Nội dung

Huber et al. BMC Musculoskeletal Disorders 21(09): https://doi.org/10.1186/s12891-019-2582-4 RESEARCH ARTICLE Open Access Green exercise and mg-ca-SO4 thermal balneotherapy for the treatment of nonspecific chronic low back pain: a randomized controlled clinical trial Daniela Huber1,2†, Carina Grafetstätter1†, Johanna Proßegger1, Christina Pichler1, Ewald Wöll3, Martin Fischer3, Martin Dürl2, Karin Geiersperger2, Melanie Höcketstaller2, Stefan Frischhut4, Markus Ritter5,6 and Arnulf Hartl1* Abstract Background: Non-specific chronic low back pain (nscLBP) has a high socio-economic relevance due to its high incidence, prevalence and associated costs. Therefore, it is essential to evaluate effective therapeutic strategies. This study examines the effects of moderate mountain exercise and spa therapy on orthopedic and psychophysiological parameters. Based on a three-armed randomized controlled trial, guided mountain hiking tours and balneotherapy in thermal water were compared to a control group. Methods: Eighty patients with diagnosed nscLBP were separated into three groups: The two intervention groups GE (green exercise) and GEBT (green exercise and balneotherapy) undertook daily mountain hiking tours, whereas the GEBT group got an additional treatment with baths in Mg-Ca-SO4 thermal water. The third group (CO) received no intervention. GE and GEBT group were treated for 6 days; all groups were followed up for 120 days. Results: Compared to GE and CO group, the GEBT treatment showed significant improvements of pain, some orthopedic parameters, health-related quality of life and mental well-being in patients with nscLBP. Conclusions: The results of this study confirmed a benefit of mountain hiking combined with Mg-Ca-SO4 spa therapy as a multimodal treatment of patients with nscLBP. Further studies should focus on long-term-effects of this therapeutic approach. Trial registration: ISRCTN, ISRCTN99926592. Registered 06. July 2018 - Retrospectively registered. Keywords: Chronic non-specific low back pain, Balneotherapy, Spa therapy, Magnesium-calcium-sulfate thermal water, Mountain hiking, Green exercise, Moderate altitude, Nature therapy, Alpine environment, Mountain exercise Background Socioeconomic impact of low back pain In industrialized western countries, complaints of lower back pain are one of the leading health problems regarding incidence and prevalence. This problem occurs in almost all population groups and is responsible for a considerable extent of medical and social services, and consequently also for high * Correspondence: arnulf.hartl@pmu.ac.at † Daniela Huber and Carina Grafetstätter contributed equally to this work. 1 Institute of Ecomedicine, Paracelsus Medical University, 5020 Salzburg, Austria Full list of author information is available at the end of the article macroeconomic costs. An Austrian health survey conducted in 2015 showed that chronic back pain is the most frequent chronic illness in Austria: 1.8 million people were affected, which accounts for 24.4% of the population [1]. Furthermore, a characteristic sex- (women 25.8%, men 22.9%) and age-dependency was observed. In 2015 there were 37,463 acute hospital stays due to complaints in the back with an average length of stay of 6.9 days, with 139 persons under 15 years, 5907 persons aged 15 to 44 years, 11,984 persons aged 45 to 64 years and 19,433 persons aged 65 years and older [1]. The Austrian federal pension fund © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Huber et al. BMC Musculoskeletal Disorders 21(09): (Pensionsversicherungsanstalt, PVA) approved 55,807 orthopaedic inpatient rehabilitation proposals in 2016, of which 1548 were related to chronic back pain. In Austria, there is also the possibility to apply for cure vacations, which is a stay at a health resort with the main goal of maintaining the working ability, where the costs are partly covered by the insurance. In 2016, the PVA granted 87,640 orthopaedic cure applications, 5909 of them for chronic back pain. According to the PVA, a curative stay in 2016 costed in average about 1900 Euros and a stationary rehabilitation program approximately 3600 Euros, which in turn allows conclusions to the total costs [2]. Additionally to the direct costs of hospitalization and rehabilitation, low back pain also incurs considerable indirect costs due to incapacity, disability and early retirements. The drug treatment of pain patients causes further annual costs of more than 1.6 billion Euros and analgesics comprise one of the most frequently prescribed groups of medicines in Germany [3]. These data demonstrate the socio-economic relevance of this topic. Etiology and duration of low back pain Discomfort of the lower back, located centrally and paravertebrally to the spine - caudal to costal arch