Preprint Initial findings from the DecodeME genome-wide association study of myalgic encephalomyelitis/chronic fatigue syndrome, 2025, DecodeMe Collaboration

[Adrian]

My understanding is that there are concerns around quality and methods of sampling from these sources. DecodeME specifically used the same sampling technology as the UK Biobank to avoid those kind of issues when using the Biobank for controls.

Is there something that could be done statistically to see if the data was consistent given a higher degree if error? It may depend on whether the sampling methods are biased or errors at random?

 

[Evergreen]

[Since this is relevant to both this thread and the media thread, I'm posting it in both threads.]

I’ve been trying to work out the strongest arguments against the criticism that participants were not diagnosed by a medical professional.

Here’s a recap of the criticism:

[Reuters] Scientists who were not involved in the study said using volunteers who self-reported chronic fatigue syndrome rather than restricting participation to those with a diagnosis from a medical professional somewhat weakened its conclusions. https://www.reuters.com/business/he...ked-with-chronic-fatigue-syndrome-2025-08-06/

[Carson SMC] Diagnosis was done by questionnaire and this has a significant error rate. The National Institute of Health in the US has suggested diagnostic error rates of around one third.

[bobbler's transcript of what Lucy Beresford said is in post #130 above] I think what people are very worried about with this particular piece of research is that I’m quite a lot of the people who were part of the study had actually self-reported this condition so they hadn’t necessarily had a formal diagnosis and that just unfortunately kind of dilutes the effect of this research.

It’s true that DecodeME participants were not recruited through health professionals, but speculation that they did not have a formal diagnosis of ME/CFS is...purely speculative. I don’t think it is a strong response to say that cases were all diagnosed by a health professional, because we cannot verify that. We only know that they answered “Yes” to this question:

Have you been given a diagnosis of ME, CFS, ME/CFS or CFS/ME by a health professional?

When people are objecting to self-report (ah, the irony), then countering with another self-report won't work.

I think the strongest response is something like:

• Good-quality GWAS studies need huge numbers of participants in order to have high enough statistical power to detect robust associations that would be missed in smaller studies.
• It would not have been possible to recruit 26K participants with ME/CFS indirectly through health professionals.
• By recruiting directly, DecodeME could reach the number of participants they needed, but then they needed to make sure that they were reaching people who really did have ME/CFS as currently defined. So they took extra steps that other studies in ME/CFS have not taken.
• DecodeME participants didn’t just report they had ME/CFS, they also reported that they had been diagnosed with it by a health professional, that they had post-exertional malaise and fulfilled either CCC or IOM criteria or both.
• Over 5000 potential participants (5281) were excluded from DecodeME because they did not meet these criteria:

[From the preprint] After launch, 26,901 people (84% female) completed the online or paper participant entry questionnaire and consented to take part. Of these, 21,620 (85% female) met our study criteria of having a diagnosis of ME/CFS from a health professional, having post-exertional malaise as a symptom, and having symptoms consistent with Canadian Consensus and/or Institute of Medicine / National Academy of Medicine diagnostic criteria (5,26) (Supplementary Methods). We sent all cases a saliva DNA collection kit (Fig. 1A).

A point should probably be added in there about whether GWAS studies in other diseases that have found robust associations have recruited directly or through health professionals. I don't have the knowledge of those studies to make the point, but others on here probably do. If studies with similar statistical power to DecodeME have managed to recruit through health professionals, then it would be worth explaining why this wasn't possible for ME/CFS in the UK.

 

[Evergreen]

I think the way the info was presented in the preprint contributed to the criticism, and that could be rectified in the future published version. This is the "Participant recruitment and data collection" section on p.3:

We recruited participants between 12 September 2022 and 31 January 2024. We used podcasts, webinars, blog posts, media interviews, social media, and articles in ME/CFS charity members’ print magazines to build trust and understanding about DecodeME and to recruit to the study. Participants completed a phenotype questionnaire which they could complete either online or on paper. We gave the most severely ill telephone support, if needed. We made it as easy as possible for people with ME/CFS, especially those housebound or bedbound, to give us their DNA, by using a ‘spit and post’ design to collect saliva samples (Supplementary
Methods).

That makes it look like everyone who filled out the questionnaire provided DNA.

There's a whole section on "Genotyping and analysis" before it returns to "Case and control definitions" and readers who think everyone who filled out a questionnaire provided saliva get the info they need:

For genetic analyses, we used case inclusion and exclusion criteria based on the Canadian Consensus criteria and/or US Institute of Medicine / National Academy of Medicine criteria for ME/CFS (5,26). Both require post-exertional malaise as a symptom (Supplementary Methods). To meet these criteria, we did not include participants as cases if they reported that their illness had started within the previous six months, or had lasted their whole life. Additionally, cases needed to live in the UK, be at least 16 years old, and to have received a diagnosis of ME, CFS, ME/CFS or CFS/ME from a health professional. Also, those with an ME/CFS diagnosis after SARS-CoV-2 infection needed not to have been hospitalised, nor to
have heart or lung damage, because of this infection.

