Systematic Review and Meta-Analysis Medicine ®

OPEN

Effectiveness of spa therapy for patients with

chronic low back pain
An updated systematic review and meta-analysis
∗
Ruixue Bai, MPHa,b,c,d, Chihua Li, MPHe,f, Yangxue Xiao, BSca,b,c, Manoj Sharma, PhDg, Fan Zhang, PhDa,b,c, ,
∗
Yong Zhao, MSca,b,c,

Abstract
Background: Low back pain (LBP) is a major health problem around the world. Two previous meta-analyses showed that the spa
therapy has a positive effect on reducing pain among patients with LBP based on studies published before 2006 and studies
published between 2006 and 2013. In recent years, more studies reported the effect of spa therapy on treating chronic low back pain
(CLBP). Our study aimed to update the meta-analysis of randomized controlled trials (RCTs) about the effect of spa therapy on
treating CLBP and to examine the effect of spa therapy based on different interventions.
Methods: PubMed, Embase, Web of Science, and Cochrane Library were searched until May 2018 to identify RCTs about spa
therapy among patients with CLBP. Summary effect estimates were calculated by using a random-effects model. The quality of each
eligible study was evaluated by Jadad checklist.
Results: Twelve studies met the inclusion criteria for the systematic review and were included in meta-analysis. There was a
significant decrease in pain based on visual analogue scale (VAS) (mean difference [MD] 16.07, 95% confidence interval [CI] [9.57,
22.57], P < .00001, I2 = 88%, n = 966), and lumbar spine function in Oswestry disability index (ODI) (MD 7.12, 95% CI [3.77, 10.47],
P < .00001, I2 = 87%, n = 468) comparing spa therapy group to control group. Methodological assessment for included studies
showed that the study’s quality is associated with lacking blinding.
Conclusion: This updated meta-analysis confirmed that spa therapy can benefit pain reliving and improve lumbar spine function
among patients with CLBP. Physiotherapy of subgroup analysis indicated that it can improve lumbar spine function. However, these
conclusions should be treated with caution due to limited studies. More high-quality RCTs with double-blind design, larger sample
size, and longer follow-up should be employed to improve the validity of study results.
Abbreviations: CI = confidence interval, CLBP = chronic low back pain, LBP = low back pain, MD =mean difference, ODI =
Oswestry disability index, RCTs = randomized controlled trials, VAS = visual analogue scale.
Keywords: chronic low back pain, meta-analysis, spa therapy, systematic review

1. Introduction
Editor: Dennis Enix.
R.B. and C.L. contributed equally to this work. Low back pain (LBP) is a major health problem around the
This work was supported by 2012 Chinese Nutrition Society (CNS) Nutrition world, with an estimated prevalence of around 7.0%.[1,2] The
Research Foundation—DSM Research Fund. majority of adults (60%–80%) have medical complaints on LBP
The authors have no conflicts of interest to disclose. at some time point in their lives,[3,4] and 5% to 10% of them will
Supplemental Digital Content is available for this article. further develop chronic low back pain (CLBP).[5] CLBP patients
a
School of Public Health and Management, b Research Center for Medicine and
can have symptoms of LBP for over 3 months,[6–8] and elder
Social Development, c Innovation Center for Social Risk Governance in Health, people, women, domestic workers, and people with higher body
d
Health Management (Physical Examination) Center, The Second Affiliated mass index are more likely to have CLBP.[9–14] Patients with
Hospital, Chongqing Medical University, Chongqing, e Zhengzhou Central CLBP may face heavy burden and suffer from long time
Hospital Affiliated to Zhengzhou University, Henan, China, f Department of
incapacity, which is accompanied by repeated treatment and
Epidemiology, Mailman School of Public Health, Columbia University, New York,
NY, g Department of Behavioral and Environmental Health, Jackson State social support.[2,15–18]
University, Jackson, MS. Different methods can be applied for treatment and manage-
∗
Correspondence: Fan Zhang and Yong Zhao, No. 1, Yixueyuan Road, Yuzhong ment of CLBP, including pharmacological and nonpharmaco-
District, Chongqing 400016, China (e-mails: [email protected]; logical treatments.[19] Spa therapy is a nonpharmacological and
[email protected]). widely used treatment,[20] in which patients bath in natural spring
Copyright © 2019 the Author(s). Published by Wolters Kluwer Health, Inc. water with a temperature over 20°C and rich mineral contents for
This is an open access article distributed under the Creative Commons 20 to 30 minutes (min). In a broad sense, spa therapy comprises
Attribution License 4.0 (CCBY), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
therapeutic modalities including balneotherapy, mud-pack
therapy, massage, and supervised water exercises in spa resorts,
How to cite this article: Bai R, Li C, Xiao Y, Sharma M, Zhang F, Zhao Y.
Effectiveness of spa therapy for patients with chronic low back pain. Medicine adding other benefits such as a pleasant climate, relaxing natural
2019;98:37(e17092). scenery, and clean air.[21,22] It is an ancient way to treat rheumatic
Received: 24 June 2019 / Received in final form: 9 August 2019 / Accepted: 16 and musculoskeletal disorders which can relieve the pain and
August 2019 improve the function in musculoskeletal disorders,[23–26] but the
http://dx.doi.org/10.1097/MD.0000000000017092 mechanism has not been clearly illuminated.[27,28] It may

