Rethinking childhood adversity in chronic fatigue syndrome

[Cheshire]

James E. Clark, Sean L. Davidson, Laura Maclachlan, Julia L. Newton and Stuart Watson

Background: Previous studies have consistently shown increased rates of childhood adversity in chronic fatigue syndrome (CFS). However, such aetiopathogenic studies of CFS are potentially confounded by co-morbidity and misdiagnosis particularly with depression.

Purpose: We examined the relationship between rates of childhood adversity using two complimentary approaches (1) a sample of CFS patients who had no lifetime history of depression and (2) a modelling approach.

Methods: Childhood trauma questionnaire (CTQ) administered to a sample of 52 participants with chronic fatigue syndrome and 19 controls who did not meet criteria for a psychiatric disorder (confirmed using the Structured Clinical Interview for DSM-IV). Subsequently, Mediation Analysis (Baye’s Rules) was used to establish the risk childhood adversity poses for CFS with and without depression.

Results: In a cohort of CFS patients with depression comprehensively excluded, CTQ scores were markedly lower than in all previous studies and, in contrast to these previous studies, not increased compared with healthy controls. Post-hoc analysis showed that CTQ scores correlated with the number of depressive symptoms during the lifetime worst period of low mood. The probability of developing CFS given a history of childhood trauma is 4%, a two-fold increased risk compared to the general population. However, much of this risk is mediated by the concomitant development of major depression.

Conclusions: The data suggests that previous studies showing a relationship between childhood adversity and CFS may be attributable to the confounding effects of co-morbid or misdiagnosed depressive disorder.

https://sci-hub.cc/http://www.tandf...1080/21641846.2018.1384095?journalCode=rftg20

 

[user9876]

The problem is retrospective childhood adversity studies are unreliable because people can look back and find things if they try. Many people have childhood adversity and are not ill.

 

[markielock]

Agreed with @user9876 . What I respect and love about science is that people want to explore the many different angles of a topic and find true enjoyment in doing so. It gives us a thorough perspective on what's going on (and what isn't going on). At least they're adding to this body of research and adding a counter position to what's come before. I don't know about you, but even on an instinctive level it seems wrong to attribute CFS so 'directly' to childhood trauma, which other research papers seemed to do (making this comment from my memory, I may be incorrect over how enthusiastic certain researchers were, ha ha).

 

[Londinium]

The problem is retrospective childhood adversity studies are unreliable because people can look back and find things if they try. Many people have childhood adversity and are not ill.

Agreed - I’ve certainly had one GP trawl my past looking for incidents that could ‘explain’ my illness and only the fact that I’ve had a pretty charmed life meant they couldn’t find anything; many of my non-ill friends would have had something turned up had they faced a similar inquisition. However, I don’t think that is what this study is driving at, if I’m interpreting it correctly, because both ME/CFS patients and controls receive the same questioning: instead, it shows that patients and controls have similar rates of childhood adversity provided you screen out depression from the ‘ME’ patients. My knowledge of Bayes Rule is a bit rusty but if I’m following it correctly they seem to be saying that

• Of the c.35% of the population that have suffered childhood adversity, about a third will develop a major depressive disorder, which is roughly double the wider population’s prevelence of MDD at 16%
• Of those with childhood adversity and a major depressive disorder, 8% will get a diagnosis of CFS
• This is approximately four times the lifetime prevelence of CFS diagnoses amongst the wider population (at 2%, which IMHO screams over-/mis-diagnosis).
• However, amongst the two-thirds of those suffering childhood adversity but who don’t suffer from a depressive disorder, just under 2% get diagnosed with CFS, in line with the wider population.

