Physical activity intensity but not sedentary activity is reduced in CFS and is associated with autonomic regulation, 2011, Newton et al

@Hilda Bastian, please read this research report and the commentary here. This is the sort of thing that we are up against. And the lead author is one of the people you have put on the Cochrane review writing team. Perhaps Julia Newton would not repeat the issues that we see in this 2011 paper now, maybe she would. But this is one of the people that we are now relying on to properly assess the exercise research and produce an accurate report. Maybe you can understand why ether are concerns? If you can't see the problems with this paper, please say, and we can try to explain better.

 

retagging as I didn't get the tag right in the last post. @Hilda Bastian

 

A short critique of this paper would be good to be included as one of the pieces of evidence for the Cochrane Exercise review - it illustrates a number of the problems that are common in BPS research suggesting that exercise is useful in treating ME/CFS.

 

I started looking at the paper closely and I've got to say that it's a mixed bag. The discussion is a lot more nuanced than the abstract. Here are some notes I started. I'll park them here in case I don't get back to the job.

Selection criteria
The paper uses the Fukuda criteria to select the CFS cohort. In this criteria, post-exertional malaise (PEM), the defining feature of ME/CFS, is not required. Therefore it is impossible to know how many of the participants did in fact have ME/CFS. Although this paper was written in 2011, and the centrality of PEM in ME/CFS symptomology has been better described since then, researchers are still using the Fukuda criteria. The AHRQ in 2015 (from memory) noted that papers with Fukuda selected cohorts did not provide useful information about ME/CFS.


Short time period for activity monitoring
While wearable activity monitors are much better than a recall survey, this study only had participants record activity for 7 days. Monitoring, even if only a step count, over a much longer time period, perhaps 3 months, is required to ensure that any Hawthorne effect is minimised. It is likely that people who are monitored make an effort to be more active than normal initially, but this will not be sustained. People with ME/CFS are particularly unlikely to be able to sustain a high level of activity.


Fatigue Impact Scale (FIS)
This scale is likely to have resulted in people with a wide range of fatigue severities being assessed as having the same amount of fatigue. This scale gives options from 0 ‘no problem’, to 4 ‘extreme problem’ for 40 questions. Most people with ME/CFS who are able to participate in a study are unlikely to respond to many questions with a ‘4’, as there is still some capacity. However, to receive a diagnosis of CFS, it is necessary to have a substantial reduction in capacity (50% in some criteria), and so they are unlikely to respond to many questions with ‘0’ either.

​

Furthermore, someone only moderately incapacitated is likely to feel the incapacity is a significant problem compared to their previous level of function and may score the incapacity the same as someone more profoundly disabled. For example, with the question ‘I have to reduce my workload or responsibilities’, someone who can now only go to work three days a week might choose 3 ‘a big problem’. Another person who has given up working entirely, but is still able to help in the care of their child might also choose ‘3’, comparing it to a period when they were completely bedbound.

​

The scale therefore lacks sensitivity and does not accurately measure the severity or frequency of fatigue. It is likely that the participants’ scores were clustered, and not indicative of the level of fatigue.

It is good to see that the authors also realised that the FIS was not a good measure of fatigue

In addition, there are issues with some questions not, or not necessarily, relating to fatigue. For example, a question about physical discomfort increasing says very little about fatigue and much more about pain. A question about clumsiness might be rated highly if there are balance issues.



Relationship between fatigue and activity level
The issues with the activity monitoring period potentially being too short and the fact that the FIS doesn't provide a good measure of fatigue are enough to explain why the study found no relationship between fatigue and activity level.

But, there is another problem. There is an implicit assumption that CFS is, as the name implies, all about fatigue and that people who have more fatigue will have lower levels of activity . Actually, the illness involves fatigability and PEM.

If a person reduces their activity level below a certain level, they may be able to avoid 'fatigue'. Therefore, a person with a low level of activity may have a low level of fatigue, while a person with a high level of activity may have a high level of fatigue. It is also possible that someone experiencing PEM or a period of a flare may have a high level of fatigue and a very low activity level.

It is therefore unrealistic to expect a close relationship between fatigue and activity across a cohort of people with ME/CFS.

 

Inference of ‘they brought it on themselves’; a lack of grounding comparisons

The abstract states ‘Thirty-seven percent of the CFS group were overweight and 20% obese.’ This, together with the conclusion that low levels of physical activity is a perpetuating factor, leads the reader to conclude that these people are not helping themselves to get well.