1
Bai et al. Medicine (2019) 98:37 Medicine

associate with hydrostatic pressure, mineral composition, and names, and publication time, reported study characteristics:
temperature.[23,29,30] Immersing in warm water may contribute course of the treatment, overall follow-up duration, character-
to an analgesic effect by thermal effect and hydrostatic pressure of istics of the thermal water: geographical area, composition,
water on the skin according to the “Gate control theory of mineral concentration, and temperature, intervention and
pain.”[31] And due to a lower specific heat, mud-pack therapy control group: method of therapy, duration, and frequency,
elevates the body-core temperature more efficiently.[32] In observing parameters: visual analogue scale (VAS), Schober test,
addition, exercise or physical activity is vital for CLBP patients and Oswestry disability index (ODI), outcome measurements: the
to help them complete their daily activities by enhancing muscle evaluation of outcome.
strength, increasing aerobic capacity of lumbar muscles, and
promoting local blood flow.[23,24,33,34] Using spa therapy for
2.4. Methodological quality assessment
managing CLBP is a Grade B recommendation.[35]
In 2006, Pittler et al[20] performed a meta-analysis about the Jadad checklist was used to evaluate included studies on
effect of spa therapy among patients with LBP, and concluded different aspects, including treatment methods relevant to the
that spa therapy has a positive effect in pain relieving based on 5 description of randomization, double-blind structure, and with-
studies. A later systematic review summarized the studies drawals/dropouts.[39] The range of quality score is from 0 to 5
published between 2005 and 2013 and reported the positive (the lowest to highest). Studies with a score of or over 3 were
effects of spa therapy in treating CLBP.[36] Considering different regarded as having a good quality. Two reviewers (R.B. and C.L.)
additional intervention methods may affect therapeutic effects; assessed the quality of included studies independently. Disagree-
therefore, we conducted a systematic review and meta-analysis to ments were resolved through discussion until reaching a
provide an updated overview of the literature in this area and to consensus.
further assess short-term effect of spa therapy in patients with
CLBP with a more detailed classification on intervention methods 2.5. Statistics analysis
of 3 subgroups: balneotherapy, balneotherapy with mud pack,
and balneotherapy with physiotherapy. VAS, Schober test, and ODI evaluate the intensity of pain,
lumbar spine mobility, and lumbar spine function respectively,
and they were chosen as main outcome measures for meta-
2. Materials and methods analysis. In some included studies, these measures were
This study was performed according to the statement, preferred examined for several times at different time points. The data
reporting items for systematic reviews and meta-analyses at the first time point after treatment and/or in the rest condition
(PRISMA)[37] and recommendations of the Cochrane Collabora- were used for analysis. All the quantitative data were converted
tion.[38] All analyses were conducted based on previously published into millimeter unit. The random effects model was applied to
studies, so no ethical approval and patient consent are required. generate summary estimates. Heterogeneity was assessed by I2
test. When I2 < 25%, it means no heterogeneity; when 25%
2.1. Search strategy I2 < 50%, it means moderate heterogeneity. The heterogeneity
is acceptable; when I2≥50%, it means strong heterogeneity.
The study used the following words as search terms: “spa Subgroup and sensitivity analysis were used to examine the
therapy,” “balneotherapy,” “balneology,” “hot spring,” and source of heterogeneity.[40] Funnel plots were used to assess
“geothermal spring” combine with “low back pain” and publication bias. All statistical analyses were conducted in
“lumbago” in PubMed, Embase, Web of Science, and Cochrane Review Manager (version 5.2).
Library. Each database was searched from its inception to May
2018. Two authors (R.B. and C.L.) screened independently. The
search strategy applied a combination of title and abstract, and 3. Results
used the Mesh Term. Hand searching is performed by reviewing 3.1. Study selection
the references of included studies.
A total of 327 studies were initially retrieved from databases, and
12 RCT studies met the eligibility criteria and were included, and
2.2. Study selection
their data were assessed in the meta-analysis (Fig. 1).
Titles and abstracts of identified articles were reviewed by 2 authors
(R.B. and C.L.) independently. When 2 reviewers could not reach a 3.2. Study characteristics
consensus, disagreements and uncertainties were resolved through
discussion. The including criteria were: patients who were The characteristics of included studies were summarized in
diagnosed with CLBP, treated with spa therapy in a randomized Table 1. Based on their intervention methods, 3 were
way [randomized clinical trials (RCTs)], clinical trials whose main balneotherapy,[41–43] 2 were balneotherapy with mud-pack
objectives included the effectiveness of spa therapy, intervention for therapy,[44,45] and 7 were balneotherapy with physiothera-
spa therapy applied as a combination of balneotherapy with py.[46–52] The length of treatment in most trials was around 3
physiotherapy, mud-pack, publications in English only. Exclusion weeks.[41,42,44–46,48–51] The follow-up efficacy of spa-therapy was
criteria were: the mineral water was not natural spring, spa therapy observed in 8 trials.[41–45,47,50–52] Most of them reported a
intervention lasted for more than 3 months. significant improvement in pain relief, lumbar flexibility,
functional capacity, and quality of life. No adverse events were
reported in all included trials. These studies were performed
2.3. Data extraction
in Hungary,[42,43,50,51] Turkey,[46–48,52] France,[41,44,45] and
Two reviewers (R.B. and C.L.) extracted the following data from Croatia[49] and the temperature of spa therapy was between
all included studies independently: article information: authors’ 31°C and 38°C.

