Protocol

Technology supported High Intensity Training in chronic non-specific low back pain (the Techno-HIT trial): study protocol of a randomised controlled trial

Abstract

Chronic low back pain (CLBP) is one of the most common chronic musculoskeletal disorders worldwide. Guidelines recommend exercise therapy (ET) in CLBP management, but more research is needed to investigate specific ET modalities and their underlying mechanisms. The primary goal of this study is to evaluate the short-term and long-term effectiveness of a time-contingent individualised high-intensity training (HIT) protocol on disability compared with a time-contingent moderate-intensity training (MIT) as used in usual care, in persons with severely disabling CLBP. Additionally, the effectiveness on central effects, the added value of prolonged training at home and technology support, and the cost-effectiveness are evaluated. In this randomised controlled trial, CLBP patients will be randomly divided into three groups of 56 participants. Group 1, ‘TechnoHIT’, receives HIT with technology-support in the home-phase. Group 2, ‘HIT’, receives HIT without technology support. Group 3, ‘MIT’, receives MIT, reflecting training intensity as used in usual care. The primary outcome is patient-reported disability, measured by the Modified Oswestry Disability Index. Secondary outcomes include quantitative sensory testing, psychosocial factors, broad physical fitness, quality of life, cost-effectiveness, adherence and usability of technology. Trial registration number NCT06491121.

Why this study is important

  • This pre-registered three armed double blinded multicentred randomised controlled trial will be the first to evaluate the short-term and long-term effectiveness of a time-contingent individualised high-intensity training (HIT) protocol on disability level compared with a time-contingent moderate-intensity training used in usual care, with a long-term follow-up period of 18 months in persons with chronic non-specific low back pain (CNSLBP).

What this study adds

  • This project will provide high-quality, methodologically standardised data on the effectiveness of HIT, while also focusing on underlying fundamental mechanisms, as well as clinical implementation through the evaluation of cost-effectiveness and the potential benefits of extended home-based training with technological support.

How this study might affect research, practice or policy

  • This project will provide innovative insights on how to optimise multimodal exercise therapy as a clinical treatment strategy for patients with CNSLBP.

Background

Chronic low back pain (CLBP) is one of the most common chronic musculoskeletal disorders worldwide with a prevalence of 20%.1 Up to 85% of low back pain (LBP) diagnoses are non-specific, meaning no specific cause can be defined and management can be challenging.2 CLBP is currently ranked as the number one cause of disability.1 However, the magnitude of disability related to CLBP is not merely determined by the direct impact of pain and discomfort, but also by limitations in daily activities and societal participation.3 4 For instance, CLBP contributes to 13% of all causes of work absenteeism.5 In persons with CLBP, sleep disturbances and psychological factors such as anxiety and stress play a role.6 7 Since psychosocial and emotional factors are strong predictors of LBP chronicity, persons with CLBP may end up in a lifelong vicious circle characterised by invalidating pain, work absenteeism, physical deconditioning, sedentary lifestyle and comorbidities such as obesity or depression.2 8–10 As a result, CLBP has major socioeconomic implications and creates a burden on our healthcare system with global cost estimations rising substantially each decade.11 Optimised CLBP management through innovative research efforts is therefore essential.12

State-of-the-art guidelines recommend implementing exercise therapy (ET) in CLBP management.13 Nevertheless, treatment effect sizes in CLBP remain modest.13 Different factors may explain this.

First, many studies providing ET in CLBP started from the idea that the back should be treated carefully.14 15 This often results in exercise intensities below the required level for optimal treatment success and reinforcing anxious thoughts about loading the spine.16 Nevertheless, recent findings have demonstrated that high-intensity training (HIT) can be a valuable method to improve the effect sizes of concerning short-term and long-term disability and exercise capacity in persons with CLBP.17 18 However, the short-term and long-term effects of HIT in a large population with severe CLBP need to be investigated.

