https://www.ncbi.nlm.nih.gov/pubmed/30063870
I wasn't sure whether to put this here or in the BPS forum so mods, please move it if BPS would be more suitable.
-
New Forum Software Installed (Still a few issues but reopening the forum) More details.
Association between cytokines and psychiatric symptoms in chronic fatigue syndrome and healthy controls (2018), Groven et al
- Thread starter hixxy
- Start date
https://www.ncbi.nlm.nih.gov/pubmed/30063870
I wasn't sure whether to put this here or in the BPS forum so mods, please move it if BPS would be more suitable.
Hoopoe
Senior Member (Voting Rights)
I looked at one of the questionnaires they used, the SCL-90-R which gives scores in categories such as "somatisation", "anxiety", "obsessive-compulsive" and so on.
If you report symptoms such as headache, weakness, restlessness, dizziness, decreased sexual desire, chest pain, etc this counts towards one of these categories. Apparently their thinking is that these are strictly psychiatric symptoms. At this level of intellect, this group is unlikely to produce anything useful.
If you report symptoms such as headache, weakness, restlessness, dizziness, decreased sexual desire, chest pain, etc this counts towards one of these categories. Apparently their thinking is that these are strictly psychiatric symptoms. At this level of intellect, this group is unlikely to produce anything useful.
Last edited:
Small study, lots of correlations tested, a few reached significance. So is this a case of p-hacking? At least they don't try to contort the conclusion to pretend they have found any useful connection between ME and psychological problems. And they do conclude there is a different immune picture in ME, so it's biological.
There are issues with the HADS scale in terms of measuring depression. The questions are not great.
What may be interesting is to look at correlations between the cytokines and particular questions and what that may say about levels of disability. Rather than trying to place bad semantics of sets of question answers,
What may be interesting is to look at correlations between the cytokines and particular questions and what that may say about levels of disability. Rather than trying to place bad semantics of sets of question answers,
Yes, seeing 'twenty patients' and 'subgroups' together had my head in my hands. Pretty certain this isn't the first study finding TNF-alpha differences though, will need to have a hunt...
Woolie
Senior Member
The number of participants are so small that there was probably insufficient power to detect anything in the first place. And indeed, there's a good chance that even the positive findings might turn out to be spurious.
Notice too that one of the "psychiatric" measures was "somatisation", but the scale they used doesn't really measure somatisation (that's probably not even a thing), but in fact counts the number symptoms you report experiencing. So let me fix that for you:
I also shudder when I see patients from two distinctly different clinical groups - in this case PwMEs and matched controls - pooled together to perform a single correlation analysis. No, no, NO!
I'm figuring they didn't measure IL1beta because its present in such small concentration in plasma that its hard to detect.
Notice too that one of the "psychiatric" measures was "somatisation", but the scale they used doesn't really measure somatisation (that's probably not even a thing), but in fact counts the number symptoms you report experiencing. So let me fix that for you:
Seems much easier to interpret now.
I also shudder when I see patients from two distinctly different clinical groups - in this case PwMEs and matched controls - pooled together to perform a single correlation analysis. No, no, NO!
I'm figuring they didn't measure IL1beta because its present in such small concentration in plasma that its hard to detect.
Last edited:
Thanks @Woolie for that - I thought it seemed a very odd thing to do, and completely meaningless.
Woolie
Senior Member
Yea, Wot @strategist said.
Luther Blissett
Senior Member (Voting Rights)
Aren't they also common side effects of medications too? (especially anti-depressants?)
Snow Leopard
Senior Member (Voting Rights)
The only meaningful cytokine association regularly reported in the literature is TGF-Beta, and well, they didn't test for that in this study. White even did a systematic review highlighting TGF-Beta so its absence is notable.
Mithriel
Senior Member (Voting Rights)
'RESULTS:
Increased plasma levels of TNF-α in CFS/ME patients almost reached significance compared to healthy controls (p = .056). When studying the CFS/ME and control groups separately, there was a significant correlation between TNF-α and The Hospital Anxiety and Depression Scale (HADS) depressive symptoms in controls only, not in the CFS/ME group. A correlation between IL-10 and psychoticism was found in both groups, whereas the correlation for somatisation was seen only in the CFS/ME group. When looking at the total population, there was a significant correlation between TNF-α and both the HADS depressive symptoms and the SCL-90-R cluster somatisation. Also, there was a significant association between IL-10 and the SCL-90-R cluster somatisation when analyzing the cohort (patients and controls together).'