(Worth noting that the study is small and thus the confidence interval for the stats above is likely to be wide. Also worth noting the next paragraph is my own speculations)

This suggests that childhood adversity itself doesn’t increase the risk of ME/CFS per se, contradicting previous findings. But that such adversity increases the risk of a major depressive disorder which in turn increases the chance of a *diagnosis* of CFS. Which could be because depression is a genuine risk factor for ME/CFS (perhaps via some kind of neuro-inflammatory route?), or (more likely, IMHO) because having depression and the fatigue symptoms associated with it are more likely to land you with an inappropriate CFS diagnosis.

 

[Woolie]

We've got ourselves into such a mess with childhood adversity.

Here are some of the other problems:

Open Interviewing (interviewer not blind to person's illness):
Obviously, this won't work. As @Londinium points out, if you question a patient enough, its not difficult to find something negative in their past. If you expect to find it, you will question harder till you find the evidence you seek.

But even systematic studies have problems.

Cherry picking:
Researchers will often collect a lot of different measures of adversity in the hope that if they throw enough mud, some of it'll stick. In addition to physical and sexual abuse, they might ask about neglect, feeling unloved, or feeling overly criticised, or having 'overly controlling' parents.

Of course, this sort of approach increases your chances of finding something significant just by chance alone.

Its also theoretically impoverished. You shouldn't just have a theory that 'bad stuff" in you childhood makes people ill - we don't care what it is - that's a pathetic theory. Your theory should predict that specific types of bad stuff will be important, whereas other types of bad stuff will not be. You should be looking for a pattern of results that fit your case, not just as many ticks a you can gather.

You can also see from the examples I gave above, that some of the 'adversity' types considered are almost self-contradictory. Parenting that is too loose, or too controlling, can these really be expected to affect people in the same way?

Demographic Factors:
The whole enterprise involves making value judgements about what a "healthy" upbringing is, based on today's standards.

But sixty years ago, parenting was a lot different. Most parents would tick some of the abuse boxes. So age must be controlled for. Also, certain types of adversity will be gender dependent. And parenting varies a lot with social class and culture. If you don't control properly for all these things in your control group, you'll just end up measuring demographics and not abuse at all.

Parenting was a pretty rudimentary business for most of history. Kids commonly lost their parents early, and many poor kids would have fended for themselves from a young age. Even rich kids might have been shipped off to boarding school quite young. Everybody hit their kids. If childhood hardship caused illness, few people would have been healthy.

Reporting biases:
This is the big whammy, as @user9876 mentioned. our recollection of past experiences, especially negative ones, is strongly influenced by our current state of mind. Even people with confirmed organic disease report more adverse life events than healthy individuals.

Also, patients with unexplained illnesses may be unusually likely to have searched their life history for possible explanations, which could further bias their reports.

 

[Woolie]

I'd add that even in depressed persons, we can't assume childhood adversity had anything to do with causing their depression. They're depressed, FFS, of course they put a negative slant on things - whether it be the past, present or future. That's literally what depression is.

 

[Misfit Toy]

Yes, in the narcissistic support groups people talk about AI diseases being caused by childhood trauma. But...it doesn't make sense because so many are well with childhood trauma.

I was sick from the moment I was born with being born in normal time but being so tiny I was a preemie (put in preemie ward and no fingernails) and then I was allergic to all formulas (MCAS) from the get go...then CVID as a child and IC...also as I had my urethra stretched by the time I was 5 twice due to constantly peeing. At this point, my childhood was awesome, so it doesn't add up for me.

 

[Woolie]

Its also theoretically impoverished. You shouldn't just have a theory that 'bad stuff" in you childhood makes people ill - we don't care what it is - that's a pathetic theory. Your theory should predict that specific types of bad stuff will be important, whereas other types of bad stuff will not be. You should be looking for a pattern of results that fit your case, not just as many ticks a you can gather.

Oh, and researchers, don't try to excuse yourself for not having a proper theory. You can't just say 'we didn't need a theory, because we were only doing an exploratory study'. That's two faced. You've chosen to look at childhood adversity - as opposed to, say, how many chicken pies the patient has eaten in their life, or some other random life event. There was a reason for that choice. It was because you had a theory that childhood adversity might cause illness.