and cranial to os coccygis -, with or without radiations, is called low back pain (LBP). According to causality, a distinction is made between specific LBP, describing pain of a diagnosable genesis, and non-specific LBP (nscLBP), in which no clear diagnostic indications of a specific cause, i.e. a central pathomechanism or an irritated structure, can be detected [4]. About 85% of the affected patients suffer from nscLBP [5]. In relation to the course of time, acute pain (less than 6 weeks), subacute (less than 12 weeks), and chronic or chronically recurrent (more than 12 weeks) back pain are differed [6]. Acute non-specific back pain is usually self-limiting, with a rate of convalescence of 90% within the first 6 weeks. Chronification of back pain can be assumed after 12 weeks (subacute phase) of therapeutic intervention without relief of symptoms, which is the case in two to 7 % of patients [5]. Using the hypothesis of the biopsychosocial pain model as an explanation for nscLBP, the complexity of this disease pattern becomes clear: The origin and duration of back pain is not only depending on physical (e.g. diminished muscle function, impairment of tissue repair), but also psychological (e.g. self-efficacy claims, solution competence) and social factors (e.g. work history, family expectations) [7]. However, in respect to the heterogeneous patient group, a precise etiology and Page 2 of 18 pathogenesis of this kind of chronic musculoskeletal pain is still elusive [8]. Therapy options of non-specific LBP Due to multimodal causes and complex interactions of biological, psychological and social factors, nscLBP cannot be cured persistently, but various conservative therapy options to reduce pain and impairment are available, addressing the consequences of long-term pain [9]. Preference is given to conservative non-pharmacological treatments [9], with a wide range of recommendations. There are more than 50 different potential therapies promising pain relief or even healing, but only few have been thoroughly evaluated by evidence-based methods [10]. For patients already suffering from back pain, recurrences and chronification can be prevented best by multimodal programs, which means the combination of different sorts of therapy [11]. Two-thirds of systematic reviews covering this topic emphasize the need for new high-quality therapeutic studies [12]. Based on these recommendations, we investigated the following therapy concept for the treatment of patients with nscLBP: green exercise (active therapy combined with experience of nature) and balneotherapy. Physical activity in natural environments such as meadow, forests or alpine pastures, is called green exercise. This nature therapy addresses patients in many ways [13–16]. Current evidence on green exercise refers to three main areas: regulation of immunological and physiological (stress) responses, improvement of psychological states and facilitation of health-promoting behavior [17]. Although, valid data elevating effects of green exercise in the treatment of nscLBP is still missing. In addition to exercise, balneological treatment can be seen as a meaningful therapeutic option. Despite limited available data, there is encouraging evidence that balneo or spa therapy may be effective in the treatment of nscLBP [18]. As trials have reported, patients with nscLBP show improvements in pain, functionality and psychological parameters after balneological intervention [19]. However, further randomized controlled trials (RCTs) with high scientific quality are necessary to investigate this kind of therapy more accurately [20]. This RCT was conducted to find out whether green exercise - in this case moderate mountain hiking - combined with Mg-Ca-SO4 thermal balneotherapy is an effective and economic nature therapy to reduce symptoms (like pain, physical disability, quality of life, depression) and improve wellbeing of patients with nscLBP. Furthermore, we investigated whether green exercise or the combination of green exercise and spa therapy in Mg-Ca-SO4-type water discloses specific effects in intergroup comparisons. This RCT examines the hypothesis, whether a multimodal therapy (spa therapy and green exercise) can reduce the symptoms and Huber et al. BMC Musculoskeletal Disorders 21(09): improve the spinal mobility of patients suffering from nscLBP, in comparison to an intervention group without spa therapy and a non-intervention control group. Materials and methods Subjects Eighty patients (35 men, 45 women, 19 to 65 years old) were included in this study. The participants were primarily recruited all over Austria through communication via the Wasser Tirol web page (www.albenbad.at), advertisements in newspapers and by physicians. Requirements for study inclusion were diagnosed nscLBP pain and