Second, it is striking to observe the lack of correlation between patients’ improvements in pain and disability after ET on the one hand and results on physical outcome measures such as endurance or strength on the other hand.19 This shows that ET effects in CLBP might not be directly or solely attributable to these changes in the musculoskeletal system.19 Substantial evidence now points out that ET provides multiple central effects such as psychological, (neuro)physiological and autonomic adaptations.20 21 In this regard, literature shows a positive effect of HIT on central sensitisation,22 mental well-being23 and adaptations of the cardiovascular system.24 Furthermore, there are indications that the implementation of sufficient exercise intensity and duration is related to better effects on pain.25 26 To date, it is unknown how ET can be optimised to improve and retain these central modulation effects maximally.

Third, many ET interventions used a pain-contingent approach (progress based on pain), as opposed to a time-contingent approach (progress over time regardless of pain) that is now advised in chronic pain,27 reinforcing again the message that the spine is vulnerable and that the exercise should be adapted or stopped in case of an increase in pain.28 A time-contingent HIT programme leads to significant short-term effect sizes to reduce disability in persons with CLBP.17

Additionally, long-term exercise programmes can lead to larger effect sizes for pain reduction and to an improved endogenous pain processing.29 30 Furthermore, it takes weeks to years to achieve behavioural change.31 From a feasible, cost-efficient perspective, longer ET protocols are preferably executed at home. However, such a training set-up requires clear guidance and support, as treatment compliance is low without it.32 33 For that reason, technology might support the treatment by encouraging persons with CLBP to keep performing their exercises in their own environment and prolong the rehabilitation process.34 For example, a video-based protocol is more effective in improving pain, function, kinesiophobia, expectations and several other factors than usual exercise practice at home.35 Still, it is unclear which duration leads to optimal improvement and the additional value of technology should be further investigated.36

Finally, although individualised exercises combined with education are recommended in chronic pain,14 37 clinicians often fail to implement this in clinical practice.19 Evidence shows altered brain structure and function in CLBP patients, and an approach including pain neuroscience education to cognitively prepare patients for ET is recommended.38 Additionally, exercises are often too uniform given the heterogeneity of CLBP patients.19 Therefore, many patients fail to adhere to their exercises, leading to poor treatment outcomes.39 Furthermore, the exercise intensity of these non-individualised exercises can, again, be too low for the specific individual. Individualised interventions and personalised guidance to increase patients’ adherence and to adapt the intensity accordingly need to be investigated to enhance therapy success.39

Therefore, the primary goal of this study is to evaluate the short-term and long-term effectiveness of a time-contingent individualised HIT protocol preceded by pain science education (PSE) on disability compared with moderate-intensity training (MIT) as used in usual care, in persons with severely disabling CLBP. Secondary goals entail evaluating (1) the short-term and long-term effectiveness of HIT on central effects such as psychological, (neuro)physiological and autonomic adaptations, and on broad physical fitness; (2) the additional effects of prolonged HIT at home; (3) the added value of technology through a mobile application that offers support during HIT home training and (4) cost-effectiveness of (technology supported) HIT compared with MIT.

Methodology

This protocol is reported according to the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) guidelines 2013: Explanation and Elaboration: guidance for protocols of clinical trials.34

Patient and public involvement

Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.

Study design and study setting

A double-blind (outcome assessor and statistical analysis) superiority three-armed multicentre randomised controlled trial will be conducted. Patients with CLBP will be randomly divided (see‘Randomisation procedure’ section below) into three groups of each 56 participants (Group 1: ‘TechnoHIT’; Group 2: ‘HIT’; Group 3: ‘MIT’) and kept naive. For each group, the trial involves a 24-week exercise intervention with a total of 52 rehabilitation sessions (4 educational sessions in the biopsychosocial programme in the first 2 weeks and two physical therapy sessions each week for 24 weeks (n=48 in total)) in three phases (two therapy phases and one follow-up phase). This study will be a collaboration between Universitair ziekenhuis Antwerpen (UZA), Jessa Ziekenhuis, Universiteit Hasselt (UHasselt) and Universiteit Antwerpen (UAntwerpen). An overview of the study design can be found in figure 1.

