Hypertension is a global problem that typically presents as an asymptomatic, chronic cardiovascular disease.1 However, blood pressure goals are achieved by only 25% to 40% of patients who take antihypertensive drug treatments.2 Successful treatment of high blood pressure is based on a combination of nonpharmacological and pharmacological therapies.3 Massage therapy is 1 of the most widely accepted complementary and alternative medicine therapies, and it has become a multibillion-dollar industry in the United States, with 8.3% of adults receiving at least 1 massage treatment in 2007.4–64–64–6
Although massage therapy can help reduce blood pressure, previous studies have presented conflicting results. A meta-analysis showed that massage therapy can efficiently reduce blood pressure,7 but the effectiveness of massage in managing hypertension remains unclear, as are the optimal type and regimen of massage therapy for lowering blood pressure. Therefore, a meta-analysis on the effects of massage therapy on blood pressure in patients with hypertension and prehypertension was performed in this study.
The records of 6 electronic databases were searched from their inception until October 31, 2013. The databases examined were the Cochrane Library (Issue 6 of 12, 2013), PubMed, OVID-MEDLINE, CINAHL, Scopus, and EMBASE.
Inclusion and Exclusion Criteria
Original randomized controlled trials were included in this review based on the following criteria: (1) participants—studies involving patients with hypertension or prehypertension were included, with no limitations on participant age, sex, or nationality; (2) intervention—research where massage therapy was used as an intervention, using either the hands or mechanical devices; (3) comparison—studies involving specific control interventions were included (ie, use of a placebo, no treatment, standard care, or any active treatment unrelated to massage therapy); and (4) outcomes—studies were included where the main outcomes of interest involved systolic blood pressure (SBP) and diastolic blood pressure (DBP). Trials were excluded if they were not reported in English.
Two reviewers independently screened the titles and abstracts of potentially eligible articles by using a predetermined search strategy. Subsequently, the full-text manuscripts of candidate studies were retrieved and reviewed. Finally, 9 studies meeting the inclusion criteria were included in the meta-analysis.
Two reviewers extracted data independently according to the following criteria: first author, study year, sample size, blood pressure group (hypertension or prehypertension), sex, mean participant age, experimental and control group intervention, massage type, massage area, treatment dose, follow-up time, and main results. Any discrepancy was discussed until the reviewers reached consensus.
Methodological Quality Assessment
The methodological quality of randomized controlled trials was assessed independently by the 2 raters based on the Cochrane collaboration tool for assessing risk of bias, using the reporting criteria: low, high, or unclear risk of bias. This tool considers the presence of bias resulting from random sequence generation, allocation concealment, blinding of participants and personnel, incomplete outcome data (because of high discontinuation rates, analysis type, or missing data), selective reporting, and other biases.8 The 2 raters independently assessed the risk of bias; when necessary, consensus was determined with the help of a third rater as the arbitrator.
Data Synthesis and Analysis
The mean change in outcomes (ie, SBP and DBP) from baseline to posttest was calculated and the differences between experimental and control groups were estimated. Hedges g and 95% confidence interval (CI) were calculated to determine the effect size. In studies involving multiple control groups, the researchers restricted their analysis to the massage therapy group and a single control group. The I2 statistics and Cochran Q test were used to assess heterogeneity, where an I2 value of 50% or more represented substantial heterogeneity.9 To improve the accuracy of the results, we used a random effect model using variation factors among studies as a correction weight. We performed Egger tests to detect the existence of publication bias among the included studies, with the significance level set at .05.10 Potential publication bias was also examined using a funnel plot. An asymmetric inverted funnel-shape scatter plot of treatment effects against a measure of study size would indicate potential publication bias or the existence of systematic heterogeneity.11 Subgroup analyses according to type of massage, sex, type of control group (active or inactive), whether blinding was involved, treatment sessions, and the massaged area of the body were conducted. Control groups were classified into inactive and active control groups. Inactive controls referred to any condition that was thought to exert no active treatment effect, such as placebo or usual care controls. Active controls were defined as alternative interventions that were believed to exert active treatment effects on the outcome of interest, such as other forms of relaxation treatments. Comprehensive Meta-Analysis Version 2.0 was used to perform the meta-analysis (Biostat Inc, Englewood, New Jersey).