repeated medical treatment because of nscLBP during the past 3 years. Therefore, a medical certificate from the respective treating doctor was mandatory for inclusion. Exclusion criteria were as follows: malignant diseases, previous operations in the lumbar spine area, suspected disc herniation, acute pain, confirmed osteoporosis, contraindications of balneotherapy (cardiovascular dysfunction, such as unstable hypertension, angina pectoris, thrombosis; pulmonary dysfunction; endocrine disorders like hyperthyroidism and hyperparathyroidism; other uncontrolled metabolic disorders like diabetes mellitus, active infectious diseases, incontinence etc.), hernia or pregnancy. No sample size calculation was possible with the available data at the time of study proposal for the selected outcomes. The Ethics Committee of Salzburg (415-E/1487/4– 2012) approved the study protocol and all participants gave written consent. This study adheres to CONSORT guidelines for reporting clinical trials. Study design/interventions The study was set up as a randomized controlled clinical trial with three arms. Except for the non-intervention group (control/CO, n = 27), all participants spent 8 days at the village of Grins (Tyrol, Austria, 47°08′30.1″N 10° 30′55.2″E), hosted in comparable hotels and receiving the same meals. The study procedure was carried out according to the following structure: Medical examinations were performed at arrival (day 0, T1), after the Page 3 of 18 intervention (day 8, T2) and after 4 months (day 120, T3). The arrival of the patients took place on Saturday followed by medical measurements and an anamnesis of medical history at the same day. From Sunday to Friday all persons undertook daily guided hiking tours in the mountains. The structured movement program included a daily 5 h hiking tour in and around the municipality of Grins. The weekly hiking program is shown in Table 1. The first group (n = 27) only participated in green exercise/GE, whereas the second group (n = 26) got an additional treatment with baths in thermal water in the Albenbad (green exercise and thermal balneotherapy/ GEBT). The baths in a tub lasted 20 min. Thereafter, the patients could relax in a heated (room temperature) resting tent. The Albenbad houses two bathtubs; after each patient, the water was omitted, the tub was carefully cleaned, disinfected and newly filled with warm (37 °C) thermal water. On the following Saturday all patients took part in identical medical assessments as at the arrival day and departed after finishing them. This procedure was repeated five times with about ten patients per week. Since blinding of modes of (non-)intervention has not been possible (specific smell of the thermal water), the inclusion of a control group (CO) without any intervention was necessary. The participants of this group received no intervention and stayed at home, maintaining their usual lifestyle. In return for their participation the CO group participants received an amount of 150 Euros per person to cover their study-prone expenses. The patients of the GE and GEBT group got a one-week free holiday including food and lodging in hotels in the village of Grins in Tyrol, Austria. Patients were free to discontinue participation in the study at any time without giving reasons. The study was executed from September 2013 to January 2014 for the first two groups (GE and GEBT) and from September 2014 to January 2015 for the third group (CO) to exclude seasonal effects. Due to personnel restrictions, a parallel assessment of all three study groups was not feasible. Orthopedic, physiological and psychological data from day 0 and 8 were collected Table 1 Intervention program Sunday Sea level starting point (= end point) Sea level highest point Elevation gain Kilometers 1459 1842 386 7.85 Monday 1003 1242 434 13.18 Tuesday 1000 1528 632 15.20 Wednesday 1002 1358 377 6.92 Thursday 989 1609 657 8.17 Friday 1025 2290 921 9.61 3407 60.93 Total [sea level] = m a. s. l. Huber et al. BMC Musculoskeletal Disorders 21(09): on-site in a medical field laboratory in Grins. Follow-up examinations on day 120 of the intervention-group members as well as all examinations of the CO group members were conducted at the Paracelsus Medical University of Salzburg, Austria. The participants were stratified by the Korff assessment (pain related disability) [21] and were separated into the three groups by “Random Allocation Software 2.0” via block randomization. The schematic chronological sequence of this controlled clinical examination is shown in Fig. 1. Environmental parameters of the Albenbad water This randomized controlled study was performed in Grins, which is located 1015 m above sea level in the district of Landeck (Tyrol, Austria) in the submontane of the Lechtal Alps. In 2007, this