Tuesday, June 21, 2016

The Therapeutic Effect of Balneotherapy: Evaluation of the Evidence from Randomised Controlled Trials


There is widespread popular belief that balneotherapy is effective in the treatment of various diseases. Balneotherapy is defined as the use of baths containing thermal mineral waters from natural springs at a temperature of at least 20 °C and with a mineral content of at least 1 g/l. More so, it is more common that the temperature of the thermal water is approximately 34°. Balneotherapy has been used not only in the ancient years in the treatment of various illnesses but also in modern times.

Under the broad term balneotherapy (in contrast to the strict definition given above), various treatment methods are included: Dead Sea salt or mineral baths, sulphur baths, radon-carbon dioxide baths. Balneotherapy is either provided as the solitary component of the therapeutic approach or in the context of spa therapy.

Spa therapy additionally to balneotherapy employs various modalities such as physiotherapy, and even the change in environment and lifestyle per se may contribute to the changes seen in patient outcome measurements, i.e. the therapeutic result may not be attributed to the balneotherapy alone.

Thus, while some researchers have regarded balneotherapy and spas as more or less interchangeable terms, others disagree. More so one has to acknowledge that the composition of the mineral water differs in its content in cations and anions, and thus assessing which is the specific therapeutic component is difficult. We sought to review the existing evidence regarding randomised controlled trials (RCTs), examining the clinical effects of balneotherapy.


We searched the PubMed, Scopus and Cochrane library (the period examined was 1950-2006) for RCTs, examining the effect of balneotherapy (both as a solitary approach and in the context of spa) on various diseases. The following search terms were used: 'balneotherapy', 'spa', 'skin diseases', 'disease', 'disorders', 'respiratory', 'cardiovascular', 'rheumatic', 'gynaecology', 'allergic', 'gastrointestinal' in various combinations. Our review, from the methodological standpoint, did not seek to look into a specific disease, but to retrieve the data from the literature regarding the effect of balneotherapy (if any) in human disease in general.

Study Selection Criteria

Included studies were RCTs, comparing the effectiveness of balneotherapy vs. therapeutic modalities, not comprising balneotherapy for the treatment of rheumatic, cardiovascular, gynaecological, allergic, gastrointestinal. RCTs written in other languages than English, German, Italian and French were excluded from this systematic review.

We excluded studies regarding the effect of hydrotherapy (generally employing tap water). We also excluded studies regarding thalassotherapy only when the temperature was below the defined for balneotherapy (balneotherapy was not a separate treatment arm in these studies) or studies examining the effect of balneotherapy when applied once a week or examining the effect of balneotherapy in patients with ≥ 2 concurrent diagnoses, studies reporting on the use of thermal water with a temperature lower than 34 °C or interim analyses of the same research group (not included to avoid reporting of duplicate data) or a dropout rate of > 15% of the patients in one of the comparator arms or application of one of the comparators for a period of time with over 1 month difference or the conduct of the RCT in over two different geographical areas while concurrently they recruited a trial group with < 50 per site (25 patients and 25 controls per site) or technical errors in allocation.

For the enrolment of patients in the trials considered for inclusion in this systematic review, a diagnosis of the disease should be based primarily on established clinical criteria, whether or not further supported by radiological or other laboratory criteria.

Data Extraction

Two reviewers (PIR and EZ) independently evaluated all retrieved articles, on the basis of title and abstract, for eligibility and for inclusion in the systematic review. We also reviewed the references of the relevant articles in our attempt to identify additional publications of potential interest. Full-text papers of possibly relevant articles were further reviewed. Selected articles for inclusion in this review are presented in Table 1, on the basis of first author and year of the relevant publication.

Data referring to the characteristics of each trial were extracted, including geographical area (country) and thermal water composition, especially regarding the temperature and electrolyte composition of the thermal water, type of study, number of randomised patients and their category of disease (Table 1), treatment arms and the time of the assessments, the methods employed for the evaluation of the different outcomes and findings of the study regarding the outcomes (Table 2). Any differences in the extracted data between the two reviewers were resolved in the meetings of all authors.


The primary comparison studied in this review, included balneotherapy vs. other therapeutic modalities in the treatment of various diseases. The primary effectiveness outcome of this review was clinical success, comprising substantial improvement of symptoms by clinical criteria and assessment tools related to the specific disease, of the patients. The time of determination of the primary effectiveness outcome was the time of the assessment of the primary clinical efficacy endpoint, which was used in each trial.

Validity Assessment

A review of the quality of each RCT included in our meta-analysis was performed by using the Oxford Quality Scale (OQS), which examines whether there is randomisation, blinding and information on withdrawals from the study, and evaluates the appropriateness of randomisation and blinding, if present.

One point was awarded for the presence of each of the former three criteria, whereas the latter two criteria could be awarded the values of -1 (inappropriate), 0 (no data) and +1 (appropriate). Thus, the maximum score for a study was five, and a score higher than two points denotes a good quality RCT according to this methodology.

