• 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.

(Ongoing UK study) "Investigating the epidemiology of CFS/ME in children using the ALSPAC cohort"

Dolphin

Senior Member
Messages
17,567
I think it would have been much more interesting and the results would likely be more reliable if it involved some biological testing.

Extract:
"The CFS/ME research community will benefit from bringing new leaders of research from obesity, activity, sleep, psychology, psychiatry in to CFS/ME"

-----
http://gtr.rcuk.ac.uk/project/92F6B9E6-B13E-4BC5-AD74-F2245F8B5D54

Funded Value: £295,826

Funded Period: Oct 13 - Mar 16

Funder: MRC

Project Status: Active

Project Category: Research Grant

Project Reference: MR/K020269/1

Principal Investigator: Esther Crawley

Health Category: Other (100%)

Research Activity: 2.4 Surveillance and distribution (100%)

Investigating the epidemiology of CFS/ME in children using the ALSPAC cohort.


Lead Research Organisation: University of Bristol Department Name: Social Medicine Go back


Overview

Organisations

People


Abstract


CONTEXT: Chronic Fatigue Syndrome/Myalgic Encephalomyelitis (CFS/ME) in adolescents is relatively common and causes significant suffering.

1% of secondary school children miss one day a week because of CFS/ME, and 0.1% are so severely affected they are unable to attend school at all. More than half of children are bed bound at some stage, losing on average one academic year of schooling. Despite the importance of paediatric CFS/ME to those affected, their families and society, little is known about how common CFS/ME is in adolescents, what causes it and what the chances are of an adolescent recovering without treatment. In addition, little is known about the different types of CFS/ME in adolescents. If there are different types, these may suggest different causes and may need different treatments. So far, researchers have been hampered in their research in to the causes of CFS/ME because this needs reliable information about children before they develop CFS/ME.These type of studies also need to be large enough to answer these types of questions. This lack of knowledge prevents the development of interventions to prevent or treat paediatric CFS/ME in adolescents. The MRC understands the importance of understanding the mechanisms of CFS/ME in children and knowing more about the different types, and have asked researchers for studies to investigate these areas in the recent CFS/ME highlight notice.


The Avon Longitudinal Study of Parents and Children (ALSPAC) provides a unique opportunity to investigate the risk factors for CFS/ME. It has enough data so that we can define CFS/ME at 3 ages (13, 16 and 17). Information on possible risk factors has been collected before the onset of CFS/ME from pregnancy to adolescence, and the number of children is sufficiently large that there is enough power to explore the causes of CFS/ME in adolescence in unprecedented detail.


AIM: to describe CFS/ME in adolescence in order to understand more about the mechanisms that cause it and develop ideas on how to treat and prevent CFS/ME in this age group.


OBJECTIVES

1. Find out how common CFS/ME is at 15 and 17 and whether it is more common in girls or boys. Describe whether CFS/ME in adolescents gets better if untreated.

2. Investigate the different types of CFS/ME in adolescents and whether it overlaps with chronic pain disorders which have similar features.

3. Investigate risk factors for CFS/ME at 13, 16 and 17 years. We will investigate factors that have been shown to be associated with CFS/ME in either adults or children such as activity, patterns of sleep, anxiety and depression and see whether they increase the risk of developing CFS/ME or are secondary to it.

4. Investigate "maintaining factors". We will compare children who have persistent fatigue between the time points with those that recover, to understand whether there is a difference in risk factors between the two groups.


POTENTIAL APPLICATIONS AND BENEFITS:

Clinicians and commissioners need to know how common adolescent CFS/ME to be able to plan CFS/ME services. Knowing whether CFS/ME gets better without treatment is important for clinicians, adolescents with CFS/ME and their carers who need to decide whether to have treatment.

Information on the different types of CFS/ME is important as it may help us understand more about the mechanisms involved in different types. If there are different types of CFS/ME, it is important that researchers know this so they can study the same type of illness if they need to. Understanding what the risk factors are for CFS/ME should help researchers know more about the mechanisms that cause CFS/ME and prevent recovery. This should help researchers develop better treatment and prevention trials.


