...discrepancies between objective and subjective measurement of the physical activity level in female patients with [CFS], 2021, Vergauwen et al

Discussion in 'Research methodology news and research' started by Andy, Mar 15, 2021.

  1. Andy

    Andy Committee Member

    Messages:
    23,032
    Location:
    Hampshire, UK
    Full title: An exploratory study of discrepancies between objective and subjective measurement of the physical activity level in female patients with chronic fatigue syndrome

    Highlights
    • There is a discrepancy between self-report and objective activity measures in CFS.

    • The activity diary cannot replace objective monitoring systems, i.e. accelerometry.

    • Age is negatively associated with the discrepancy in both cases and controls.

    • No demographic factors, illness-related complaints or HRQOL explain the discrepancy.

    Abstract
    Objective
    To explore the ability of a self-report activity diary to measure the physical activity level (PAL) in female patients with chronic fatigue syndrome (CFS) and whether illness-related complaints, health-related quality of life domains (HRQOL) or demographic factors are associated with discrepancies between self-reported and objectively measured PAL.

    Methods
    Sixty-six patients with CFS, recruited from the chronic fatigue clinic of a university hospital, and twenty matched healthy controls wore an accelerometer (Actical) for six consecutive days and registered their activities in an activity diary in the same period. Participants' demographic data was collected and all subjects completed the CFS Symptom List (illness-related complaints) daily and Short-Form-36 (HRQOL domains) during the first and second appointment.

    Results
    A significant, but weak association between the activity diary and Actical was present in patients with CFS (rs = 0.376 and rs = 0.352; p < 0.001) and a moderately strong association in healthy controls (rs = 0.605; and rs = 0.644; p < 0.001) between week and weekend days, respectively. A linear mixed model identified a negative association between age and the discrepancy between the self-reported and objective measure of PA in both patients with CFS and healthy controls.

    Conclusion
    The activity diary showed limited ability to register the PAL in female patients with CFS. The discrepancy between measures was not explained by illness-related complaints, HRQOL domains or demographic factors. The activity diary cannot replace objective activity monitoring measured with an accelerometer, but may provide additional information about the perceived activity.

    Paywall, https://www.sciencedirect.com/science/article/abs/pii/S0022399921000623
     
  2. Jonathan Edwards

    Jonathan Edwards Senior Member (Voting Rights)

    Messages:
    15,175
    Location:
    London, UK
    Does this mean that the results of the PACE trial are unreliable?
     
    sebaaa, Amw66, lycaena and 19 others like this.
  3. Lidia Thompson

    Lidia Thompson Senior Member (Voting Rights)

    Messages:
    123
    Location:
    Leeds
    Ha ha. Touché!
     
  4. cassava7

    cassava7 Senior Member (Voting Rights)

    Messages:
    1,051
    Out of all the journals this study could have been published in, I am surprised that it appeared in the Journal of Psychosomatic Research. And it comes just days after the other case study on blinding in a homeopathy trial for CFS.

    Are we going to see more articles on evidence quality in this journal that would debunk prior work from its advisory board members (Fink, Rosmalen, White, Sharpe, Stone - and the only French associate editor, Lemogne, supports GET for CFS)?

    @Woolie mentioned that the new editor of the journal, Jess Fiedorowicz, is pushing for more rigorous reviews of BPS papers. If it is truly happening, then this looks like a (cautiously) promising start.

    ETA: In order to provide reliable evidence, investigators of trials of GET for CFS should have conducted this study before starting any such trial. Instead, they decided without evidence that subjective reports of physical activity would be valid, even though they were aware of concerns around this issue. This further questions their scientific integrity.
     
    Last edited: Mar 15, 2021
  5. Milo

    Milo Senior Member (Voting Rights)

    Messages:
    2,138
    No it means that middle-age women are not to be trusted when reporting their symptoms :bucktooth:
     
    ukxmrv, MSEsperanza, MEMarge and 4 others like this.
  6. Colin

    Colin Established Member (Voting Rights)

    Messages:
    92
    Location:
    Brisbane, Australia
    Doesn't that mean that there is less of a discrepancy between the self-reporting and the measurement in the older women? So the older women are more accurate in their perception/reporting? And was the length of time that they'd had ME/CFS taken into account?
    And why was this a single-sex study?
     
