Effects of a 10-Minute Back Rub on Cardiovascular Responses in Healthy Subjects

Effects of a 10-Minute Back Rub on Cardiovascular Responses in Healthy Subjects

Tommy Boone

(Accepted for publication July 28, 2000)

Abstract: This study determined the cardiovascular responses to a 10-minute back rub. Twelve healthy, college-age males and females (mean age = 22 years) volunteered to participate as subjects. Using an ABA design, the subjects tested for 10 minutes (Control #1) on a padded plinth lying on one side. During the Treatment period, the back rub was administered, which was followed by Control #2. Oxygen consumption ([VO.sub.2]) was determined via the Medical Graphics CPX/D metabolic analyzer, which also estimated cardiac output (Q) using the [CO.sub.2] rebreathing (equilibrium) method. A repeated measures ANOVA was performed to statistically compare the cardiovascular responses across the three periods. The back rub, when compared to Control #1, had no significant effect on [VO.sub.2], but the central and peripheral components of [VO.sub.2] were changed. Cardiac output was decreased as a result of the decreased stroke volume (SV), as a function of the increased peripheral vascular resistance (PVR). We also found an increase in the extraction of oxygen (a-v[O.sub.2] diff) in the peripheral tissues. These results indicate that the [VO.sub.2] response during the back rub was achieved by reciprocal central (SV,Q) and peripheral (a-v[O.sub.2] diff) adjustments. Following the back rub, (i.e., Control #2 vs. Treatment), the decrease in [VO.sub.2], [VCO.sub.2] [Ve, and a-v[O.sub.2] diff appears to indicate that it was effective in inducing relaxation. Since HR, SV, and Q were unchanged, the [VO.sub.2] response was a result of the decreased a-v[O.sub.2] diff. Hence, the findings suggest certain positive implications for the health care industry.

Massage is assumed to have developed from Chinese Folk medicine. By 2000 BC, its uses for healing purposes were known throughout India, Egypt, Persia, and Japan. The Greeks prescribed massage for their athletes, and Hippocrates left behind prescriptions for massage (Tappan, 1988). The benefits of massage and back rubs are said to include a decrease in pain, relaxation, and improvement in circulation. However, the physiological reports in the literature are conflicting.

Researchers report a decrease in heart rate (Madison, 1973), an increase in heart rate (Tyler et al., 1990), no change in heart rate or blood pressure (Kaufmann, 1964; Longworth, 1982; Bauer and Dracup, 1987), and a decrease in blood pressure and heart rate (Fakouri and Jones, 1987). In spite of the discrepant results, massage and back rubs are examples of touching by caregivers that is perceived as communicating care and concern (Barnett, 1972; McCorkle, 1974). Touch and back massages are postulated to bring about relaxation (Harrison, 1986; Carnahan, 1988), which is hypothesized to result in a decrease in sympathetic nervous system activity and an increase in parasympathetic activity (Benson et al., 1974).

Because the cardiovascular findings on back rubs are not conclusive enough to guide caregivers, and because of the need to address the central (heart rate, HR; stroke volume, SV; cardiac output, Q) and peripheral (arteriovenous oxygen difference, a-v[O.sub.2] diff) components of the change in oxygen consumption ([VO.sub.2]), the present study was therefore designed to compare [VO.sub.2] and the central and peripheral components that contribute to it at rest with and without the back rub. We also investigated the effects of the back rub on selected hemodynamic responses (systolic blood pressure, SBP; diastolic blood pressure, DBP; mean arterial pressure, MAP; and double product, DP) at rest. Our null hypothesis was that in healthy adults, without inherent or induced cardiovascular dysfunction, the responses during rest would not change with the back rub.