Technical Summary


We will use data ALSPAC, a birth cohort (www.alspac.bris.ac.uk).

OBJECTIVE 1. Describe the prevalence and persistence of CFS/ME in adolescence We will estimate the prevalence (total and by gender) of CFS/ME at 16 and 17 years by identifying adolescents with CFS/ME at 16 years. We will combine data with prevalence at 13 years and define adolescents with persistent fatigue at > 2 time points. We will examine the receiver-operator characteristics curves, of the Chalder fatigue scale to determine a threshold for severe fatigue.


OBJECTIVE 2. Investigate the heterogeneity of CFS/ME in adolescence We will use latent class analysis (fitted with MPlus software) to identify symptom based phenotypes of CFS/ME at 17, compare these with a specialist clinic cohort and investigate whether associations of risk factors for CFS/ME differ between phenotypes. We will describe how many 17 year olds with CFS/ME fit the definition for chronic pain syndrome.


OBJECTIVE 3. Investigate risk factors for incidence of CFS/ME at 13,

16 and 17 years

We will investigate risk factors for CFS/ME at 13, 16 &17 years. We will use: logistic regression to relate each of the potential risk factors adjusted for age and sex; directed acyclic graphs (DAGs) to identify confounding and mediating factors; fractional polynomials to investigate the linearity of the relationship between activity and CFS/ME; linear regression to examine the continuous outcome (Chalder fatigues score) at age 16. For exposures measured at multiple times, we will create categorical variables coded as 0, 1, 2 or 3+ episodes, identify critical periods that are biologically plausible and compare regression models using likelihood ratio tests.


OBJECTIVE 4. We will use logistic regression to investigate predictors of persistent fatigue Missing data: Where multiple imputation is appropriate, we will generate 50 imputed datasets, and combine estimates using Rubin's rules and we will explore whether weighting is helpful


Planned Impact


This study will tackle two research areas that the MRC considers to be "important and tractable for research" according to the recent MRC highlight notice. Most of this research grant will be used to "improve the unerstanding of the mechanisms that lead to the early onset of the disease" by investigating risk factors for CFS/ME at 13, 16 and 17. We will also use this study to "improve the sub-phenoytping and stratification of CFS/ME" in adolescence as we agree that this will eventually lead to "better targeting of treatment".


The main research outputs will be: prevalence data for CFS/ME at 16 and 17, information on the natural history of CFS/ME and a detailed understanding of the risk factors for CFS/ME. Additional outputs will be exploration of the phenotypes at 17 and analyses of the appropriate threshold for the Chalder Fatigue Scale used in the diagnoses and treatment of adolescent CFS/ME in the UK. The main beneficiaries are adolescents with CFS/ME and their families/carers. Other beneficiaries are, policy makers, commissioners, clinicians and clinical service providers, CFS/ME researchers and the wider research community.


Adolescents with CFS/ME (and their families) will benefit:

a) In the short term (2014-'15) by increased awareness of how common CFS/ME and what the natural history is. Understanding the prevalence is helpful for adolescents and their families. Understanding the natural history will enable adolescents and their carers make informed choices about treatment.

b) In the medium/long term (2018-2022) understanding the aetiological risk factors and maintaining factors will enable us to understand more about disease mechanisms in CFS/ME. This should lead to the development and evaluation of new treatments, before they are assessed in large-scale clinical trials. Understanding what the risk factors are for the development of CFS/ME will enable us to explore whether strategies to prevent the development of CFS/ME are feasible before testing them in large scale trials. Because the health resource use and societal costs of this illness are so large, interventions to prevent the development of CFS/ME or improve treatment outcomes in children are likely to have a significant impact on the nation's wealth.


Paediatricians and clinicians that either work in specialist services or see children with CFS/ME in paediatric services including paediatric physiotherapists, occupational therapists and psychologists will benefit in the short term from:

1. understanding how common CFS/ME is, and which children are most likely to have CFS/ME. This will enable clinicians and commissioners to target services more appropriately.

2. Understand what the natural history is and what maintaining factors prevent children with CFS/ME getting better. This will enable clinicians to adapt treatment more appropriately.

