Evidence-based balneotherapy in Hungary. Updated systematic review and meta-analysis of the Hungarian clinical trials

Abstract

In 2014 the authors published a review and meta-analysis on the effectiveness of balneotherapy in Hungary based on articles by Hungarian authors issued until 2012. Now they prepared a new review and meta-analysis based on the updated database of the Hungarian articles published after 2012. A total of 18 articles were included in the new meta-analysis; with a subset having been previously incorporated in the earlier analyses. In such a way the results became more convincing. Regarding the updated systematic review (period: 2012–2025), the number of papers was 14. Regarding the meta-analysis (period: 1989–2025), the number of papers was 18 (8 + 10 = 18). The meta-analysis showed a significant reduction of pain at rest and pain on load evaluated by visual analogue scale (VAS) score and of the quality of life evaluated by the EuroQol-5D (EQ-5D) questionnaire and indexed by the EQ-VAS and the EQ-5D index. Even accounting for the heterogeneity of the studies, which is a source of bias, based on the results we can say the Hungarian studies’ results are promising regarding the positive effect of balneotherapy, especially in the treatment of degenerative locomotor diseases.

Introduction

Balneotherapy means the use of natural mineral waters, natural peloids, muds, and natural sources of different gases for medical purposes: prevention, treatment, and rehabilitation (Gutenbrunner et al. 2010). The definitions of these waters are based on the sum of the cations and anions (inorganic elements), but some authors have recently mentioned the role of organic substances (Varga 2016). Hungary ranks among the leading countries that are rich in thermal mineral springs used for medical purposes. Under two-thirds of the Hungarian national territory can be found a huge amount of thermal reservoirs in sand sediments and karst formations. The balneological research is very active in Hungary, and balneotherapy is an important part of physiotherapy. Calculating the number of articles per million inhabitants related to water-based treatments, Hungary is one of the top leading countries, and looking at the number of trials performed with natural source water, Hungary is in the top five, according to the recently published article (Güneri et al. 2022). In 2014 we prepared and published a systematic analyses of the balneological clinical trials authored by Hungarian researchers, accompanied by a meta-analysis evaluating the changes in pain during movement and at rest (Bender et al. 2014). Because of the large number of clinical studies that have been published in the past 10 years in Hungary, our intention was to prepare a new article as an updated review and meta-analysis. Previously Turkish and Israeli authors published a systemic review on this topic. (Karaguelle MZ and Karagulle M 2004, Katz et al. 2012) In the medical database, numerous publications on balneology have been published in recent decades. The quality of the communications varied greatly from no control to randomised double-blind controlled studies with appropriate statistical power. In addition to the Italian, Turkish, French, Spanish, German, and other international papers, Hungarian papers also contribute to the growing the recognition of balneology. Several reviews have recently been published on the positive effects of balneotherapy, mostly in osteoarthritis (Montvydaitė-Kreivaitienė et al. 2025; Protano et al. 2023; Ma et al. 2021; Boopalan et al. 2024; Maier et al. 2024; Kardes 2021; Corvillo et al. 2020; Tenti et al.2015; Fiorvanti et al. 2017), in fibromyalgia (Garcia-Lopez et al. 2024) and in low back pain (Karagulle M. and Kagulle MZ. 2015; Forestier et al. 2022a, b). According to these reviews, in some studies the authors emphasised a serious risk of bias, and inconsistency sometimes was highlighted (Aribi et al. 2025). All reviews called for an enlarged number of cases and more precise methodology, according to the authors of the reviews. Many summaries deal with the effects of balneotherapy in the treatment of patients with musculoskeletal disorders (Fioravanti et al. 2024; Antonelli et al.2021; Maccarone et al. 2023; Maccarone and Masiero 2021). One of the most recent reviews mentions that since 1990 only 11% of cases have been issued by the double-blinded method (Szendi et al. 2025). Unfortunately, there are still no international recommendations in balneotherapy, although the number of studies shows an increasing tendency. Balneotherapy was last mentioned in international guidelines in 2014 as a possible treatment for osteoarthritis (Mc Alindon et al. 2014). The publications originate mainly from countries that have thermal water, such as France, Italy, Spain, Hungary, Russia, Serbia, Romania, Poland, Japan, Germany, Portugal, Lithuania, Israel, Iran, and so on, where balneotherapy is traditionally used in the treatment of rheumatological diseases (de Oliveira et al. 2023). The majority of the original publications have been published from these countries. Besides the original publications, there are more and more reviews and meta-analyses about the use of balneotherapy in medicine (Cheleschi et al. 2022; Antonelli et al. 2024; Kamioka et al. 2020; Ferrara et al. 2025; Galvez et al. 2024a; Rapoliené et al. 2025b; Fioravanti et al. 2024, Kardes 2021; Harari 2024). Unfortunately, cost-effectiveness studies have hardly been done in this area, but balneotherapy seems to be an effective adjunct therapy to other physiotherapy treatments.