The reviewers calculated the score of each study independently. In addition, we proceeded to use the Oxford Pain Validity Scale (OPVS),[7] which takes into account blinding (score range 0-6), size of the group trials (score range 0-3), pre hoc desirable outcomes (score range 0-2), baseline level of pain/outcomes and internal sensitivity (score range 0-1) and data analysis consisting of definitions of outcomes, data presentation, statistical testing and handling of dropouts (score range 0-4) (Table 1).


We identified initially 203 potentially relevant articles and out of it 174 articles were excluded. Finally, 29 RCTs regarding balneotherapy in various medical fields were analysed (Addendum). In Table 1 and Table 2 , we present the characteristics and outcomes, respectively, of the selected and reviewed trials.

The majority of the RCTs [25 out of 29 (89.6%)] pertained to the use of balneotherapy in rheumatological and other musculoskeletal diseases: eight in osteoarthritis, six in fibromyalgia, four in ankylosing spondylitis, four of them regarding in rheumatoid arthritis and three regarding chronic low back pain.

A total of 1720 patients were evaluated in these studies. Four studies were also found from other medical fields, specifically, three in psoriasis and one in Parkinson's disease. Nine of the 29 RCTs were performed in Israel (Dead Sea area), seven in Turkey, five in Hungary, four in France, one in Italy, one in the UK, one in Germany and one in Austria/Netherlands.

Balneotherapy did result in more clinical improvement (as assessed by various indices) in patients with rheumatological/musculoskeletal diseases in comparison to the control group in the 25 RCTs examined ( Table 1 and Table 2 ).

This positive effect did not always refer to the same outcome measurement characteristic and lasted for different periods of time (follow-up periods were different in the RCTs). Pain was the most common outcome examined in these 25 RCTs investigating the role of balneotherapy in rheumatological/musculoskeletal diseases. Pain was improved (with statistical difference) in the treatment arm of balneotherapy more than in the comparator treatment arm in 17 (68%) of the 25 RCTs ( Table 1 and Table 2 )

In another eight studies, pain was improved in the balneotherapy treatment arm, but this improvement was statistically not different than that of the comparator treatment arm(s). Specifically, the beneficial effect lasted for 10 days in one study, 2 weeks in one study, 3 weeks in one study, 12 weeks in two studies, 3 months in 11 studies, 16-20 weeks in one study, 24 weeks in three studies, 6 months in three studies, 40 weeks in one study and 1 year in one study.

Morning stiffness was among the outcomes in eight RCTs: in one RCT, statistically significant improvement was present at the end of treatment; in seven of them, improvement was noted in all comparator arms.

The number of tender points and the Fibromyalgia Impact Questionnaire (FIQ) were outcomes examined in common in five of the six RCTs investigating the role of balneotherapy in fibromyalgia. In three of five of these RCTs,[20,26,33] more improvement (statistically significant) was noted regarding the number of tender points and the FIQ in the balneotherapy treatment arm than in the comparator.

This improvement lasted from 3 weeks up to 6 months. In the remaining two studies, improvement was noted in both comparators.

Analgesic consumption was the outcome in five RCTs; in two of them analgesic consumption was decreased in the balneotherapy arm in comparison to the comparator, in the other two studies improvement occurred in both comparators, while in one study no improvement was noted in either of the comparators.

The Arthritis Impact Measurement Scale was the outcome in two RCTs and was improved more (statistically significant) in the balneotherapy treatment arm. Quality of Life was assessed in four RCTs and found to be improved with statistical difference in all of them at the last follow up.

A sum of 25 of the 29 RCTs (86.2%) did refer to the same therapeutic area (identical location) for the comparator groups. In four of the RCTs as shown in Table 1 , there was difference in the location of the therapeutic intervention. Regarding the three RCTs examining the role of balneotherapy in psoriasis, no statistical differences could be observed regarding the skin lesions.

We proceeded to a subset analysis of the studies with the highest methodological scores on the OQS and the OPVS. Nine of the studies had an OQS score of ≥ 4 and in these studies patient blinding was performed by the use of water with same physical characteristics (i.e. temperature, colour and odour) in the comparator arms.

The same plus another four studies had an OPVS score of ≥ 12. In total, 10 of the 13 studies that scored higher on the OQS and OPVS referred to rheumatological/musculoskeletal disease while three to psoriasis. An analysis of these 13 studies which scored higher showed: the six studies regarding osteoarthritis showed at least one favourable outcome significantly different over the control group.

[pain using a Visual Analogue Scale (pVAS) at 2 weeks, Nottingham Health Profile pain score and tenderness score at 12 weeks]

Western Ontario and McMaster activity, pain and total scores at 3 months, pVAS at 3 months, non-steroidal anti-inflammatory drugs and analgesic consumption during 24 weeks, improvement of pain and quality of life at 24 weeks, night pain at 20 weeks and severity of knee osteoarthritis at 16 weeks, pain at movement, pressure sensitivity at end of treatment (~3 weeks).

