• Welcome to Phoenix Rising!

    Created in 2008, Phoenix Rising is the largest and oldest forum dedicated to furthering the understanding of and finding treatments for complex chronic illnesses such as chronic fatigue syndrome (ME/CFS), fibromyalgia (FM), long COVID, postural orthostatic tachycardia syndrome (POTS), mast cell activation syndrome (MCAS), and allied diseases.

    To become a member, simply click the Register button at the top right.

New paper on ME in Psycho Research by Esther Crawley Jan 2018

Murph

:)
Messages
1,799
https://www.ncbi.nlm.nih.gov/pubmed/29275782

J Psychosom Res. 2018 Jan;104:29-34. doi: 10.1016/j.jpsychores.2017.11.007. Epub 2017 Nov 8.
Chronic fatigue syndrome (CFS/ME) symptom-based phenotypes and 1-year treatment outcomes in two clinical cohorts of adult patients in the UK and The Netherlands.
Collin SM1, Heron J2, Nikolaus S3, Knoop H3, Crawley E2.
Author information
Abstract

OBJECTIVE:
We previously described symptom-based chronic fatigue syndrome (CFS/ME) phenotypes in clinical assessment data from 7041 UK and 1392 Dutch adult CFS/ME patients. Here we aim to replicate these phenotypes in a more recent UK patient cohort, and investigate whether phenotypes are associated with 1-year treatment outcome.

METHODS:
12 specialist CFS/ME services (11 UK, 1 NL) recorded the presence/absence of 5 symptoms (muscle pain, joint pain, headache, sore throat, and painful lymph nodes) which can occur in addition to the 3 symptoms (post-exertional malaise, cognitive dysfunction, and disturbed/unrefreshing sleep) that are present for almost all patients. Latent Class Analysis (LCA) was used to assign symptom profiles (phenotypes). Multinomial logistic regression models were fitted to quantify associations between phenotypes and overall change in health 1year after the start of treatment.

RESULTS:
Baseline data were available for N=918 UK and N=1392 Dutch patients, of whom 416 (45.3%) and 912 (65.5%) had 1-year follow-up data, respectively. 3- and 4-class phenotypes identified in the previous UK patient cohort were replicated in the new UK cohort. UK patients who presented with 'polysymptomatic' and 'pain-only' phenotypes were 57% and 67% less likely (multinomial odds ratio (MOR) 0.43 (95% CI 0.19-0.94) and 0.33 (95% CI 0.13-0.84)) to report that their health was "very much better" or "much better" than patients who presented with an 'oligosymptomatic' phenotype. For Dutch patients, polysymptomatic and pain-only phenotypes were associated with 72% and 55% lower odds of improvement (MOR 0.28 (95% CI 0.11, 0.69) and 0.45 (95% CI 0.21, 0.99)) compared with oligosymptomatic patients.

CONCLUSIONS:
Adult CFS/ME patients with multiple symptoms or pain symptoms who present for specialist treatment are much less likely to report favourable treatment outcomes than patients who present with few symptoms.

Copyright © 2017 The Authors. Published by Elsevier Inc. All rights reserved.

KEYWORDS:
Chronic fatigue syndrome; Latent class analysis; Phenotypes; Symptom profiles; Treatment outcomes

PMID:
29275782
DOI:
10.1016/j.jpsychores.2017.11.007
 

Murph

:)
Messages
1,799
So this is Crawley, and perhaps I'm being too kind to her, but I think this is probably good research. (I'm sure someone will explain why I am being too kind).

The more symptoms you have the harder it is to get better? makes good sense to me. Also, chronic pain is famously resistant to curing so the fact the pain phenotype is hard to improve also makes sense.
 

Murph

:)
Messages
1,799
Biosystems. 2017 Dec 23. pii: S0303-2647(17)30243-5. doi: 10.1016/j.biosystems.2017.12.010. [Epub ahead of print]
A test of the adaptive network explanation of functional disorders using a machine learning analysis of symptoms.
Melidis C1, Denham SL2, Hyland ME3.
Author information
Abstract

The classification and etiology of functional disorders is controversial. Evidence supports both psychological and biological (disease) models that show, respectively, that functional disorders should be classified as one (bodily distress syndrome) and many (e.g., irritable bowel syndrome (IBS), fibromyalgia syndrome (FMS), and chronic fatigue syndrome (CFS)). Two network models (symptom network and adaptive network) can explain the specificity and covariation of symptomatology, but only the adaptive network model can explain the covariation of the somatic symptoms of functional disorders.

The adaptive network model is based on the premise that a network of biological mechanisms has emergent properties and can exhibit adaptation. The purpose of this study was to test the predictions that symptom similarity increases with pathology and that network connection strengths vary with pathology, as this would be consistent with the notion that functional disorder pathology arises from network adaptation. We conducted a symptom internet survey followed by machine learning analysis. Participants were 1751 people reporting IBS, FMS or CFS diagnosis who completed a 61-item symptom questionnaire.