Figure 1
Figure 1

Study design showcasing the different study groups and study phases. HIT, high-intensity training; MIT, moderate-intensity training.

Timeline

Recruitment will start in May 2024 and the recruitment phase will last approximately 3 years.

Randomisation procedure

The randomisation functionality of the castor data management software40 will be used to randomise patients into three groups. Participants will not be informed about the approach of the different groups, but they will be informed that the study investigates three different active rehabilitation programmes for CLBP. A rater cross-over assessment protocol will be performed to ensure the assessors are blinded. The assessors will perform baseline measurements and guide the treatment at one clinical site, and follow-up measurements at the other.

Participants

Adult participants17–64 can be included in the study in case they are diagnosed with severe non-specific CLBP, defined as chronic primary musculoskeletal pain below the costal margin and above the inferior gluteal folds for more than 12 weeks41 whereby fluctuations in pain can be present. Remission phases can alternate the pain. Participants must have ≥20% on the Modified Oswestry Disability Index (MODI) to be categorised as ‘severe’. All inclusion and exclusion criteria can be found in table 1.

Table 1
|
Inclusion and exclusion criteria

Recruitment

Participants who ought to be eligible for the study during consultation at UZA/Jessa, will be informed about this study by the physician or one of the researchers. If they are interested, a study flyer and an approval form for further contact (via email and/or telephone according to the preference of the possible participant) will be issued by the physician. The researchers contact the potential participant within 2–7 days, answer initial questions, review the inclusion criteria and provide informed consent (online/hard copy according to the preference of the potential participant). Patients who sign and return the informed consent within 2 weeks will be contacted for possible enrolment.

Sample size

A sample size calculation (power analysis) was performed with JMP Pro V.14.1 in collaboration with the Center for Statistics (CenStat, UHasselt) based on the therapy effect of a HIT compared with an MIT programme on disability measured with the MODI in persons with CLBP after a 12-week intervention protocol17 (see online supplemental appendix 1). As the primary aim of this project is to be able to show significant short-term and long-term differences between HIT and MIT (at T1, T2, T3 and T4) on the primary outcome disability assessed by MODI, a sample size calculation for significance level α=0.05 and power level β=0.80 was performed to test the null-hypothesis that MODI outcome is equal for both groups. As a reference for variability, the highest value between estimated values for T2 was taken to ensure that actual project results will be correct for lower variance values. Additionally, the same number of patients was taken for the other group, resulting in a total sample size of 83. Accounting for a maximum potential total 50% loss-to-follow-up after T4 and an allocation ratio of 1, this results in a sample of 168 patients to be included in this project divided into three groups of 56 participants (Group 1: ‘TechnoHIT’; Group 2: ‘HIT’; Group 3: ‘MIT’).

Intervention

At the start of the intervention, all participants will receive four PSE sessions (60 min/session, 2×/week) as part of a biopsychosocial therapy model in groups of 2–5 participants. These sessions will be organised by researchers and occupational therapists at the hospitals, comprising various topics in a PowerPoint presentation. The content of these sessions was developed using scientific literature42 43 and two online web tools called ‘Retrain Pain’44 and ‘Pain Revolution’.45 Topics vary from neurophysiological pain mechanisms to activity management and debunking ergonomic myths, all aiming to provide insights into the biopsychosocial components of pain, movement and activities of daily living. All three groups will then follow a 12-week training programme in-centre under the supervision of a trained physiotherapist (phase 1) and a 12-week training programme at home (phase 2) (for a detailed display, see online supplemental appendix 1).

