Reflexology is considered to be a form of complementary and alternative medicine (CAM). CAM refers to treatments used either as an adjunct to, or instead of conventional medical care. The House of Lords Select Committee for Science and Technology [
Reflexology is one of the top six forms of CAM used in the UK [
While the general population is spending large sums of money on CAM, health care professionals are reluctant to promote any benefits for such treatments. Reflexology has come under much criticism based predominantly on the dearth of high quality evidence supporting a clear scientific mechanism of action for the treatment. Most of the research carried out in this area has investigated the psychological outcomes from reflexology focusing on qualitative outcomes. Researchers have repeatedly shown that reflexology has a positive effect on quality of life, stress, anxiety, and pain [
To date six literature reviews of reflexology have been carried out [
The exact mechanism of action of reflexology has yet to be confirmed; however, various theories have been proposed and Tiran and Chummun [
This systematic review aimed to assess the quality of evidence from RCTs that have tested changes in physiological or biochemical outcome parameters as a result of reflexology.
The Cochrane guidelines for conducting systematic reviews were followed throughout this review. An inclusion and exclusion protocol was devised to determine which studies were included for evaluation, and the risk of bias (ROB) of each eligible RCT was assessed independently by two members of the review team. The PRISMA guidelines [
GRADE evaluation guidelines [
Study limitations | The quality of the evidence was downgraded if serious study limitations existed such as a lack of allocation concealment, lack of blinding, large loss to follow-up, or randomized trials stopped early for benefit or the selective reporting of outcomes. |
|
|
Inconsistency | The quality of the evidence was downgraded if there was inconsistency in the results, for example, if studies showed varying or different effects of the same intervention. |
|
|
Indirectness | The quality of the evidence was downgraded if there was a level of indirectness in the studies, for example, if interventions had not been compared directly to one another or if the studies investigated a restricted version of the main review question in terms of population, intervention, or outcomes. |
|
|
Imprecision | The quality of the evidence was downgraded if the studies were imprecise in any respect, for example, if they included few participants and few events and thus had wide confidence intervals. |
|
|
Publication bias | The quality of the evidence was downgraded if some element of reporting bias was evident, for example, authors failed to report all the outcomes they set out to or perhaps only reported the positive findings of their study. |
Systematic literature search PRISMA flow diagram.
The following databases were searched from their inception to December 2013: AMED, CAM Quest, CINAHL Plus, Cochrane Central Register of Controlled Trials, Embase, Medline Ovid, Proquest, and Pubmed. Search terms used in various combinations were “reflexology, blood, urine, saliva, plasma, electrolyte, hormone, neurotransmitter, neuroregulator, immune response, lymph, lymphatic system, respiratory function, respiratory function tests, blood pressure, heart rate, foetal heart rate, fetal heart rate and endorphin”. Hand searches of relevant journals and reference lists along with citation tracking were undertaken to ensure comprehensive coverage of all relevant literature. The references of all articles were hand searched. Identified publications were read either as abstracts or full texts.
A total of 19337 articles were retrieved via the initial database and hand searches. Twenty-three studies remained after the exclusion of irrelevant studies (19193), those with no biochemical or physiological outcome measure (47), those articles not available in English (27), duplicates (45), and those where the full text was not available (2). Another five studies were excluded as they were not RCTs, and one further RCT was excluded as it used a combination of reflexology plus back massage for the intervention group [
Risk of bias (ROB) analysis.