Subjects and Experimental Design

Twelve physically healthy college-age students (6 male and 6 female) volunteered to participate in this study. Their physical characteristics are mean age = 22.4 [+ or -] 2 years, height = 177 [+ or -] 7 cm, and body mass 78 [+ or -] 10 kg. The experimental design was explained to all subjects, and informed consent was obtained. On Day One of the two-day protocol, the subjects went to the Exercise Physiology Laboratory where they were familiarized with the data collection procedures and experimental design. The monitors were turned on and calibration checks were completed. During Day Two, the subject and investigators went to the laboratory to collect data. The room temperature was measured and recorded during the calibration check.

Using an ABA design, the subject was assisted to a comfortable position on a padded plinth lying on one side so that the back was accessible for the back rub. The subject rested for 10 minutes (Control #1). The investigator then immediately administered a 1 O-minute back rub using warm unscented massage oil as the lubricant. The back rub (Treatment) consisted of basic Swedish massage strokes (Jensen-Nelson, 1948). At first, the subject’s skin was stroked with the hands facing palms downward (effleurage). Then, during the last minute of the rub, the subject’s skin was gently lifted with a C-shaped motion of the hand (petrissage). Following the treatment, the subject remained motionless for a 10-minute quiet period (Control #2).

Measurements and Calculations

During the three, 10-minute periods, breath-by-breath measurements of [VO.sub.2], carbon dioxide production ([VCO.sub.2]), and expired minute ventilation (Ve) were made using an automated system (Medical Graphics, model CPX/D, St. Paul, MN). The CPX/D was calibrated by using the system’s standard computer programs and precision-grade gas mixtures. The Med-Graphics prevent[TM] pneumotach (designed to measure flow and volume directly at the mouth) was calibrated by using a 3-liter calibration syringe. Multiple syringe strokes were performed over a range of flows to check the linearity of the pneumotach response. Calibrations were performed with the pneumotach attached to the non-rebreathing valve.

Oxygen consumption and respiratory exchange ratio (RER) data were averaged across the second, five minutes of each 10-minute period. Heart rate data were derived from the electrocardiogram during the last 15 seconds of each minute, using the Physio-Control Lifepack 7. The data were averaged across the last five minutes of each period. Cardiac output (Q) was estimated during the 10th minute of each period using the indirect [CO.sub.2] rebreathing procedure (Beekman et al., 1984; Collier, 1956; Kirby, 1985). Arterial [CO.sub.2] (Pa[CO.sub.2]) was derived from the end-tidal pulmonary [CO.sub.2] (Pet[CO.sub.2]) (Boone et al., 1990; Defares, 1958). Mixed venous pulmonary [CO.sub.2] (Pv[CO.sub.2]) was derived from the [CO.sub.2] rebreathing (bag) procedure (Ferguson et al., 1968). The CPX/D displayed the [CO.sub.2] signal graphically to ensure the Pv[CO.sub.2] equilibrium.

During the rebreathing maneuver, each subject was instructed to breathe in fully from the rebreathing anesthesia bag via a recirculation circuit in accordance with an instructed rate and depth to optimize the mixed venous [CO.sub.2] lung-bag concentration. Within 10 to 15 seconds of rebreathing Q was determined and recorded (Godfrey and Wolf, 1972).

Arteriovenous oxygen difference was estimated by dividing [VO.sub.2] by Q. Stroke volume (SV) was estimated by dividing Q by HR. Mean arterial pressure (MAP) was calculated by adding one-third of the pulse pressure (SBP-DBP) to the diastolic blood pressure (DBP). Peripheral vascular resistance (PVR) was estimated by dividing MAP by Q. Double product (DP) was determined by multiplying .01 x HR x SBP (Kitamura et al., 1972; Fox, 1996).

Statistical Analysis

The results are given as means [+ or -] SD; P [is less than] 0.05 was taken to represent statistical significance (Table 1). A repeated measures ANOVA was performed to statistically compare the physiological responses during Control #1, Treatment, and Control #2. When significant F-ratios were obtained, the post hoc Scheffe comparisons for treatment were used to identify significant differences between paired group means.