2
Bai et al. Medicine (2019) 98:37 www.md-journal.com

Figure 1. Study selecting flowchart.

3.3. Meta-analysis 3.5. Sensitivity analyses

[41–48,50–52]
Overall, 11 studies were included in meta-analysis. In Lower heterogeneity was observed in results of spa therapy ODI
Figure 2, 966, 808, and 468 patients with data on VAS, Schober (I2 = 87–54%) after excluding the study of Kulisch.[51] After
tests, and ODI respectively were included in data synthesis. In excluding 1 study from the balneotherapy and physiotherapy of
respect of effectiveness of spa therapy for CLBP, there was subgroup, the heterogeneity decreased dramatically in all
statistical significance between treatment and control group in outcomes indicators, shown in Supplementary Figure 4, http://
VAS (mean difference [MD] 16.07, 95% confidence interval [CI] links.lww.com/MD/D231.
[9.57, 22.57], P < .00001, I2 = 88%, n = 966), and ODI (MD
7.12, 95% CI [3.77, 10.47], P < .00001, I2 = 87%, n = 468). No 3.6. Quality assessment
statistically significance was found in Schober test (MD 2.94,
95% CI [ 0.75, 6.63], P < .00001, I2 = 97%, n = 808). For quality evaluation, 6 studies have good quali-
ty[42,44,47,48,50,51] with only 1 trial having a full score.[51] Other
3.4. Subgroup analyses trials showed low quality: 3 scored 1 point,[43,46,49] and 3 had
2 points.[41,45,52]
According to treatment method, 3 subgroups were divided:
balneotherapy group, balneotherapy with mud-pack therapy
group, and balneotherapy with physiotherapy group. No ODI
4. Discussion
data were collected in balneotherapy group, results shown in Our meta-analysis examined the effect of short-term spa therapy
Supplementary Figure 1, http://links.lww.com/MD/D231, het- on pain relief and lumbar spine function improvement among
erogeneity of VAS, and Schober test were still huge. No ODI data patients with CLBP. Although spa therapy has been widely used
were collected in balneotherapy with mud-pack therapy group. in the world, especially in Europe, comprehensive and concrete
Only VAS had statistical significance (MD 23.99, 95% CI [18.33, evidence is still needed to verify its effectiveness for CLBP.
29.66], P < .00001, I2 = 0%, n = 340), shown in Supplementary Compared to previous meta-analysis and review publica-
Figure 2, http://links.lww.com/MD/D231. Supplementary Fig- tions,[20,36] the present meta-analysis included more studies,
ure 3, http://links.lww.com/MD/D231 showed results of balneo- examined more outcome measurements including lumbar spine
therapy with physiotherapy group. However, heterogeneities in mobility (Schober test) and lumbar spine function (ODI), and
VAS, Schober test, and ODI were significant. stratified analysis based on different intervention methods:

3
 Table 1
Characteristics of the included studies.
Geographical area and
Authors thermal water Length of the Intervention Control Follow-up Main outcome Jadad
and year composition treatment group group Main outcomes duration analysis scores
Balneotherapy
Balogh et al 2005 GA: Budapest, Hungary 6 d/wk for 15 d Balneotherapy: 30 min Tap water Pain intensity (10-cm Baseline, 15 d, Intervention group: VAS, 1
TW: 748 mg/L total thermal water bath, VAS); Oswestry 3 mo follow-up Schober test (all with
mineral content, the 36°C disability index (ODI): a P < .01) were
Bai et al. Medicine (2019) 98:37

dominant cationic ability to attend to significantly improved

content was sulfide personal needs, Compare with
residue (sulfur ions=2.4 lifting, standing, baseline. Comparator
mg/L), and rich in free walking; Extension group: P < .01,
carbonic acid and and lateral flexion; P > .3 for VAS,
metasilicic acids Analgesic dose Schober test score
requirements respectively compare
with baseline
Guillemin et al 1994 GA: Bains-les-Bains, 6 d/wk for 3 wk Balneotherapy: 15 min Waiting list Pain intensity (100-mm Baseline, 3 wk, VAS and Schober test 2
France TW: < 500 mg/L underwater high VAS); Schobertest; 9 mo follow-up (all with a P < .0001)
total mineral, mainly pressure showers and Finger-floor distance; were significantly
contains sulfate and 3 min showers with drug consumption improved compared
sodium various pressures, with 2 groups.
31–36°C
Tefner et al 2012 GA: Matraderecske, 5 d/wk for 3 wk Balneotherapy: 30 min Tap water Pain intensity (100-mm Baseline, 3 wk, 3 wk VAS, ODI, and Schober 3