Fig 1A also makes it look like the 5000 participants were cut due to saliva eligibility criteria:

Whereas the text re "Results" tells a different story:

After launch, 26,901 people (84% female) completed the online or paper participantentry questionnaire and consented to take part. Of these, 21,620 (85% female) met our study criteria of having a diagnosis of ME/CFS from a health professional, having post-exertional malaise as a symptom, and having symptoms consistent with Canadian Consensus and/or Institute of Medicine / National Academy of Medicine diagnostic criteria (5,26) (Supplementary Methods). We sent all cases a saliva DNA collection kit (Fig. 1A).

So I'd say @Chris Ponting & co will be able to make it clearer and less open to misinterpretation in the future published version.

 

[Ariel]

Sorry, can someone explain the "European ancestry" thing? I am not even sure what this means. Thank-you.

 

[Utsikt]

Sorry, can someone explain the "European ancestry" thing? I am not even sure what this means. Thank-you.

People with genes that trace back to Europe. If you immigrated from outside Europe, or your parents, etc. did, you wouldn’t be included. (I have not checked the cutoff they used)

It’s just a way to make sure the genes of the participants are as similar as possible, except for the difference we’re trying to explain (in this case it’s ME/CFS).

The differences between different continents (or countries) exist because of limited travelling and migration throughout most of human history. The continents themselves are just arbitrary borders if you look at Europe vs Asia, but they can still be useful.

 

[Jonathan Edwards]

Sorry, can someone explain the "European ancestry" thing? I am not even sure what this means. Thank-you.

Ancestry will carry with it all sorts of irrelevant gene variants. Social factor could easily mean that patients sending in samples were of a slightly different racial mix than controls. With big numbers slight systematic differences in populations can easily come up as significant.

 

[Ariel]

People with genes that trace back to Europe. If you immigrated from outside Europe, or your parents, etc. did, you wouldn’t be included. (I have not checked the cutoff they used)

It’s just a way to make sure the genes of the participants are as similar as possible, except for the difference we’re trying to explain (in this case it’s ME/CFS).

The differences between different continents (or countries) exist because of limited travelling and migration throughout most of human history. The continents themselves are just arbitrary borders if you look at Europe vs Asia, but they can still be useful.

Thanks; I would be curious about this because I am a participant and one of my parents did immigrate but one did not. Presumably this means I would be included but I cannot remember what questions I answered. I don't remember the questions being asked or what I wrote.

I am not sure in general though - how long does one's family have to be here to have not "immigrated"? I don't understand this criterion. I get it about comparing similar genes but it would be good to know if I was actually included in the GWAS portion of the study and what this means for others.

 

[Evergreen]

Thanks; I would be curious about this because I am a participant and one of my parents did immigrate but one did not. Presumably this means I would be included but I cannot remember what questions I answered. I don't remember the questions being asked or what I wrote.

I am not sure in general though - how long does one's family have to be here to have not "immigrated"? I don't understand this criterion. I get it about comparing similar genes but it would be good to know if I was actually included in the GWAS portion of the study and what this means for others.

I think it was done on the basis of your genes rather than answers to any question. The rationale was "Not accounting for ancestry could influence ME/CFS risk-association and/or confound case-control genetic associations."

There's a flowchart on p.40 of the preprint:

 

[Utsikt]

Thanks; I would be curious about this because I am a participant and one of my parents did immigrate but one did not. Presumably this means I would be included but I cannot remember what questions I answered. I don't remember the questions being asked or what I wrote.

I am not sure in general though - how long does one's family have to be here to have not "immigrated"? I don't understand this criterion. I get it about comparing similar genes but it would be good to know if I was actually included in the GWAS portion of the study and what this means for others.

I don’t know if there’s any practical way for us to determine who was actually included in this analysis. The chart Evergreen shared gives some info, but there’s always the chance that something went wrong with the sample or the analysis of it, so it had to be cut even though it was initially eligible.

The results of the GWAS won’t have anything to say for any one individual, because there are no ME/CFS genes per se. Most people with the genes won’t have ME/CFS.

The value of the GWAS lies in that it points us to the processes that are involved in ME/CFS, because the only way a gene can increase or decrease the risk of a disease is if it’s actually somehow involved in the disease mechanism(s), or in creating some of the circumstances that leads to ME/CFS like possibly increasing the risk of infection.