Phase 1: in-centre therapy (months 1–3)

Experimental groups (‘TechnoHIT’ and ‘HIT’)

Groups 1 and 2 will perform a 12-week time-contingent protocol, previously published,17 encompassing two 1.5-hour ET sessions weekly in the hospitals. These sessions include cardiorespiratory interval training, general resistance training and core muscle strength training, all at high intensity. The cardiorespiratory protocol will be individualised based on a maximal cardiopulmonary exercise test, and will be executed on a cycle ergometer. On a screen, patients will be able to see their heart rate (HR) and repetitions per minute (RPM). The protocol will consist of five 1 min bouts (110 RPM at 100% VO2max) alternating with 1 min of active rest (75 RPM at 50% VO2max). General resistance training includes three upper-body and three lower-body exercises that will be executed on fitness equipment. All exercises will be performed at 80% of the individual 1 repetition maximum (RM) and start at eight repetitions. Individual progressions will be implemented.46 Core muscle training includes six static core exercises. Exercises will be chosen in function of their ability to load the core muscles at an intensity of >60% of the individual maximum voluntary contraction. Participants will have to perform one set of 10 repetitions of a 10 s hold alternating with 5 s rest.

Control group ‘MIT’

Group 3 will perform a similar programme, but at a lower intensity. A continuous cardiorespiratory protocol of 14 min will be executed on a cycle ergometer (90 RPM at 60% VO2max). During the first 12 sessions, the duration will be increased by 1 min 40 s every two sessions up to 22 min 20 s. From sessions 13 to 24, the protocol is repeated with an increase in workload (+5% Wmax). General strength exercises and core strength training are identical to the HIT protocol, with the exception of the intensity and repetitions. Participants will perform 15 repetitions of the strength exercises at 60% of 1 RM. The workload will progress every 2 weeks by 5%. Regarding the core exercises, participants will have to perform one set of 10 repetitions of a 10 s hold. The static hold time will be increased every six sessions when executed correctly, and the posture will become more demanding when the core is stable for the indicated time for two consecutive sessions.

Phase 2: at-home therapy (months 3–6)

After completing the first 3 months of the training programme, all three groups will follow a comparable 12-week programme in their home setting. They are asked to perform a 60 min training session two times a week. Groups 1 and 2 receive a fitness bike, a smartwatch, a polar HR sensor and a training mat during this phase. Group 1 will also be guided by a mobile application (figure 2). Training sessions, corresponding to the actual progress of the patient in the HIT training programme, are scheduled in the mobile application in a weekly programme. Instructions with visual representations (drawings or pictures) are given for the specific exercises, and guidance or feedback by the application is momentarily provided while the patient performs the exercises (eg, a countdown with respect to the expected frequency of execution, a timer or a visualisation of a patient’s HR). After performing the exercise, the patient reports the completion level for that exercise. This way, the application calculates adherence to the training programme over time. Motivational feedback is given based on the patient’s performance. Based on the achieved adherence, the application will attempt to predict whether additional guidance is needed and whether alternative strategies to increase adherence are recommended.

Figure 2
Figure 2

Example of the smartphone application screen.

A web-based dashboard application supports the researchers, who will be able to monitor every individual and send out personalised feedback or tips. The dashboard allows managing patient inclusion with an authorised account. Individual patient files can be retrieved to change the training programme (so the set of exercises and their parameters such as difficulty level) based on suggestions by the system.

Groups 2 and 3 receive an exercise schedule on paper and a diary to log their training sessions. Group 3 will be asked to replace the cardiorespiratory protocol on the bike by performing a brisk walking modality. As the goal is to represent usual care as much as possible, this group will not be provided with a bicycle or exercise mat.

Phase 3: follow-up (month 6 until 18 months follow-up)

All participants are advised to continue their exercise programme at home. The researchers do not contact them during this phase. Participants of the Technology supported High Intensity Training (Techno-HIT) group are allowed to continue using the app at their initiative. At months 6 and 18 after the intervention, the participants will be reassessed to collect follow-up data.