4
Hungary TW: 10,900 thermal water bath, VAS); Schober test; follow-up, 10 wk test (P < .01, P < .05
mg/L total mineral rich 31°C Oswestry Disability follow-up and P < 0.01
in sodium (2800 mg/L), Index (ODI); range of respectively) were
hydrogen carbonate lateral flexion of the significantly improved
(4728 mg/L), chloride lumbar spine; Short- compared with 2
(1860 mg/L), sulfate Form 36 Health groups.
(640 mg/L), lithium (6.9 Survey (SF 36)
mg/L), and bromide
(9.4 mg/L), also
containing abundant of
iodine (0.85 mg/L),
fluoride (0.86), and
metaboric acid (44 mg/
L)
Mud-pack therapy
Constant et al 1998 GA: Vittel, France TW: 6 d/wk for 3 wk Balneotherapy: 10 min Waiting list Pain intensity (100-mm Baseline, 3 wk, P < .01, P=.22 for VAS, 2
treatment group 1 has thermal water bath, VAS); Duke health 3 mo follow-up Schober index
510 mg/L little total 36°C + 15 min mud profile(QoL); Finger- respectively compared
mineral; treatment application, 45°C and floor distance; with 2 groups.
group 2 has 1585 mg/L 20 min massage Schober index;
large total mineral under flowing water Disability
contains sulfate, questionnaire
chloride, sodium
(continued )
Medicine
 Table 1
(continued).
Geographical area and
Authors thermal water Length of the Intervention Control Follow-up Main outcome Jadad
and year composition treatment group group Main outcomes duration analysis scores
Constant et al 1995 GA: Saint Nectaire, France 6 d/wk for 3 wk Balneotherapy: 10 min Waiting list Pain intensity (100-mm Baseline, 3 wk, P < .0001 for VAS, 3
TW: 8073 mg/L total thermal water bath, VAS); Finger-floor 3 mo follow-up P = .38 for Schober
mineral mainly contains 36°C+ 20 min mud distance; Schober index compared with
bicarbonate, chlorine, application 45°C and test; Disability 2 groups.
Bai et al. Medicine (2019) 98:37

and sodium 2.5 min high-pressure questionnaire; Drug

shower consumption
Physiotherapy
Demirel et al 2008 GA: Afyon, Turkey TW: 5 d/wk for 3 wk Balneotherapy: 20–25 Exercise Pain intensity (10-cm Baseline, 3 wk Intervention group: 2
containing sodium, min thermal water program VAS); Oswestry P = .003, P = .002,
bicarbonate, sulfate, bath, 36–38°C + disability index (ODI); and P = .119 for VAS,
calcium, magnesium, Exercise program: 10 SF36; Symptom ODI, and Schober
iron, aluminum, min for each exercise Checklist-90-Revised index respectively
chlorine, and session and took 2 (SCL-90-R); The compared with
metasilicate min rest between Hospital Anxiety and baseline. Comparator
each exercise Depression Scale group: P = .01,
(HAD); Spine Joint P = .011, and
Mobility Tests; Spine P = .452 for VAS,
Joint Mobility Tests ODI, and Schober

5
index respectively
compared with
baseline.
Dogan et al 2011 GA: Sivas, Turkey TW: 5 d/wk for 3 wk Balneotherapy: 20 min Physiotherapy 10-cmVAS; Schober Baseline, 3 wk Intervention group: VAS, 1
3454 mg/L rich in thermal waterbath,?°C test; ROI score; right Schober test, ROI
sodium, bicarbonate, + Physiotherapy: 6 and left lateral flexion score (all with a
chloride, calcium, and min ultrasound (US) P < .0001) were
magnesium, also at a dose of 1.5 W/ significantly improved
containing fluoride, cm2 and a frequency compared with
sulfate and silicate of 1 MHz and applied baseline. Comparator
20 min group: P < .0001,
transcutaneous, P = .03, P < .0001
electrical nerve for VAS, Schober test,
stimulation (TENS), ROI score,
20 min hot pack and respectively compared
standard exercise with baseline.
Gáti et al 2017 GA: Budapest, Hungary 5 d/wk for 3 wk Balneotherapy: 20 min Physiotherapy Pain intensity (100-mm Baseline, 3 wk, Intervention group: 3
TW: 1080 mg/L mineral thermal water bath, VAS); Oswestry 3 mo follow-up P < .001 for both
rich in calcium, 38°C + disability index (ODI); VAS and ODI
magnesium, sodium Physiotherapy: EuroQol Five respectively. VAS and
bicarbonate with high physical therapy, Dimensions ODI (all with a
hardness (total hardness massage, TENS and Questionnaire (EQ-5D) P < .0001) were
259 CaO mg/L, 25.9 ultrasound treatments significantly improved
nkf) compared with 2
groups.
www.md-journal.com

(continued )
 Table 1
(continued).
Geographical area and
Authors thermal water Length of the Intervention Control Follow-up Main outcome Jadad
and year composition treatment group group Main outcomes duration analysis scores
Kesiktas et al 2012 GA: Karaali, Turkey TW: 5 d/wk for 2 wk Balneotherapy: 30 min Physiotherapy 100-mm VAS, manual Baseline, 2 wk, Intervention group: 3
581.54 mg/L total thermal water bath, muscle tests (MMT), 3 mo follow-up P = .000, P = .01,
mineral content, 36°C + Modified Schober and P = .01 for VAS,
typically a calcium Physiotherapy: method (MS), Schober test, and ODI
Bai et al. Medicine (2019) 98:37

bicarbonate character, exercise program, Oswestry disability respectively

but sodium chloride transcutaneous index (ODI), Short- Comparator group:
content is also high, electrical nerve Form 36 Health VAS P = .000,
and also containing stimulation (TENS), Survey (SF 36), P > .05, and P > .05
fluoride, sulfate and ultrasound, infrared Paracetamol dose for VAS, Schober test,
magnesium radiation, and and ODI respectively
exercise.
Kulisch et al 2009 GA: Celldömölk, Hungary Daily for 3 wk Balneotherapy: 20 min Tap water and Pain intensity (100-mm Baseline, 3 wk, Intervention group: VAS, 5
TW: 3350 mg/L total thermal water bath, electrotherapy VAS); Oswestry 3 mo follow-up Schober test, ODI (all
mineral rich in sodium, 34°C + disability index (ODI); with a P < .01) were
hydrogen carbonate, Electrotherapy: 3 min SF 36; Schober test; significantly improved.
and chloride, also additional left and right lateral Comparator group:
containing fluoride and electrotherapy forward flexion P < .01 for VAS,
iodides (3 d/wk) Schober test, and ODI