Study | Type of study | Adequate sequence generation | Allocation concealment | Adequate blinding-participant | Adequate blinding- clinician | Adequate blinding-outcome assessor | Incomplete outcome data assessment | Selective reporting bias | Other bias | Risk of bias |
---|---|---|---|---|---|---|---|---|---|---|
Jones et al., 2013 [ |
RCT | Low | Low | Low | High | Low | Low | Low | Low | Low |
Hodgson and Lafferty, 2012 [ |
Pilot | Unclear | Unclear | High | High | Low | Low | Low | Low | Unclear |
Jones et al., 2012 [ |
RCT | Low | Low | Low | High | Low | Unclear | Low | Low | Low |
Ruiz-Padial et al., 2012 [ |
RCT | Unclear | Unclear | Low | High | Unclear | Low | Low | Low | Unclear |
Sliz et al., 2012 [ |
RCT | Low | Low | High | High | Unclear | Low | Low | Low | Low |
Hughes et al., 2011 [ |
Pilot | Low | Low | Low | High | Low | Low | Low | Low | Low |
Lu et al., 2011 [ |
RCT | Unclear | Unclear | Low | Unclear | Unclear | Low | Unclear | Unclear | Unclear |
Moeini et al., 2011 [ |
RCT | Low | High | High | High | High | Unclear | Low | Low | Unclear |
Green et al., 2010 [ |
RCT | Low | Low | High | High | Unclear | Low | Low | Unclear | Low |
Holt et al., 2009 [ |
RCT | Low | Low | Low | High | Low | Low | Low | Low | Low |
Mackereth et al., 2009 [ |
RCT | Low | Low | High | High | Unclear | High | Low | Unclear | Unclear |
Hodgson and Andersen, 2008 [ |
RCT | High | High | High | High | Low | Unclear | High | Unclear | High |
Gunnarsdottir and Jonsdottir, 2007 [ |
Pilot | Unclear | Unclear | Unclear | High | High | Unclear | Low | Low | Unclear |
Mc Vicar et al.,2007 [ |
Pilot | Unclear | Unclear | High | High | High | High | High | Low | High |
Wilkinson et al., 2006 [ |
RCT | Unclear | Unclear | High | High | Unclear | High | High | High | High |
Mollart, 2003 [ |
Unclear | High | High | High | High | High | High | Unclear | High | |
Frankel, 1997 [ |
Pilot | Unclear | Low | Unclear | High | Unclear | Low | Low | Low | Unclear |
All results were screened by two independent reviewers and differences in opinion were resolved through discussion to reach a consensus. Exclusions were applied using the criteria in Table
Systematic literature search inclusion and exclusion criteria.
Inclusion criteria | Exclusion criteria |
---|---|
Foot reflexology treatment only | Self-treatment only |
A quantitative biochemical outcome measure | Qualitative outcome measure only |
A quantitative physiological outcome measure | Full text not available |
Randomised controlled trials | Full article not available in English |
Pilot studies | Duplicate |
Owing to the consistent use of blood pressure (BP) and heart rate as outcome measures in the studies collated, a meta-analysis was conducted using Review Manager 5.2 to gain further insight into the effects of reflexology across a wide range of populations (Figures
Meta-analysis and forest plot of systolic blood pressure.
Meta-analysis and forest plot of diastolic blood pressure.
Meta-analysis and forest plot of heart rate.
Twelve randomised controlled trials and five feasibility or pilot randomised controlled trials, involving a total of 697 participants, were reviewed (Table
Table of study characteristics.
Study | Participants | Intervention | Comparison | Dropouts | Outcome measure | Adverse effects | Results | Comments |
---|---|---|---|---|---|---|---|---|
Jones et al., 2013 [ |
12 patients with stable chronic heart failure | 4.5 mins reflexology to heart reflex area (active heart point) (Ingham Method) | 4.5 mins reflexology on gross heel area | No dropouts |
Beat-to-beat cardiovascular parameters HR, BP, stroke index (SI), cardiac output (CO), cardiac index (CI), total peripheral resistance (TPR), baroreceptor up/down events (BarUpEv)/barDwEv), and heart rate variability (HRV) | None | No sig. difference for any outcome for either group | The authors state that participants medication may have masked any potential benefit |
|
||||||||
Hodgson and Lafferty, 2012 [ |
18 older cancer survivors in nursing homes |
|
|
No dropouts |
Salivary cortisol | None reported | Sig. change for both groups, |
No details regarding whether cortisol levels returned to baseline during the washout period were given |
|
||||||||
Jones et al. |
16 healthy volunteers | 4.5 mins reflexology to heart reflex area (active heart point) (Ingham method) | 4.5 mins reflexology on gross heel area | 1 due to data collection issue | Beat-to-beat cardiovascular parameters HR, BP, stroke index (SI), cardiac output (CO), cardiac index (CI), total peripheral resistance (TPR), baroreceptor up/down events (BarUpEv)/(BarDwEv), heart rate variability (HRV) | None reported | Sig. decrease in CI for intervention group, |