3. If we identify modifiable risk factors, clinicians will be able to target these in treatment.


In the long term, identifying risk factors and the mechanisms of adolescent CFS/ME may lead to more effective different treatments or strategies to prevent the development of CFS/ME.


The CFS/ME research community will benefit from bringing new leaders of research from obesity, activity, sleep, psychology, psychiatry in to CFS/ME as well as training a post-doctoral researcher who will be encouraged to apply for a post-doctoral fellowship in CFS/ME at the end of the study.

The wider research community will benefit from the collaborations that will result.
---


Organisations


University of Bristol, United Kingdom (Lead Research Organisation)


People


Esther Crawley (Principal Investigator)

Jonathan Sterne (Co-Investigator)

Jonathan Tobias (Co-Investigator)

Carol Joinson (Co-Investigator)

Kate Tilling (Co-Investigator)

Charlotte Elizabeth Clark (Co-Investigator) Paul Gringras (Co-Investigator) Stephen Stansfeld (Co-Investigator) Andrew Robert Ness (Co-Investigator) Peter S Blair (Co-Investigator) Glyn Lewis (Co-Investigator) Julian HamiltonShield (Co-Investigator)
 
Last edited:
Messages
13,774
Are they diagnosing CFS just on the basis of Chalder Fatigue Scores?

The lack of specifics makes it hard to say much, but this doesn't look great.
 

Bob

Senior Member
Messages
16,455
Location
England (south coast)
...there is enough power to explore the causes of CFS/ME in adolescence in unprecedented detail.
Can this study really determine the 'causes' of CFS/ME if no biological investigations are involved?
As usual, they are conflating 'cause' with 'association'.
No doubt they will report that certain cognitive and behavioural associations (i.e. "activity, patterns of sleep, anxiety and depression") are precipitating and perpetuating factors, which they will then claim are the 'causes' of CFS.

Why, oh why, have the MRC given funds?
 
Last edited:

user9876

Senior Member
Messages
4,556
We will combine data with prevalence at 13 years and define adolescents with persistent fatigue at > 2 time points. We will examine the receiver-operator characteristics curves, of the Chalder fatigue scale to determine a threshold for severe fatigue.

I'm pretty sure this doesn't make sense as the Chalder fatigue questionnaire is a random set of questions about fatigue rather than a scale. The questions themselves are quite confusing and there is an arbitrary weighting between mental and physical fatigue.

But more importantly RoC curves are about classification but you can't classify someone as having ME or not based on a fatigue survey. A whole range of different things such as ruling out other diseases needs to be taken into account to perform a classification and say yes or no. I feel it is quite dangerous to suggest that the CFQ could be used as the input to a classifier.

OBJECTIVE 2. Investigate the heterogeneity of CFS/ME in adolescence We will use latent class analysis (fitted with MPlus software) to identify symptom based phenotypes of CFS/ME at 17, compare these with a specialist clinic cohort and investigate whether associations of risk factors for CFS/ME differ between phenotypes. We will describe how many 17 year olds with CFS/ME fit the definition for chronic pain syndrome.

My guess is this is run a series of questionnaires about symptoms and see how they cluster. I don't understand how it helps to cluster symptoms from a questionnaire without an understanding of mechanism we have no idea whether they have any meaning. If they were to compare these to blood tests, b cell functioning, NK cells, response to exercise such as in the Light's study it might become interesting.


OBJECTIVE 3. Investigate risk factors for incidence of CFS/ME at 13,

16 and 17 years

We will investigate risk factors for CFS/ME at 13, 16 &17 years. We will use: logistic regression to relate each of the potential risk factors adjusted for age and sex; directed acyclic graphs (DAGs) to identify confounding and mediating factors; fractional polynomials to investigate the linearity of the relationship between activity and CFS/ME; linear regression to examine the continuous outcome (Chalder fatigues score) at age 16. For exposures measured at multiple times, we will create categorical variables coded as 0, 1, 2 or 3+ episodes, identify critical periods that are biologically plausible and compare regression models using likelihood ratio tests.