6
were not significantly
improved. VAS and
Schober test had no
significantly difference
between 2 groups,
ODI was significantly
improved compared
with 2 groups.
Nemcić et al2013 GA: Croatia TW: Consisting 5 d/wk for 3 wk Balneotherapy: 45 min Same physiotherapy but Modified Schober Baseline, 3 wk There was a statistically 1
mainly of sodium, thermal water, 36°C exercise program method (MS); left and significant
calcium, hydrogen + Physiotherapy: based on land right lateral flexion; improvement in
carbonate, and sulfate under water exercise Physical Disability lumbar mobility and
program, including Index (PDI) physical disability
warming up through compared with initial
the water in the pool; values. Results
active range of showed improvement
motion of the joints of due to the treatment
the upper and lower (significant main effect
extremities; of the treatment) but
stretching; not significant
strengthening interaction effects
exercises for hips, between the types of
knees, arms, elbows exercise therapy
and wrists; and tested before and after
cooling- down, TENS the treatment.
Medicine

(5 d/wk), and under

(continued )
 Table 1
(continued).
Geographical area and
Authors thermal water Length of the Intervention Control Follow-up Main outcome Jadad
and year composition treatment group group Main outcomes duration analysis scores
water massage (2 d/
wk)
Onat et al 2014 GA: Ankara, Turkey TW: 5 d/wk for 3 wk Balneotherapy: 20 min Physiotherapy Pain intensity (10-cm Baseline, 3 wk Intervention group: 3
2595 mg/L total thermal water bath, VAS); Oswestry P < .001 for both
Bai et al. Medicine (2019) 98:37

mineral contains 38°C + Disability Index (ODI); VAS and ODI, and all
sodium, bicarbonate, Physiotherapy: 20 min Fingertip-to-floor the domains of SF-36
fluoride, and chlorine hot pack (HP), 20 min distance; Short-Form were found
transcutaneous 36 Health Survey (SF significantly improved
electrical stimulation 36) respectively
(TENS) (50–100 Hz), Comparator group:
and 5 min P < .001 for VAS,
ultrasonography (US) P < .01 for ODI, only
at a dose of 1 W/cm2 2 domains (P = .088
and a frequency of 1 for mental status role
MHz, and a home- (MSR), P = .118 for
based standardized mental score (MS)) of
exercise program SF-36 were not found
significantly improved

7
www.md-journal.com
Bai et al. Medicine (2019) 98:37 Medicine

Figure 2. Effect estimates on included studies comparing thermal water with control.

balneotherapy,[41–43] balneotherapy with mud-pack,[44,45] and control group. The treatment durations of both studies in this
balneotherapy with physiotherapy.[46–52] group were relatively short, which were around 10 minutes
In subgroup and sensitivity analysis, VAS’s improvement in thermal mineral water bath. Generally, mean duration of
treatment group was significantly higher than control group, balneotherapy was 20 minutes to 30 minutes. Thus, this
which is consistent with the findings reported by Pittler et al.[20] experimental design may lead to an incomplete demonstration
Significant heterogeneity was observed in balneotherapy and in the effectiveness of spa therapy. Otherwise, although ODIs
balneotherapy with physiotherapy group. In sensitivity analyses, were significantly decreased in subgroup analysis and sensitivity
heterogeneity decreased, Schober tests variations increased in analysis, significant heterogeneities could not be neglected.
both subgroups. In balneotherapy group, heterogeneity may be Indeed, ODI is a patient self-rated scale with greater subjectivity,
explained by differences in study design between Guillemin while Schober test is more objective. In addition to pain
et al[41] and others.[42,43] In balneotherapy with physiotherapy alleviating in patients with CLBP, spa therapy also improves
group, Gáti et al’s and Kulisch et al’s studies[50,51] were lumbar mobility.
conducted in Hungary, and other studies[46–49,52] were conducted We evaluated the short-term spa therapy effect. Eight trials
in Turkey and Croatia, all of latter were Mediterranean evaluated the follow-up efficacy,[41–43,45,47,50,51] and most of
countries. In addition, Kulisch’s study[51] had the full scores of them lasted for 3 months, except one was 6 months[44] and the
Jadad check list, the heterogeneity across included studies other was 9 months.[41] After follow-up, most studies have
decreased after its exclusion, which may be due to the overall observed that VAS significantly decreased compared with the
methodology inconsistencies. Specifically, this trial used tap baseline levels, and the Schober index and ODI also improved
water combined with physiotherapy was performed as control significantly. There were significant differences in the drop of
group, different from other physiotherapy studies. However, in VAS scores between the intervention and control group. In
balneotherapy with mud-pack group, there was no significant Guillemin et al,[41] the authors used spa therapy as an
difference in Schober test variation between treatment and intervention group, while the control group only allowed