Outcome measures

Outcome measures will be collected at baseline (T−1 and T0) and at 12 weeks (T1), 24 weeks (T2), 1 year (T3) and 2 years (T4). The primary outcome is the change in the MODI score from baseline to follow-up. The MODI is a self-report questionnaire comprising 10 items scored on a 5-point Likert scale. The total score will be noted on a scale of 0–100 with higher scores indicating higher levels of disability. It has good clinimetric properties to evaluate disability experienced by people in their daily activities due to CLBP.17 Secondary outcome measures including quantitative sensory testing, psychosocial outcome measures, broad physical fitness, usability of technology, adherence and motivation, cost-effectiveness and quality of life are listed in table 2 (for a detailed display, see online supplemental appendix 1).

Table 2
|
Overview of secondary outcome measures

All assessments will be conducted by trained researchers at two location sites, namely at the MOVANT Research Labs, UAntwerpen and REVAL Research Labs, UHasselt (Belgium). Physical assessments are performed following standardised protocols. Participants will be asked to fill in questionnaires in a quiet room. To limit the cognitive burden as much as possible, starting at T0, all questionnaires will be divided into two parts, separated by a physical assessment. Participants also have the option to take a 15 min break. An in-detail schedule of enrolment, intervention and assessments can be found in table 3, according to the SPIRIT 2013 guidelines.47 All assessments will be conducted predetermined, as shown in figure 3.

Table 3
|
Schedule of enrolment, interventions and assessments in accordance with the SPIRIT 2013 guidelines
Figure 3
Figure 3

Chronological overview assessments. B-IPQ, Brief Illness Perception Questionnaire; BPI, Brief Pain Inventory; BRS-NL, Brief Resilience Scale-NL; CPET, cardio-pulmonary exercise test; CRP, C reactive protein; EARS, Exercise Adherence Rating Scale; EQ5D, EuroQOL; ETR, expectations to recover; FACS, Fear Avoidance Components Scale; HbA1C, haemoglobin A1C; HRV, heart rate reliability; IEQ, Injustice Experience Questionnaire; IMI, Intrinsic Motivation Inventory; IPAQ-sf, International Physical Activity Questionnaire Short Form; IPCQ, iMTA Productivity Cost Questionnaire; MODI, Modified Oswestry Disability Index; PANAS, positive and negative affect schedule; PSQI, Pittsburgh Sleep Quality Index; QST, quantitative sensory testing; SEE, self-efficacy for exercise; SUS, System Usability Scale; TSK, Tampa Scale of Kinesiophobia; UMARS, User Version of the Mobile Application Rating Scale; WAI, Work Ability Index.

Data analysis and statistics

Data will be collected and logged consistently through the Castor data management software40 to ensure full data traceability throughout the study. Data analysis will be performed in JMP Pro (V.14.0, SAS Institute, Cary, USA). It will first be checked whether the data are normally distributed to determine whether parametric or non-parametric analyses should be performed.

For baseline assessment analyses, T−1 to T0 PSE session effects (PRE-POST) will be analysed with a dependent t-test/Wilcoxon signed-rank test to evaluate the impact of education on pain processing in this population. Second, baseline T0 data will be analysed to determine descriptive statistics for the different outcome measures for each group. Third, correlation analysis (linear multiple regression analysis) will be performed to determine associations between disability and pain processing.

For longitudinal assessment analyses, to evaluate the effectiveness of the HIT versus MIT intervention on T1 and the effectiveness of the Techno-HIT versus HIT versus MIT intervention on T2, different versions of linear mixed models will be considered (random intercept; random intercept, random slope with different covariance structure; dependent errors; transformed versions of responses; with unstructured times) and the version with the best fit to the data will be used for the analysis. Multiple comparisons will be executed to evaluate group (baseline differences), time (within-group differences) and interaction effects (between-group differences). For all tests of significance, an α-level=0.05 will be used. To account for bias due to deviations from intended interventions (drop outs), both an intention-to-treat (to evaluate the effect of assignment to the interventions at baseline, regardless of whether the interventions are received as intended) and a per-protocol analysis (to evaluate the effect of adhering to the interventions as specified in the trial protocol) will be performed. The intention-to-treat analysis will use a multiple imputation technique under the assumption of values missing at random. To check for selective drop-out, differences between participants completing the trial and drop-outs will be examined (independent t-tests, Mann-Whitney U tests, χ2 tests). The researcher performing these statistical analyses will be blinded as he/she will only receive a coded version of the data containing no personal identification data.