village obtained the official therapeutic and thermal water status for its water. The water produced by the three springs derives predominantly from a deep-seated aquifer with a subordinated admixture of young water, which duration of dwell time can be estimated to be less than 50 years. From age modeling the residence time of the old groundwater component was estimated to be between 10′000 and 30′000 years. The water of the wells is very well protected by the surrounding. It is of the rare Mg-Ca-SO4 – type with a 1.2 ratio of Ca/Mg. An overview of the water analysis can be seen in Table 4 in Appendix. This composition makes this water very interesting for medical application [22]. In 2010, a new bathing facility called Albenbad was established, in which the balneological part of the study was carried out. In the Albenbad, people can experience the water in form of undergoing Kneipp cures, drinking cures or bath cures. Additionally, a regional network of hiking trails, the Albigen paths, was set up in 2011, Page 4 of 18 which also constituted the basic infrastructure for the present investigation. Assessments At the beginning and end of the one-week intervention, as well as 4 months after the intervention, the functional spinal mobility was measured by parts of the Back Performance Scale, the Spine-Check Score MediMouse® and per trunk rotation measurements. Prior to the start, directly after the intervention week and also during the observation period, chronicity, pain, physical and psychological impairment as well as disability due to cLBP (chronic low back pain) of all subjects were assessed via different validated questionnaires (Oswestry Low Back Disability Index, Medical Outcomes Study Short Form 36, modified Visual Analogue Scale, World Health Organization Well-Being Index). In a pain diary, the use of pain medication was documented during the whole study period. Furthermore, the days of incapacity to work and the number of medical consultations due to cLBP in the last months were assessed. These three parameters were collected two times (day 0 and day 120). Back performance scale (BPS) The BPS is an assessment of mobility-related activities in patients with back pain. It is seen as a reliable and valid instrument for measuring functional spinal mobility and to detect relevant clinical changes [23, 24]. The BPS includes five movements, which are based on the activities of daily life (ADL): sock test, lifting test, pick-up test, roll-up test and finger-to-floor test [25]. In this study, only the first two movements (sock test on both sides and lifting test) were performed. The worst measured score during the assessment was recorded – the fewer points the patients reached, the better the Fig. 1 Time Schedule. Schematic chronological process of the controlled clinical trial Huber et al. BMC Musculoskeletal Disorders 21(09): functional mobility of their back (maximum possible value per test: 3 points). Spine-check score© The MediMouse® (Idiag, Switzerland) is a computer-supported skin surface device for measuring lumbar, thoracic and sacral spinal curvature and sagittal range of motion. The MediMouse® is guided along the spine of the patient from the spinous process of the 7th cervical vertebra to the spinous process of the 3rd sacral vertebra, recording the length and contour of the spine. The electronic 3D sensors simultaneously detect the local inclination in all three levels of the room. This information is transmitted via Bluetooth to the PC and evaluated by the software. By intersegmental angles this test calculates the Spine-Check Score© which includes the three criteria of posture, mobility and postural stability. The examination is carried out in the upright position, in flexion and a loaded recording of the spine column (Matthias test) [26]. Since the validity and quality of the recording depends strongly on the applied pressure of the instrument, the instruction and variations of body positioning [27], the MediMouse® test was carried out a total of three times and the mean value of these three measurements was used for further calculation. The overall outcome of the Spine-Check Score© was calculated according to the following weighting: mobility 40%, postural stability 40% and posture 20%. The results were interpreted as follows: the lower the score, the more abnormalities exist. With great range of movement, values of degrees were higher. Measurement of torso rotation For the elevation of the individual trunk rotation degree, the patients were sitting on a treatment bed, while their feet had ground contact. A bar was placed over the shoulders behind the cervical spine and fixed with both hands. The patients were then asked to turn the trunk as far as possible to one side (right or left). It was important that the