8
Bai et al. Medicine (2019) 98:37 www.md-journal.com

painkillers. After 9 months of follow-up, the authors observed Second, heterogeneity in results was considerable. We
that results based on VAS and Schober test significantly ascribed this to poor designs and excessive time gap of
improved. But, considering the control group did not receive included studies. Most studies reported unclear randomiza-
any treatment in 9 months, symptoms might become more severe. tion and insufficient double-blind design. Further research
The authors believed that this may lead to an overestimation in with high-quality RCTs was required. Furthermore, the
the long-term therapeutic effect of spa therapy. While the short- sample size in all the included studies was small (<100 per
term effect of spa therapy is well known, its long-term benefit is treatment arm). The small number of studies and participants
still under discussion because of the paucity evidence. Compared included would result in an underpowered analysis. These
to the baseline, 7 trials observed that the VAS of spa therapy included studies’ published time spanned over 24 years, and
group was significantly decreased.[43,46–48,50–52] The effect of spa the excessive time gap that might reduce the homogeneity of
therapy on Schober index and ODI is controversial: most participants. In addition, more parameters are needed to
researchers suggested that the spa therapy could ameliorate the evaluate and verify the efficacy of spa therapy, and the long-
lumber function or mobility after the treatment,[43,46–48,50,51] term efficacy should be confirmed.
although other researchers did not find the improvement.[52]
Meanwhile, some trials used therapeutic methods in control
5. Conclusion
groups because of ethical reasons, such as hydrotherapy,[42,43,51]
physiotherapy.[46,47,50,51] In these studies, VAS was also In conclusion, this updated systematic review and meta-analysis
significantly lower than the baseline. Hydrotherapy, exercise demonstrated that spa therapy may have short-term beneficial
therapy, as well as the physiotherapy, also has therapeutic effects on pain reliving and lumbar spine mobility improvement
effects.[53] These designs will influence results of the studies. in patients with CLBP. This meta-analysis provides recommen-
Although Tefner et al[42] observed that the VAS and range of dations for future research: more rigorous study design, longer
motion significantly improved and differed between groups, there follow-up period, and bigger sample size to provide more
was no statistical difference in Kulisch et al’s study.[51] convinced evidence in spa therapy to treat CLBP.
No adverse events were reported in included studies and
adverse events in spa therapy are rarely reported. Previous studies Acknowledgments
pointed out that the most common adverse event was respiratory
tract infections (8%), which were more common among patients Thanks for Hongtao Tie and Guochao Zhong of Chongqing
with chronic respiratory failure and chronic bronchitis.[35,54] Medical University who modified the manuscript and gave us
Other common adverse events include mild neurological some advice, and Xinran Lai of Monash University who modified
disorders (6%), pain exacerbation (5%), skin diseases (2%), language and grammar for this study.
falls (1%), urinary tract infections (<1%), cardiovascular
disorders, and erysipelas (0.005%) and should also be paid
Author contributions
attention to.
Up to now, there is no guideline about spa therapy. According Conceptualization: Ruixue Bai, Chihua Li, Fan Zhang, Yong
to designs of included studies, we recommend that the duration of Zhao.
spa therapy should longer than 30 minutes; temperature should Methodology: Ruixue Bai, Chihua Li, Fan Zhang, Yong Zhao.
be higher than 38°C. Besides, patients with following conditions Data curation: Ruixue Bai, Chihua Li, Yong Zhao.
are not suggested to receive spa therapy: acute infection, Validation: Ruixue Bai, Chihua Li, Yangxue Xiao, Fan Zhang.
pregnancy, cardiovascular diseases (such as heart failure, Writing – review & editing: Ruixue Bai, Chihua Li, Yang-xue
unstable hypertension, angina pectoris), respiratory insufficiency, Xiao, Manoj Sharma, Fan Zhang, Yong Zhao.
uncontrolled liver disorders, uncontrolled and unstable metabol- Conceptualization: Manoj Sharma, Fan Zhang, Yong Zhao.
ic disorders, epilepsy, and uncontrolled epilepsy.[17,45,52] Data curation: Ruixue Bai.
Interestingly, all included studies in our review were conducted Formal analysis: Ruixue Bai, Chihua Li.
in Europe (Hungary, Turkey, France, and Croatia). The first trial Funding acquisition: Yong Zhao.
about spa therapy who used double blind and tap water control Methodology: Fan Zhang, Yong Zhao.
was performed in Hungary, was applied among patients Project administration: Fan Zhang, Yong Zhao.
with rheumatoid arthritis. It might be because that in other Software: Ruixue Bai.
countries, people go to spas not only for health but also for Supervision: Fan Zhang, Yong Zhao.
recreation and rest.[55] Writing – original draft: Ruixue Bai, Chihua Li.
As for the methodological assessment, there was only 1 full Writing – review & editing: Ruixue Bai, Chihua Li, Yangxue
marks study.[51] Interestingly, in subgroup analysis and sensitivi- Xiao, Manoj Sharma, Fan Zhang, Yong Zhao.
ty analysis, after exclusions of this study, we found the Ruixue Bai orcid: 0000-0001-6043-3877.
heterogeneity declined, maybe the inconsistencies of study
methods cause the heterogeneity, especially the missing designs
of double-blind study design. However, it is difficult to execute References
blinding because of special smell of spa water. Therefore, RCTs [1] GBD 2016 Disease and Injury Incidence Prevalence CollaboratorsGlo-
with more rigorous double-blind design are needed. bal, regional, and national incidence, prevalence, and years lived with
There are certain limitations in this meta-analysis. First, all disability for 328 diseases and injuries for 195 countries, 1990-2016: a
included studies were only published in English, whereas in systematic analysis for the Global Burden of Disease Study 2016. Lancet
2017;390:1211–59.
this filed the majority of the studies were conducted in Europe, [2] Becker A, Held H, Redaelli M, et al. Low back pain in primary care: costs
so studies published in other languages cannot be analyzed in of care and prediction of future health care utilization. Spine (Phila Pa
this meta-analysis. This may contribute to publication bias. 1976) 2010;35:1714–20.

