A randomised controlled trial of the effectiveness of an exercise training program in patients recovering from SARS, 2005 Lau et al

I also note that the training programme was 6 weeks long, with evaluation at the end. That is not long enough to identify any sustainable benefit. Those people with a post-viral fatigue syndrome/ME/CFS may be swapping out some of their tasks of daily living for the prescribed exercise.

Importantly, "effects on health-related quality of life" were not statistically significant.

VO2max wasn't actually measured. Instead, cardiorespiratory fitness was assessed using two measures where both familiarity and motivation could very easily result in better performances. The exercise group got to practice; the control group did not.

Patients had had SARS-1 and had been referred for physiotherapy, presumably because they were having ongoing problems with tasks of daily living. People who were "poorly motivated, uncooperative" were excluded from the study. Also, people with above-normal performance on the 6-minute walk test and Chester Step test were excluded. Of course, the fact that there were significant numbers of people with above-normal performance in cardiorespiratory measures but who still were fronting up to the medical system requesting help to regain function suggests that deconditioning wasn't their problem. Of the total 258 SARS patients at the clinic, 54 refused or were assessed as having 'above normal performance'. This selection of patients who only had sub-normal performance of cardiorespiratory measures is a bias. Are they suggesting that physical training should not be used in patients with post-viral impacts and above-normal walk test and step test performance?

It is reported that no patients were lost to followup, although a further 38 people were excluded for being "unable to participate in the subsequent training sessions".

 

RESULTS

Mean change 6-minute walk distance (m) Controls 614.3 + 20.7 = 635 ___ Exercise 590.7 + 77.4 = 668
Mean change Predicted VO2max (ml/kg/min) Controls 37.8 + 1.0 = 38.8 ___ Exercise 35.1 + 3.6 = 38.7

VO2 max was not actually measured, but was estimated based on a walk test and a step test, both of which have a volitional element. People could potentially push themselves harder if they were trying to show that they had improved, therapists could apply more encouragement at the end of the 6 weeks, and practice could increase speed.

There were no statistically significant differences in the two muscle strength measures (anterior deltoid and gluteus maximus) - this was not reported in the abstract.

SF–36 *** controls baseline + controls change *** exercise baseline + exercise change
No differences in mean change were reported as being statistically significant.
Each SF-36 measure is scored out of 100, and higher scores means better health.

Physical functioning ___ 71.1 + 3.7 ___ 69.6 + 3.7
Role physical ___ 35.7 + 14.6 ___ 26.8 + 14.4
Role emotional ___ 47.0 + 8.9 ___ 42.3 + 1.9
Bodily pain ___ 69.4 + -5 ___ 65.2 + 0
Social functioning ___ 61.1 + 14.2 ___ 59.5 + 12.9
General health ___ 46.6 - 2.5 ___ 51.6 - 0.76
Mental health ___ 64.0 - 0.3 ___ 67.9 - 1.7
Vitality ___47.5 + 2.5 ___ 51.2 + 1.3

 

The cardiovascular performance was measured by largely subjective measures. Although some measures of musculoskeletal performance improved more in the supervised exercise group, some weren't, even after intensive training. Given that the treatments were unblinded and there was a lot of therapist contact in the supervised exercise group, we would have expected those participants to report with a positive bias, but no measure of health reported by participants improved more in the supervised exercise group than it did in the control group.

This study does not constitute evidence of useful benefits.

 

Not that that is a requirement for citing it to say it does, who reads the references anyway

 

It's been cited again in Balancing the value and risk of exercise-based therapy post-COVID-19: a narrative review, 2023, Singh, Chalder et al.

 

I'm aware I'm giving weight to subjective SF-36 results and discounting measures that might be regarded as objective. I think in this case it's ok.

As I've said, I don't think the walk test and step test are very objective. I consider myself significantly affected by my illness, but I don't think it substantially affects my 6-minute walk performance if I have rested before it. I think my performance would be affected by how motivated I was to do well though.

I don't think a measure of hand grip strength that didn't involve a lot of repeat measurement has much to do with my illness if I have rested before it.

An increase in SF-36 items in an unblinded trial certainly may be the result of the placebo effect. But, this study found a lack of an increase in SF-36 items over the control group in an unblinded trial where participants had been specifically selected for their motivation and cooperation and where they had had 6 weeks of contact with therapists. The biases were loaded in favour of the supervised exercise therapy group reporting better SF-36 increases than the controls. But, even so, they didn't.