They plan to use complex stats techniques to try and find stuff but without thinking about the basic structure of data from things such as the CFQ probably the sf36 and HAD scales. If your working that hard with complex stats tests you are probably not finding real cause and effect or confounding factors especially from relatively small datasets. Stats should be used to find interesting stuff that help generate models that can then be tested in further work they don't explain.

This study will tackle two research areas that the MRC considers to be "important and tractable for research" according to the recent MRC highlight notice. Most of this research grant will be used to "improve the unerstanding of the mechanisms that lead to the early onset of the disease" by investigating risk factors for CFS/ME at 13, 16 and 17. We will also use this study to "improve the sub-phenoytping and stratification of CFS/ME" in adolescence as we agree that this will eventually lead to "better targeting of treatment".

I don't see how this research could help "improve the understanding of mechanisms" or even "improve sub-phenotypng" although this latter one is more arguable. Do the reviewers actually read stuff?

a) In the short term (2014-'15) by increased awareness of how common CFS/ME and what the natural history is. Understanding the prevalence is helpful for adolescents and their families. Understanding the natural history will enable adolescents and their carers make informed choices about treatment.
The MRC could help children and carers make more informed choices by insisting White released the PACE data. And that wouldn't cost almost 300000.

b) In the medium/long term (2018-2022) understanding the aetiological risk factors and maintaining factors will enable us to understand more about disease mechanisms in CFS/ME. This should lead to the development and evaluation of new treatments, before they are assessed in large-scale clinical trials. Understanding what the risk factors are for the development of CFS/ME will enable us to explore whether strategies to prevent the development of CFS/ME are feasible before testing them in large scale trials. Because the health resource use and societal costs of this illness are so large, interventions to prevent the development of CFS/ME or improve treatment outcomes in children are likely to have a significant impact on the nation's wealth.

I believe aetiological refers to causal yet nothing in this study looks at causality its just playing with stats methods that show correlation. At best they may find leading and lagging indicators. Given EC tells patients that CFS is time limited and she can cure it with a coloring chart (even when a child is still ill) then I worry about her thoughts on maintaining factors. Maybe she is just looking for ways to justify recommending sick children are locked up in psychiatric wards.

I suspect health resource use for ME are quite low. The significant impact on the nations wealth would be to do the pre-competative research into the biology of the disease so that patients can be treated and drug companies can look at making drugs. The problem is patients need to raise money for good research whilst the MRC fund this rubbish.
 

alex3619

Senior Member
Messages
13,810
Location
Logan, Queensland, Australia
On any large data set you can can find clustering by chance. It always needs further investigation. Of course the point made about association and causation is also critical - how many times must it be stressed that association does not equal causation?
 

alex3619

Senior Member
Messages
13,810
Location
Logan, Queensland, Australia
Yes, the conclusion had undoubtedly already been written. All that remains is to fill in the actual figures and find ways to make the data support that conclusion.

The modus operandi of huge areas of psychogenic research is to find supporting evidence. Its not to test the hypothesis. Alternative explanations get ignored. This isn't even science when this happens, its called nonscience, and even pseudoscience. This methodology was mostly put to rest in the mid-twentieth century, with the demise of logical positivism, but not, apparently, in psychiatry.
 

Dolphin

Senior Member
Messages
17,567
I'm not that interested in symptom phenotypes without checking the underlying biology. The intensity of my individual symptoms fluctuate. I don't think clusters at any particular point in time likely are that solid/reliable. I remember Esther Crawley previously published a study based on phenotypes at one moment in time.
 

alex3619

Senior Member
Messages
13,810
Location
Logan, Queensland, Australia
There is unpublished research I was a part of, primarily as a test subject, which shows massive shifts in symptomology. These can be rapid. Indeed, we experience this from time to time. Snapshots have a purpose, but they don't capture the dynamics of what is wrong.
 