The aim of the article

The primary objective of our study was to update our previous review of Hungarian clinical publications on balneology, which was published in 2014. That review was submitted in 2012 (published in 2014), and covered studies published between 1989 and 2012.

The secondary objective was to perform a meta-analysis of the published controlled clinical studies on musculoskeletal diseases focusing on both pain and quality of life outcomes. In our study published in 2014 the meta-analysis focused only on pain. Adding the articles published recently, we decided to point out the quality of life also, on top of the meta-analyses focusing on pain with the enlarged number of studies.

Method

Protocol registration

Our study was in accordance with the Prisma protocol (Shamseer et al. 2015), and was registered in the Prospero database (CRD420251085150).

Data extraction and study selection

Criteria for systemic review were as follows: controlled clinical studies. We took into account studies published between 2012 and 2025 by Hungarian authors and/or carried out in Hungary on hydro- and balneotherapy with or without any other treatments, in English.

Criteria for the meta-analysis were clinical randomised controlled trials on musculoskeletal diseases with relevant common endpoints suitable for meta-analyses regarding pain or quality of life. Pain was assessed using the visual analogue scale (VAS) or the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain subscale score. Quality of life was evaluated using the EuroQol-5D (EQ-5D) questionnaire, indexed by the EQ-VAS and EQ-5D index.

Two investigators independently reviewed Hungarian publications indexed in PubMed, Scopus, Web of Science, and Google Scholar (B.T. and T.I.K.). The new review focused on Hungarian studies published between 2012 and 2025. Regarding the studies between 1989 and 2012, we refer to our previously mentioned publication (Bender et al. 2014). For the current meta-analysis, we evaluated the eligible studies published between 1989 and 2025. The following words were used in the database search: balneotherapy, hydrotherapy, spa therapy, underwater traction therapy, thermal water, health resort therapy, and mud therapy. The statistical analysis was performed by a third person (H.K.).

Outcome measures for the meta-analyses included pain on load VAS or WOMAC pain score and pain at rest VAS; regarding quality of life, the EQ-5D index and EQ general health VAS. We prepared the meta-analyses calculations regarding the visits at 3–4 weeks (visit 2) and at the visits at 10–12 weeks (visit 3).

Risk of bias assessment

We used the same method with the 11-point scale score to assess the internal validity of the studies (van Tulder et al. 2003). Studies scored lower than 50% were evaluated as low-quality trials, between 50% and 75% as moderate-quality, and at least 75% as high-quality trials. Each criterion was scored 1 point if it met the criterion, 0 if it did not meet it, or was not known. Moderate and high quality trials were included in the meta-analysis. One enclosed trial (Konrad et al. 1992) was evaluated as low quality in our previous publication, as intention-to-treat analyses was not commonly used that time.

Statistical analysis

The same method as in the 2014 study was applied to ensure comparability. The standard deviations (SD) and the mean values, or test statistics, were recorded for each of the studies, and effect sizes were calculated using the standardised mean difference (SMD) method. The SMD is a scale-free measure of the ratio of the difference in mean outcomes between the groups to the SD of the outcome in the study population. The intervention effect expressed in SMD units is a standardised value rather than the original unit of measure. Sample size and effect sizes were processed with the MedCalc software package with the purpose of generating forest plots (graphical displays) of effect size for each study. Begg’s test and Egger’s test were used to assess publication bias. In case of p < 0.05 publication bias was considered to be present.

