Criticisms of DecodeME in the media - and responses to the criticisms

Utsikt said:
Could it not also have made patterns appear that aren’t really there because some non-patients had something else in common?
Click to expand...

I'm not an expert, but I don't think that is likely.

If there was a systematic error maybe there could be non-related patterns. For example if there were a lot of people with some unrelated genetic decease that could be mistaken for ME/CFS.

But if there is high error rate in the diagnosis for ME/CFS that would probably include a lot of people with various psychological illnesses such as burnout, depression and hypochondria. I don't think that would create systematic patterns in the data. It would just make the existing patterns more blurry. And I don't think blurry data would create spurious patterns. The size of the study and the statistical methods ensure that.
 
Last edited:
Liie said:
Isn't this criticism backwards? (Even if it were true, which if I have understood correctly it isn't.)
Click to expand...
Yes, it's backwards. The study results confirm the validity of what was done, that includes the diagnostic criteria, recruitment process, and analysis.

The identified regions are different from those found in GWAS of other conditions and seem to tell a coherent story about a disease initiated by some immunological factors, with the brain being the organ that appears to be most affected. That's what patients have been saying for decades. The findings don't look like random noise or stuff you would find due to some bias in the data. Even if bias is distorting the findings to some degree, it doesn't look significant.

It's always possible that some flaw will be recognized that reduces the credibility of the findings, but currently the appearance is of valid and reliable results. That said, I'm not a researcher.
 
Last edited:
Liie said:
But if there is high error rate in the diagnosis for ME/CFS that would probably include a lot of people with various psychological illnesses such as burnout, depression and hypochondria. I don't think that would create systematic patterns in the data. It would just make the existing patterns more blurry. And I don't think blurry data would create spurious patterns. The size of the study and the statistical methods ensure that.
Click to expand...
All of those conditions have genetic risk factors themselves. They are just as much physical even though they are often placed under the label «psychological».

So they would also pop up in the genetic hits if the cohorts were sufficiently contaminated.
 
Utsikt said:
All of those conditions have genetic risk factors themselves. They are just as much physical even though they are often placed under the label «psychological».

So they would also pop up in the genetic hits if the cohorts were sufficiently contaminated.
Click to expand...
Yes. If a condition only makes up a small amount of ME/CFS cases, but has a really high heritability. It could tip scales fast.
 
Liie said:
Isn't this criticism backwards? (Even if it were true, which if I have understood correctly it isn't.)

A high error rate in the diagnosis would lead to messy data which would weaken the patterns that the analysis uncovered.

But the analysis showed strong patterns.

If the data were messy but still showed strong patterns that would have meant that the patterns for a clean dataset would have been even stronger.

This criticism would have been relevant to make before the study, to warn the investigators about things they should improve to get a better result.
Click to expand...
Seems like you’re trying to apply logic to nonsense!
 
Liie said:
A high error rate in the diagnosis would lead to messy data which would weaken the patterns that the analysis uncovered.
Click to expand...

That is one argument - and valid.
But the other argument is also valid - that the 'ME/CFS/' cohort contained a significant number of people with some other problem with a genetic basis. The main argument against this is that the genes pulled out by DecodeME are not picked out in conditions likely to cause interference. The pain gene may be the same variant picked up in fibromyalgia but that would be a not unexpected overlap.
 
Kiristar said:
What about the heritabillity claim? "Heritability was 9.5% in the
@DecodeMEstudy . The heritability of depression is 40-50%, and 30-40% for anxiety disorders."
Click to expand...
You find similarly high rates in ME if you look at twin heritability studies.

And similarly modest ones to decodeME if you look at GWAS heritability in depression and anxiety.

Fundamentally he’s weaponising two very different ways to measure heritability. It’s like comparing degrees celcius to degrees farenheit.
 
What would Carson make of the observation that the following things are typical of ME/CFS but not typical of depression?

Having to lie flat at least some of the time
Delayed worsening of symptoms in response to activities
Rest in a quiet, darkened room improves symptoms.
Can have significant functional impairment and symptoms with normal mood, for long periods of time.
PEM episodes follow a very different time pattern than depressive episodes.

The healthcare system is oriented towards detecting certain symptoms typical of depression, but has not been trained to look for certain problems typical of ME. When one only looks at things like fatigue, sleep issues, cognitive impairment ME/CFS does resemble depression, but when other things are considered it begins looking different.
 
Last edited:
Utsikt said:
Carson in SMC.

Funny how he highlights depression when there were no hits on genes related to depression!
Click to expand...
Some depressive co morbidity is to be expected in long term illness. Open question as to whether it worsens non depressive features. My bet is that any study of , say, MS. or sarcoid will cover numerous with depressive states.
Depressive co-morbids in the absence of genetic signatures for depression strengthens the argument against depression as initiating factor at least. The more the merrier. For a Prof of NPs I would think this would be an obvious immediate observation.
I think the question as to constitution of cohort is valid and I wonder whether I as a sarcoid sufferer with a fatigue state illness well established before sarcoid diagnosis have shown the ME signature or not.
 
Utsikt said:
I wonder which mechanisms Carson proposes that the functional somatic syndromes work through if not the nervous system.
Click to expand...
Utsikt said:
I wonder which mechanisms Carson proposes that the functional somatic syndromes work through if not the nervous system.
Click to expand...
I think there is a difference between conceptualisation and what is going on bilogically in individuals. So, C might focus on stress handling and a few of Garner's Recovery crew might fit in there, but even the Recovery crew allow that the kind of biophysical aspects addressed by Alex Howard's clinic are varied , that treatment may be necessary and such aspects may be making people sick by direct mechanisms and as general stressors.
Why " ME is a FSD" rather than "FSD is found in ME"?
 
Someone better versed in the science than I might like to respond to Douglas Campbell under science's fb page post about the study?
 

Attachments

  • Screenshot_20250814_130648_Facebook.jpg
    Screenshot_20250814_130648_Facebook.jpg
    350.1 KB · Views: 61
You must log in or register to reply here.
Back
Top Bottom