Simon

Senior Member
Messages
3,789
Location
Monmouth, UK
I think the the idea of using a large prospective study, with a rich set of data, to investigate CFS is a good one. I'm a little concerned about how the study might be executed though. It's not clear that how they will confirm diagnosis of a difficult-to-diagnose illness, or whether they will make use of any of biosamples or physiological data collected by the ALSPAC cohort. And they appear to be looking at 'maintaining' factors in chronic fatigue rather than CFS.

As others have pointed out, if the drivers of CFS are biological, and they don't collect biological data, they will end of sifting through a lot of data with only marginal relevance to the illness - which would be a lost opportunity.

"Children of the 90s" /ALSPAC cohort
It's worth looking at the cohort they used:
  • Children of the 90s is a long-term health research project that started in the early 1990s. It recruited more than 14,000 pregnant women who were due to give birth between April 1991 and December 1992. These women and their families have been involved ever since.
  • All these methods provide a huge source of information on things that influence people’s health and well-being across the course of their lives. So far we have a million biological samples (such as blood, urine, hair, teeth) and answers to 400 million questions — a billion items of information altogether!
There were clinic visits by participants at approximate ages 7,8,9,10,11,12,13,4,16 &18. Not all children attended all clinics, with 8,000 making the first one down to 5,000 by age 18+. They also have DNA for 10,600 children and cell lines (immortalising the DNA to enable more sampling) for 7,100 children.

upload_2014-7-7_19-6-39.png


Since they are looking at 16/17 year olds, they have a maximum sample of 12,500 people (data). Assuming 1% of these have CFS (or had it at 13) that would give a sample of 125 patients, approx 13 of whom would have been bedbound/very severe [using their estimates of 1% prevalence, 0.1% too ill to attend at all). That would actually be a pretty respectable sample.

Diagnosis
Since the children were born in 91/92, they would be 16 in 2007/9. I'm not sure how they the CFS diagnosis will be asssigned eg if this was recorded at a clinic or if they will use questionnaires to retrospectively ask if they had CFS (the approach used by earlier birth cohort studies).

It's possible that some patients were diagnosed in CFS clinics were operating in Avon at that time (it's an Avon cohort), but given the emphasis in the study proposal on measuring 'recovery without treatment', it looks like most are not expected to have attended a CFS clinic.

The problem is that CFS is a diagnosis of exclusion, and if patients haven't been properly tested/assessed to rule out other illnesses that puts the findings in doubt too as they may not relate to CFS. Maybe the study has a way to provide clear diagnosis, but there is nothing in the information here that says so.

Chronic Fatigue vs CFS

Objective 4. Investigate "maintaining factors". We will compare children who have persistent fatigue between the time points with those that recover, to understand whether there is a difference in risk factors between the two groups.
This does seem to be confusing chronic fatigue with CFS.

A Simon Wessely study on chronic fatigue in adults found that most adults attending GP surgeries with chronic fatigue attributed their fatigue to psychosocial factors

All particpants were asked what cause they attributed to their fatigue:
- 40% cited psychosocial causes eg work stress
- 17% cited psychological causes
- 15% cited physical causes
- 20% didn't answer this question, 7% cited other causes and 2% cited pregnancy.
Only 0.25% cited CFS.
http://forums.phoenixrising.me/inde...-90s-that-keep-cropping-up.10274/#post-183825
A chronic fatigue sample is likely to be made up of overwhelmingly non-CFS patients, so it's hard to see the relevance of 'maintaining' factors for Chronic Fatigue to CFS.

The study says they will use the Chalder fatigue scale. The questionnaires used (which ones not stated) covered fatigue at age 13 and 16 ie just 2 time points, so it's not clear how they will do this:
we will define adolescents with persistent fatigue at > 2 time points
But then it is just a brief description of the study.


Biology vs psychological/behavioural factors
The ALSPAC cohort has a lot of physiological data as well as biosamples [as above] and presumably clinical data on other illnesses subjects have too. Yet the proposal seems to focus on those factors studied in previous birth cohorts, primarily behavioural/psychosocial factors eg activity (though it appears they used activity monitors at ages 13 & 16, as well as an activity diary at age 10 which should reduce self-report bias).