We identified 144 abstracts from 6 databases. Among these, 18 articles were retrieved and reviewed in full, and a further 9 articles were excluded. Articles were excluded only if they did not meet the inclusion criteria (eg, case report, participants were healthy, not relevant). Thus, 9 randomized controlled trials were considered eligible, all of which were reported in English.12–2012–2012–2012–2012–2012–2012–2012–2012–20Figure 1 depicts a summary of the study selection process.
The key data of the 9 included studies are summarized in Table 1. In total, 278 participants were included. The per-group sample size ranged from 8 to 28 participants, and 5 trials12–1612–1612–1612–1612–16 involved only female participants. The number of massage therapy sessions ranged from 1 to 20, with session durations of 10 to 60 minutes. The total exposure over the study period was 20 to 300 minutes. Five studies involved using Swedish massage therapy,12,13,15,17,1812,13,15,17,1812,13,15,17,1812,13,15,17,1812,13,15,17,18 1 study involved therapeutic chair massage therapy,16 1 study used light-touch and stroking-type massage,19 and 2 studies were unclear about the massage type.14,2014,20 The areas massaged were categorized as the whole body, the head, the neck, and back area (including the face, neck, shoulders, upper chest, and back). In terms of control conditions, 4 studies used active control groups16–18,2016–18,2016–18,2016–18,20 and 5 studies used inactive control groups.12–15,1912–15,1912–15,1912–15,1912–15,19Table 2 shows a summary of the methodological quality of results. We minimized the potential for publication bias by conducting a thorough literature search. The generated funnel plots and Egger test results for the primary outcomes exhibited symmetry and indicated no bias.
Effects of Massage Therapy on Blood Pressure
The mean change in outcomes (ie, SBP and DBP) from baseline to posttest was calculated, and the differences between experimental and control groups were estimated. One study did report insufficient data for calculating the mean difference.17 The mean difference was thus calculated using the Vernier Caliper by measuring the figure that showed the baseline and posttest data for the massage group and the comparison group. Overall, the pooled results from the 9 studies indicated that massage contributed to a significantly greater reduction in both SBP (mean difference, −7.39 mm Hg) and DBP (mean difference, −5.04 mm Hg) when compared with the control treatment.
As shown in Figure 2A and B, massage therapy had a significant effect on SBP and DBP. The effect size of SBP was −0.728 (95% CI, −1.182 to −0.274; P = .006), and that of DBP was −0.334 (95% CI, −0.560 to −0.107; P = .004). Heterogeneity was observed among the studies for SBP (χ2 = 27.599, df = 8, P = 0.001; I2 = 71%). For DBP, the test for heterogeneity showed homogeneity among studies (χ2 = 5.202, df = 8, P = 0.736; I2 = 0%).
Subgroup Analysis: Systolic Blood Pressure
Massage therapy significantly improved SBP in patients in the following: (1) 5 studies using Swedish massage therapy; (2) 5 studies that included all-female samples; (3) 4 studies that included both female and male samples; (4) 4 studies that included active control groups; (5) 5 studies that included inactive control groups; (6) 4 studies that blinded outcome assessors; (7) 7 studies involving multiple massage sessions; and (8) 6 studies of massage therapy involving the head, neck, or back. Moderator analysis indicated that the studies that blinded outcome assessors yielded a larger effect size than did those that did not involve blinding (P < .001); studies that used an inactive control group had a significantly larger effect size than did those that included an active control group (P = .003). The results are illustrated in Figure 3A–H and summarized in Table 3. Meta-regression revealed that the effect of massage on SBP was not significantly associated with total massage time (P = .058).
Subgroup Analysis: Diastolic Blood Pressure
Massage therapy significantly improved DBP in patients in the following: (1) 4 studies that used non-Swedish massage therapy; (2) 4 studies that enrolled both female and male participants in the sample; (3) 4 studies involving active control groups; (4) 5 studies with blinded outcome assessors; (5) 7 studies reporting multiple treatment sessions; (6) 3 studies that involved whole body massage; and (7) 6 studies that reported head, neck, or back massage therapy. As shown in Table 4, moderator analysis was nonsignificant for all moderators except for the moderating effect of control conditions. Studies that used an active control group reported significantly larger effect sizes than did those involving an inactive control group (P = .011). Meta-regression results revealed a nonsignificant association between doses and effect size (P = .735).