One study regarding rheumatoid arthritis showed significantly improved pain intensity and Arthritis Impact Measurement Scale over the control group at 6 months, and one study regarding ankylosing spondylitis showed a significantly improved Pooled Index of Change (a combined measure of primary outcomes) over the control group at 40 weeks.

Two studies regarding low back pain showed a favourable outcome (one of them over the control group at 3 months in physical health and mental health, anxiety and depression, pain duration, pain intensity and functional disability). Three of these studies reported on psoriasis, two showed a favourable outcome (non-significant difference over the control) and one showed no effect.

Five of the 29 studies made a distinction between primary and secondary outcomes. Three among them achieved a higher OQS and OPVS; two of these studies examining patients with osteoarthritis and rheumatoid arthritis showed a significant difference of a primary outcome over the control group.


A critical view of the available evidence suggests that, despite the considerable heterogeneity of patient populations included in the reviewed RCTs, there seems to be a beneficial effect of balneotherapy in the majority of them, mainly on amelioration of pain in the mentioned rheumatological diseases and chronic low back pain.

The duration of the beneficial effect of balneotherapy varies and in the majority of the studies reviewed, it lasts for at least 3 months.

In interpreting these data, one must take into account that periods of follow up are not identical. The scarce data from the use of balneotherapy in non-rheumatological diseases do not allow us to draw conclusions regarding its true impact in the fields of psoriasis vulgaris and Parkinson's disease.

The mechanism of action of balneotherapy is largely unknown. Various inflammatory mediators seem to be increased in patients with fibromyalgia, and balneotherapy was shown to decrease the levels of prostaglandin E2 as well as interleukin-1 and leukotriene B4.

An interesting finding is that balneotherapy reduces the levels of catalase, superoxide dismutase, malondialdehyde protein and glutathione peroxidase. In addition, it has been postulated that some penetration of minerals in the body may confer some of the therapeutic result.

In a comprehensive recent review, regarding the effect of balneotherapy in the treatment of rheumatoid arthritis, it was concluded that balneotherapy may improve symptoms of rheumatoid arthritis, but most of the randomised controlled studies performed in this field could be of higher quality.

Additionally, the effect of various confounders such as the environment per se and psychological factors including the avoidance of everyday stress, as is the case in spas, may influence the balneotherapy effect.

In the majority of the studies reviewed herein, however, the therapeutic location was the same for the comparator groups, and this means that balneotherapy has an effect beyond the one attributed to the environment per se.

Adherence to methodological criteria should be followed when conducting and reporting on an RCT. The main drawback in some of the studies reviewed is that not enough details are given regarding the blinding (i.e. unconvincingly double-blind studies or single-blinded).

One should acknowledge that it is difficult to perform blinding of the intervention, i.e. whether patients receive the active treatment or not. However, blinding may be accomplished when physical characteristics of thermal water in the active treatment group of patients and water used in the control group (i.e. temperature, colour and odour) are the same.

Only occasionally is information available regarding the allocation of concealment. The size of the trial groups is another limiting factor in trying to draw firm conclusions. Only four studies reported on more than 100 patients. The total number of patients reviewed herein is 1720 patients.

In contrast to an excellent cohort study regarding the effects of balneotherapy, the NAIADE survey in Italy, 23,680 patients with eight disease groups received entry and return inquiry after 1 year. Description of the interventions and reporting of the results of the herein reviewed trials was appropriate considering the complexity in evaluating the treatment of musculoskeletal diseases.

Information on potential adverse effects in the reviewed studies is practically absent. This is not to be ignored, as even RCTs on the management of well-known diseases do not always take into consideration this information.

Not withstanding the methodological limitations, one cannot ignore that there is evidence that balneotherapy improves many of the symptoms associated with rheumatic diseases. One must stress that the multitude of methods were used to assess objectively and subjectively the effect of balneotherapy on patients. More so, quality of life scores are examined, although not in all of the studies.

One should acknowledge the lack of standard evaluation of evidence and the challenges arising in evaluating these therapies. A limitation of our review is that a significant number of studies written in the Russian language were excluded. Also, it should be emphasised that limited data exist in the literature regarding the effect of balneotherapy in diseases other than rheumatological.


In conclusion, there is possibility that balneotherapy is associated with clinical improvement in rheumatological disease mainly such as osteoarthritis, fibromyalgia, ankylosing spondylitis, rheumatoid arthritis and in chronic low back pain. However, existing research is not sufficiently strong to draw firm conclusions.

More RCTs are needed to help draw firm conclusions regarding the effectiveness of balneotherapy in various medical fields, especially on dermatological, cardiovascular, respiratory, gastrointestinal and allergic and gynaecological skin diseases.

M. E. Falagas; E. Zarkadoulia; P. I. Rafailidis

No comments:

Post a Comment