It just wasn't a very well specified theory.

(Apologies for quoting self... but it was the best way to highlight the point).

 

[Hutan]

That's a great list @Woolie.

One more problem is the correlation is not causation issue. I think the famous Dunedin study which has followed a cohort of babies through, so far, into their forties, and has done a lot of good work, has fallen into this trap sometimes.

A poor family in a wealthy society is more likely to be stressed, in poor quality crowded housing, rely on cheap poor quality food, have issues with crime and drugs, have parents unable to work due to chronic illness, have smokers routinely in the house, have poor access to medical services and be a single-parent family than a family that is not poor.

So, a child in a poor family is more likely to have experienced childhood adversity. And that child is more likely to have health issues when they are an adult.

But the child has also experienced a whole range of biological factors that could account for that increased risk of poor health as an adult: poor nutrition, repeated infections with communicable diseases, exposure to cigarette smoke, exposure to lead from the car fumes from their house beside a busy street, bad genes from their chronically ill parents, mouldy cold houses and so on.

Some studies say that they control for those factors, but I doubt that the control is always perfect. Other studies seem very happy with a psychological link and don't think beyond that.

 

[Woolie]

A poor family in a wealthy society is more likely to be stressed, in poor quality crowded housing, rely on cheap poor quality food, have issues with crime and drugs, have parents unable to work due to chronic illness, have smokers routinely in the house, have poor access to medical services and be a single-parent family than a family that is not poor.

So, a child in a poor family is more likely to have experienced childhood adversity. And that child is more likely to have health issues when they are an adult.

But the child has also experienced a whole range of biological factors that could account for that increased risk of poor health as an adult: poor nutrition, repeated infections with communicable diseases, exposure to cigarette smoke, exposure to lead from the car fumes from their house beside a busy street, bad genes from their chronically ill parents, mouldy cold houses and so on.

Absolutely, good point. Its a total and utter discombobulated mess.

Social class is the elephant in the room in any study looking at lifestyle/health practices and health outcomes. The health practices that are supposed to be linked to better health outcomes - cheese, red wine, coffee - are also often those heavily associated with posh people. That makes me super suspicious.

Many studies claim to "factor out' the effects of social class, but I think what they do is generally not adequate to achieve this. They measure social class in some appropriate way and then take out some of the variance tied to that measure. But even if the proxy measure is a really good one, its still only going to remove small proportion of the actual variance due to social class. Just because no measure captures anything perfectly. The rest of the variance will remain.

 

[Snow Leopard]

The problem is retrospective childhood adversity studies are unreliable because people can look back and find things if they try. Many people have childhood adversity and are not ill.

Yes, the danger of retrospective studies is people try to look for "reasons and causes" for things. The bias is greater when there is a cultural belief that xyz is a cause.
I dare say this childhood adversity reporting bias may be increased amongst people suffering from depression - the hypothesis being the very same group of people may report less adversity when not depressed.

Even Simon Wessely has said that retrospective studies are junk.

 

[Forbin]

I can't say for all studies, but the highly publicized 2009 Reeves CDC study* used the Childhood Trauma Questionnaire (CTQ).

There are CTQ sample questions posted here (although I'm not certain that they are from the "official" CTQ):
http://childhoodtraumarecovery.com/tag/childhood-trauma-questionnaire-sample-questions/

This CTQ does not look like it's measuring "I never got a pony" type of trauma.

I scored ZERO on these questions, but that didn't keep me from getting ME/CFS.


2009 Reeves CDC study
*Childhood trauma and risk for chronic fatigue syndrome: association with neuroendocrine dysfunction.
http://www.researchgate.net/publica...e_Association_with_Neuroendocrine_Dysfunction

 

[Woolie]

I can't say for all studies, but the highly publicized 2009 Reeves CDC study* used the Childhood Trauma Questionnaire (CTQ).