 

[Andy]

and fulfilled either ICC or IOM criteria.

*CCC and/or IOM

 

[Ariel]

I think it was done on the basis of your genes rather than answers to any question. The rationale was "Not accounting for ancestry could influence ME/CFS risk-association and/or confound case-control genetic associations."

Thank-you, this makes more sense. I guess it's about removing samples not relevantly similar to the control group after analysing them?

Do you (or does anyone) know if this means having one parent who is not "white"/European and one who is basically means you'd be too dissimilar?

 

[Evergreen]

Thank-you, this makes more sense. I guess it's about removing samples not relevantly similar to the control group after analysing them?

That's my understanding, but others will know better.

Do you (or does anyone) know if this means having one parent who is not "white"/European and one who is basically means you'd be too dissimilar?

I don't know, but others might. I think that even if you weren't included in this particular analyses that has been written up as a preprint, your data is likely to be included in some other analyses in the future.

 

[Evergreen]

*CCC and/or IOM

Thanks Andy. I've added "or both" to both posts.

 

[Andy]

Thanks Andy. I've added "or both" to both posts.

And correcting from ICC to CCC?

 

[Amw66]

[Since this is relevant to both this thread and the media thread, I'm posting it in both threads.]

I’ve been trying to work out the strongest arguments against the criticism that participants were not diagnosed by a medical professional.

Here’s a recap of the criticism:





It’s true that DecodeME participants were not recruited through health professionals, but speculation that they did not have a formal diagnosis of ME/CFS is...purely speculative. I don’t think it is a strong response to say that cases were all diagnosed by a health professional, because we cannot verify that. We only know that they answered “Yes” to this question:

When people are objecting to self-report (ah, the irony), then countering with another self-report won't work.

I think the strongest response is something like:

• Good-quality GWAS studies need huge numbers of participants in order to have high enough statistical power to detect robust associations that would be missed in smaller studies.
• It would not have been possible to recruit 26K participants with ME/CFS indirectly through health professionals.
• By recruiting directly, DecodeME could reach the number of participants they needed, but then they needed to make sure that they were reaching people who really did have ME/CFS as currently defined. So they took extra steps that other studies in ME/CFS have not taken.
• DecodeME participants didn’t just report they had ME/CFS, they also reported that they had been diagnosed with it by a health professional, that they had post-exertional malaise and fulfilled either ICC or IOM criteria or both.
• Over 5000 potential participants (5281) were excluded from DecodeME because they did not meet these criteria:

A point should probably be added in there about whether GWAS studies in other diseases that have found robust associations have recruited directly or through health professionals. I don't have the knowledge of those studies to make the point, but others on here probably do. If studies with similar statistical power to DecodeME have managed to recruit through health professionals, then it would be worth explaining why this wasn't possible for ME/CFS in the UK.

Here we have no specialist clinics at all, and a lot of GPs resist a CFS ( never mind ME) diagnosis -it's post viral fatigue in a lot of places. If anything you may have under representation - particularly at the milder end of the spectrum.

 

[Yann04]

Same here. In a region of two million people I can count on two hands the number of people online saying they have ME/CFS.

 

[Kitty]

Do you (or does anyone) know if this means having one parent who is not "white"/European and one who is basically means you'd be too dissimilar?

I don't know, but it probably depends how you defined your ancestry on the form. If you ticked boxes indicating white European heritage that's how your sample would have been categorised; if you ticked mixed heritage, it's possible might have been separated in some of the analyses.

All of the viable samples will have been sequenced, though, so people of mixed and non-European heritage are represented in the data and can be compared in genetic studies.

 

[Ariel]

I don't know, but it probably depends how you defined your ancestry on the form. If you ticked boxes indicating white European heritage that's how your sample would have been categorised; if you ticked mixed heritage, it's possible might have been separated in some of the analyses.

All of the viable samples will have been sequenced, though, so people of mixed and non-European heritage are represented in the data and can be compared in genetic studies.

I don't remember what I put on the form. It seems it's separated out by objective analysis though according to the figure above? It wouldn't really make sense to rely on what people have put on forms (I never know how to answer these forms)

 

[MelbME]

Great to see it published. Some interesting hits.

CA10 stands out to me

 

[Evergreen]

Realised I haven't yet said thank you so much to @Chris Ponting, @Andy, @Simon M and everyone else involved in this project. It feels like a gamechanger, a bit like the starting gun of a race. I'm really hoping that researchers will be intrigued and that we'll start getting some really meaty studies to discuss in the next few years as a result.

I'm sure this week is both very exciting and really exhausting, whether healthy or not. Know that we appreciate it all.