Physical activity before and after exercise in women with chronic fatigue syndrome, 1998, Sisto et al.

[ME/CFS Skeptic]

Abstract
We measured physical activity after strenuous exercise in 20 women with chronic fatigue syndrome (CFS), compared to 20 sedentary healthy volunteers who exercised no more than once per week.

Activity was measured for 2 weeks using a portable waist-worn vertical accelerometer. After the first week of activity monitoring, all participants returned for a maximal treadmill test, followed by continued activity monitoring for the second week. Five activity measures were derived from the data: (i) average activity; (ii) total activity; (iii) duration of waking day; (iv) duration; and (v) number of daily rests.

A repeated measures ANCOVA was used to determine post-treadmill group differences accounting for pre-treadmill differences. There was a significant reduction in overall average activity after the treadmill test, with the greatest decrease on days 12 through 14. This reduction was accompanied by a significant increase in the duration of the waking day and number of daily rests.

Thus, marked exertion does produce changes in activity, but later than self-report would suggest, and are apparently not so severe that CFS patients cannot compensate.

Link: https://pubmed.ncbi.nlm.nih.gov/9797929/

 

[ME/CFS Skeptic]

An old study from 1998 by the research team of Benjamin Natelson. They measured activity levels before and after an exercise test.

The exercise test had remarkably little effect on activity levels. There appears to be no anticipation or reduction of activity before the test and activity levels remained relatively constant afterwards. Interestingly there was a reduction in activity and more rest but only 5 days after the exercise test.

The results of this study indicate that the effect of a maximal exercise test does not produce immediate dramatic effects on recorded average activity or on the number of daily rests for the CFS group, but does so 5 days later.

[....[

It is important to point out that despite both more rests and a decrease in average activity on days 5 through 7 following exertion for the CFS group, total daytime activity did not change at this time. This indicates that CFS patients appear to compensate to post-exertional consequences by lengthening their active period.

 

[Hutan]

Use of technology
Impressive to be using an activity monitor as far back as in the 1990s, when there are researchers telling us that even now there is no suitable technology to produce useful objective outcomes. @sarahtyson

Recording activity with an accelerometer provides such an objective measure of activity. Accelerometers have been used to measure physical activity as early as 1960, when vertical acceleration was measured while participants walked on a treadmill.4 Haskell and associates5 reported that physical activity motion sensors have a major use in quantifying physical activity. ...The moderate to high validity correlation and the subjects’ favourable response to wearing the accelerometer supported its validity and utility as an objective method for monitoring activity in field settings.

That quote may be useful sometime e.g. when commenting on the PACE trial suggesting that activity monitoring was too onerous, or when responding to current day researchers who suggest activity monitoring is not feasible in trials.


Problems in activity monitoring
However, I think the length of monitoring is way too short to draw useful conclusions - undoubtedly there will have been a Hawthorn effect going on here. I expect there were a range of thought patterns affecting activity levels. The drop off in activity in the last couple of days may have been a common reaction to activity levels sustained at a level over the fortnight higher than normal.

I think we need people to understand the impact of accumulated activity better. There's a lot of focus on one off exercise events, but I think adrenalin or something can get us through those sometimes, especially if they are quite a brief burst of activity.


Participant selection
There are possible issues with the characterisation of the sample:

CFS patients were identified from advertisements for individuals complaining of fatigue that were placed in community and support group newsletters, news- papers and from letters of request sent to surrounding primary care physicians.

They were screened (physical examination and blood work) to rule out other causes for the fatigue.

Using this strategy, we identified 40 women volunteers — 20 healthy sedentary controls and 20 with physician- diagnosed CFS which fulfilled the 1988 CFS working case definition modified to include only those patients reporting a three or greater intensity on five point symptom severity scales in the month prior to recruitment.

The diagnostic criteria is the CDC 1988 Holmes criteria

 

[Hutan]

There was no significant difference in the V ̇O2 max (ml/kg/min) between the two groups, indicating that both groups were of similar fitness levels

After 7 days of wearing the waist activity monitor (WAM), subjects performed a tread- mill exercise test which consisted of a progressive ramping of the treadmill speed and grade for 3-min workloads until the participant could no longer continue. Each participant was encouraged to reach her age-predicted maximal heart rate which was determined by the equation (210-[0.65×age]).14 Following the treadmill test, the participants wore the activity monitor for an additional seven days. A prior study indicated 5–7 days of activity were needed to represent stable daily physical activity data.15 The activity monitors were collected after the 14-day period.