9
Bai et al. Medicine (2019) 98:37 Medicine

[3] Kolber MJ, Beekhuizen K. Lumbar stabilization an evidence-based [30] Melzack R, Wall PD. Pain mechanisms: a new theory: a gate control
approach for the athlete with low back pain. Stength Cond J 2007;29: system modulates sensory input from the skin before it evokes pain
26–37. perception and response. Pain Forum 1996;5:3–11.
[4] Lara-Palomo IC, Aguilar-Ferrándiz ME, Matarán-Peñarrocha GA, et al. [31] Rahmann AE. Exercise for people with hip or knee osteoarthritis: a
Short-term effects of interferential current electro-massage in adults with comparison of land-based and aquatic interventions. Open Access J
chronic non-specific low back pain: a randomized controlled trial. Clin Sports Med 2010;1:123–35.
Rehabil 2013;27:439–49. [32] Maeda T, Kudo Y, Horiuchi T, et al. Clinical and anti-aging effect of
[5] Manchikanti L, Singh V, Falco FJ, et al. Epidemiology of low back pain in mud-bathing therapy for patients with fibromyalgia. Mol Cell Biochem
adults. Neuromodulation 2014;17:3–10. 2017;444:87–92.
[6] Spitzer WO, Skovron ML, Salmi LR, et al. Scientific monograph of the [33] Gordon R, Bloxham S. A systematic review of the effects of exercise and
Quebec Task Force on Whiplash-Associated Disorders: redefining “whip- physical activity on non-specific chronic low back pain. Healthcare
lash” and its management. Spine (Phila Pa 1976) 1995;20(8 suppl):1S–73S. (Basel) 2016;4:22–40.
[7] Bogduk N. Management of chronic low back pain. Med J Aust [34] Smeets RJ, Severens JL, Beelen S, et al. More is not always better: cost-
2004;180:79–83. effectiveness analysis of combined, single behavioral and single physical
[8] van Tulder MW, Koes B, Malmivaara A. Outcome of non-invasive rehabilitation programs for chronic low back pain. Eur J Pain 2009;13:
treatment modalities on back pain: an evidence-based review. Eur Spine J 71–81.
2006;15(suppl 1):S64–81. [35] Forestier R, Erol-Forestier FB, Francon A. Current role for spa therapy in
[9] Andersson GB. Epidemiological features of chronic low-back pain. rheumatology. Joint Bone Spine 2017;84:9–13.
Lancet 1999;354:581–5. [36] Karagülle M, Karagülle MZ. Effectiveness of balneotherapy and spa
[10] Mohseni Bandpei MA, Ehsani F, Behtash H, et al. Occupational low therapy for the treatment of chronic low back pain: a review on latest
back pain in primary and high school teachers: prevalence and associated evidence. Clin Rheumatol 2015;34:207–14.
factors. J Manipulative Physiol Ther 2014;37:702–8. [37] Panic N, Leoncini E, Belvis GD, et al. Evaluation of the Endorsement of
[11] Minghelli B, Oliveira R, Nunes C. Non-specific low back pain in the Preferred Reporting Items for Systematic Reviews and Meta-Analysis
adolescents from the south of Portugal: prevalence and associated (PRISMA) statement on the quality of published systematic review and
factors. J Orthop Sci 2014;19:883–92. meta-analyses. PLoS One 2013;8:e83138.
[12] Durlov S, Chakrabarty S, Chatterjee A, et al. Prevalence of low back pain [38] Higgins JPT, Thompson SG, Deeks JJ, et al. Measuring inconsistency in
among handloom weavers in West Bengal, India. Int J Occup Environ meta-analyses. BMJ 2003;327:557–60.
Health 2014;20:333–9. [39] Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of
[13] Shmagel A, Foley R, Ibrahim H. Epidemiology of chronic low back pain randomized clinical trials: is blinding necessary? Control Clin Trials
in US adults: National Health and Nutrition Examination Survey 2009- 1996;17:1–2.
2010. Arthritis Care Res (Hoboken) 2016;68:1688–94. [40] Strauss-Blasche G, Ekmekcioglu C, Vacariu G, et al. Contribution of
[14] Silva MR, Badaró AF, Dall’Agnol MM. Low back pain in adolescent and individual spa therapies in the treatment of chronic pain. Clin J Pain
associated factors: a cross sectional study with schoolchildren. Braz J 2002;18:302–9.
Phys Ther 2014;18:402–9. [41] Guillemin F, Constant F, Collin JF, et al. Short and long-term effect of spa
[15] Balague F, Mannion AF, Pellise F, et al. Clinical update: low back pain. therapy in chronic low back pain. Br J Rheumatol 1994;33:148–51.
Lancet 2007;369:726–8. [42] Tefner IK, Németh A, Lászlófi A, et al. The effect of spa therapy in
[16] Dagenais S, Caro J, Haldeman S. A systematic review of low back pain chronic low back pain: a randomized controlled, single-blind, follow-up
cost of illness studies in the United States and internationally. Spine J study. Rheumatol Int 2012;32:3163–9.