Results

Main findings

The data of 977 patients from 14 studies were analysed in this new study for review. Dropout was 30 patients (mainly because of lack of compliance, intercurrent infection, arrhythmia in one case, surgery in one case, etc.). No side effects were registered. (Fig. 1; Tables 1 and 2) All included RCTs followed appropriate national ethical guidelines.

Fig. 1.

Flowchart of the review (2012-2025). Flowchart of the meta-analyses of trials (1989-2025)

Table 1.

Controlled clinical studies (2012–2025) relevant for the updated meta-analyses (1989–2025)

Trials	Study design	Diagnosis	Internal validity Score	Treatment	n	Outcome parameters	Follow up period	Result
Kulisch A et al. 2014	Single blind RCT	Knee OA	9	Group 1: immersion in thermal mineral water (34 °C) of lake Hévíz with carbonate, sulphur, calcium, magnesium, bicarbonate, low radon content for 30 min 15 times in 3 weeks plus underwater jet massage Group 2: the same as treatment group using tap water	38 39	pain at rest and on exertion VAS, opinion of patient and investigator VAS, active flexion degree, knee circumference, stair-climb time, WOMAC, EQ-5D	15 weeks	Comparison of the two groups revealed a statistically significant difference in pain visual analogue scale scores (P < 0.01), active flexion degree (P < 0.01), physical function components of WOMAC (P < 0.05), and EQ-5D scores (P < 0.05) even after 15 weeks.
Hanzel A et al. 2018	Double blind RCT	Knee OA	7	Group 1: immersion in thermal mineral water (34 °C) with sodium chloride, bicarbonate contant 30 min 15 times in 3 weeks Group 2: the same as treatment group using tap water	26 24	range of movement of the involved joints, WOMAC, pain VAS, SF-36	15 weeks	Treatment with the thermal mineral water of Szigetvár significantly improved ROM, WOMAC scores, and SF-36-scored quality of life of the patients.
Hanzel A et al.2019	Double blind RCT	Knee OA	7	Group 1: immersion in thermal mineral water (34 °C) with sodium chloride, bicarbonate contant for 30 min 15 times in 3 weeks Group 2: the same as treatment group using tap water Group 3: the same as treatment group using organic fraction water	26 24 24	range of movement of the involved joints, WOMAC, pain VAS, SF-36	15 weeks	Treatment with the redissolved organic fraction water significantly improved ROM, WOMAC, and SF-36 scores compared to the tap water.
Kovács Cs et al. 2016	Single blind RCT	Hip OA	9	Group 1: immersion in thermal mineral water with high sulfide ion content for 20 min 15 times in 3 weeks plus exercise therapy Group 2: only exercise therapy	21 20	WOMAC, EQ-5D The main endpoint was achievement of Minimal Clinically Important Improvement (MCII) at 12 weeks, defined as ≥ 7.9 points in a normalized WOMAC function score.	12 weeks	At 12 weeks, 17 (81%) balneotherapy group patients had Minimal Clinically Important Improvement and 6 (30%) of controls (p = 0.001). At the end of the treatment significant between group difference in the WOMAC stiffness score only, whereas after 12 weeks, significant between group difference regarding he WOMAC pain, stiffness, function, and total scores in favour of the balneotherapy group. Significant between group difference after 12 weeks in point of EQVAS score, too.