execution took place without pain and that the motion was not carried out abruptly. The mobility was measured with a digital goniometer (Stabilized Compass AndroidApp, Anagog Software, Tel Aviv, ISR). This survey was repeated three times for each side, the mean value was used for further calculation. Oswestry low Back disability index (ODI) The ODI contains ten items related to pain behavior and daily activities of living that may be affected by cLBP [28]: pain intensity, personal care (e.g. washing or dressing), lifting, walking, sitting, standing, Page 5 of 18 sleeping, social life, traveling and changing degree of pain. The lower points achieved, the lower the impairment in everyday life. A combination of high reliability (on intra-class consistency correlation and on the test-retest) and validity (especially on construct validity) was found in the ODI and it appeared to react sensitive with meaningful clinical changes [29–31]. Modified visual analogue scale (mVAS) A modified visual analogue scale for pain and status of health assessed the subjective pain intensity of the patients. The Visual Analogue Scale is a unidimensional measuring instrument, often used for the pain monitoring of adults [32]. The patients were asked to report their “current” pain intensity and subjective status of health - on one scale each. On the front of these scales there is a verbal question (e.g. “How do you judge your current health status?”), as well as a comic strip with six faces that provides an orientation for the intermediate state of the two parameters. A revised faces pain scale supplemented the mVAS. On the back there is a scale of 0 to 100 mm to quantify the momentary situation of the patients in a valid and reliable way, whereby a higher value indicates a better clinical result (lower pain intensity as well as a better health status) [33–35]. Medical outcomes study short form 36 (SF-36) The SF-36 is a health status scale for measuring the subjective health-related quality of life. It includes the following eight dimensions of health: physical function, role behavior due to physical impairment, pain, general health, vitality and physical energy, social functioning, role behavior due to mental impairment, as well as mental function [36]. The SF-36 questionnaire includes some of the most frequently measured health concepts and is used as a valid health measure for documenting the burden of a disease [37, 38]. World Health Organization well-being index (WHO-5) The WHO-5 is a brief self-contained questionnaire including five positively formulated phrases, which refer to positive mood, vitality and general interests for screening depression in chronic illness: good spirits, relaxation, being active, waking up fresh and rested, and being interested in things are topics of it [39]. The answers of this test result in a sum value with a low total value corresponding to a low level of well-being. WHO-5 Index is seen as a sensitive, specific [40] and effective assessment [39], which is applied successfully in various study fields [41]. Huber et al. BMC Musculoskeletal Disorders 21(09): Page 6 of 18 Statistical analysis Results Statistical analyses were performed using the SPSS software (IBM Corp. Released 2016. IBM SPSS Statistics for Windows, Version 24.0. Armonk, NY: IBM Corp). Statistical significance was set at the 0.05 probability level for all tests and is expressed as p ≤ 0.05 (*), as p ≤ 0.01 (**) or as p ≤ 0.001 (***). For the per protocol analysis, missing values were not replaced. Evaluation of the data distribution was performed by Kolmogorov-Smirnov (with Lilliefors correction of significance) and Shapiro-Wilk normality test and yielded that more than 60% of the generated data follows normal distribution. Due to the smaller sample sizes in the three subgroups, the Shapiro-Wilk test was added for the individual evaluation of normal distribution in the GEBT, GE and CO group [42]. To identify significant differences between the groups at baselines, ANOVA was performed for parametric data and Kruskal-Wallis for non-parametric data, both including Bonferroni correction. Linear mixed models (LMM) with treatment and time or treatment, time and the interaction of treatment and time as fixed factors were used to analyze the effects of the treatments over time. To account for individual differences, the patient ID was set as random effect for all models. In LMM1 the treatments GEBT and GE vs. CO, time and the interaction of treatment and time were included. In LMM2 the treatments GE vs GBT, time and the interaction of treatment and time were evaluated. In LMM3, only treatment (GEBT and GE vs. CO) and time were included. For measurements concerning claim of medical care (pain medication, physicians’ consultations, status of employee’s illness), taken at