Eleven symptom clusters were identified. Differences in symptom clusters between IBS, FMS and CFS groups decreased as overall symptom frequency increased. The strength of outgoing connections between clusters varied as a function of symptom frequency and single versus multiple diagnoses. The findings suggest that the pathology of functional disorders involves an increase in the activity and causal connections between several symptom causing mechanisms. The data provide support for the proposal that the body is capable of complex adaptation and that functional disorders result when rules that normally improve adaptation create maladaptive change.

KEYWORDS:
Irritable bowel syndrome; chronic fatigue syndrome; complexity; fibromyalgia; network

PMID:
29278731
DOI:
10.1016/j.biosystems.2017.12.010
 

dreampop

Senior Member
Messages
296
My original comment referred to another study @Murph posted on the same day (machine algorithim & network analysis).

However @Deepwater is exactly right about how BPS, especially Wessley has a history of justifiying psychosomatism of these kinds of studies.
 
Last edited:

alex3619

Senior Member
Messages
13,810
Location
Logan, Queensland, Australia
Guess what, people really sick, have more sick symptoms converging in fatigue and pain.
Shocking, absolutely shocking. Who would have thunk it? :eek:

Without modelling specific biochemical and physiological mechanisms this approach can at best be only suggestive.

I wonder what would happen if you did this will all oncological disorders merged in, or all infectious diseases, or all genetic disorders?

To be fair I have not read the paper. It may be that there is more to it than we can get from the abstract. However I really distrust symptom clusters that are not tied to pathology.

I do think this kind of thinking can lead to major mistakes, like the notion that depression is a disorder. Depression is a symptom, anything else is a category mistake. Throw in other general symptoms like fatigue and pain and you just create a big muddle of category mistakes. At some point you need to show close correlation, at least, between these symptoms and physiology, using established biochemical mechanism, or novel mechanisms if you run an appropriate study for them.

Machine learning is very powerful. Its greatest benefit however is when it is based on concrete measurable data, and reliable data at that. Even then you have to be careful. Machine learning can find patterns when they do not exist.
 

Deepwater

Senior Member
Messages
208
So this is Crawley, and perhaps I'm being too kind to her, but I think this is probably good research. (I'm sure someone will explain why I am being too kind).

The more symptoms you have the harder it is to get better? makes good sense to me. Also, chronic pain is famously resistant to curing so the fact the pain phenotype is hard to improve also makes sense.

Perhaps you are being too kind, @Murph, or perhaps I just have a suspicious mind, but I fear this research may not be as innocent as it appears. I don't know the first two authors but H. Knoop is a Dutch psychologist and leading proponent of the BPS model of CFS, so one would have to wonder about the purpose of this study from the authors' point of view.

Remember this old gem from Simon Wessely - “The greater the number of symptoms and the greater the perceived disability, the more likely clinicians are to identify psychological, behavioural or social contributors to illness….If the chronic fatigue syndrome did not exist, our current medical and social care systems might force us to invent it” (ref Ann Intern Med 2001:134:9S:838-, from Margaret Wlliams' current article reviewing Wessely's career) - or his claim to have "shown" that "the only determinant of outcome in this condition is strength of belief in a solely physical cause. . . " (ref letter of 1993 to Mansel Aylward of Department for Social Security, from same source).

The BPS school have also been extremely scathing about chronic pain, particularly in people with ME or FM. Because the cause of it is not immediately obvious, they have insinuated that our brains are manufacturing it by over-reacting to normal stimuli. I was lectured in this at an NHS physio clinic little more than a year ago.

Why do the authors rely on reported symptoms rather than test results? The results from such questions would lack objectivity even if properly elicited, but I'm afraid I've attended two UK ME clinics in my time; in both I was asked whether I had these kinds of symptoms, and in both the doctor answered for me - 'Yes. . . Yes. . .' in a disparaging tone before I had a chance to open my mouth, rather as if it is a practice in which they have been schooled and encouraged. And, no, they didn't actually look at my throat.
Another problem regarding the objectivity of the study is to know what the authors mean when they claim their patient group all have PEM. Whilst I wholly agree that post-exertional malaise is the cardinal sign of ME, it is poorly defined, if at all, by either NICE or Esther Crawley's "research" papers, so that (unless it is properly defined in the full version of this paper, which I haven't been able to access) then it could mean anything the authors wanted it to mean, including perfectly normal post-exertional fatigue. It may therefore be impossible to know whether some, or all, of the patients with fewer symptoms had any clinical illness at all.