Király et al. 2020	Double blind RCT	Knee OA	11	Group 1: hot mud pack therapy (42 °C) for 30 minutes 15 times in 3 weeks with mud originated from Tiszasüly (inorganic peloid) Group 2: same treatment but with mud originated from Kolop (inorganic peloid)	29 31	VAS for knee pain, WOMAC, KOOS, Lequesne index, EQ-5D	12 weeks	both kinds of mud had similar strong effect, no significant difference between the groups, the non-inferiority of Tiszasüly mudpack is proved
Tefner IK et al. 2013	Double blind RCT	Knee OA	10	Group 1: hot mud pack therapy (42 °C) for 30 min 10 times in 2 weeks with mud originated from Neydharting (mixed peloid) Group 2: same treatment but with artificially produced substance with similar physical properties	27 26	WOMAC, EQ-5D, analgesic and NSAID requirement	12 weeks	both groups improved, clear tendency in favour of the mud treated group but no significant between group differences
Gyarmati N et al. 2017	Single blind RCT	Hand OA	9	Group 1: hot mud pack therapy (42 °C) (mixed peloid) for 20 min 15 times in 3 weeks Group 2: same treatment but with nylon gloves on hands	23 24	pain at rest and on exertion VAS, opinion of patient and investigator VAS, handgrip, number of tender and swollen joints, morning stiffness, EQ-5D, HAQ	16 weeks	Both groups improved similarly. The pain at movement VAS, the examiner’s opinion VAS and the swollen joint’scount both side improved more in group 1 compared to the control (at 3rd visit) Controversial regarding the EQ-5D VAS and HAQ the improvement was only observed in the control group (at 2nd and 3rd visit).
Gáti et al. 2018	single blind RCT	Chronic low back pain	8	Group 1: immersion in thermal-mineral water (38 °C) with calcium-magnesium-bicarbonate content 15 times in 3 weeks plus usualcare Group 2: usualcare	52 53	Pain at rest and during activity, Oswestry index, EQ-5D, medicine requirement (number of patients taking medicine for LBP)	12 week	All parameters improved significantly in group 1, significant between group differences in short and longterm
Tefner IK et al. 2023	single blind RCT	Chronic low back pain	8	Group 1: immersion in thermal-mineral water (42 °C) with radon, natrium, calcium, bicarbonate content for 20 min per day, 15 occasions in 3 weeks plus usual care Group 2: usual care	68 69	pain VAS at rest and during activity, Oswestry index, EQ-5D, Patient Acceptable Symptom State, medicine requirement	12 week	All parameters improved significantly in group 1, significant between group differences in short and longterm
Tefner et al. 2015	Single blind RCT, multicentre	chronic shoulder pain	9	Group 1: immersion in 32 °C mineral water with sodium, calcium, hydrogen carbonate, sulfate content for 30 min 15 times in 4 weeks plus exercise therapy plus TENS 10 times Group 2: exercise therapy plus TENS 10 times	23 23	pain VAS at rest and at movement, SPADI, EQ-5D, SF-36, active and passive range of motion	3 months	The SPADI pain, function, and total scores and the VAS scores at rest and on movement significantly improved in both groups after treatments. A great improvement was observed in the balneotherapy group compared to the control goup; regarding some parameters (VAS score on movement and SPADI function score at visit 2; VAS score at rest at visits 3 and 4), the difference between the groups was significant. The improvement of SF-36 and EQ-5D quality of life scores and the active range of motion was more pronounced in the balneotherapy group, the difference between the groups was not significant, except for EQ-5D index at visit 2. Improvement of passive range of motion was not significant.