two instants of time, Friedman test (non-parametric data) was applied. Means and corresponding standard deviations can be found in Table 6 in Appendix. Baseline characteristics expressed as mean ± SD are shown in Table 5 in Appendix. Demographics and patients’ characteristics This randomized controlled clinical study comprised 80 persons with diagnosed nscLBP consisting of 35 men and 45 women (for full demographics see Table 5 in Appendix). The CO group has a bias in age and is slightly younger compared to the GEBT and GE group. With the exception of the parameters physician’s consultations, employee’s illness, Spine-Check Score© postural stability, spine rotation on both sides and mVAS pain, there are no statistically significant baseline differences between the three treatment groups (Table 2, Table 5 in Appendix). Patients could cancel their participation in this clinical trial at any time without giving reasons. (Fig. 2). All dropouts were voluntarily decided by the participants. There was no injury during the intervention and follow-up period. Functional spine mobility With respect to the criteria mobility, LMM1 analysis of the Spine-Check Score© displays a significant increase of values in the GEBT group (p = 0.033) on day 8, reflecting a higher mobility (Fig. 3, Table 3). Furthermore, LMM3 analysis shows a significant difference over time in both intervention groups on day 8 (p = 0.044) (Table 3). Regarding the postural stability, LMM1 analysis observes significantly lower values in both intervention groups on day 120 (GEBT: p = 0.001; GE: p = 0.039) (Fig. 3). LMM3 analysis of the total value of the Spine-Check Score© detects changes over time on day 8 (p = 0.035) (Table 3). Referring to the remaining measurements of the Spine-Check Score©, no significant differences between the groups could be detected. No significant differences between the groups concerning to the items of the Back Performance Scale could be detected via LMM1, but the model Table 2 Baseline characteristics of study participants (full version see Table 5 in Appendix) Green exercise and balneotherapy Green exercise Control Number 26 27 27 p-value with Bonferroni correction Sex (female/male) 14/12 14/13 17/10 Age (years) 53.35 ± 8.26 52.85 ± 6.43 43.81 ± 12.07 < 0.000*** BMI 26.32 ± 4.47 24.78 ± 2.73 25.06 ± 3.18 0.245 Korff 1.92 ± 1.02 1.52 ± 0.75 1.63 ± 0.77 0.580 Pain medication 0.73 ± 1.04 0.74 ± 1.1 1 ± 0.89 0.368 Physicians consultations 0.88 ± 1.28 0.63 ± 0.79 1.37 ± 0.63 0.002** Status of employee’s illness 1 ± 1.67 0.3 ± 0.87 0.89 ± 0.64 0.002** Job satisfaction 74.38 ± 25.21 79.19 ± 20.59 79.96 ± 23.16 1 Data represented as mean ± SD Huber et al. BMC Musculoskeletal Disorders 21(09): Page 7 of 18 Fig. 2 Patient Recruitment. Numbers of included and excluded patients excluding the interaction of time and treatment (LMM3) shows significant changes over time in both intervention groups at all time points (day 8 p < 0.000, day 120 p < 0.000) for the lifting test (Table 3). LMM3 analysis of the trunk rotation to the right side detects significant changes over time in both intervention groups on day 8 (p = 0.001). Looking at the total trunk rotation, both intervention groups show a significant change in treatment (GEBT p = 0,001; GE p = 0,000). Pain behavior and daily activities of living Results of the mVAS indicate a clear long-term trend of a beneficial effect in the GEBT group concerning pain intensity, exhibited in a significant pain relief on day 120 (LMM1, p = 0.002). Also, the GE group showed a long-term effect on day 120 (LMM1, p = 0.035) (Fig. 4, Table 3). It is essential to mention that the patients in the GEBT group evince higher pain intensity in the baseline measurement than those in the GE and CO group, as shown in Table 3. LMM3 analysis of the mVAS pain scale detects a change over time on day 8 (p = 0.048) (Table 3). LMM1 analysis of the values from the mVAS health status scale demonstrate significant improvements of the GEBT group: day 8 p = 0.026, day 120 p = 0.015. Analysis via LMM3 also reports positive significant changes in both intervention groups (day 8: p < 0.000; day 120: p < 0.000) (Table 3). Regarding the health behavior assessed with the mVAS, there is a significant difference in favor of the GEBT group on day 120 (LMM2, p = 0.037) (Table 3). The Oswestry Disability Index (ODI), a measurement for pain behavior and activities of daily living for patients with cLBP, recorded significant changes in both intervention groups over time at all measurement points (LMM3; day 8: p < 0.000, day 120, p < 0.000). Both the ODI and the Korff graduation scale, monitoring pain related disability, elicit trends for improvement in all study arms (LMM 1), but no significant differences over time between the three study groups could be observed (Table 