Introduction

High-intensity aerobic exercise is a proven mode for improving disability, although the evidence pertains to training using a cycling ergometer. Almost every study on aerobic exercise in low back pain (LBP) patients has a form of aerobic training where one’s body weight is supported (for example, on a bike) or partially eliminated, like in aquatic environments. Unintentionally this may have caused healthcare professionals to avoid body-loading training in people with LBP, and may have fed the stigma that running is unsafe for people affected. More recently, we learned that recreational runners, compared to the general population, have lower rates of LBP (Maselli et al. 2020) and healthier spinal tissues (Belavý et al. 2017), but unfortunately, only a few randomized trials have compared running interventions to treat LBP. The current study wanted to explore whether running to tackle chronic LBP is a viable option.

Methods

To explore the efficacy of running to tackle chronic LBP, this 2-arm parallel randomized controlled trial (RCT) was set up to compare a group receiving a progressive run-walk interval training program to a waitlist control group. The RCT was labeled the ASTEROID trial, the abbreviation for “Assessing safety and treatment efficacy of running on intervertebral discs”. As such, apart from assessing the effectiveness, the current study could also examine the safety of running with LBP.

People between 18 and 45 years of age who suffered from chronic (lasting at least 3 months) nonspecific LBP were eligible candidates if they were not involved in running or running-based sports for the past 3 months and were injury-free to the lower limb.

Participants in the intervention group were instructed to follow three weekly 30-minute exercise training sessions over 12 weeks, prescribed by an exercise physiologist. These sessions were tailored to the individual, based on an initial fitness assessment where the time participants could run for 2 minutes on a treadmill was used to determine the starting level of the program. Every participant started at level 1, 2, or 3, where, respectively, 15-, 30-, and 45-second running intervals were prescribed.

The training sessions were not supervised but supported and guided by the clinicians and provided educational content. Brief 10- to 15-minute video calls were scheduled every week (during weeks 1-4) and every two weeks (weeks 6-12) to follow the participants in their journey. The participants also had the possibility to contact the clinicians via phone or text messages outside of the foreseen video calls. The training data were captured in the Runkeeper app.

Table 1 here shows how the walk-run intervals were prescribed. A participant starting at level 1, for example, had to complete 6 to 10 repetitions consisting of 15-second run intervals and 120-second walk intervals. Progressions to the next level were allowed when the individual could successfully complete the upper number of repetitions during at least 2 sessions in one week. Only one progression in level per week was allowed. During the run intervals, participants were instructed to jog at a slow-to-moderate speed of 10 km/h. The walk intervals were at a self-chosen speed.

Participants in the control group were asked to manage their LBP as usual and avoid running, but other types of exercise were allowed. This study used a so-called waitlist control, meaning that when the participants randomized to the control group completed the 12 weeks, they were offered the same exercise training program as the intervention group.

The primary outcome was the level of pain and disability, and these were measured at baseline, 6 and 12 weeks, but no specific time point was mentioned to be the primary endpoint.

• Pain intensity was measured using a 100-point visual analog scale (VAS) over the past week. The minimal clinically important difference (MCID) is 20 points
• Disability was captured using the Oswestry Disability Index (ODI). The score ranges from 0-100 points, with reductions of 10 points being clinically relevant (MCID).

Results

A total of 40 participants were included in the RCT and equally divided into the intervention and control groups. At baseline, the participants were 32.8 years old and had an average LBP intensity of 39.7 points and disability of 22 points.

Significant between-group differences were observed in the change in current pain intensity levels at 12 weeks, and average pain intensity at 6 and 12 weeks, but not in worst pain intensity levels in the intervention group. Disability improved significantly at 12 weeks in the intervention group.

Questions and thoughts

Although these results seem promising, we must acknowledge that these differences observed in the intervention group are not enough to be clinically relevant, as they did not surpass the MCIDs. We cannot say that, since the MCIDs were not surpassed, running is an effective treatment compared to waitlist control in chronic LBP, but the current trial confirms the safety of the intervention. Only nine minor adverse events were recorded, with only one record of increasing LBP.

The within-group improvements are promising and further studies could build on this RCT to improve the current run-walk interventions. Could supervised interventions of running to tackle chronic LBP be better? Also, other intensities, volumes, and progression rates could potentially lead to larger (clinically relevant?) improvements. To be continued!

We must highlight some potentially important differences at baseline between the intervention and control groups. Observing the VAS current, average, and worst pain intensities relatively large differences between the groups emerge. At baseline participants’ habitual physical activity levels were recorded and here also, large differences between people in both groups existed. All other baseline variables were comparable at the start of the RCT. People randomized to the control group reported higher levels of baseline pain, on all three subcategories. One could suppose that regression to the mean would be more present in this group due to the high baseline scores. Yet, it was the intervention group showing substantial reductions in pain, compared to the controls.

Despite the intervention not being supervised and progressions could be made by the patient when attaining a certain level, the participants did improve their performance in running distance over the weeks as seen in the graph. The participants showed good adherence to the training protocol and no attrition was observed, but this is partially due to the selection procedure likely recruiting individuals motivated to participate in a running training program for improving their LBP.

Talk nerdy to me

Although we gained some important insights, let’s just not be carried away yet. We must also acknowledge that the current findings should not be overemphasized. A difference below 20 points on a 100-point VAS scale and a reduction below 6 points on a 100-point scale seem negligible. I tend to give more meaning to the current results in changing how we think about LBP and hopefully promote the shift from being too careful and protecting the spine to confidently loading the spine progressively. The safety of the intervention and the absence of dropouts seem to indicate that the intervention was well-tolerated by the participants. This should form a base for further research to give meaning to the observed reductions in pain and disability. Meanwhile, we can take the current study as an example to progressively increase loading in these patients and tackle the perception of running being dangerous in people suffering from chronic LBP.

Responder analyses might reveal subsets of participants likely to respond differently, given that the confidence intervals indicated that some participants did achieve the MCIDs. However, as the participants were allowed to manage their LBP as usual, it could not be controlled if other treatments or combinations of treatments caused improvements. Food for thought for future trials! We must also think about the possibility that people affected by lower levels of pain intensity are potentially more eager to participate in exercise trials, whereas people with higher levels of pain and disability might be scared off.

Take-home messages

The current study explored the effectiveness and safety of running to tackle chronic LBP in adults from 18-45 years of age. Despite the absence of clinically relevant effects, an unsupervised training protocol that achieves such clear within-group reductions in pain and disability might be interesting to explore further. Safety was confirmed and high levels of adherence showed the feasibility of the training protocols. This together might indicate that there is potential for an intervention of running to tackle chronic LBP. To date, this study only confirms that chronic LBP is not worsened in people participating in such a running program. Future studies can explore other modi for delivering the intervention.

Reference

Neason C, Samanna CL, Tagliaferri SD, Belavý DL, Bowe SJ, Clarkson MJ, Craige EA, Gollan R, Main LC, Miller CT, Mitchell UH, Mundell NL, Scott D, Tait JL, Vincent GE, Owen PJ. Running is acceptable and efficacious in adults with non-specific chronic low back pain: the ASTEROID randomised controlled trial. Br J Sports Med. 2024 Oct 7:bjsports-2024-108245. doi: 10.1136/bjsports-2024-108245. Epub ahead of print. PMID: 39375007.