Sustained Improvements in EQ-5D Utility Scores and Self-rated Health Status in Patients With Ankylosing Spondylitis After Spa Treatment Including Low-dose Radon

Background Patients with ankylosing spondylitis (AS) have signicantly lower quality of life (QoL) than the general population. Holistic interventions addressing QoL include spa- or balneotherapy. Inclusion of radon in spa-therapy treatments is benecial in reducing pain and shows promising results in improving QoL in AS-patients. We aimed to explore the association of spa therapy including low-dose radon with systematically monitored QoL in AS-patients over an extended timeperiod. Methods Registry data collected for the “Radon indication registry for the assessment of pain reduction, increase of quality of life and improvement in body functionality throughout low-dose radon hyperthermia therapy” in the Austrian Gastein valley comprising data on QoL (EuroQol EQ-5D) directly before the treatment (baseline), directly, 3; 6 and 9 months after the treatment, age, sex and body mass index (BMI) were analysed. Two linear regression models explored the association between time of measurement with 1) EQ-5D utilities and 2) EuroQol visual analogue scale (VAS) score. Alterations of 0.05 (utilities) and 5.00 (VAS) were considered clinically relevant.


Introduction
Ankylosing spondylitis (AS) is the most common form of the rheumatic disease group of spondyloarthritides. It occurs in approximately 23.8 per 10,000 Europeans and is more prevalent in men than women. AS affects the axial skeleton leading to in ammatory back pain, damage to physical structures as well as impairments in physical functioning. These impairments may result in reduced participation and decreased quality of life (QoL). (1)(2)(3)(4) The growing understanding of QoL as key factor when measuring the effectiveness of healthcare interventions as well as the embracement of biopsychosocial models rather than just biological models for the evaluation of health emphasize the relevance of interventions focussing on the improvement of QoL. (5,6) Previous research has pointed out that AS-patients have signi cantly lower QoL than the general population but that pharmacological treatment is bene cial in improving their QoL. Particularly the combination of anti-TNF-α therapy in combination with physical exercise may reduce the adverse effect of AS on QoL. (7,8) However, the evidence on the effectiveness of alternative or complementary non-pharmacological interventions in improving QoL in AS patients is still limited. Common symptom-oriented interventions like physiotherapeutic treatment are effective in the reduction of disease activity and pain as well as the improvement of functional capacity. (9) Yet, holistic interventions have the potential of addressing a wider range of the AS-patient's health state including mental health and participation in daily life. (10,11) Those aspects are particularly relevant when assessing QoL from a patient's perspective.
Holistic interventions for AS regularly comprise multidisciplinary treatments including spa-or balneotherapy/speleotherapy. Still, little is known about the effect of these interventions on QoL. Kamioka et al. summarized the body of knowledge in an overview of systematic reviews with metaanalysis based on randomized controlled trials of balneotherapy and spa-therapy from 2000 to 2019 and did not identify any review focussing on spa-therapy in relation to QoL. (12) A limited number of studies speci cally addressed the effectiveness of combined spa-exercise therapy on QoL. For example, Colina et al. demonstrated that in AS patients, combining pharmacological treatment with spa-therapy resulted in signi cantly better QoL than pharmacological treatment alone six months after treatment initiation. (13) A randomized controlled trial by van Tubergen et al. showed that QoL, expressed by EuroQol-5D utilities, was signi cantly higher in patients that received spa-therapy (one with and one without radon treatment) compared to those who received usual care until 40 weeks after the treatment. In this study, the application of utilities enabled valuation of QoL from a societal perspective (i.e., utility values accounted for preferences the society has for a particular health state). (14) Among spa-therapies, treatment with low-dose radon has shown to be effective in achieving long-term pain reduction in persons with musculoskeletal diseases (including AS) (11,(15)(16)(17)(18)(19) and showed promising results with regard to improvements in functionality (20,21) as well as in QoL. (22,23). However, to the best of our knowledge, until now no data exist on the association of spa therapy including radon with systematically monitored QoL in patients with AS over an extended period of time while accounting for both, QoL from a societal perspective and individually perceived QoL. Therefore, the aim of the current study was to explore whether spa treatment including low-dose radon results in sustained signi cant and clinically relevant improvement of QoL in patients with AS.