3). Subjective health-related quality of life For the GEBT group, LMM1 analysis of the SF-36 reveals improvement in two of the three test categories of the SF-36 questionnaire, on day 8 (physical health: p = 0.023; total score: p = 0.026). The GE group showed no significant changes (Fig. 5). The linear model without interplay of time and treatment (LMM3) shows significant alteration over time on day 8 (p < 0.000) as well as on day 120 (p < 0.000) in both intervention groups concerning physical health (Table 3). Regarding the SF-36 total score, LMM3 analysis reports significant changes in both Huber et al. BMC Musculoskeletal Disorders 21(09): Page 8 of 18 Fig. 3 Spine-Check Score Mobility and Postural Stability. Functional spine mobility. Mobility and postural stability of the Spine-Check Score©. Linear mixed model of both intervention groups compared to the control group (LMM1). Data shown in percentage change from baseline (±SD), significances are indicated by asterisks. Means (±SD) of data are shown in Table 3 intervention groups (day 8: p = 0.002; day 120: p < 0.000) (Table 3). Concerning the category mental health of the SF-36 questionnaire, LMM3 analysis recorded significant changes over time in the GE and GEBT group on day 8 (p = 0.009) and day 120 (p = 0.040) (Table 3). Claim of medical care All three study groups show an improvement over time (Friedmann test; pain medication p < 0.000; physicians consultations p < 0.000; status of employee’s illness p < 0.000), but no statistically valid differences in-between the groups on day 120 (Kruskal-Wallis test) (Table 6 in Appendix). Depression in chronic illness A significant higher well-being level could be measured by LMM1 analysis in the GEBT group after the intervention week on day 8 (p = 0.001) for the WHO-5 Index (Fig. 6, Table 3). LMM3 analysis demonstrates significant changes over time on day 8 (p < 0.000) in both intervention groups (Table 3). Discussion The present controlled and randomized clinical trial addressed the question, whether green exercise (moderate mountain hiking) alone or in combination with Mg-Ca-SO4 thermal balneotherapy provides beneficial Huber et al. BMC Musculoskeletal Disorders 21(09): Page 9 of 18 Table 3 Linear mixed models Table 3 Linear mixed models (Continued) GEBT and GE vs. Control LMM1 fe LMM3 p value fe p value GEBT vs. GE GEBT and GE vs. Control LMM2 LMM1 fe p value SCS© total fe GEBT vs. GE LMM3 p value fe LMM2 p value fe GEBT x T3 0.871 n.s. GE 0.028 ** GEBT x T3 0.917 n.s. GEBT x T2 0.135 n.s. GEBT 0.249 n.s. GEBT x T2 0.903 n.s. GEBT x T3 0.002 ** 0.308 n.s. GEBT x T2 0.312 n.s. GEBT 0.237 n.s. GEBT x T2 0.837 n.s. GE x T2 0.231 n.s. T2 0.662 n.s. GEBT x T3 0.362 n.s. GE 0.085 n.s. GEBT x T3 0.923 n.s. GE x T3 0.785 n.s. T3 0.088 n.s. GE x T2 0.222 n.s. T2 0.035 * GE x T3 0.409 n.s. T3 0.584 n.s. mVAS pain SCS© mobility GEBT x T2 0.033 * GEBT 0.694 n.s. GEBT x T2 GEBT x T3 0.063 n.s. GE 0.598 n.s. GEBT x T3 GE x T2 0.242 n.s. T2 0.044 * GE x T3 0.567 n.s. T3 0.582 n.s. 0.315 n.s. 0.167 n.s. GE 0.589 n.s. GEBT x T3 GE x T2 0.105 n.s. T2 0.048 * GE x T3 0.035 * 0.166 n.s. T3 mVAS health status SCS© postural stability GEBT x T2 0.026 * GEBT 0.611 n.s. GEBT x T2 0.755 n.s. GEBT x T3 0.015 * GE 0.190 n.s. 0.394 n.s. GEBT x T3 GEBT x T2 0.341 n.s. GEBT 0.040 * GEBT x T2 0.308 n.s. GE x T2 0.054 n.s. T2 0.000 *** GEBT x T3 0.001 *** GE 0.007 ** GEBT x T3 0.200 n.s. GE x T3 0.258 n.s. T3 0.000 *** GE x T2 0.967 n.s. T2 0.325 n.s. GE x T3 0.039 * 0.940 n.s. T3 p value mVAS health behaviour SCS© posture GEBT x T2 0.198 n.s. GEBT 0.968 n.s. GEBT x T2 0.581 n.s. GEBT x T3 0.910 n.s. GE 0.547 n.s. GEBT x T3 0.822 n.s. GE x T2 0.059 n.s. T2 0.123 n.s. GE x T3 0.900 n.s. T3 0.410 n.s. GEBT x T2 0.881 n.s. GEBT 0.730 n.s. GEBT x T2 0.222 n.s. GEBT x T3 0.074 n.s. GE 0.168 n.s. GEBT x T3 0.037 * GE x T2 0.187 n.s. T2 0.184 n.s. GE x T3 0.815 n.s. T3 0.065 n.s. WHO-5 BPS sock test right GEBT x T2 0.001 *** GEBT 0.844 n.s. GE x T2 0,191 n.s. GEBT x T3 0.229 n.s. GE 0.125 n.s. GE x T3 0,247 n.s. GEBT x T2 0.870 n.s. GEBT 0.224 n.s. GEBT x T2 0.261 n.s. GE x T2 0.057 n.s T2 0.000 *** GEBT x T3 0.635 n.s. GE 0.976 n.s. GEBT x T3 0.981 n.s. GE x T3 0.999 n.s. T3 0.150 n.s. GE x T2 0.305 n.s. T2 0.307 n.s. GE x T3 0.614 n.s. T3 0.309 n.s. ODI BPS sock test left GEBT x T2 0.056 n.s. GEBT 0.198 n.s. GEBT x T2 0.101 n.s. GEBT x T3 0.534 n.s. GE 0.859 n.s. GEBT x T3 0.366 n.s. GE x T2 0.775 n.s. T2 0.987 n.s. GE x T3 0.797 n.s. T3 0.732 n.s. GEBT x T2 0.103 n.s. GEBT 0.837 n.s. GEBT x T2 0.779 n.s. GEBT x T3 0.266 n.s. GE 0.234 n.s. GEBT x T3 0.598 n.s. GE x T2 0.169 n.s. T2 0.000 *** GE x T3 0.551 n.s. T3 0.000 *** SF36 total GEBT x T2 0.026 * BPS lifting test GEBT 0.978 n.s. GEBT x T2 GEBT x T3 0.071 n.s. GE 0.669 n.s. GEBT x T3 GEBT x T2 0.584 n.s. GEBT 0.183 