12 weeks high intensity interval training versus moderate intensity continuous training in chronic low back pain subjects: a randomised single-blinded feasibility study.

This was a feasibility study of 30 people with no sample size calculation, and the authors say plainly that it was not designed to evaluate treatment effects. It cannot show which training works better. The person supervising training was also the outcome assessor, so neither the care provider nor the assessor was blinded to which group people were in. Three of the 14 people in the interval group dropped out, one of them after only two of 36 sessions. The authors say that single dropout had a strong effect on the interval group's adherence figures because the sample was so small. The interval group started with higher pain scores than the moderate group, which gave it more room to improve. The only change large enough to count as clinically important was the pain drop in the interval group, 3 points from a median of 6, against a recommended minimum of 2.4 points on that scale. The heart rate targets could not be followed exactly: participants' heart rates adapted too slowly for them to stay in the intended lower range during the recovery phases, although the burst-phase targets were reached. So the interval protocol as delivered was not quite the protocol as planned. Everyone had to be able to cycle and had to have stopped physical therapy to take part, so the findings may not apply to patients who are receiving ongoing physiotherapy or who cannot manage a bike. The authors suggest the Oswestry cut-off of 14% used for recruitment may have been too low to detect clinically important change.

The text supplied for this summary contains no funding statement and no declaration of competing interests, so it is not possible to say who paid for the work or what the authors declared.

Do not read the headline that interval training is "as feasible as" moderate training as meaning the two were the same in practice. Three of the 14 people in the interval group stopped training (21%) while nobody dropped out of the moderate group, and the difference in how many sessions people attended was not statistically significant. With 30 people and no sample size calculation, this study was not designed to show that either training programme works.