Table 2.

Controlled clinical studies that finally not selected for meta-analyses (2012–2025)

Trials	Study design	Internal validity score	Inclusion criteria	Treatment	n	Outcome parameters	Follow-up period	Results
Fritsch et al. 2022	RCT multicentre	8	Patients with SLE in remission or low disease activity	Group 1: immersion in thermal mineral water (33–35 C) for 30 min 15 times in 3 weeks plus standard of care Group 2: standard of care	16 14	LupusQoL, SF-36, WPAI-Lupus	10 weeks	Several subdomains of SF-36 physical condition improved significantly for long term. General health improved significantly short term. Fatigue worsened despite treatment.
Szenczi Á et al. 2023	Double blind RCT	8	psoriasis vulgaris	Group 1: immersion in mineral water plus complex rehabilitation plus dithranol Group 2:immersion in tap water plus complex rehabilitation plus dithranol	10 10	PASI, oxidative stress marker (malondialdehyde, MDA)	3 weeks	PASI score improved in both groups. MDA levels in group 2 increased significantly compared to group 1.
Gáti et al. 2020	Single blind RCT, multicentre	7	outpatients with non-specific low back pain that persists for at least 12 weeks, showing degenerative symptoms, and suffering from moderately reduced mobility. Patient’s pain intensity during activity minimum of 30 mm on the visual analogue scale (0–100 mm VAS).	Group 1: underwater weightbath traction therapy 15 times in 3 weeks in indifferent water (33–35 C) for 20–30 min plus NSAID Group 2: underwater weight bath traction therapy Group 3: NSAID	45 94 43	pain VAS at rest and during activity, EQ-5D, Oswestry Index	3 month	The pain VAS levels improved significantly (p < 0.05) in Group 1 and Group 2 for long term, no change in Group 3.
Horváth et al. 2017	non randomised, single blind controlled study	6	Patients between 18–75 years of age with joint contractures of the hand	Group 1: hand stretching exercises, ergotherapy supplemented with thermal and mudbaths, whirlpool therapy and softtissue massage were daily used during a three-week period. Group 2: same as Group 1 leaving out their hands	31 22	HAQ, DASH, HAI, CHFT and clinical characteristics	6 month	The complex physical therapy caused favourable changes in both the HAQ and the DASH, indicating that his particular program had some long-term beneficial effect on hand function in patients with SSc

Pain

If the paper did not specify what kind of pain was measured, we contacted the authors for clarification.

Pain, pain on load VAS, or WOMAC pain score

The overall improvement in “pain on loading” was approximately 70–75% at visit 2 (SMD 0.734, 95% CI 0.856 to − 0.613, fixed effect; SMD 0.704, 95% CI 0.997 to 0.41, random effect) and approximately 70% at visit 3 (SMD 0.711, 95% CI 0.836 to − 0.587, fixed effect; SMD 0.691, 95% CI 0.978 to 0.403, random effect). Using summarised analysis, the aggregate improvement of “pain on loading” was significant. The heterogeneity test was significant for “pain on loading.” (Figs. 2 and 3).

Fig. 2.

Pain VAS on load at visit 2

Fig. 3.

Pain VAS on load at visit 3

Pain at rest VAS

The overall improvement in “pain at rest” was approximately 80% at visit 2 (SMD 0.826, 95% CI 0.996 to 0.657, fixed effect; SMD 0.765, 95% CI 1.095 to 0.435, random effect) and approximately 80% at visit 3 (SMD 0.82, 95% CI 0.99 to 0.65, fixed effect; SMD 0.77, 95% CI 1.13 to 0.411, random effect). Using summarised analysis, the aggregate improvement of “pain at rest” was significant. The heterogeneity test was significant for “pain at rest.” (Figs. 4 and 5).

Fig. 4.

Pain VAS at rest at visit 2

Fig. 5.

Pain VAS at rest at visit 3

Quality of life

EQ VAS

The overall improvement in the EQ-VAS was approximately 50–60% at visit 2 (SMD 0.609, 95% CI 0.453 to 0.766, fixed effect; SMD 0.525, 95% CI 0.127 to 0.923, random effect) and approximately 70% at visit 3 (SMD 0.754, 95% CI 0.596 to 0.911, fixed effect; SMD 0.682, 95% CI 0.335 to 1.030, random effect). Using summarised analysis, the aggregate improvement of “EQ VAS” was significant. The heterogeneity test was significant for “EQ VAS”. (Figs. 6 and 7)

Fig. 6.

EQ-VAS at visit 2

Fig. 7.

EQ-VAS at visit 3

EQ-5D index

The overall improvement in the EQ-5D index was approximately 50–55% at visit 2 (SMD 0.578, 95% CI 0.423 to 0.736, fixed effect; SMD 0.508, 95% CI 0.165 to 0.851, random effect) and approximately 60% at visit 3 (SMD 0.641, 95% CI 0.485 to 0.796, fixed effect; SMD 0.572, 95% CI 0.294 to 0.851, random effect). Using summarised analysis, the aggregate improvement of the “EQ-5D index” was significant. The heterogeneity test was significant for “EQ-5D index”. (Figs. 8 and 9)

Fig. 8.