Methods
The current study concerns a longitudinal analysis of registry data collected for the "Radon indication registry for the assessment of pain reduction, increase of quality of life and improvement in body functionality throughout low-dose radon hyperthermia therapy" (registration ID ISRCTN67336967; https://doi.org/10.1186/ISRCTN67336967) in the valley of Gastein in Austria. The registry collects data from individuals visiting the valley of Gastein for the purpose of spa-treatment including radon for a variety of rheumatic diseases. Data are collected by means of standardized questionnaires that are completed by participants directly before commencement of the treatment (baseline), directly after the treatment and 3; 6 and 9 months after the treatment.

Population
For the current study, data provided by participants with AS were included if they completed the questionnaire at each timepoint.

Intervention
The intervention consisted of an individualized spa-treatment including radon. This so-called low-dose radon balneo/speleo therapy (LDRnBST; radon-therapy) is part of a holistic treatment program for patients with AS and is applied in terms of balneo-and/or speleotherapy. The former includes bathing in water with low activity of radon as applied by the local facilities according to standardized treatment regimens. An intervention including radon-therapy consists of approximately 10 baths with a duration of 20 minutes. Speleotherapy including radon describes the process of relaxation while being exposed to low activity of radon, high humidity and mild hyperthermia (37-41.5°C) in the healing gallery of Gastein (a former gold mine) located in moderate altitude (1280 m) above sea level for an average time of 60 minutes on alternate days.
The intervention had an average duration of 17.5 days (SD 3.5) and took place in the valley of Gastein in the Austrian Alps.

Outcomes
The EuroQol EQ-5D (© EuroQol Research Foundation. EQ-5D™) is a self-reported questionnaire consisting of two parts, a descriptive system comprising 5 dimensions of health (i.e., mobility, self-care, usual activities, pain/discomfort, anxiety/depression) and a visual analogue scale (VAS) capturing participant's self-rated health status on a 0-100 scale with higher values representing better health. Using the unique score from each of the 5 dimensions of health a utility index score can be calculated (i.e., von Neumann-Morgenstern utility value for current health). (24) Single values for each of the 5 dimensions re ect the level of problem with each dimension resulting in an individual health state. This health state can be converted into a weighted health state by applying scores from the EQ-5D preference weights extracted from the general population which can take a value from 0 (death) to 1 (full health).
The EQ-5D utility index and EuroQol VAS were used as outcome variables for the current study. In absence of Austrian population weights, German population weights were used to calculate the EQ-5D utility index. (25) Main independent variable of interest and covariates The timepoint of survey completion by the participants was used as main independent variable of interest. Covariates were chosen a priori and included age (in years), sex (men/women) and body mass index (BMI; BMI=weight[kg]/height[m] 2 ) due to their already established in uence on health and health related QoL. (26-28)

Statistical analyses
First, descriptive statistics were used to characterize the sample in terms of age, gender and BMI at baseline (i.e., directly before the intervention) and to describe the EQ-5D utility index and VAS-score for each of the timepoints of measurement. Next, two linear regression models were computed to explore the association of timepoint of measurement with a) the EQ-5D utility index and b) the EuroQol VAS-score while adjusting for age, sex and BMI. After each model, margins and their 95% con dence interval (CI) were calculated to produce speci c age, gender and BMI standardized estimates for the utility index and VAS score.
P-values ≤0.05 were considered statistically signi cant. A change of 0.05 in the EQ-5D utility index and of 5.00 in the EuroQol VAS was considered clinically relevant. (29,30)

Results
The nal sample included in the analyses consisted of 291 participants who provided complete data for all timepoints. The sample consisted of 128 women, the mean age was 52 years and the average BMI was 26. Table 1 shows the unstandardized EQ-5D utility index and VAS scores for each timepoint. Figure  1 illustrates the course of the dimensions (i.e. mobility, self-care, usual activities, pain/discomfort, anxiety/depression) based on which the utility index was calculated.  Figure 2 illustrates the age, sex and BMI adjusted course of self-reported health state based on EuroQol VAS scores and utility index score.