Suggests a link between reflexology stimulation to the heart reflex area and cardiac blood flow and circulation |
|
||||||||
Ruiz-Padial et al., 2012 [ |
41 healthy volunteers |
|
|
None reported | BP, BRS, HRV, Inter-beat interval (IBI) | Some pain reported for reflexology group | Increases in interbeat interval, HRV and BRS in all groups. Sig. increase in BP in reflexology group as a function of time | The authors state that the increase in BP in the reflexology group suggest a “co-activation of the two branches of the ANS,” namely, the sympathetic and parasympathetic pathways |
|
||||||||
Sliz et al. |
40 healthy volunteers used a mental stress test to increase stress levels |
|
|
None reported | fMRI, blood oxygen level dependent (BOLD) response (indicated blood flow to areas of activation) | None reported | Positive BOLD response in ACC and PCC brain region for reflexology, Swedish massage and control, |
The ACC and PCC regions of the brain are thought to be linked to emotional response, learning, and memory and are also involved in major depressive disorders (Dervets et al., 2008) |
|
||||||||
Hughes et al., 2011 [ |
25 healthy volunteers using a mental stress test to increase stress levels |
|
|
No dropouts occurred | BP, HR | None reported | Sig. reduction in SBP for intervention and control groups, |
|
|
||||||||
Lu et al., 2011 [ |
37 participants |
|
|
None reported | ECG, BP, HRV, PP, RRI | None reported | Sig. reduction in BP and PP for both groups, sig increase in RRI in reflexology group. |
HRV benefits lasted longer for CAD patients (60 mins) compared with controls (30 mins) |
|
||||||||
Moeini et al. |
50 CABG patients |
|
Usual care | None reported | BP, HR, respiratory rate | None reported | Sig reduction in SBP and DBP in reflexology group, |
|
|
||||||||
Green et al., 2010 [ |
183 Early stage breast cancer (6 weeks post-surgery) | 8 |
(1) Self-initiated support |
Full data sets were obtained for 87 participants, |
Blood lymphocytes (CD profiles) cytokine production (Th1, Th2), prolactin, cortisol, growth hormone | None reported | Sig. increase in CD25 + cells in reflex and massage group compared with baseline. Sig. increase in CD25 + cells between massage and SIS, |
Results for only 47.5% of the participants were reported due to a loss of blood sample in the analysis process |
|
||||||||
Holt et al., 2009 [ |
49 Women with anovulation |
|
Sham reflexology | 9 dropouts | Serum progesterone | None reported | Ovulation occurred in intervention (42%) and sham groups (46%), Pregnancy occurred in intervention (15%) and sham groups (9%), |
The authors stated that the rate of ovulation in this trial was double that expected giving rise to an idea that the sham treatment may also have had an effect on the outcome measures |
|
||||||||
Mackereth et al. |
53 MS patients |
|
Progressive muscle relaxation (PRM) training | 3 dropouts | HR and BP, salivary cortisol | None reported | Sig. decrease before and after treatment and before and after weeks 1–6 for cortisol within reflexology group, |
The variable of interest failed to return to initial levels resulting in problems relating to the ordering of the treatments and these interactions made analysis very difficult to determine |
|
||||||||
Hodgson and Andersen, 2008 [ |
21 dementia sufferers in nursing homes |
|
|
Not stated | BP, pulse, salivary |
None reported | Sig. decrease in salivary |
The authors did not consider the impact of the absence of a washout period on results. Also, no details or numbers of patients receiving hand or foot reflexology were given |
|
||||||||
Gunnarsdottir and Jonsdottir, 2007 [ |
9 Coronary artery bypass graft patients |
|
Cream application to feet (1 min) + 30 mins rest | 2 due to post surgery complication | BP, HR, respiration rate | None reported | Sig reduction in SBP in control group, |
Anxiety levels in the control group were consistently lower in the control group and authors attribute higher anxiety scores to a potential lack of validity of SAI to the Icelandic population |
|
||||||||
Mc Vicar et al., 2007 [ |
30 healthy volunteers |
|
Sitting as a group in a quiet room | Not stated | Salivary melatonin & cortisol, BP, pulse rate | None reported | Significant reduction in pulse and SBD in reflexology group, |