Table 1. Cardiovascular Responses to the 10-Minute Back Rub

Treatment (back

Control # 1 (A) rub) (B)

[VO.sub.2] 0.37 [+ or -] 0.08 0.35 [+ or -] 0.09

1/min A-C(**) B-C(**)

HR 70 [+ or -] 17 70 [+ or -] 18


SV 90 [+ or -] 35 77 [+ or -] 32

ml/beat A-B(**)


Q 6.1 [+ or -] 1.6 5.0 [+ or -] 1.3

1/min A-B(**)


a-v[O.sub.2] diff 6.0 [+ or -] 1.3 7.2 [+ or -] 1.5

ml/100 ml A-B(**) B-C(**)

PVR 13.4 [+ or -] 4 16 [+ or -] 5

mmHg/l/min A-B(**)


MAP 82 [+ or -] 5 82 [+ or -] 5


SBP 114 [+ or -] 5 114 [+ or -] 4


DBP 68 [+ or -] 6 68 [+ or -] 6


DP 77 [+ or -] 20 77 [+ or -] 22

Ve 10 [+ or -] 3 12 [+ or -] 4

l/min A-B(**) B-C(**)

[VCO.sub.2] 0.34 [+ or -] 0.9 0.33 [+ or -] 1.0

l/min A-C(**) B-C(**)


Control # 2 (C) & Prob

[VO.sub.2] 0.31 [+ or -] 0.08 6.71

l/min 0.01(*)

HR 70 [+ or -] 15 0.13

beats/min 0.88

SV 78 [+ or -] 29 2.19

ml/beat 0.14

Q 5.2 [+ or -] 1.4 3.84

1/min 0.04(*)

a-v[O.sub.2] diff 6.0 [+ or -] 1.3 4.25

ml/100 ml 0.03(*)

PVR 16 [+ or -] 5 3.56

mmHg/1/min 0.05(*)

MAP 84 [+ or -] 5 0.16

mmHg 0.82

SBP 116 [+ or -] 5 0.26

mmHg 0.68

DBP 70 [+ or -] 6 0.08

MmHg 0.92

DP 82 [+ or -] 18 0.06


Ve 10 [+ or -] 3 5.17

l/min 0.02(*)

[VCO.sub.2] 0.31 [+ or -] 0.8 3.23

l/min 0.05(*)


The physiological responses to the back rub are presented in Table 1. From Control # 1 to Treatment, there were significant (p [is less than] 0.05) increases in a-v[O.sub.2] diff, PVR, and Ve and significant decreases in SV and Q. From Treatment to Control #2, there were significant decreases in [VO.sub.2], a-v[O.sub.2] diff, [VCO.sub.2], and Ve. From Control #1 to Control #2, there were significant decreases in [VO.sub.2], SV, Q, [VCO.sub.2], and DP, and a significant increase in PVR.


We showed that [VO.sub.2] remained constant during the back rub. The energy expended during the 10-minute back rub was the same as the baseline measure (Control #1). This finding is consistent with a previous report by Boone and Cooper (1995). As might be expected, the production of carbon dioxide ([VCO.sub.2]) was therefore unchanged.

However, although the back rub had no significant effect on [VO.sub.2], the central and peripheral circulatory components of [VO.sub.2] were changed at rest. Cardiac output was decreased as a result of the decreased SV, as a function of the increased afterload (PVR). We also found an increase in the extraction of oxygen in the peripheral tissues (a-v[O.sub.2] diff). These results demonstrate that the [VO.sub.2] response during the back rub was achieved by reciprocal central (SV, Q) and peripheral (a-v[O.sub.2] diff) adjustments (Rowell, 1993).

There was no significant effect on HR (or blood pressure), which agrees with the reports by Longworth (1982) and Kaufmann (1964). However, our hemodynamic data are contrary to the results of Madison (1973), Fakouri and Jones (1987), and Tyler et al. (1990). With respect to DP (i.e., the product of SBP and HR), as an index of relative cardiac work, the back rub had no effect on coronary blood flow or myocardial oxygen consumption (Kitamura et al., 1972).