EQ-5D at visit 2

Fig. 9.

EQ-5D at visit 3

Subgroup analyses

Pain, pain on load VAS, or WOMAC pain score in trials with quasi-placebo (tap water or hot pack) controlled double or single blind design (subgroup 1).

The overall improvement of “pain on loading” was about 55–60% at visit 2 (SMD 0.55, 95% CI 0.712 to 0.388, fixed effect; SMD 0.586, 95% CI 0.926 to 0.246, random effect). Using summarised analysis, the aggregate improvement of “pain on loading” was significant. The heterogeneity test was significant for “pain on loading.” (Fig. 10).

Fig. 10.

Pain VAS on load at visit 2 in subgroup 1

Pain, pain on load VAS, or WOMAC pain score in trials evaluating the additional effects of balneotherapy on the top of a standard treatment received in both groups (balneotherapy plus standard treatment versus standard treatment alone) (subgroup 2).

The overall improvement of “pain on loading” was about 90–100% at visit 2 (SMD 1.021, 95% CI 1.241 to 0.801, fixed effect; SMD 0.892, 95% CI 1.610 to 0.174, random effect). Using summarised analysis, the aggregate improvement of “pain on loading” was significant. The heterogeneity test was significant for “pain on loading.” (Fig. 11).

Fig. 11.

Pain VAS on load at visit 2 in subgroup 2

Publication bias

Begg’s test (p = 0.6769) and Egger’s test (p = 0.36) showed no publication bias for pain VAS on load at visit 2. The funnel plot graph indicated symmetrical study distribution. (Fig. 12)

Fig. 12.

Funnel plot for pain VAS on load at visit 2

Discussion

In our meta-analysis of pain on loading, 18 studies fulfilled the inclusion criteria (Konrad et al. 1992; Kovacs and Bender 2002; Balogh et al. 2005; Balint et al. 2007; Kulisch et al. 2009; Horvath et al. 2012; Tefner et al. 2012; Kovacs et al. 2012; Tefner et al. 2013; Kulisch et al. 2014; Tefner et al. 2015; Kovács et al. 2016; Gyarmati et al. 2017; Gáti et al. 2018; Hanzel et al. 2018; Hanzel et al. 2019; Király et al. 2020; Tefner et al. 2023).

Pain improved significantly in 11 out of 18 studies at visit 2, and 8 out of 17 at visit 3 (the study of Kovacs and Bender 2002 was only assessed at visit 2). At visit 2, in eleven trials, “pain on loading” improved significantly relative to the control group (Konrad et al. 1992; Kovacs and Bender 2002; Balint et al. 2007; Tefner et al. 2012; Kovacs et al. 2012, Kulisch et al. 2014; Tefner et al. 2015; Gáti et al. 2018; Hanzel et al. 2019; Király et al. 2020; Tefner et al. 2023). In the remaining seven trials, the improvement of “pain on loading” was significant also in the control group (Balogh et al. 2005; Kulisch et al. 2009; Horváth et al. 2012; Tefner et al. 2013; Kovacs et al. 2016; Gyarmati et al. 2017; Hanzel et al. 2018). (Figs. 2 and 3) This finding can be explained by the cumulative effects of warm water or hot packs, along with the concomitant interventions such as electrotherapy, magnetotherapy, and exercise therapy. At visit 3, the improvement became significant in one study (Kovács et al. 2016) and became nonsignificant in three studies (Konrád et al. 1992; Bálint et al. 2007; Király et al. 2020). The study of Király et al. is a non-inferiority trial aiming to prove the similar therapeutic effects of two hot muds originating from the same place.