Authors stated that carry over effects and order of treatments due to study design may have affected results. They also, suggest that sitting in a room as a group may have resulted in anxiety for some control participants |
|
||||||||
Wilkinson et al. |
20 Chronic Obstructive Pulmonary Disease (COPD) patients |
|
Friendly visits | 19 participants did not complete all of the study | BP, HR, respiration rate, oxygen saturation, FVC, FEV, vital capacity, peak flow | None reported | Significant pre-postdecrease in HR within reflexology group, |
Peak flows were self-reported and as only one participant completed all of the study, these results are open to bias |
|
||||||||
Mollart, 2003 [ |
69 Pregnant women 30 weeks + gestation with foot oedema |
|
(1) Relaxing reflexology |
Only 20 participants completed all 3 sessions | BP, ankle, and foot circumference measurements | None reported | Nonsignificant reduction in BP for all groups, nonsignificant decreases in ankle and foot measurements | Results from the first treatment session only were analysed due to dropouts |
|
||||||||
Frankel, 1997 [ |
24 healthy participants |
|
(1) Foot massage |
None reported | Baroreceptor reflex sensitivity (BRS), BP, sinus arrhythmia (SA) | None reported | Nonsignificant reduction in BRS for reflexology and FM (60%) compared with no treatment (50%), |
Author suggests a “Neuro theory” may explain the mechanism of action as BRS is under ANS control |
ACC: anterior cingulate cortex; ANS: autonomic nervous system; BarDwEv: baroreceptor down events; BarUpEv: baroreceptor up events; BOLD: blood oxygen level dependent; BP: blood pressure; BRS: baroreceptor reflex sensitivity; CABG: coronary artery bypass graft; CAD: coronary artery disease; CHF: chronic heart failure; CI: cardiac index; CO: cardiac output; COPD: chronic obstructive pulmonary disease; DBP: diastolic blood pressure; ECG: electrocardiogram; FEV: forced expiration volume; FM: foot massage; fMRI: functional magnetic resonance imaging; FVC: forced vital capacity; HR: heart rate; HRV: heart rate variability; IBI: interbeat interval; PCC: posterior cingulate cortex; PEF: peak expiratory flow; PMR: progressive muscle relaxation; PP: pulse pressure; RCT: randomised controlled trial; RRI: R-R interval; SA: sinus arrhythmia; SAI: Spielbergers State Anxiety Inventory; SBP: systolic blood pressure; SI: Stroke Index; SIS: self-initiated support; TPR: total peripheral resistance.
In total, 34 physiological or biochemical outcome measures were analysed in the 17 included studies. Significant within reflexology group changes were recorded for 11 outcome measures. These were blood pressure in five studies, heart rate in three studies, cortisol in two studies, salivary amylase, lymphocyte production, heart rate variability (HRV), R-R interval, pulse pressure, cardiac output, cardiac index (CI), and blood oxygen level dependant (BOLD) response, in one study each (Table
The strength of the evidence presented by the 17 included RCTs was assessed using the GRADE criteria which resulted in the quality of evidence being rated as very low (Table
Assessment of quality using the GRADE system.
Number of studies and participants | Study limitations | Consistency of results | Directness of the evidence | Precision | Reporting bias | Overall quality of the evidence |
---|---|---|---|---|---|---|
17 RCTs and pilot studies (697 participants) start score = 4 | −2 serious limitations due to problems with blinding | −2 serious inconsistency in results between studies | −1 some indirectness as most studies not comparable | −1 some imprecision due to low participant numbers | Unlikely as positive and negative effects found | Very low |
The physiological parameter most commonly investigated within reflexology studies was BP, with 13 studies including this as an outcome measure. BP responded in a positive manner to reflexology in half of the studies: Mollart [
Contrary to all of the other studies in this review Ruiz-Padial et al. [
Changes in heart rate (HR) were investigated in nine studies and positive results were demonstrated by Mc Vicar et al. [
Jones et al. [
fMRI involves the use of MRI to detect visible changes that occur in the brain as a result of some external stimuli. Sliz et al. [
A total of seven studies investigated a biochemical response to reflexology. These studies tended to focus on stress hormones and were correlated with qualitative data on stress and anxiety levels.
