, measures by venous occlusion plethysmography may underestimate flow values owing to the simultaneous compression of both limb veins and conduit arteries with cuff inflation(8,17).

Another method commonly used to assess massage-induced muscle blood flow changes has been 133Xe clearance(7,9,12,20). While the clearance of133 Xe may provide an estimate of muscle blood flow, it tends to overestimate these flow values owing to local hyperemia following the trauma of injection (14).


By using beat-by-beat measures of brachial and femoral artery MBV, together with direct measures of arterial diameter, we have measured the time course of blood flow responses to three different types of manual massage on both a small and large muscle mass. The results do not support the hypothesis that massage elevates limb blood flow at any time during the massage treatment. However, even light voluntary exercise significantly elevated flow over rest. These data indicate that, if an elevated blood flow is the desired therapeutic effect, then light exercise would be beneficial whereas massage would not.

Figure 1-Brachial artery mean blood velocity at rest and during 5 min of effleurage, petrissage, and tapotement forms of massage and following a brief voluntary forearm contraction. Effleurage and Petrissage massage altered the variability of the blood velocity response compared with rest but did not alter the average blood velocity response (:
P > 0.05). The voluntary contraction significantly elevated blood velocity over rest( P < 0.05). Tracings are the mean of 10 subjects with second-by-second resolution. Arrow indicates onset of massage or exercise.
Figure 2-Femoral artery mean blood velocity at rest and during 5 min of effleurage, petrissage, and tapotement forms of massage and following a brief voluntary quadriceps contraction. Effleurage and Petrissage massage altered the variability of the blood velocity response compared with rest but did not alter the average blood velocity response (:
P > 0.05). The slight reductions in blood velocity with petrissage and tapotement were not statistically different from rest ( P > 0.05). The voluntary contraction significantly elevated blood velocity over rest ( P < 0.05). Tracings are the mean of 10 subjects with second-by-second resolution. Arrow indicates onset of massage or exercise.


1. Butler, G. C., X. Huacheng, and R. L. Hughson. Cardiovascular response to 4 hours of 6° head-down tilt or of 30° head-up tilt bed rest. Aviat. Space Environ. Med. 61:240-246, 1990.
2. Cafarelli, E. and F. Flint. The role of massage in preparation for and recovery from exercise. Sports Med. 14:1-9, 1992.
3. Callaghan, M. J. The role of massage in the management of the athlete: a review. Br. J. Sports Med. 1993:28-33, 2796.
4. Chauveau, M., B. Levy, J. F. Dessanges, E. Savin, O. Bailliart, and J. P. Martineaud. Quantitative Doppler blood flow measurement method and in vivo calibration. Cardiovasc. Res. 19:700-706, 1985.
5. Guldvog, I., M. Kjaernes, M. Thoresen, and L. Walloe. Blood flow in arteries determined transcutaneously by an ultrasonic doppler velocitymeter as compared to electromagnetic measurements on the exposed vessels. Acta Physiol. Scand. 109:211-216, 1980.
6. Gupta, S., A. Goswami, A. K. Sadhukhan, and D. N. Mathur. Comparative study of lactate removal in short-term massage of extremities, active recovery and a passive recovery period after supramaximal exercise sessions. Int. J. Sports Med. 17:106-110, 1996.
7. Hansen, T. I. and J. H. Kristensen. Effect of massage, shortwave diathermy and ultrasound upon 133Xe disappearance rate from muscle and subcutaneous tissue in the human calf. Scand. J. Rehab. Med. 5:179-182, 1973.
8. Hiatt, W. R., S. Y. Huang, J. G. Regensteiner, et al. Venous occlusion plethysmography reduces arterial diameter and flow velocity.J. Appl. Physiol. 66:2239-2244, 1989.
9. Hovind, H. and S. L. Nielsen. Effect of massage on blood flow in skeletal muscle. Scand. J. Rehab. Med. 6:75-77, 1974.
10. Lehn, C. and W. E. Prentice. Massage. In:Therapeutic Modalities in Sports Medicine, W. E. Prentice (Ed.). St. Louis: Mosby Year Book, Inc., 1994, pp. 335-363.
11. Levy, B. I., W. R. Valladares, A. Ghaem and J. P. Partineaud. Comparison of plethysmographic methods with pulsed doppler blood flowmetry. Am. J. Physiol. 236:H899-H903, 1979.
12. Linde, B. Dissociation of insulin absorption and blood flow during massage of a subcutaneous injection site. Diabetes Care 9:570-574, 1986.
13. Lubbers, J., P. J. L. M. Bernink, G. J. Barendsen, and J. W. Van Den Berg. A continuous wave doppler velocimeter for monitoring blood flow in the popliteal artery, compared with venous occlusion plethysmography of the calf. Pflugers Arch. 382:241-248, 1979.
14. Marcus, M. L., C. J. Bischoff, and D. D. Heistad. Comparison of microsphere and xenon-133 clearance method in measuring skeletal muscle and cerebral blood flow. Circ. Res. 48:748-761, 1981.
15. Micco, A. J. CW and Pulse Doppler Diagnostic System. U.S. Patent 4,819,622. April 22, 1989; Int. Cl. A61B 10/00. U.S. Patent 4,819,652. April 22, 1989.
16. Tiidus, P. M. and J. K. Shoemaker. Effleurage massage, muscle blood flow and long-term post-exercise strength recovery. Int. J. Sports Med. 16:478-483, 1995.
17. Tschakovsky, M. E., J. K. Shoemaker, and R. L. Hughson. Beat-by-beat forearm blood flow with Doppler ultrasound and straingauge plethysmography. J. Appl. Physiol. 79:713-719, 1995.
18. Van Leeuwen, B. E., J. Lubbers, G. J. Barendsen, and L. De Pater. Calf blood flow and posture: Doppler ultrasound calibrated by plethysmography. J. Appl. Physiol. 72:1668-1674, 1992.
19. Walloe, L. and J. Wesche. Time course and magnitude of blood flow changes in the human quadriceps muscles during and following rhythmic exercise. J. Physiol. 405:257-273, 1988.
20. Wyper, J. and D. R. McNiven. Effects of some physiotherapeutic agents on skeletal muscle blood flow.Physiotherapy 62:83-85, 1976.
21. Zananiri, F. V., P. C. Jackson, M. Halliwell, et al. A comparative study of velocity measurements in major blood vessels using magnetic resonance imaging and Doppler ultrasound. Br. J. Radiol. 66:1128-1133, 1993.


©1997The American College of Sports Medicine