What is particularly interesting is the effect of the back rub on [VO.sub.2] during Control#2. Our data demonstrate that the post-back rub [VO.sub.2] response was decreased, as was also [VCO.sub.2]. Because there was no increase in Q after the back rub, there was also a decrease in a-v[O.sub.2] diff. The central components (HR, SV, Q) of the [VO.sub.2] response were unchanged from Treatment to Control #2, which was also the case with the hemodynamic data (SBP, DBP, MAP, DP).

The implication of the decrease in metabolism after the back rub is interesting, particularly since it appears to have resulted in a post-treatment 11% decrease in metabolism. The decrease in [VO.sub.2] appears to be consistent with the relaxation response as a direct function of the decrease in tissue extraction of oxygen, which is further supported by the decrease in [VCO.sub.2] and Ve.

But, to be sure that the decrease in [VO.sub.2] was not a function of how long the subjects were in the lying position, the investigators evaluated 10 subjects who were not part of the original experiment. Their physical characteristics and metabolic data are presented in Table 2. Note that, when using the same research design of three, 10-minute (connected) periods of lying on the side, the subjects’ [VO.sub.2] and Ve responses were unchanged (p [is greater than] 0.05). In other words, without the back rub, the metabolic data were reproducible from the first, 10-minute period to the second and to the third. These findings further support the fact that the back rub had a significant effect on the subjects’ metabolism while lying on their side for 30 minutes.

Table 2. Subject Characteristics (A) and Metabolic Responses (B) to

Three, Consecutive 10-Minute Periods without the Back Rub (N=9)

(A) Age (yr) Weight (kg)

21 [+ or -] 2 74 [+ or -] 10

(B) Minutes Minutes

(6-10) (16-20)

[VO.SUB.2] 0.23 [+ or -] 0.04 0.22 [+ or -] 0.04


Ve 7.8 [+ or -] 2 8.0 [+ or -] 1


(A) Height (cm)

173 [+ or -]8

(B) Minutes F-ratio

(26-30) & Prob

[VO.SUB.2] 0.24 [+ or -] 0.05 1.75

l/min 0.21

Ve 8.6 [+ or -] 3 0.98

l/min 0.40

Hence, aside from touch perceived by the subjects as communicating care and concern (Barnett, 1972; McCorkle, 1974), it appears that it (i.e., in the form of a back rub) may also contribute to inducing a type of relaxation. This finding is consistent with the decrease in the subjects’ SV, Q, and resting metabolism ([VO.sub.2]) during the back rub even though HR and SBP were unchanged during and after the back rub.

Although it might be expected that PVR would decrease during and following the back rub (to support the notion of relaxation), given the decrease in [VO.sub.2], that is not the case in this study. It appears that PVR increased to keep Q from decreasing beyond a certain point since the subjects’ sympathetic drive (HR) was unchanged. The slight increase in vasoconstriction seen in these subjects is a direct reflection of a positive peripheral adjustment to maintaining normal cardiovascular function.

In conclusion, this study demonstrates statistically significant changes in the central and peripheral components of [VO.sub.2] during the back rub. There were no significant changes in the subjects’ hemodynamic responses. Following the back rub, the decrease in [VO.sub.2] appears to indicate that it was effective in inducing relaxation (which is in agreement with the decrease in Ve). The findings suggest certain positive implications for the health care industry. Clearly, the back rub (or similarly, massage therapy) is an inexpensive and cost-effective intervention with tangible results.


[1.] Beekman, R.H., V. Katch, C. Marks and A.P. Rocchini. Validity of [CO.sub.2]-rebreathing cardiac output during rest and exercise in young adults. Medicine and Science in Sports and Exercise 16: 306-310, 1984.

[2.] Boone, T., D. Redondo and C. Cortes. Reproducibility of the [CO.sub.2] rebreathing method to estimate cardiac output at rest and during exercise. Abstract. Southeast Chapter of the American College of Sports Medicine February 1-3, 1990.