n.s. GEBT x T2 0.882 n.s. GE x T2 0.108 n.s. T2 0.002 ** GEBT x T3 0.299 n.s. GE 0.256 n.s. GEBT x T3 0.410 n.s. GE x T3 0.557 n.s. T3 0.000 *** GE x T2 0.675 n.s. T2 0.000 *** GE x T3 0.751 n.s. T3 0.000 *** 0.229 n.s. SF36 physical health GEBT x T2 0.023 * Spine rotation right GEBT x T2 0.072 n.s. GEBT 0.000 *** GEBT x T2 0.796 n.s. GEBT x T3 0.321 n.s. GE 0.000 *** GEBT x T3 0.995 n.s. GE x T2 0.121 n.s. T2 0.001 *** GE x T3 0.328 n.s. T3 0.410 n.s. GEBT 0.874 n.s. GEBT x T2 GEBT x T3 0.058 n.s. GE 0.439 n.s. GEBT x T3 GE x T2 0.106 n.s. T2 0.000 *** GE x T3 0.104 n.s. T3 0.000 *** 0.449 n.s. 0.794 n.s. SF36 mental health Spine rotation left GEBT x T2 0.597 n.s. GEBT 0.157 n.s. GEBT x T2 0.539 n.s. 0.535 n.s. GEBT x T2 0.104 n.s. GEBT 0.663 n.s. GEBT x T2 0.829 n.s. GEBT x T3 0.168 n.s. GE 0.723 n.s. GEBT x T3 0.058 n.s. GE x T2 0.009 ** 0.161 n.s. T2 Huber et al. BMC Musculoskeletal Disorders 21(09): Page 10 of 18 Table 3 Linear mixed models (Continued) GEBT and GE vs. Control LMM1 GEBT vs. GE LMM3 fe p value fe GE x T3 0.550 n.s. T3 LMM2 p value fe p value 0.040 * P-values of three linear mixed models (LMM). LMM1: GEBT and GE vs. CO, time and interaction of treatment and time; LMM2: GEBT vs. GE, time and interaction of treatment and time; LMM3: GEBT and GE vs. CO, time; Day 8 (T2) and day 120 (T3). Level of significance ≤0.05 *, level of significance ≤0.01 **, level of significance ≤0.001 ***. Fe (Fixed effect), SCS© (Spine-Check Score©), BPS Back performance scale, SCS© (Spine-Check Score©), mVAS (modified Visual Analogue Scale), WHO-5 (World Health Organization Well-Being Index), ODI (Oswestry Disability Index), SF-36 (Medical Outcomes Study Short Form 36) effects for the treatment of non-specific chronic low back pain. As indicated by our results, it can be assumed that green exercise has the greatest therapeutic influence, besides the social component of the group constellation as an important contributing factor. Commonly used in multimodal treatments of the rehabilitative care sector, the impact of positive group dynamics is a well-known influencing aspect [6]. This assumption is justified by the fact, that hardly any significant differences could be found in statistical analysis of our data by means of the two-armed linear mixed models (LMM2, GEBT group compared to GE group), but many signifiers in comparison to the control group (LMM1, LMM3). Most data feature a high standard deviation. This could be attributed to the fact, that stratification by the Korff assessment achieved an optimal distribution of the patients with regard to pain related disability, also reflected in subjective health-related quality of life, mobility and further assessments. Since a pathophysiological change of pain perception occurs in the brains of chronic pain patients (so-called pain memory) [43], a persistent significant reduction of pain after 1 week is not realistic to assume. However, cortical plasticity, which is important in chronic pain conditions, offers potential rehabilitation goals to be achieved by motoric, cognitive behavioral and sensory strategies. In our study, a green exercise approach (moderate mountain hiking) was adopted and an additional sensory component (relaxation and pleasant temperature during spa therapy) was used in the GEBT group. Regarding pain relief and health behavior after 1 week of intervention, the results in the GEBT group are particularly to be emphasized. Additionally, the surveys of subjective health-related quality of life as well as depression in chronic illness are indicators of the multidimensionality of the clinical pattern of nscLBP and showed a corresponding reaction to both interventions, but especially in the GEBT group. Interestingly, the results of the functional spinal mobility are partly contradictory (e.g. day 8/ GEBT group: Spine-Check Score© mobility and trunk rotation total versus sock test right). A possible explanation for this could be avoidance mechanisms or movement patterns reflected in these tests, which chronic pain patients have appropriated over years (e.g. methods/tools to pull on socks). Through different approaches like sight [44, 45], noise and smell during green exercise [45], natural environment has a direct and positive influence on physiological [46], as well as on psychological parameters [47, 48]. Unlike indoor exercises, physical outdoor activities offer opportunities which are associated with better health, like exposure to sunlight for adequate vitamin D levels [49] or a setting of a Fig. 4 Pain and Health Status. Pain behavior and health status of the patients assessed with the mVAS. Statistical analysis calculated with Linear mixed model (LMM1). Asterisks indicate significances. Data shown in percentage change from baseline (±SD). Data means (±SD) are shown in Table 3