I can see two possible uses for this study by the BPS school:
1) They can claim that the more entrenched the false illness belief/ hysteria (as evidenced by the number of symptoms reported), the more resistant the patient will be to treatment (note that they found women more likely to report more symptoms);
2) If scientific developments make this unavoidable, it could be used to allow them to take the credit for "discovering" that the sicker patients may actually have a biological illness after all (one can see hints in this direction in recent statements by Crawley and Hammond).
 

Countrygirl

Senior Member
Messages
5,429
Location
UK
I think, in layman's terms, Esther has made the shattering discovery that patients diagnosed with ME who actually have ME don't respond well to her 'treatment' (aka abuse), in comparision to another group of ME-diagnosed patients who..............don't have ME.

Give the girl a Damehood! Who'd have thunk it! :bang-head:


https://www.ncbi.nlm.nih.gov/pubmed/29275782

J Psychosom Res. 2018 Jan;104:29-34. doi: 10.1016/j.jpsychores.2017.11.007. Epub 2017 Nov 8.
Chronic fatigue syndrome (CFS/ME) symptom-based phenotypes and 1-year treatment outcomes in two clinical cohorts of adult patients in the UK and The Netherlands.
Collin SM1, Heron J2, Nikolaus S3, Knoop H3, Crawley E2.
Author information

Abstract
OBJECTIVE:
We previously described symptom-based chronic fatigue syndrome (CFS/ME) phenotypes in clinical assessment data from 7041 UK and 1392 Dutch adult CFS/ME patients. Here we aim to replicate these phenotypes in a more recent UK patient cohort, and investigate whether phenotypes are associated with 1-year treatment outcome.

METHODS:
12 specialist CFS/ME services (11 UK, 1 NL) recorded the presence/absence of 5 symptoms (muscle pain, joint pain, headache, sore throat, and painful lymph nodes) which can occur in addition to the 3 symptoms (post-exertional malaise, cognitive dysfunction, and disturbed/unrefreshing sleep) that are present for almost all patients. Latent Class Analysis (LCA) was used to assign symptom profiles (phenotypes). Multinomial logistic regression models were fitted to quantify associations between phenotypes and overall change in health 1year after the start of treatment.

RESULTS:
Baseline data were available for N=918 UK and N=1392 Dutch patients, of whom 416 (45.3%) and 912 (65.5%) had 1-year follow-up data, respectively. 3- and 4-class phenotypes identified in the previous UK patient cohort were replicated in the new UK cohort. UK patients who presented with 'polysymptomatic' and 'pain-only' phenotypes were 57% and 67% less likely (multinomial odds ratio (MOR) 0.43 (95% CI 0.19-0.94) and 0.33 (95% CI 0.13-0.84)) to report that their health was "very much better" or "much better" than patients who presented with an 'oligosymptomatic' phenotype. For Dutch patients, polysymptomatic and pain-only phenotypes were associated with 72% and 55% lower odds of improvement (MOR 0.28 (95% CI 0.11, 0.69) and 0.45 (95% CI 0.21, 0.99)) compared with oligosymptomatic patients.

CONCLUSIONS:
Adult CFS/ME patients with multiple symptoms or pain symptoms who present for specialist treatment are much less likely to report favourable treatment outcomes than patients who present with few symptoms.

Copyright © 2017 The Authors. Published by Elsevier Inc. All rights reserved.

KEYWORDS:
Chronic fatigue syndrome; Latent class analysis; Phenotypes; Symptom profiles; Treatment outcomes

PMID:
 

Countrygirl

Senior Member
Messages
5,429
Location
UK
https://www.ncbi.nlm.nih.gov/pubmed/29278731


Biosystems. 2017 Dec 23. pii: S0303-2647(17)30243-5. doi: 10.1016/j.biosystems.2017.12.010. [Epub ahead of print]
A test of the adaptive network explanation of functional disorders using a machine learning analysis of symptoms.
Melidis C1, Denham SL2, Hyland ME3.
Author information

Abstract
The classification and etiology of functional disorders is controversial. Evidence supports both psychological and biological (disease) models that show, respectively, that functional disorders should be classified as one (bodily distress syndrome) and many (e.g., irritable bowel syndrome (IBS), fibromyalgia syndrome (FMS), and chronic fatigue syndrome (CFS)). Two network models (symptom network and adaptive network) can explain the specificity and covariation of symptomatology, but only the adaptive network model can explain the covariation of the somatic symptoms of functional disorders. The adaptive network model is based on the premise that a network of biological mechanisms has emergent properties and can exhibit adaptation. The purpose of this study was to test the predictions that symptom similarity increases with pathology and that network connection strengths vary with pathology, as this would be consistent with the notion that functional disorder pathology arises from network adaptation. We conducted a symptom internet survey followed by machine learning analysis. Participants were 1751 people reporting IBS, FMS or CFS diagnosis who completed a 61-item symptom questionnaire. Eleven symptom clusters were identified. Differences in symptom clusters between IBS, FMS and CFS groups decreased as overall symptom frequency increased. The strength of outgoing connections between clusters varied as a function of symptom frequency and single versus multiple diagnoses. The findings suggest that the pathology of functional disorders involves an increase in the activity and causal connections between several symptom causing mechanisms. The data provide support for the proposal that the body is capable of complex adaptation and that functional disorders result when rules that normally improve adaptation create maladaptive change.