The text notes that there were no significant differences in the BMI between the CFS and healthy control cohorts – but given that the healthy controls were specifically selected to match the CFS participants on BMI, this is not a surprise. There is no context given for the levels of BMI in the CFS cohort.

The 2019 UK adult population is reported as consisting of 36% of people in the overweight range and 28% of people in the obese range. Therefore, with 57% of people with CFS in this study in the overweight or obese range and 64% of the general population in that category, the CFS cohort’s BMIs are hardly remarkable. And yet, the inference from the mention of people with CFS being overweight is that that is a causal factor.

This tactic is used frequently in papers (and particularly abstracts) aiming to show that people with CFS have depression, anxiety, flawed behaviour or personalities, or histories of trauma. Incidences in people with CFS are not compared with incidences in the general public, much less with incidences in people with chronic health conditions. Typically when investigated, incidences in people with CFS are normal.

 

Ok, coming back to this to try again to understand the study's findings about activity levels. I found its logic quite muddled on a first read.

Most of the people with ME/CFS were actually relatively active, making conclusions that people with ME/CFS need to increase habitual physical activity unsubstantiated
People with ME/CFS did an average of just over 7000 steps, while the healthy controls did an average of just over 10000 steps. Have a look at these quotes

https://www.caba.org.uk/help-and-guides/information/ways-reach-10000-steps-day

https://www.10000steps.org.au/articles/counting-steps/. (These categories are for adults without a chronic disease condition)

Now I don't believe for a minute that the number of steps the ME/CFS cohort or the healthy controls recorded in the week when they had monitoring equipment is representative of their normal amount of steps - for sure, they were more active than normal. Nor do I think that there aren't lots of people with ME/CFS doing less than 5000 steps per day. But, for goodness sakes, a finding of an average of over 7000 steps in the people with ME/CFS in this study is not evidence that these people are sick because they are not being active enough. This is another example of a lack of a grounding comparison.

This line in the discussion part of the paper is not in accord with information provided by the '10000 steps' organisation. Most of the people in the ME/CFS cohort of this study would not be classed as 'sedentary' according to their scale and are well above the national average in the number of steps that they do.

This study presents no evidence to support a conclusion that people with CFS need to be encouraged to increase their physical activity. This particular cohort are already, on average, making substantially more steps per day than the average UK person is. Most people with ME/CFS that I have come into contact with express frustration that they cannot be more active. This paper neither presents evidence that encouraging people with ME/CFS to increase their daily activity will have any lasting effect on their daily activity nor that an increase in activity will bring net benefits. Therefore its conclusion that it is vital to encourage such an increase is not supported by its findings.

tbc

 

One point on weight of PWME is that Amitryptiline seems to be commonly prescribed for insomnia/pain. Increased appetite/weight gain is a common side effect.

 

The correlation is not causation problem

The last line of this paragraph is a nice example of the 'correlation is not causation' problem that we see in nearly all BPS papers. Yes, people with multiple sclerosis who exercise regularly do report being more well than those who don't. But, people with multiple sclerosis who are more well are more able to participate in regular exercise. The fact that wellness and participation in regular exercise are correlated does not mean that the primary causal direction in multiple sclerosis is Exercise causes Wellness.

It's such a basic error, and yet it happens so often when people are determined to make data fit their preconceived ideas about how the world works.

Here's a table from a paper about the number of steps people with MS take:


You can see that people with more severe forms of the disease walk a lot less steps than people with less severe forms. Has the laziness of those with progressive MS caused their disease to be worse? I guess we can all speculate according to our prejudices. But disease duration is more clear cut - no positive personality and commitment to exercise can stop the passage of time. Those who have been sick for less than 10 years walk an average of nearly 1000 steps more per day more than those who have had the disease for more than 10 years. Aging may have contributed a bit to the reduction, but the mean age of 46 years across all of the participants in the study had quite a tight standard deviation around it. I think we can conclude, given the generally progressive nature of the illness, that being more ill has a major causal effect on the number of steps walked.

 

Yes, it is.

 

Thanks @Nightsong. I can't make those images big enough to read. Personally I have found that an increase in the POTS measure (increase in HR on standing) is associated with being in PEM.

That is a problem.

Re the strengthening of core muscles. If I had to hypothesise a mechanism, I'd go for 'By maintaining firm muscle support around the abdominal area, it minimises fluid retention in that area - the splanchnic pooling'. There is some evidence for the splanchnic pooling in people with orthostatic intolerance issues, but I don't know of any evidence suggesting that good core muscles help. I think the excess blood pools in the gastrointestinal tract, so, would toned core muscles really stop that? I doubt it.