The results of this study indicate that the effect of a maximal exercise test does not produce immediate dramatic effects on recorded average activity or on the number of daily rests for the CFS group, but does so 5 days later. An unexpected outcome was that neither total daily activity nor duration of rests changed following maximal treadmill testing. This outcome is quite different from that predicted by the fear avoidance model of post-exertional fatigue in CFS and thus does not support that model.

Thus, the data do not support the frequent report of CFS patients that exertion puts them to bed 1–2 days later.19 This is important because maximal exercise tests are often done to track studies of gentle physical conditioning. It may be that the brief aerobic exertion is not what produces their increase in fatigue reported by CFS patients, but that an extended submaximal exercise test might produce this out- come. However, our own ongoing work treating CFS by gentle physical conditioning along with the work of Fulcher and White20 does not support this possibil- ity either. In that study, participants only rarely reported a severe increase in symptoms following submaximal exertion.

Another explanation is that the CFS patient some- how misinterprets the way she feels after exertion. Although the duration of rest does not change after exertion, the CFS patient does rest more frequently after strenuous exercise. CFS patients may interpret their need for more, albeit brief, rests, as decreased activity after exertion. There may also be a problem with the patient’s perception of post-exertional fatigue. In our last study20 when a simple Likert scale of fatigue was administered after the same treadmill exercise test, patients reported a significant increase in fatigue which remained for a 2-week period. When a more complex questionnaire which had questions about fatigue embedded with questions about other feelings was administered 4 days after the treadmill test, the same significant increase was not seen. Thus, the patients’ expectations that they should be fatigued following exertion rather than a striking increase in fatigue due to exertion may be responsible for this apparent discrepancy.

Another possibility may relate to patients mislabel- ling the time of their diminished activity. Patient self- report is that this occurs 1–2 days after the end of exertion. Average activity did fall, but not within this time frame. Instead it fell 5–7 days after the end of exertion. At this same time, the number of rests for the CFS patients was also increasing. Thus, our results showing a 30% decrease in average activity after exertion support the patient self-report of an activity decrease, although it is delayed beyond the point that the patient usually reports. It is possible that CFS patients recall that this occurs several days after exertion but instead are incorrectly remembering the time that the decrease in activity actually begin. These data support the need for further research on post-exertional fatigue using objective measures such as the actigraph rather than relying on patient self- report.

The bolding in the last quote shows that people did understand the need for objective measures 25 years ago.

 

[Hutan]

It is important to point out that despite both more rests and a decrease in average activity on days 5 through 7 following exertion for the CFS group, total daytime activity did not change at this time. This indicates that CFS patients appear to compensate to post-exertional consequences by lengthening their active period. This trend toward a longer active period in the latter half of the post-treadmill week is apparent in Figure 3. The fact that CFS patients can compensate for reduced average activity by remaining active for longer periods implies that the delayed effects of strenuous exercise on activity are not dangerous to the patient. Thus, these data support patient self-report of diminished activity but do not provide objective evidence for the additional self- report that these effects are extreme.

They are noting that the average activity decreased in the second half of the post-exercise week for the CFS participants. But, that the CFS participants were active for more hours, resulting in similar total daytime activity to that in the first week. And then they conclude that "the delayed effects of strenuous exercise on activity are not dangerous for the patient".

That of course is a ridiculous conclusion.
This so called increase in the mean length of the day was very small, with enormous variability, and with the mean CFS day substantially shorter than the healthy day throughout.

How they can conclude from one exercise exposure that the delayed effects of strenuous activity are not dangerous, I do not know. Again, they are missing the important issue of cumulative effects, and the sustainability of the activity levels in these two weeks.

Here's the chart of total daily activity.

There's a great deal of variation, and there is a comment that the CFS data has much more daily variation than the healthy data - the authors mention a rollercoaster pattern. CFS daily activity levels are, on average, considerably lower than those of the sedentary healthy controls. There is therefore likely to be much less discretionary activity going on prior to the exercise event, and so, after the exercise event they may try to essentially "soldier on" with activities of daily living and essential obligations such as family care and work.

I've watched my son attempt to go back to full-time school, and it's a slow motion crash - it took a few weeks for activity levels to start to drop off, with the decline continuing over months.

 

[Hutan]

We must point out a number of shortcomings with the nature of the activity data collected. First, it was expected that a one-week baseline period would be enough to generate stable baseline data. While that was true for the healthy controls, it was not always true for the CFS patients. Whether this baseline variability represented anxiety concerning the upcom- ing treadmill test or variability of activity innate to CFS is not clear. This baseline variability may provide objective support for the clinical impression that patients with CFS tend to do too much on one day, causing them to be far less active the next day, thus having a ‘roller-coaster’ effect on activity. However, this baseline instability may have reduced our ability to understand the effects of exertion on measures of activity, since our statistical analysis was done after accounting for average baseline data. Therefore, sub- sequent studies of activity in CFS patients need to evaluate baseline activity for longer than 7 days to either improve stability of the baseline activity meas- ures or to document any inherent variability in CFS activity which in itself may be meaningful.