KEYWORDS:
Irritable bowel syndrome; chronic fatigue syndrome; complexity; fibromyalgia; network

PMID:

29278731

DOI:

10.1016/j.biosystems.2017.12.010
 

Learner1

Senior Member
Messages
6,305
Location
Pacific Northwest
We conducted a symptom internet survey followed by machine learning analysis. Participants were 1751 people reporting IBS, FMS or CFS diagnosis who completed a 61-item symptom questionnaire. Eleven symptom clusters were identified. Differences in symptom clusters between IBS, FMS and CFS groups decreased as overall symptom frequency increased.
This is ridiculous. What is the point here?

The patients are ill. These illnesses have a variety of triggers and underlying drivers. This does nothing to understand why each patient is ill or provide any insight to fix them.

You could take a group of patients with headache, swollen lymph nodes, and elevated temperature. Does this tell you what's actually the matter or how to treat it or how well equipped the patient is to fight the illness?

Could be meningitis, could be the flu or the common cold, could be food poisoning...
 

ljimbo423

Senior Member
Messages
4,705
Location
United States, New Hampshire
I immediately thought of the Cell Danger Response when I read this.

I thought the exact same thing. The CDR is a healthy adaptation the mito have to threats. It's only when the CDR is chronic that it becomes a problem and can create poor health or disease.

Robert Naviaux-

The cell danger response (CDR) is the evolutionarily conserved metabolic response that protects cells and hosts from harm.

It is triggered by encounters with chemical, physical, or biological threats that exceed the cellular capacity for homeostasis
.
http://www.sciencedirect.com/science/article/pii/S1567724913002390
http://www.sciencedirect.com/science/article/pii/S1567724913002390
Jim
 

ljimbo423

Senior Member
Messages
4,705
Location
United States, New Hampshire
I wasn't sure what a functional disorder was, so I looked it up.:)

Functional disorder definition-

A functional disorder is a medical condition that impairs the normal function of a bodily process, but where every part of the body looks completely normal under examination, dissection or even under a microscope. This stands in contrast to a structural disorder (in which some part of the body can be seen to be abnormal) or a psychosomatic disorder (in which symptoms are caused by psychological or psychiatric illness).
https://en.wikipedia.org/wiki/Functional_disorder
https://en.wikipedia.org/wiki/Functional_disorder
Jim
 

ljimbo423

Senior Member
Messages
4,705
Location
United States, New Hampshire
@Ijimbo423 Functional disorder in the UK is a term used by doctors when they believe that their patient's symptoms are 'all in their head' and they cast their patient into the waste bin.

Hi Countrygirl - There seems to be more than one definition of functional disorder-

functional disorder
functional disorder in Medicine
functional disorder n.
A physical disorder in which the symptoms have no known or detectable organic basis but are believed to be the result of psychological factors such as emotional conflicts or stress. Also called functional disease .
http://www.dictionary.com/browse/functional-disorder

Of course most doctors are going to call it a psychological problem because they don't have any other frame of reference. :rolleyes:

Things seem to be moving fast in CFS research though and with a little luck there will soon be a bio-marker and we can point our doctors in the right direction when that is found!:D:snigger:

Jim
 

Countrygirl

Senior Member
Messages
5,429
Location
UK
Hi Countrygirl - There seems to be more than one definition of functional disorder-

http://www.dictionary.com/browse/functional-disorder

Of course most doctors are going to call it a psychological problem because they don't have any other frame of reference. :rolleyes:

Things seem to be moving fast in CFS research though and with a little luck there will soon be a bio-marker and we can point our doctors in the right direction when that is found!:D:snigger:

Jim

I think that is saying the same thing Jim. A UK doctor would be quick to show you the door if they thought you had a stress related condition.
 

alex3619

Senior Member
Messages
13,810
Location
Logan, Queensland, Australia
I'm not sure how a machine learning is going to be better at pattern recognition than a normal scientist using conventional statistical methods and instinctive pattern finding and application of logic, could perhaps just be a marketing trick to get funding.
Given reliable objective data, machine learning can find patterns that are almost impossible for even most scientists to see. Even if those patterns do not actually exist. Given unreliable and subjective data, the chances of anything useful being found plummet. To be really useful they have to tie their findings to pathophysiology then confirm that experimentally.