    Last edited: Mar 16, 2021
    sebaaa, MSEsperanza, MEMarge and 5 others like this.
  7. Jonathan Edwards

    Jonathan Edwards Senior Member (Voting Rights)

    Messages:
    15,175
    Location:
    London, UK
    Perhaps the new editor is taking the name of the journal seriously - for research that is really all in the minds of the researchers.
     
  8. Hutan

    Hutan Moderator Staff Member

    Messages:
    29,374
    Location:
    Aotearoa New Zealand
    Yep, and not just PACE.
    (edit to add) If (daily?) activity diaries aren't reliable records of how much activity was done, how much worse must questionnaires asking about levels of activity after a month of an intervention be?

    What are they actually reporting here? I can't make sense of it - what have the week and weekend days got to do with things? It will be interesting to see the paper.

    Regardless, the headline finding must fundamentally change future trials of BPS ME/CFS interventions. Now we just have to argue that six days of activity monitoring isn't nearly enough to show a sustainable change in activity levels resulting from real health improvement.
     
    Last edited: Mar 16, 2021
  9. Sean

    Sean Moderator Staff Member

    Messages:
    8,064
    Location:
    Australia
    Well, certainly the primary outcomes. :laugh:
     
  10. Jonathan Edwards

    Jonathan Edwards Senior Member (Voting Rights)

    Messages:
    15,175
    Location:
    London, UK
    This puzzled. I guess they are simply reporting that weekday activities and weekend activities are statistically more similar to other days in the same subset than in the other.
     
  11. Hutan

    Hutan Moderator Staff Member

    Messages:
    29,374
    Location:
    Aotearoa New Zealand
    Yes, but if they are saying
    1. there was a significant but weak association between the activity diary and the activity monitor for CFS patients (and presumably no significant association between the activity diary and the activity monitor for healthy controls), and
    2. there was a moderately strong association in healthy controls between activity on week days and activity on weekend days (and presumably no significant association in people with CFS between activity on week days and activity on weekend days)

    then, they haven't really reported things properly in the highlights (i.e. that CFS patients are somewhat better than healthy controls at accurately recording activity, but no one is great at it). And what use is the 'respectively' serving the sentence?
     
  12. rvallee

    rvallee Senior Member (Voting Rights)

    Messages:
    13,659
    Location:
    Canada
    To (probably mis)quote something I saw not long ago: I do not question their scientific integrity, I reject its existence.
     
  13. Jonathan Edwards

    Jonathan Edwards Senior Member (Voting Rights)

    Messages:
    15,175
    Location:
    London, UK
    There are two r values for CFS and two for controls so isn't that they looked for one correlation amongst all weekdays and another for all weekend days, for each subject group, presumably because they thought the correlations might be rather different? So the correlations were better for healthy controls.
     
  14. Hutan

    Hutan Moderator Staff Member

    Messages:
    29,374
    Location:
    Aotearoa New Zealand
    Oh yes, you are right, somehow I missed those pairs of r2 values. ( I should have been asleep at the time.)
     
    Last edited: Mar 16, 2021
  15. Joan Crawford

    Joan Crawford Senior Member (Voting Rights)

    Messages:
    701
    Location:
    Warton, Carnforth, Lancs, UK
    Yes
     
    MSEsperanza, Amw66, Mithriel and 2 others like this.
  16. MSEsperanza

    MSEsperanza Senior Member (Voting Rights)

    Messages:
    2,947
    Location:
    betwixt and between
    It would be interesting to see the full text and know in which way the relationship of objective and subjective measurements exactly differed in each group and between the groups.

    I have no reason to doubt the correctness of their conclusion:
    I'd like to know though whether there was a clear tendency towards over- or underestimating activity.

    Also, what kind of activities did they investigate?

    Is it possible that the accelerometers that were applied don't measure particular movements/ activities/ inactivities with the same accuracy as other movements/ activities? So that the differences in accuracy of perceived activity might rather reflect an inaccuracy of the objective measures due to the fact that patients just have different movement patterns?
     
    Last edited: Jun 22, 2021
  17. Amw66

    Amw66 Senior Member (Voting Rights)

    Messages:
    6,769
    To me it suggests people are working during the week?
     
    Peter Trewhitt likes this.

Share This Page