When we examined pain at rest, 5 out of 8 studies showed significant improvement at visit 2 and 5 out of 8 at visit 3 (the study of Kovacs and Bender 2002 only assessed at visit 2). At visit 2, in five trials, pain at rest improved significantly relative to the control group (Tefner et al. 2012; Horváth et al. 2012; Kulisch et al. 2014; Gáti et al. 2018; Tefner et al. 2023). In the remaining three trials, the improvement of pain at rest was significant also in the control group (Kulisch et al. 2009; Tefner et al. 2015; Gyarmati et al. 2017). At visit 3, the improvement became significant in one additional study (Tefner et al. 2015) and reverted to being nonsignificant in one study (Horváth et al. 2012). (Figs. 4 and 5)

At visit 2, in three trials, EQ-VAS improved significantly relative to the control group (Tefner et al. 2012; Gáti et al. 2018; Tefner et al. 2023). In the remaining five trials, the improvement of EQ-VAS was significant also in the control group (Kovacs et al. 2012; Tefner et al. 2013; Kulisch et al. 2014; Tefner et al. 2015; Kovacs et al. 2016; Gyarmati et al. 2017; Király et al. 2020). At visit 3, two additional studies reached statistical significance(Kulisch et al. 2014; Tefner et al. 2015). (Figures 6 and 7)

Similar to the EQ-VAS, the EQ-5D index also improved in the treatment group compared to the control group. The improvement was significant in 5 out of 10 studies at visit 2, and 4 out of 10 at visit 3. At visit 2, in five trials, the EQ-5D index improved significantly relative to the control group (Tefner et al. 2012, 2015; Kulisch et al. 2014; Gáti et al. 2018; Tefner et al. 2023). In the remaining five trials, the improvement of the EQ-5D index was significant also in the control group (Kovacs et al. 2012; Tefner et al. 2013; Kovacs et al. 2016; Gyarmati et al. 2017; Király et al. 2020). At visit 3, the improvement became nonsignificant in one more study (Tefner et al. 2015). (Fig. 8 and 9)

Reducing pain is one of the most important issues for patients with chronic musculoskeletal disorders (Bender et al. 2005). In our previous review communication, we evaluated and showed the favourable effect of balneotherapy on pain, findings that have been reaffirmed in the present study with a larger dataset. Morever, this analyses allowed us to register improvements in quality of life also.

Indeed, before the 2000 s, balneotherapy was considered by many as an alternative treatment. However, a growing body of published research has since documented its therapeutic benefits, particularly in the management of osteoarthritis. The number of papers on the treatment of inflammatory musculoskeletal diseases remains limited compared to osteoarthritis (Forestier et al. 2022a, b; Fernandez-Gonzales et al. 2021, Santos et al. 2016). Regarding the possible biological (anti-inflammatory, neuromodulatory, psychological) mechanisms reviewed recently (Cheleschi et al. 2022), there are some new publications addressing this topic. After mud treatment, serotonin levels decreased, contributing to an anti-inflammatory effect. (Galvez et al. 2024b) In post-COVID patients, balneotherapy reduces anxiety and depressive symptoms.(Constantino et al. 2024)

In fibromyalgia, low back pain, and knee OA treatment, balneotherapy has reached the level of evidence of other nonpharmacological treatments, although—as already mentioned—there is no official recommendation yet. Countries without access to thermal water and balneotherapy have less faith in this kind of therapy. The rise of evidence based medicine (EBM) has also had an impact on balneotherapy, enhancing the need for scientific reports. The quality of published articles indexed in databases continues to show gradual improvement.

Seasonality also can influence the effect balneotherapy treatment. From a cardiovascular perspective, prolonged summer treatments are more strenuous (Wang et al. 2023), while winter treatments appear to have a much more stress-reducing effect (Rapoliené et al. 2025).

In Hungary, balneology and medical meteorology are closely related fields of medicine that study and apply the effects of natural healing factors. In Hungarian spas and sanatoriums, balneotherapy is often supplemented with climatotherapy, i.e. treatments that utilize the beneficial effects of air, temperature, and sunlight. Hungary has officially certified health resorts, which, according to the law, are areas where, in addition to natural healing factors (healing waters, healing mud, healing climate), adequate infrastructure and healthcare facilities are available. Health resort certification is issued by the National Public Health Center. In 2025, there are 37 health resorts in Hungary. We want to highlight, that in our analysed studies the balneological treatments were carried out on an outpatient basis, not changing the patient’s usual everyday life, and both groups were under the same meteorological conditions. This way we tried to reduce the effects of the beneficial change of climatic, meteorological and geological factors and assess the specific effects of balneotherapy more accurately.