The CFS/ME research community will benefit from bringing new leaders of research from obesity, activity, sleep, psychology, psychiatry in to CFS/ME
Might have been nice to draw in, say, immunologists and endocrinologists, neurologists even ahead of obesity researchers. It looks unbalanced to me, for a study that puts so much weight on mechanisms:
Most of this research grant will be used to "improve the understanding of the mechanisms that lead to the early onset of the disease" by investigating risk factors for CFS/ME at 13, 16 and 17.

Other biological data they might like to include:
- Sudden vs gradual onset
- Any illnesses/vaccinations etc that appeared to trigger the illness
For adolescents, I would have thought glandular fever would be a key one to investigate, as something like 10% of adolescent glandular fever cases go on to CFS - and GF diagnosis would be documented in medical records for verification.
- while the sample is probably too small to do much in the way of DNA studies, it might be interesting to follow-up the Dubbo study findings of that those who developed CFS post GF were more likely to have highly-active versions of interferon-gamma genes. This would be a relatively easy way to follow-up on the Dubbo findings, the sort of collaborative work that would allow this new study build on existing research on biological findings in CFS.

What is 'better'/'recovery'?
The study outline said:
Knowing whether CFS/ME gets better without treatment is important for clinicians, adolescents with CFS/ME and their carers who need to decide whether to have treatment.
That's very true, but what definition of 'recovery' or 'better' will the study used? The principal investigator erroneously claimed the PACE trial showed a 40% recovery rate, but even the 22% recovery rate later claimed by the PACE trial doesn't match real world recovery (13% of PACE participants met either fatigue or function 'recovery' criteria - or both - before receiving any treatment).
 
Messages
43
This study is a waste of money - spending scare resources on a psychiatric study of prevalence and risk factors whereby the study will be biased by the bias of the investigators who will use an out-dated Oxford Criteria , risk factors will be drawn from biased previous studies done by psychiatrists, and data collection will be biased by how risk factors are understood and defined.

Knowing how many children suffer pain or fatigue wont reveal much - and we may just find that due to the biases of this study far more children are deemed CFS patients requiring CBT then first thought.

Patients and ME supports groups MUST make their voice heard by writing to the MRC to object to this use of funds when biological factors are not funded.

30-40 years of psychiatry has not helped us understand ME/CFS much - indeed its diverted attention away from biological and immunological factors that require careful study.

No point complaining on here: action must be taken to show the MRC patients are not happy with how tax payers money is spent.
 

Dolphin

Senior Member
Messages
17,567
This study is a waste of money - spending scare resources on a psychiatric study of prevalence and risk factors whereby the study will be biased by the bias of the investigators who will use an out-dated Oxford Criteria , risk factors will be drawn from biased previous studies done by psychiatrists, and data collection will be biased by how risk factors are understood and defined.

Knowing how many children suffer pain or fatigue wont reveal much - and we may just find that due to the biases of this study far more children are deemed CFS patients requiring CBT then first thought.

Patients and ME supports groups MUST make their voice heard by writing to the MRC to object to this use of funds when biological factors are not funded.

30-40 years of psychiatry has not helped us understand ME/CFS much - indeed its diverted attention away from biological and immunological factors that require careful study.

No point complaining on here: action must be taken to show the MRC patients are not happy with how tax payers money is spent.
10-15 years ago, I remember some individuals and patient groups in the UK were calling for epidemiological research. I'm not saying this had anything to do with this study, but I thought I'd make the point now I made then that epidemiological research is very often done by those of the CBT/rehab school of thought.

One of the reasons people seemed to want epidemiological research is because it would demonstrate the scale of the problem and the need for biological research. However, it is well-accepted that a condition like Fibromyalgia is quite common, probably affecting in the region of a million people in the UK (possibly a lot more), but that doesn't mean it gets a lot of biomedical research in the UK. We already have an idea of the scale of the problem.

Some of these people and other people said it was hard to plan services without knowing exactly how many people have it. However, epidemiological research in CFS tends to involve questionnaires and so doesn't necessarily give us very accurate figures. If one develops some diagnostic tests first, it makes looking for cases much easier (don't just work by exclusion) and it should make it more accurate.