The authors recognise that the monitoring period was too short.

This study was not a terrible start to the use of activity monitoring to better understand ME/CFS. It was a start. What is terrible is that, 25 years later, we really don't have much more of an understanding about what is going on. And we have researchers actively rejecting the use of objective outcomes in studies.

(Sorry, I hope my posts here make some sense. Thinking is hard work at the moment.)

 

[Jonathan Edwards]

Cannot see how you are going to recruit people with significant PEM from an advertisement for people with fatigue. When I had Long Covid mildly there is no way I would have volunteered for a study like this. Anyone I can think of with significant ME/CFS I would expect to keep well clear of this sort of study.

And performance will have been modified by all sorts of expectations on the subjects' parts.

It is interesting that researchers actually tried this sort of documentation of PEM thirty years ago but I don't see any interpretable data coming out of one off maximal exertion tests.

 

[ME/CFS Skeptic]

Interestingly there was a reduction in activity and more rest but only 5 days after the exercise test.

Something similar was found in this study that used graded activity with activity meters. Patients were able to increase their activity levels but only for a couple of days. After 4–10 day, their activity level dropped indicating they were unable to keep up this level.
Time course of exercise induced alterations in daily activity in chronic fatigue syndrome - PMC (nih.gov)

 

[Eleanor]

What does a vertical accelerometer measure? Would it record the difference between, say, resting in a chair with your feet up while reading/working/doing something manual (eg knitting) and resting in bed with the curtains closed doing absolutely nothing?

 

[ME/CFS Skeptic]

A Dutch study did the same thing but only up to 5 days after the exercise test. Here the actometer results showed no notable decrease before or after the exercise test.

Impact of a maximal exercise test on symptoms and activity in chronic fatigue syndrome - PubMed (nih.gov)

 

[bobbler]

A Dutch study did the same thing but only up to 5 days after the exercise test. Here the actometer results showed no notable decrease before or after the exercise test.
View attachment 22841
Impact of a maximal exercise test on symptoms and activity in chronic fatigue syndrome - PubMed (nih.gov)

This is interesting but it just leaves me with a lot of questions, some of which might be answered if I could see more of the article.

I get the impression the actometer axis isn't 'number of steps in a day' [on basis of it being eg 40], so would like to know what is actually in it. I'm guessing the test itself isn't included here, but the steps to get there would be. And whilst maybe some HCs and maybe some CFS might have had to do other things on the same day it might have taken up most of the day and it is one where both might have similar 'committments'.

So it is perhaps interesting to use this as an anchor and note that is a 'low activity' day for those who are HCs vs CFS it is consistent with their max days. All I can tell from sure really from this is that the CFS group might equal but don't really do 'more' than the level on that test day, where the control is the same amount as the test day leading up then it varies above as well as below this level.

As well of course I have questions about knowing more about the CFS participants - looking at the dip at day 2, vs the little dip straight after for the HCs I suspect maybe some did have PEM but who knows if they all did. And how severe were they? what other committments did they have?

And of course the slight issue is rolling 20 people's data together when onset of PEM could be 12-48hrs mightn't be such an issue if you were looking at people with more severe ME/CFS who all had PEM that lasted quite a long time, but if you've a mixed bag of more mild, maybe some 'chronic fatigue' instead of ME/CFS and add in differing commitments..

Anyway I suspect people don't realise that they really need to be profiling and getting a lot more information about their ME/CFS participants with these things, whether they are controlling them or not. Because of floor and ceiling effects.

In effect we are 'tethered' around what we have to do to manage. Not our 'baseline' but the compromise we have with a world that often doesn't allow us what we need.

Someone who is more ill might not even do much on a well day and then their PEM might involve needing to wee a lot. Someone who isn't severe might have committments they can't avoid even when they are in PEM like caring responsibilities, having a medical appointment, work, so the most they can dip has to be 'planned'.

 

[Hutan]

A Dutch study did the same thing but only up to 5 days after the exercise test. Here the actometer results showed no notable decrease before or after the exercise test.

Which again goes to the problem of 'is that exercise test enough to cause PEM or general deterioration, which would be the reasons why activity would reduce?'.