It was not our aim to recommend the ideal patients profile, optimal treatment duration, frequency or temperature of intervention. These questions are with high importance, and regarding the fact that in lack of evidence there is no consensus regarding these issues, these could be of importance for possible future researches in balneotherapy. These parameters are needed to be standardized with large homogenous patient group studies. We urgently need therapeutic recommendations and evidence levels from international scientifical organisations.

There are some novelties and differentiation from the 2014 publication. This observation is originated from the fact that there is a higher number of included studies with better methodological qualities, e.g. using intention-to-treat analyses, including higher patient numbers, with longer follow-up period. The new publication included the analyses of quality of life also in addition to pain. The results of the two publications interact, making the results more promising.

We have to highlight the high importance of further double-blind possible placebo controlled trials with a long-term (6–12 months) follow-up, appropriate methodology and sample size. International trials with similar design would be welcome.

Hungarian thermal waters and muds have demonstrated beneficial effects in alleviating pain as well as in improving the quality of life. These data are consistent with internationally published clinical studies and reviews. Most of these sources – including our own – emphasize the need for further high-quality, well-controlled trials to substantiate the efficacy of balneotherapy. In the updated review we displayed four clinical articles, so they were not included in the meta-analyses (Horvath et al. 2017; Szenczi et al. 2023; Gati et al. 2020; Frtisch et al. 2022). Beyond the articles that evaluate the effect of balneotherapy on locomotor diseases, we also highlight additional Hungarian studies in dermatology (Szabó et al. 2024, Péter et al. 2017), experimental physiology (Gerencsér et al. 2019; Varga et al. 2015; Tékus et al. 2018), papers about the effects of transcutaneous carbonic acid (Németh et al. 2018; Kreska et al. 2018), microbiome modulation (Tamas Bender et al. 2023; Kulisch et al. 2023), and a mineral water ingestion case study (Czirok et al. 2025), indicating that balneotherapy is widely used in Hungary. Our study supports this tendency. While our study is not without limitations (see below), we believe It contributes meaningfully to the growing recognition of balneotherapy.

Limitation of the study

The heterogeneity test regarding all the calculated parameters was statistically significant and clinically relevant. There is a true difference in the magnitude of the effects across the studies. It can arise from the differences in the study populations especially the intervention protocols and the design of the studies. Namely, the treatment in the control group was different in the studies, e.g., hot water or hot pack therapy, serving as quasi-placebos, or the identical treatments as the study group but without active intervention; even a non-inferiority trial was included, etc. The studies were different regarding the type, intensity, and duration of treatments, the methodology, and also the timing of assessments. Balneotherapy treatment is not internationally standardised. In our analyses only a small number of studies employed double-blind methodology, primary due to technical difficulties. Several good Hungarian trials, such as the first tap water-controlled double-blind trial involving patients with rheumatoid arthritis, were published in Hungarian, so we did not include them in the review.

Conclusion

In our new review and meta-analysis, we consider that the Hungarian thermal waters and muds, regardless of concentration or mineral and organic content, seem to reduce pain at rest and load of degenerative musculoskeletal disorders and improve quality of life over the study period (12 weeks). We can say that although we are aware of the heterogeneity of the studies, which may cause bias, we confirmed the results of our previous study. Balneotherapy is not a panacea but can be a useful adjunctive therapy for the treatment of musculoskeletal disorders. We have to highlight the high importance of further double-blind possible placebo controlled trials with a long-term (6–12 months) follow-up, appropriate methodology and sample size. International trials with similar design would be welcome.

Author contributions

B.T. had the idea for the article. B.T. and T.I.K. contributed to the conception of the article, the data collection and design of the study, and the manuscript writing. H.K. analysed the date and calculated the statistics.

Funding

The authors declared they received no external funding.

Declarations

Conflict of interest

The authors declared that they have no conflict of interest.