2008;8:8–20. [43] Balogh Z, Ordogh J, Gasz A, et al. Effectiveness of balneotherapy in
[17] Konrad K, Tatrai T, Hunka A, et al. Controlled trial of balneotherapy in chronic low back pain—a randomized single-blind controlled follow-up
treatment of low back pain. Ann Rheum Dis 1992;51:820–2. study. Forsch Komplementarmed Klass Naturheilkd 2005;12:196–201.
[18] Krismer M, van Tulder M. Strategies for prevention and management of [44] Constant F, Collin JF, Guillemin F, et al. Effectiveness of spa therapy in
musculoskeletal conditions. Low back pain (non-specific). Best Pract Res chronic low back pain: a randomized clinical trial. J Rheumatol
Clin Rheumatol 2007;21:77–91. 1995;22:1315–20.
[19] Marienke VM, Rubinstein SM, Verhagen AP, et al. Exercise therapy for [45] Constant F, Guillemin F, Collin JF, et al. Use of spa therapy to improve
chronic nonspecific low-back pain. Best Pract Res Clin Rheumatol the quality of life of chronic low back pain patients. Med Care
2010;24:193–204. 1998;36:1309–14.
[20] Pittler MH, Karagülle MZ, Karagülle M, et al. Spa therapy and [46] Dogan M, Sahin O, Elden H, et al. Additional therapeutic effect of
balneotherapy for treating low back pain: meta-analysis of randomized balneotherapy in low back pain. South Med J 2011;104:574–8.
trials. Rheumatology (Oxford) 2006;45:880–4. [47] Kesiktas N, Gun K, Gun N, et al. Balneotherapy for chronic low back
[21] Fraioli A, Mennuni G, Fontana M, et al. Efficacy of spa therapy, mud- pain: a randomized, controlled study. Rheumatol Int 2012;32:3193–9.
pack therapy, balneotherapy, and mud-bath therapy in the management [48] Onat SS, Tasoglu O, Guneri FD, et al. The effectiveness of balneotherapy
of knee osteoarthritis. A systematic review. Biomed Res Int in chronic low back pain. Clin Rheumatol 2014;33:1509–15.
2018;2018:1042576. [49] Nemcic´ T, Budisin V, Vrabec-Matkovic´ D, et al. Comparison of the
[22] Bender T, Karagulle Z, Balint GP, et al. Hydrotherapy, balneotherapy, effects of land-based and water-based therapeutic exercises on the range
and spa treatment in pain management. Rheumatol Int 2005;25:220–4. of motion and physical disability in patients with chronic low-back pain:
[23] Verhagen AP, Bierma-Zeinstra SM, Boers M, et al. Balneotherapy for single-blinded randomized study. Acta Clin Croat 2013;52:321–7.
rheumatoid arthritis. Cochrane Library 2004;1:CD000518. [50] Gati T, Tefner IK, Kovacs L, et al. The effects of the calcium-magnesium-
[24] Karagülle M, Kardeş S, Karagülle MZ. Real-life effectiveness of spa bicarbonate content in thermal mineral water on chronic low back pain:
therapy in rheumatic and musculoskeletal diseases: a retrospective study a randomized, controlled follow-up study. Int J Biometeorol
of 819 patients. Int J Biometeorol 2017;61:1945–56. 2018;62:897–905.
[25] van Tubergen A, van der Linden S. A brief history of spa therapy. Ann [51] Kulisch A, Bender T, Nemeth A, et al. Effect of thermal water and
Rheum Dis 2002;61:273–5. adjunctive electrotherapy on chronic low back pain: a double-blind,
[26] Gutenbrunner C, Bender T, Cantista P, et al. A proposal for a worldwide randomized, follow-up study. J Rehabil Med 2009;41:73–9.
definition of health resort medicine, balneology, medical hydrology and [52] Demirel R, Ucok K, Kavuncu V, et al. Effects of balneotherapy with
climatology. Int J Biometeorol 2010;54:495–507. exercise in patients with low back pain. J Back Musculoskelet
[27] Falagas ME, Zarkadoulia E, Rafailidis PI. The therapeutic effect of 2008;21:263–72.
balneotherapy: evaluation of the evidence from randomised controlled [53] Hall J. Hydrotherapy: principles and practice. Physiotherapy
trials. Int J Clin Pract 2010;63:1068–84. 1997;83:601.
[28] Silva A, Queiroz SS, Andersen ML, et al. Passive body heating improves [54] Nguyen C, Boutron I, Rein C, et al. Intensive spa and exercise therapy
sleep patterns in female patients with fibromyalgia. Clinics (Sao Paulo) program for returning to work for low back pain patients: a randomized
2013;68:135–40. controlled trial. Sci Rep 2017;7:17956.
[29] Guidelli GM, Tenti S, Nobili ED, et al. Fibromyalgia syndrome and spa [55] Bender T, Bálint G, Prohászka Z, et al. Evidence-based hydro- and
therapy: myth or reality? Clin Med Insights Arthritis Musculoskelet balneotherapy in Hungary—a systematic review and meta-analysis. Int J
Disord 2012;5:19–26. Biometeorol 2014;58:311–23.

10