Clinical relevance, strengths and limitations

Clinical relevance

HIT has already been found feasible and effective in decreasing disability in moderate CLBP18 and in several other chronic disorders such as chronic neck pain, axial Spondyloarthritis, multiple sclerosis, and chronic lung and heart diseases.65–69 However, this is the first clinical study to evaluate the impact of HIT in a larger spectrum of persons with severe CLBP in a rehabilitation centre setting, and this study is essential to increase the external validity of HIT as a general rehabilitation strategy. Results will be transferable to different chronic pain populations, and this could potentially be a big step forward in the future biopsychosocial treatment of chronic pain.

Furthermore, this study will investigate the added value of longer training programmes and technological support during the follow-up training phase at home.

This study will fill the gap in the literature on the underlying working mechanisms of HIT. HIT shows promising results in CLBP patients.17 However, it is unclear how. One hypothesis is a change in psychosocial factors. For instance, general mental health increases after ET in numerous chronic musculoskeletal disorders.19 70 This is an important finding, as factors such as patients’ expectations to recover and self-efficacy to cope with a disorder, are clear therapy success modulators in CLBP.71 72 Furthermore, HIT causes an increase in self-efficacy,73 supporting this theory. Another possibility is a change in (neuro)physiological factors including improved anti-inflammatory factors that accompany the increased physical demands of HIT.74 On the other hand, the role of ET in activating the endogenous pain system, often dysfunctional in persons with chronic pain,20 75 76 has been displayed in various populations.77 78 Research is necessary to improve and retain these central effects.

Strengths and limitations

The study population will include adult severe CLBP patients, diagnosed by their general practitioner or the physician at the Department of Physical Medicine and Rehabilitation. These eligible patients will be referred to and contacted by the researchers. This multidisciplinary diagnostic process is a major strength of this study. One of the challenges of this study is insufficient patient enrolment. However, CLBP is one of the most prevalent musculoskeletal disorders,1 and disability seems to be the most determinant factor in seeking help and consulting a doctor to manage their pain.79 Additionally, patient recruitment will occur at different centres in Belgium (UZA and Jessa) and there is a possibility to contact other hospitals if necessary.

As a consequence to the long duration period and the extensive assessments during baseline and follow-up measurements, the risk for drop-out can be high. However, there are indications that sufficient duration is related to improved effects on pain,25 26 and patients will be motivated by the physiotherapists to finalise the full treatment programme and follow-up measurements. In previous research, we showed that time-contingent HIT leads to substantially higher short-term effect sizes to reduce disability in comparison to therapy as usual in a randomised controlled trial consisting of persons with CLBP with mild/moderate disability and good psychosocial health.17 We expect that patients who experience good treatment results will be encouraged to complete the full programme, especially considering the long follow-up period and the technology supported motivational programme in the TECHNO-HIT group. Nevertheless, a potential loss-to-follow-up of 50% was calculated in our sample size calculation.

Due to the nature of the therapy, the physiotherapist who gives the treatment cannot be blinded. The variability of the caregivers supervising the treatment could influence the standardised working method of the study. However, this increases the transferability to the clinical practice, since preferred working methods can differ between different caregivers in the clinical setting. Moreover, rehabilitation programmes in a hospital setting are generally guided by multiple physiotherapists.

To optimise the standardisation, a manuscript with the intervention protocol, exercises and progressions will be provided to the caregivers. A detailed script with assessment protocols will additionally be set up by the researchers who perform all measurements to ensure a standardised working method.