Yes, you're quite right. I should have said, I'd already moved on the generalisations about the whole enterprise without explaining.

Still, some of the CTQ questions are a bit, well grey. Things like whether your parents demeaned you.

 

[Forbin]

I think also, with this CTQ, these are deeply personal questions that you probably have to be highly motivated to answer honestly to if they apply to you. Positive results could simply reflect the fact that people desperate to find a "cause" for their disease are more motivated to answer the questions honestly than controls, rather than indicate that there is any difference in frequency between patients and controls.

 

[Snow Leopard]

This CTQ does not look like it's measuring "I never got a pony" type of trauma.

Did you feel you were not close to your family, that they did not support you and that they did not love you or regard you as special?

AKA My parents never bought me a pony!!!!!!

The actual questions and scoring of the Bernstein CTQ are discussed here:
https://www.researchgate.net/post/Does_anyone_have_childhood_Trauma_Questionnaire_scoring_manual

 

[user9876]

Agreed with @user9876 . What I respect and love about science is that people want to explore the many different angles of a topic and find true enjoyment in doing so. It gives us a thorough perspective on what's going on (and what isn't going on). At least they're adding to this body of research and adding a counter position to what's come before. I don't know about you, but even on an instinctive level it seems wrong to attribute CFS so 'directly' to childhood trauma, which other research papers seemed to do (making this comment from my memory, I may be incorrect over how enthusiastic certain researchers were, ha ha).

I think what this paper seems to be doing (but I've not read it) is attributing depression to childhood adversity. What I can believe is the case (in the UK) is childhood adversity reflects being able to cope with chronic illness. I say in the UK because I suspect childhood adversity may relate to poverty in many cases (but I've not looked at the questions). If someone doesn't have money or family in a position to help out then they can have a very hard time with the DWP trying to get benefits and the stress that they create. Then there are issues such as poor housing/bad landlords etc.

So I think any model that tries to look at childhood adversity needs to control for socioeconomic factors.

 

[Cheshire]

I completely agree with all the limitations of childhood trauma causality for ME, depression, or any other diseases that you've talked about.

But I think this study, even with the bias associated with the search of childhood trauma, as they couldn't find more occurrence of it in CFS patients without depression than in normal population, could be to a certain degree a useful rebuttal of some of the BPS model.

But maybe I'm plain wrong and it's too biased to even achieve this? (naïve question)

 

[Woolie]

No, you're right, @Cheshire.

It shows that on top of all the usual limitations of this type of design, depression is a further problematic confounding factor that may be driving positive effects.

 

[Londinium]

But I think this study, even with the bias associated with the search of childhood trauma, as they couldn't find more occurrence of it in CFS patients without depression than in normal population, could be to a certain degree a useful rebuttal of some of the BPS model.

Agreed, that's how I read this as well. It's effectively saying that any study that claims childhood adversity is linked to ME/CFS without controlling for co-morbid lifetime major depressive disorder as a confounding factor is unreliable. Which, I think, is basically all of them. (No surprise, if you're a researcher who think ME/CFS is basically an anxiety disorder, why would you bother to filter out depression? Your supposition about the aetiology of ME/CFS then creates a vicious circle in which your trial design is such that it confirms your hypothesis/bias).

 

[RuthT]

Agreed, that's how I read this as well. It's effectively saying that any study that claims childhood adversity is linked to ME/CFS without controlling for co-morbid lifetime major depressive disorder as a confounding factor is unreliable. Which, I think, is basically all of them. (No surprise, if you're a researcher who think ME/CFS is basically an anxiety disorder, why would you bother to filter out depression? Your supposition about the aetiology of ME/CFS then creates a vicious circle in which your trial design is such that it confirms your hypothesis/bias).

The hermeneutical circle is very hard to escape & v few seem to be aware of the limitations this brings. Assumptions = conclusion. What about requiring work/research to disprove one’s own hypotheses: thought that was part was an essential part of the scientific/philosophic method.