[3.] Barnett, K. A survey of the current utilization of touch by health team personnel with hospitalized patient. International Journal of Nursing Studies 9:195-209, 1972.

[4.] Bauer, W.C. and K.A. Dracup. The influence of back massage on autonomic functions. Physical Therapy 50: 1679-1691, 1987.

[5.] Benson, H., J.F. Beary and M.P. Carol. The relaxation response. Psychiatry 37: 37-46, 1974.

[6.] Boone, T. and R. Cooper. The effect of massage on oxygen consumption at rest. American Journal of Chinese Medicine 23: 37-41, 1995.

[7.] Carnahan, B. A gentle touch — Massage therapy for people in painful times. Boulder, CO: Boulder County Hospice, 1998.

[8.] Collier, C.F. Determination of mixed venous [CO.sub.2] tension by rebreathing. Journal of Applied Physiology 9: 25-29, 1956.

[9.] Defares, J. Determination of Pv[CO.sub.2] from the exponential [CO.sub.2] rise during rebreathing. Journal of Applied Physiology 13: 159-164, 1958.

[10.] Fakouri, C. and P. Jones. Relaxation Rx: Slow stroke back rub. Journal of Gerentological Nursing 13: 32-35, 1987.

[11.] Ferguson, R.J., J.A. Faulkner, S. Julius and J. Conaway. Comparison of cardiac output determined by [CO.sub.2] rebreathing and dye-dilution methods. Journal of Applied Physiology 25: 450-454, 1968.

[12.] Fox, S.I. Human physiology. Chicago, IL: Wm. C. Brown, 1996.

[13.] Godfrey, S. and E. Wolf. An evaluation of rebreathing methods for measuring mixed venous [PCO.sub.2] during exercise. Clinical Science 42: 345-353, 1972.

[14.] Harrison, A. Getting the massage, Nursing Times 82: 34-35, 1986.

[15.] Jensen-Nelson, K.L. Massage in nursing (3rd ed). New York, NY MacMillan, 1948.

[16.] Kaufmann, M.A. Autonomic responses as related to nursing comfort measures. Nursing Research 13: 45-55, 1964.

[17.] Kirby, T.E. The [CO.sub.2] rebreathing technique for determination of cardiac output: Part I. Journal of Cardiac Rehabilitation 5: 97-101, 1985.

[18.] Kitamura, K., C.R. Jorgensen, F.L. Gobel, H.L. Taylor and Y. Wang. Hemodynamic correlates of myocardial oxygen consumption during upright exercise. Journal of Applied Physiology 32: 516-522, 1972.

[19.] Longworth, J.C.D. Psychophysiological effects of slow stroke back massage in normotensive females. Advances in Nursing Science 4: 44-61, 1982.

[20.] Madison, A.S. Psychophysiological responses of female nursing home residents to back massage: An investigation of the effect of one type of touch. (Doctoral dissertation, University of Maryland). Dissertation Abstracts International 35:941B, 1973.

[21.] McCorkle, R. Effects of touch on seriously ill patients. Nursing Research 23: 125-132, 1974.

[22.] Rowell, L.B. Human cardiovascular control. New York, NY: Oxford University Press, 1993.

[23.] Tapan, F. Healing massage technique: Holistic, classic, and emerging methods. Norwalk, CT: Appleton and Lange, 1968.

[24.] Tyler, D.O., E.H. Winslow, A.P. Clark and K.M. White. Effects of a 1-minute back rub on mixed venous oxygen saturation and heart rate in critically ill patients. Heart & Lung 19: 562-565, 1990.

Tommy Boone, Michelle Tanner and Angela Radosevich Department of Exercise Physiology, The College of St. Scholastica, 1200 Kenwood Avenue, Duluth, MN 55811

COPYRIGHT 2001 Institute for Advanced Research in Asian Science & Medicine

COPYRIGHT 2008 Gale, Cengage Learning