Overview of Findings
When the results of all 9 studies in this review were pooled, we observed that massage therapy was associated with a significant reduction in SBP and DBP in patients with hypertension and prehypertension. Massage therapy contributed to a 7–mm Hg reduction in SBP and 5–mm Hg reduction in DBP as compared with control conditions. A recent meta-analysis reported that administering blood pressure-lowering treatment did not contribute to a significant reduction in cardiovascular risk among patients with mild hypertension.21 However, linear associations were observed between blood pressure and angina, myocardial in farction, or subarachnoid hemorrhage.22 Treatments that reduce blood pressure, therefore, can potentially reduce the risk of various cardiovascular diseases. Moreover, a lifetime burden of hypertension has been reported. Compared with participants with normal blood pressure, those with blood pressure equal to or higher than 140/90 mm Hg exhibited increased overall cardiovascular disease at 30 years of age, and they developed cardiovascular disease 5 years earlier.22 A 7–mm Hg reduction in SBP and 5–mm Hg reduction in DBP are thus potentially of clinical significance.
Subgroup analyses of the type of massage revealed potentially paradoxical findings for SBP and DBP. Swedish massage therapy effectively reduced SBP, whereas non-Swedish massage therapy effectively reduced DBP. Raised SBP is modulated partially by a change in balance of sympathetic activity and parasympathetic activity.23 Swedish massage is known for its therapeutic relaxation effects, and hypertension is associated with stress.12 Swedish massage is classified as a superficial type of massage, consisting of 5 primary stroking actions that stimulate the circulation of blood through the soft tissues of the body.24,2524,25 The significant reduction in heart rate induced by Swedish massage supports the link between enhanced parasympathetic activity and reduced SBP.25 On the other hand, DBP was influenced by peripheral vascular resistance26,2726,27; the mechanical techniques of the Swedish massage may not have direct effect on changes in peripheral vascular resistance. Of note, 2 studies14,2014,20 that we classified as using non-Swedish massage therapy reported inadequately detailed information to determine the exact treatment components. Thus, misclassification of massage type might have occurred.
Significant reductions were observed in both SBP and DBP when massage therapy was compared with active control treatments. The superior effect of massage therapy in comparison with active control treatments further supports the notion that massage therapy exerts specific treatment effects in reducing SBP and DBP. Furthermore, our results provide evidence of moderating factors that could influence the effectiveness of treatment effects. For example, massage therapy involving the head, neck, or back was effective in reducing both SBP and DBP. Significant reductions in SBP and DBP were observed for both sexes. Multiple sessions of massage therapy were more effective than 1 session alone. When considering the total exposure of massage (in minutes) as a continuous variable, the treatment effect was not significantly associated with dose. However, a marginally significant association between the dose and effect size was observed for SBP (P = .058). Further studies are needed to investigate the optimal dose of massage therapy.
Previous meta-analyses or systematic reviews have assessed the effect of massage therapy28–3028–3028–30; however, those reviews were limited in scope, and the participants and outcome measures differed from those discussed in this study. Previous reviews have involved healthy men, hospice patients, post–myocardial infarction patients, breast cancer patients, or shoulder pain patients. The outcome measures in those studies included pain, anxiety, depression, stress, fatigue, and physiological components of relaxation. Only 1 meta-analysis concluded that applied massage therapy reduced blood pressure; however, the participants reviewed in that study included healthy adults, hypertensive adults, healthy women, patients undergoing cardiac catheterization, and soldiers.7
For SBP, heterogeneity was observed among the reviewed studies. To explore the factors contributing to heterogeneity, we examined whether the type of massage, the sex of participants, control group (active or inactive), quality of studies (blinding of outcome assessors), treatment sessions, and the massaged area of the body affected the effectiveness of the treatment. As expected, moderator analyses revealed that the beneficial effect of massage therapy for SBP was greater for studies involving inactive control groups compared with those that had active control groups, and larger for studies with better quality (blinding assessors). Numerous potentially interesting moderator variables remain unexamined in primary massage therapy research, and only 1 study has determined change in blood pressure and associated factors.31 Systolic blood pressure seemed to decrease after Swedish massage therapy, but the association between changes in blood pressure and the massaged area of the body, duration of the massage, or pressure applied was not statistically significant.31 Previous studies lack a clear definition of the types of massage therapy, and they have a limited number of participants. Unfortunately, the use of different massage therapy techniques by different practitioners may affect the associated therapeutic outcomes.
