Six minute walking test

Did it though?

Here are the PACE trial results
Group: Baseline distance (m): 52 week distance (m)

Adaptive Pacing: 314: 334
CBT: 333: 354
GET: 312:379
Standard care as usual: 326: 348

The 6 minute walking test in the PACE trial claimed to show that GET increased physical function, although improvements relative to other treatments were trivial. I don't think we can distinguish between 'none of the treatments made much of a difference' and 'this measure does not have sufficient discriminatory power to identify changes in physical fitness'.


Those numbers suggest to me that the following is possible:

1. There's a ceiling. The distances really don't change very much. Individual variation might be more related to stride length or interpretation of the rules than physical fitness. A measure where most of the participants are at their ceiling at baseline is not worth much.

2. There's a learning effect. The distances increase in all the groups in the second test. I think a learning effect is very likely, and that is why it's so bad that the aquatic exercise study concentrated on the change over time in the intervention group and ignored the fact that there was no significant difference between the two groups in post-trial distance.

3. There's scope for small variation according to motivation. I think it's highly likely that participants in the GET group were encouraged to push themselves harder, perhaps even thought to themselves and were encouraged to think 'I am in the GET group, I can do this'. (Also of course, the dropouts in the GET group probably were more likely to be the ones who were struggling physically.)

I'm not sure how translatable distances walked are from one trial to another. There could be differences in implementation.

• distance between turns
• walk around the end markers or just touch and turn?;
• is the end point marked with a cone, or a mark on the ground?;
• ground surface making it easy to pivot or hard?;
• what shoes are people wearing (do they come prepared?);
• a chance to watch others doing it so as to pick up ideas on how to do it better?
• how close to running were participants allowed to get?
• what happened if the person jogged for a few steps, or a lot of steps? Or if they did not quite touch the end point? Was the person required to start again if they did it wrong? Did they get tired if they had to start again? Bias could easily creep into quality control.
• differences in the encouragement given (again subject to bias)

I don't think the 6 minute walk test in mild ME/CFS measures changes in physical disability well. It's not discriminatory enough and it's too easy to bias. The people with a gross deterioration that the test could pick up will mostly have already dropped out. The rest will continue to be able to walk for 6 minutes, and will probably do better on the re-test because they have experience in it.

If there's a walking test, it needs to be for 30 minutes or an hour. Or otherwise, it needs to be a well-defined physical action that is more demanding and will result in quicker tiring.

 

Simon, I think you added that last bit about the fitness test after I started writing a response.

So, if a fitness test showed no improvement, but the PACE trial 6 minute walking test showed small improvements in all the treatment arms, with the improvement being a bit bigger in the GET group, then doesn't that suggest that the 6 minute walking test is not a useful test of change in fitness in mild ME/CFS? And suggest that it is too open to bias?

 

The 6 minute walk test in is uniterpretable in PACE because too many didn't do it at the 12 month stage.

 

My question would be, what diseases were the subject of those clinical trials? People who aren't constrained when walking for 6 minutes won't be able to increase their walking distance much. People who are constrained may be able to increase their distance after treatment. I think that most people with mild ME/CFS walk enough that they can produce distances in the normal range for a 6 minute walk test.

I think it may be useful in cohorts of people whose ability to walk for 6 minutes is substantially impaired. For example, serious COPD. I don't think this would have been true for most of the PACE trial participants, provided they weren't in PEM.

And, I don't think the results are very helpful in revealing PACE's failure. As I said, without knowing the details we can't know if many of the participants were close to their healthy distance (and/or none of the interventions improved capacity) . I recall there were concerns expressed by the researchers that the corridor used for the PACE 6 minute walk was too short. In that case, we would expect the distances walked to be shorter than in other tests that use a 30 m track. It's possible that if you took relatively sedentary people matched on age and sex, they might not have recorded better distances under the PACE testing conditions.

I think the opposite is true, longer shuttle distances boost walking distance. You have time to get into rhythm and build up speed. With short distances before turns, there is a lot of acceleration and deceleration. There do seem to be differences in distances used in various tests (and probably in those other bullet pointed factors I mentioned) so I don't think we can compare distances reported in different trials without knowing a lot about the test protocol.

I think too much about the 6 minute walk test remains subjective for it to be much use except in identifying gross impairment (e.g. as in the frail elderly or in rehabilitating amputees).