"almost reached significance" is meaningless This may not be a BPS study, but it is something the BPSers do - it leaves the impression something of scientific interest has been found when it hasn't.
Increased plasma levels of TNF-α in CFS/ME patients almost reached significance compared to healthy controls (p = .056). When studying the CFS/ME and control groups separately, there was a significant correlation between TNF-α and The Hospital Anxiety and Depression Scale (HADS) depressive symptoms in controls only, not in the CFS/ME group. A correlation between IL-10 and psychoticism was found in both groups, whereas the correlation for somatisation was seen only in the CFS/ME group. When looking at the total population, there was a significant correlation between TNF-α and both the HADS depressive symptoms and the SCL-90-R cluster somatisation. Also, there was a significant association between IL-10 and the SCL-90-R cluster somatisation when analyzing the cohort (patients and controls together).'
"almost reached significance" is meaningless This may not be a BPS study, but it is something the BPSers do - it leaves the impression something of scientific interest has been found when it hasn't.
Dolphin
Senior Member (Voting Rights)
The full text of this is available for free here:
https://www.tandfonline.com/doi/full/10.1080/08039488.2018.1493747
https://www.tandfonline.com/doi/full/10.1080/08039488.2018.1493747
Snowdrop
Senior Member (Voting Rights)
I am stepping out a little beyond my knowledge base/ability to function here so someone will correct me if I've got even this little bit wrong.
So the BPS devotees come to realise that cytokines are of interest in the ME community as a possible avenue of interest.
Et voila -- a paper that shows the association between cytokines and psychiatric symptoms in ME.
Never mind that the evidence of psychiatric symptoms comes from the laughable Chalder fatigue questionnaire.
I think that this shows an avenue for us that needs following up. We must continue to hammer home to all and sundry that the CFQ does not do what the BPS suggest it does and that it is not fit for purpose. Every medical researcher should be made aware of it's abject failings in this regard.
I do think it would benefit us to demolish the CFQ thoroughly so that it is shown to have no validity at all when used for ME (and possibly elsewhere) so that it would also invalidate findings based on it.
Is there in fact a way to do this?
So the BPS devotees come to realise that cytokines are of interest in the ME community as a possible avenue of interest.
Et voila -- a paper that shows the association between cytokines and psychiatric symptoms in ME.
Never mind that the evidence of psychiatric symptoms comes from the laughable Chalder fatigue questionnaire.
I think that this shows an avenue for us that needs following up. We must continue to hammer home to all and sundry that the CFQ does not do what the BPS suggest it does and that it is not fit for purpose. Every medical researcher should be made aware of it's abject failings in this regard.
I do think it would benefit us to demolish the CFQ thoroughly so that it is shown to have no validity at all when used for ME (and possibly elsewhere) so that it would also invalidate findings based on it.
Is there in fact a way to do this?
Tom Kindlon
Senior Member (Voting Rights)
The Science for ME forum also made a submission.
The Chalder Fatigue Questionnaire was subsequently dropped
Last edited:
Snowdrop
Senior Member (Voting Rights)
I read your post and followed the links. I am very glad this is being addressed (and appreciate the efforts made)but my cognitive functioning is such that everything about this is so unfamiliar to me that it was beyond me to parse what it all means exactly.
As an example, the CFQ was subsequently dropped from what? I know the answer is obvious but I cannot hold onto information (memory?) when presented inside a bulk of unknown info (I don't know if that makes sense at all).
How does this affect the use of CFQ in the paper that is the subject of this thread?
Again, thank you for all your efforts in pursuing this and other issues.
Tom Kindlon
Senior Member (Voting Rights)
CDE stands for Common Data Elements. Researchers in the US will be encouraged to use scales that are mentioned under the different categories. This could perhaps influence what researchers in other countries do.