Limitations of the Review
Some limitations were encountered while conducting this systematic review. This review may have been affected by the heterogeneity in massage therapy characteristics, including frequency, duration, number of sessions, and massage technique. The number of included studies was small, with each study also having a relatively small sample. Bias was also possible, and blinding subjects and therapists for treatments is difficult. However, assessor blinding and allocation concealment are critical for reducing bias. Only 4 studies have used blinding in their outcome assessments.
Moderator analyses were performed to explain gradients in treatment effect among subgroups. However, because moderator variables cannot be assumed to be statistically independent,32 inferences drawn from these analysis results should be interpreted with caution.
Methodological Recommendations for Future Studies
Future studies considering the mechanism of massage therapy are necessary. To strengthen our evidence, future studies should consider extending this review after the publication of new eligible randomized controlled trials concerning the effect of massage therapy on hypertension. However, future research must include larger samples to increase their explanatory power. In addition, blinding the researcher to the data to prevent possible bias will be critical. Moreover, most studies measured BP immediately before and after massage. Future studies evaluating the intermediate or long-term effects of massage on blood pressure are needed. Finally, future designs should consider the massage type, regimen, massaged area, patient type (prehypertension versus established hypertension), and application of massage therapy for nonpharmacological control of hypertension.
Significant differences between the control and massage therapy groups in SBP and DBP were identified, with massage therapy shown to reduce both parameters. Massage therapy is a simple nonpharmacological stress-reduction method that appears to positively impact blood pressure. This review of massage therapy for hypertension provides crucial guidelines for future research.
What’s New and Important
- Massage therapy significantly reduces SBP and DBP compared with control.
- Swedish massage therapy causes the largest reduction in SBP in patients with hypertension and prehypertension.
- Future studies of higher quality are needed, with a particular emphasis on Swedish massage therapy.
1. Chockalingam A. World Hypertension
Day and global awareness. Can J Cardiol
. 2008; 24(6): 441–444.
2. Burnier M. World Health Organization/International Society of Hypertension
, Guidelines: blood pressure
control and the implementation of guidelines in clinical practice: can we fill the gap? J Hypertens
. 2002; 20(7): 1251–1253.
3. Glynn LG, Murphy AW, Smith SM, Schroeder K, Fahey T. Self-monitoring and other non-pharmacological interventions to improve the management of hypertension
in primary care: a systematic review. Br J Gen Pract
. 2010; 60(581): e476–e488.
4. Suzuki S, Ohta T. Non-pharmacological treatment of hypertension
in the elderly. Nihon Rinsho
. 2005; 63(6): 1010–1015.
5. Moraska A, Pollini RA, Boulanger K, Brooks MZ, Teitlebaum L. Physiological adjustments to stress measures following massage
therapy: a review of the literature. Evid Based Complement Alternat Med
. 2010; 7(4): 409–418.
6. Prasad K, Sharma V, Lackore K, Jenkins SM, Prasad A, Sood A. Use of complementary therapies in cardiovascular disease. Am J Cardiol
. 2013; 111(3): 339–345.
7. Moyer CA, Rounds J, Hannum JW. A meta-analysis
therapy research. Psychol Bull
. 2004; 130(1): 3–18.
8. Higgins JPT, Green S (Eds). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration
. 2011. http://www.cochrane-handbook.org
. Accessed October 31, 2013.
9. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ
. 2003; 327: 557–560.
10. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis
detected by a simple, graphical test. BMJ
. 1997; 315: 629–634.
11. Sterne JA, Egger M. Funnel plots for detecting bias in meta-analysis
: guidelines on choice of axis. J Clin Epidemiol
. 2001; 54(10): 1046–1045.
12. Supa’at I, Zakaria Z, Maskon O, Aminuddin A, Nordin NA. Effects of Swedish massage
therapy on blood pressure
, heart rate, and inflammatory markers in hypertensive women. Evid Based Complement Alternat Med
. 2013; 2013: 171852.