Significant within group decreases in cortisol were found by Mackereth et al. [
A meta-analysis was performed on seven papers investigating BP and HR [
The focus of this paper was to review the evidence available from RCTs investigating any quantitative physiological or biochemical outcome measure as a result of reflexology as there has been minimal evaluation to date. Seventeen studies were included for review from a total of 19337 articles identified. A notable limitation, however, is the exclusion of studies not available in English, owing to a lack of availability of translation of these papers, along with these not having sufficient data to perform a meta-analysis; this would have undoubtedly further informed the review outcomes.
Only three RCTs in this review represented a significant change between the reflexology intervention and the control group [
Within group significant changes were observed for eight outcome measures across a range of ten studies, and a further four studies resulted in no significant changes for either the intervention or control group (Table
Reflexology is an area which has seen much growth within the private sector; however, little work has been carried out to determine a possible “mode of action” or how it may be best incorporated into mainstream medical care from a measurable, quantitative perspective informed by high quality evidence. This review highlights that while reflexology has seen minimal investigation over the past 20 years, the hypothesised mechanism of action has been focused on the modulation of the ANS. However, the scope of the research has been very broad, therefore making it difficult to draw any firm conclusions due to the lack of consistency in participant groups, outcome measures, methodologies, and evaluation techniques.
The studies considered in this review were undertaken in a range of countries demonstrating that reflexology is considered to be worthy of investigation and also a socially acceptable form of treatment globally. Nine studies were carried out in the UK, six of which were performed in National Health Service (NHS) hospitals. In total, ten trials were completed in hospitals, four in universities, and three in nursing homes in the following countries: UK (nine studies), USA (two studies), and one each in the following countries: Australia, Canada, Iceland, Iran, Spain, and Taiwan. Although half of the studies carried out in the UK had a low ROB, neither country nor location had any discernible effect on the ROB.
To date relatively small studies have been carried out, with the mean number of participants per study being 41; in a well-designed, 3-armed trial this would result in less than 14 participants per group. Only five of the named studies carried out power calculations prior to recruitment. Mackereth et al. [
Studies involving reflexology have investigated a wide range of outcome parameters using a range of measurement methods. Treatments have been applied to various groups of different ethnicity and gender and with different illnesses, using a wide range of experimental designs, measurements, and analyses. This has led to a very low quality evidence as stated earlier, demonstrating that the RCTs that have been carried out in this area and the results should be treated with caution. It has also demonstrated that overall, with respect to reflexology, low quality studies have been carried out and those where the ROB was low were small and yielded few statistically significant results. However, it is important to note that high ROB is largely due to the lack of participant and therapist blinding. Importantly ROB analysis has demonstrated that studies involving reflexology, and indeed any touch therapy, are complex and difficult to design in order to maintain adequate blinding status for the participants, therapists, and assessors alike. Ultimately, if statistically significant between groups differences from low ROB studies are analysed, this would give rise to only one study [
In the six studies with a low ROB for participant blinding (Table
Indeed it is a limitation of any study involving reflexology to employ a suitable sham treatment that will allow the participants to remain blind to the intervention but have no therapeutic effect, as even gentle pressure on the feet may give rise to stimulation of a reflex point of interest. Eleven of the RCTs used a control treatment which also involved touching the feet. The results of those studies, which used sham reflexology, foot hold, cream application, nonprofessional massage, or Swedish massage, showed fewer differences between groups. Conversely studies where the control measure involved no touch showed more statistically significant changes between groups and more positive changes in outcome measures suggesting that touch plays an integral part in the response to reflexology.
The number of treatments participants received ranged from one 4.5-minute treatment to seven separate one-hour treatments. The duration and frequency of treatments did not have any effect on the results. Likewise, the type of reflexology performed did not appear to have any effect.