The people with (ME)/CFS who volunteer for these studies are not put off by the prospect of 8 minutes or so of cycling, because they know that they can probably accommodate that within their usual energy envelope (for want of a better term). Perhaps the usual housework and meal preparation doesn't get done. If you look at these studies, the day of the exercise test is not a great spike of exertion (however that is measured).

I think these studies are asking the wrong question.

 

[bobbler]

This is interesting but it just leaves me with a lot of questions, some of which might be answered if I could see more of the article.

I get the impression the actometer axis isn't 'number of steps in a day' [on basis of it being eg 40], so would like to know what is actually in it. I'm guessing the test itself isn't included here, but the steps to get there would be. And whilst maybe some HCs and maybe some CFS might have had to do other things on the same day it might have taken up most of the day and it is one where both might have similar 'committments'.

So it is perhaps interesting to use this as an anchor and note that is a 'low activity' day for those who are HCs vs CFS it is consistent with their max days. All I can tell from sure really from this is that the CFS group might equal but don't really do 'more' than the level on that test day, where the control is the same amount as the test day leading up then it varies above as well as below this level.

As well of course I have questions about knowing more about the CFS participants - looking at the dip at day 2, vs the little dip straight after for the HCs I suspect maybe some did have PEM but who knows if they all did. And how severe were they? what other committments did they have?

And of course the slight issue is rolling 20 people's data together when onset of PEM could be 12-48hrs mightn't be such an issue if you were looking at people with more severe ME/CFS who all had PEM that lasted quite a long time, but if you've a mixed bag of more mild, maybe some 'chronic fatigue' instead of ME/CFS and add in differing commitments..

Anyway I suspect people don't realise that they really need to be profiling and getting a lot more information about their ME/CFS participants with these things, whether they are controlling them or not. Because of floor and ceiling effects.

In effect we are 'tethered' around what we have to do to manage. Not our 'baseline' but the compromise we have with a world that often doesn't allow us what we need.

Someone who is more ill might not even do much on a well day and then their PEM might involve needing to wee a lot. Someone who isn't severe might have committments they can't avoid even when they are in PEM like caring responsibilities, having a medical appointment, work, so the most they can dip has to be 'planned'.

I agree with @Hutan noting their son returning to school eventually and slowly collapsing, having tried to do so over a space of time. And note that it is interesting the top article talks about 12-14 days being the actual point they noted a reduction.

I think this serves more as a warning for methodology focused only on actimeter, rather than all the blurb put around these 'findings'. Because like other ill people we can only reduce our activity to the minimum we have to do. Until it catches up with us and permanently reduces our threshold or enforces a 'world on hold' crash. And I guess even child-caring has to drop, it becomes an utter can't and a whole new struggle of life. But these measures call themselves longitudinal and have no idea. The harms claims is ridiculous

 

[bobbler]

Which again goes to the problem of 'is that exercise test enough to cause PEM or general deterioration, which would be the reasons why activity would reduce?'.

The people with (ME)/CFS who volunteer for these studies are not put off by the prospect of 8 minutes or so of cycling, because they know that they can probably accommodate that within their usual energy envelope (for want of a better term). Perhaps the usual housework and meal preparation doesn't get done. If you look at these studies, the day of the exercise test is not a great spike of exertion (however that is measured).

I think these studies are asking the wrong question.

Does the day of the exercise test actually include the test?

I assumed it included walking to and from, and agree that you've probably taken the day off work or it is a weekend where they might have been doing something similar going by the graph.

I agree on the moving other tasks, and we've an issue - which might now be better accomodated for with the latest tech - of all the other tasks like showering that wouldn't show on an actimeter that I consider probably are in many pwme more the hallmarks of how someone is feeling than just pure footsteps. How much you have to recline/many hours you are having to lie down is another one - if you are well enough you can make food in PEM then the walk to the kitchen is the same (or maybe more if you are coming from the bedroom).

This doesn't really show for example if any of them (back in those days) might have in a normal week been going to the gym or had a few days a week at work that had to be reined back on to keep things only to essential. Because it isn't comparing it with any other weeks. And even those who might know they would be affected could have taken the week off work to do this and fill in all the forms.

As there is a floor effect (everyone will go to the loo and that will involve steps, they might have to answer the door or get food etc), we've no sense at all of what I assume is the real issue of interest which is whether CFS 'dipped more' in what they did vs HCs. Because we don't have the standard week comparator.

To me this doesn't feel like it was really an open-minded test, but maybe I'm being harsh given how knowledge has moved on with time, but one that was only really open to 'ruling out' unless there was some way a dramatic difference could be shown