13. Givi M. Durability of effect of massage
therapy on blood pressure
. Int J Prev Med
. 2013; 4(5): 511–516.
14. Ju MS, Lee S, Bae I, Hur MH, Seong K, Lee MS. Effects of aroma massage
on home blood pressure
, ambulatory blood pressure
, and sleep quality in middle-aged women with hypertension
. Evid Based Complement Alternat Med
. 2013; 2013:403251.
15. Moeini M, Givi M, Ghasempour Z, Sadeghi M. The effect of massage
therapy on blood pressure
of women with pre-hypertension
. Iran J Nurs Midwifery Res
. 2011; 16(1): 61–70.
16. Jefferson LL. Exploring effects of therapeutic massage
and patient teaching in the practice of diaphragmatic breathing on blood pressure
, stress, and anxiety in hypertensive African-American women: an intervention study. J Natl Black Nurses Assoc
. 2010; 21(1): 17–24.
17. Olney CM. The effect of therapeutic back massage
in hypertensive persons: a preliminary study. Biol Res Nurs
. 2005; 7(2): 98–105.
18. Olney CM. Determining long term effects and dosage of back massage
for persons with elevated blood pressure
: a clinical trial. South Online J of Nurs Res
. 2008; 8(2): 66–76.
19. Plaugher G, Long CR, Alcantara J, et al. Practice-based randomized controlled-comparison clinical trial of chiropractic adjustments and brief massage
treatment at sites of subluxation in subjects with essential hypertension
: pilot study. J Manipulative Physiol Ther
. 2002; 25(4): 221–239.
20. Hernandez-Reif M, Field T, Krasnegor J, Theakston H, Hossain Z, Burman I. High blood pressure
and associated symptoms were reduced by massage
therapy. J Psychosom Obstet Gynaecol
. 2000; 4: 31–38.
21. Diao D, Wright JM, Cundiff DK, Gueyffier F. Pharmacotherapy for mild hypertension
. Cochrane Database Syst Rev
. 2012; 8.
22. Rapsomaniki E, Timmis A, George J, et al. Blood pressure
and incidence of twelve cardiovascular diseases: lifetime risks, healthy life-years lost, and age-specific associations in 1.25 million people. Lancet
. 2014; 383(9932): 1899–1911.
23. Biaggioni I. Interventional approaches to reduce sympathetic activity in resistant hypertension
: to ablate or stimulate? Hypertension
. 2012; 59(2): 194–195.
24. Aourell M, Skoog M, Carleson J. Effects of Swedish massage
on blood pressure
. Complement Ther Clin Pract
. 2005; 11(4): 242–246.
25. Cowen VS, Burkett L, Bredimus J. A comparative study of Thai massage
and Swedish massage
relative to physiological and psychological measures. J Bodyw Mov Ther
. 2006; 10: 266–275.
26. Pinto E. Blood pressure
and ageing. Postgrad Med J
. 2007; 83: 109–114.
27. Guo X, Pantoni L, Simoni M, et al. Blood pressure
components and changes in relation to white matter lesions: a 32-year prospective population study. Hypertension
. 2009; 54(1): 57–62.
28. Kong LJ, Zhan HS, Cheng YW, Yuan WA, Chen B, Fang M. Massage
therapy for neck and shoulder pain: a systematic review and meta-analysis
. Evid Based Complement Alternat Med: eCAM
. 2013; 2013: 613279.
29. Smith JM, Sullivan SJ, Baxter GD. Massage
therapy: more than a modality. N Z J Physiother
. 2010; 38(2): 44–51.
30. Labyak SE, Metzger BL. The effects of effleurage backrub on the physiological components of relaxation: a meta-analysis
. Nurs Res
. 1997; 46(1): 59–62.
31. Cambron JA, Dexheimer J, Coe P. Changes in blood pressure
after various forms of therapeutic massage
: a preliminary study. J Altern Complement Med
. 2006; 12: 65–70.
32. Lipsey MW. Those confounded moderators in meta-analysis
: good, bad, and ugly. Ann Am Acad Pol Soc Sci
. 2003; 587: 69–81.
Keywords:Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved
blood pressure; hypertension; massage; meta-analysis