A major difficulty for researchers and therapists alike is the various different types of reflexology used. This current review demonstrated this with six different types of reflexology employed, including the Bayly method, Ingham method, Father Josef method, a gentle method developed at Anglia Ruskin University, UK, by Mc Vicar et al. [
A further potential confounder for these studies may lie in the number of therapists providing reflexology treatments during trials. During the research for this review one study boasted 32 therapists taking part; however, the authors did not cite this as a possible limitation of the study. A single reflexologist delivering treatments may generally use the same technique and employ the same treatment type and schedule per client; however, in only seven of the 17 RCTs in this review was it stated that a single therapist carried out all the treatments. Small fluctuations in treatment type or even the mannerisms of the therapist may have an important effect on treatment outcomes.
Only two studies stated whether music was played during treatments [
While this review focused on the physiological and biochemical outcomes recorded for reflexology interventions, the literature available clearly shows that for all of the articles evaluated, whether a significant change was identified or not, reflexology had a positive effect on the health and well-being, quality of life, stress, anxiety, and pain levels of the participants involved. Hodgson and Lafferty [
This review is the first to carry out a meta-analysis of papers investigating BP and HR. These were the most commonly analysed outcome measures, likely due to the ease and noninvasive nature of recording these parameters. While the forest plots appear to show positive benefits in favour of reflexology, cautious interpretation of the results is needed. The clinical heterogeneity of the studies, the mix of healthy and non-healthy populations, the variation in control interventions and the low number of participants (124) would result in a low quality evidence. Furthermore the confidence intervals for all results cross zero and, therefore, the results must not be viewed as significant at this stage until more data becomes available and further analyses can be carried out.
Overall, the review indicates that only three studies resulted in significant between group differences [
While no firm scientific evidence for the effective and efficacious use or “mode of action” of reflexology has been established, it is nonetheless currently being used in healthcare settings around the world including hospices, nursing homes, and maternity departments. In many countries reflexology is associated with the beauty industry or traditional unorthodox medicine. However, attitudes to CAM therapies are shifting towards their use as secondary medical healthcare and integrating them into mainstream medicine. Ten studies in this review were carried out in hospitals suggesting a more positive attitude of health professionals towards the use of CAM therapies as potential adjuncts to mainstream medical healthcare. Thus, this review has implications globally for all health professionals seeking innovative and novel methods for patient care.
There were no serious adverse effects related to any of the treatments for any of the studies. Ruiz-Padial et al. [
Overall reflexology has a positive effect on health, reducing physiological and psychological stress. However, it is as yet unclear how this specialised foot massage exerts its action and whether physiological stress parameters are reduced due to a reduction in psychological stress or vice versa.
This systematic literature review is the first, to our combined knowledge, to specifically analyse the existing data available from RCTs investigating the physiological and biochemical changes associated with reflexology, and it has demonstrated that a range of positive effects can be attributed to the treatment, specifically a reduction in stress parameters. This will inform health care professionals of the evidence base for known benefits and will enhance evidence based decision making at clinical level. It is important to note that, in all of the studies included in this review, where psychological parameters were assessed, a significant improvement in health and well-being was determined and this factor alone had a positive effect on disease outcomes, prognosis, and rehabilitation. None of the studies in this review investigated any long-term effects through follow-up with participants. Therefore, this is an aspect of CAM study design that must be addressed in the future.
It is still unclear from this review precisely how reflexology impacts physiological and biochemical parameters. It illustrates the need for further research into the use, efficacy, and mode of action of reflexology with well designed, high quality RCTs, if indeed RCTs are a suitable mode of investigation. Also, this review highlights the need for further research into the measurable physiological and biochemical effects of reflexology in order to address the concerns of healthcare professionals and thus allowing all patients to benefit from any positive outcomes afforded by this inexpensive, noninvasive, and nonpharmacological therapy. It is more than likely, however, that a number of factors are at work, both of a physiological and psychological nature and that reflexology is what it attests to be, a treatment that seeks to enhance and harmonise the mind, body, and spirit.
No conflict of interests has been declared by the authors.