Discussion
To our knowledge, this is the rst time that systematically collected registry data have been used to explore the association between spa-therapy including radon with alterations in QoL in AS patients over a period of 9 months. Signi cant improvements in QoL were seen immediately and were sustained until 9 months after the intervention. These improvements were clinically relevant until 6 months after the intervention in case of the EQ-5D utility index and until 9 months after the intervention in case of the EQ-VAS. Generally, these ndings are in agreement with other studies focussing on the bene ts of spa therapy for QoL in AS patients. (13,14,22,23).
Some differences were found in the course of the utility index compared to the course of the VAS-score.
The utility index showed the largest improvement directly after the intervention while the VAS score was highest 3 months after the intervention. The latter is in agreement with other publications focussing on symptom relief and alterations in QoL through spa-therapy including radon which show a delayed therapy response. For example, van Tubergen et al. found the same delay when focussing on the EQ-5D utility index (14). However, in our study the delay was only observed in the VAS-score. A possible explanation might be found in the different population preferences accounted for in the calculation of utilities in the current study (German preference weights) compared to the study of van Tubergen et al.
(Dutch preference weights). Yet, the unadjusted illustration of the 5 dimensions of health prior to the application of preference weights shows the same course of improvement suggesting that another explanation is more likely. Selection bias might be one: The observation may be attributable to the speci c population included. Provision of data for the radon registry is voluntarily and participants included in the current study had provided complete data at all timepoints. This might indicate high motivation attributable to favourable treatment effects that are more precisely represented by the utility index than by the VAS score. An interesting side nding was that men had signi cantly higher EQ-VAS scores compared to women independent of their age, BMI or the timepoint of measurement. Previous research showed, that women with AS have less improvement in AS related outcome measures compared to men. However, the reason for this phenomenon remained unclear (31). In the current study the difference between men and women was not clinically relevant and only occurred in case of the VAS but not the utility index, which might suggest that perception of health plays a relevant role.
Clinically relevant changes attributable to spa therapy including radon have, to our knowledge, not been addressed by previous studies. The current evidence points out that in our cohort of AS-patients clinically relevant improvements in QoL can be sustained until 6 months (utility index) or even 9 months (VAS) after intervention. From a clinical perspective, this indicates the bene ts of a repetitive treatment pattern.
To achieve stable results, a periodic intervention should be scheduled every 6 months.

Limitations and Strengths
As in all studies based upon registry data limitations arise from the fact that data collection is not monitored or performed by the researcher and that data on confounders is somewhat limited. (32) In the current study data on the frequency of interventions prior to the rst timepoint of measurement were not systematically collected. This might have resulted in biased baseline values as participants who have received the intervention repeatedly likely have a better baseline health state than those who receive the intervention for the rst time leading to a potential underestimation of the improvement in rst-time participants. Confounders in the association of QoL with AS have been identi ed in previous literature and might have affected the current analyses as well. For example, a lower level of education and being a smoker is associated with lower QoL but this information was not available. (33) Strengths of the study include a relatively large study sample with complete data over an extended period of time as well as the independence of data collection. Since data on the effectiveness of spa therapy including radon on the improvement of QoL in AS patients is still scarce, the current study provides relevant insights and opportunities for further research among other patient populations, and in comparison with usual care.

Conclusion
In conclusion, the current study reveals that spa-therapy including low-dose radon is bene cial in improving QoL in patients with AS and that these improvements are sustained for up to 9 months. It may be considered a valuable (complementary) treatment option for this purpose. Extrapolation of the results may support the decision of policy makers and insurances to refund bi-annual spa therapy including radon for patients with AS.

Declarations
Ethics approval and consent to participate The datasets used and/or analysed during the current study are publicly available from the open data storage platform Zenodo using the following link: https://doi.org/10.5281/zenodo.5926209.

Competing interest
The authors have no con icts of interest to disclose.

Funding
This research did not receive any speci c grant from funding agencies in the public, commercial, or notfor-pro t sectors.
Author contribution statement AvZ conceptualized the current study, conducted and interpreted the analyses, wrote the initial draft of the manuscript and nalized its current version.
VS; HD and JF prepared and organized the data before analysis, were involved in the conceptual framework and protocol for the original data collection and reviewed and critically commented on the manucript. JU; WF; MK; SE and MO were involved in the recruitment of participants, reviewed and critically commented on the manuscript. MR, BH and MG were involved in the concept of the current study and the design for the original data collection, reviewed and critically commented on the manuscript. All authors agreed on the nal version before submission.
Abbreviations AS, Ankylosing spondylitis; BMI, Body Mass Index; CI, con dence interval; EQ, EuroQol; QoL, Quality of life; RnIR, Radon indication registry for the assessment of pain reduction, increase of quality of life and improvement in body functionality throughout low-dose radon hyperthermia therapy; VAS, Visual Analogue Scale Figure 1 Unadjusted average course of health based on EQ-5D health dimensions (dimension score range 1-5 with lower scores representing better health) Figure 2 Age, sex and BMI adjusted course of self-reported health state based on EuroQol VAS scores and utility index score (utility index range 0-1; VAS-score range 0-100 with higher scores representing better health)