Defining a High-Quality Myalgic Encephalomyelitis/Chronic Fatigue Syndrome cohort in UK Biobank, 2025, Samms & Ponting

https://openresearch.nihr.ac.uk/articles/5-39/v1

Research Article
Defining a High-Quality Myalgic Encephalomyelitis/Chronic Fatigue Syndrome cohort in UK Biobank
[version 1; peer review: awaiting peer review]

Gemma L. Samms1, Chris P. Ponting

Abstract
Background
Progress in Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) research is being slowed by the relatively small-scale studies being performed whose results are often not replicated. Progress could be accelerated by analyses of large population-scale projects, such as UK Biobank (UKB), which provide extensive phenotype and genotype data linked to both ME/CFS cases and controls.

Methods
Here, we analysed the overlap and discordance among four UKB-defined ME/CFS cohorts, and additional questionnaire data when available.

Results
A total of 5,354 UKB individuals were linked to at least one piece of evidence of MECFS, a higher proportion (1.1%) than most prevalence estimates. Only a third (36%; n=1,922) had 2 or more pieces of evidence for MECFS, in part due to data missingness. For the same UKB participant, ME/CFS status defined by ICD-10 (International Classification of Diseases, Tenth Revision) code G93.3 (Post-viral fatigue syndrome) was most likely to be supported by another data type (72%); ME/CFS status defined by Pain Questionnaire responses is least likely to be supported (43%), in part due to data missingness.

Conclusions
We conclude that ME/CFS status in UKB, and potentially other biobanks, is best supported by multiple, and not single, lines of evidence. Finally, we raise the estimated ME/CFS prevalence in the UK to 410,000 using the most consistent evidence for ME/CFS status, and accounting for those who had no opportunity to participate in UKB due to being bed- or house-bound.


Plain Language Summary

Plain English summary

Myalgic Encephalomyelitis / Chronic Fatigue Syndrome (ME/CFS) is a highly debilitating and relatively common disease whose causes are unknown. Most of its research compares features (symptoms, molecules, cells or genes, for example) from small numbers of people with ME/CFS, against these features from others without the disease. Studies using small sample sizes have not yielded replicated discoveries that would quicken the pace of ME/CFS research towards effective therapies.

Alternatively, ME/CFS research could take advantage of population-scale bioresources, such as the UK Biobank, which holds diverse genetic, molecular, cellular, imaging and questionnaire data on nearly 500,000 individuals. This has the added advantage of being cheaper than recruiting a new cohort, and generating new data.

For ME/CFS, this approach raises the difficulty of how to best categorise people with this disease, for example by questionnaire response or through linked electronic health records. In the UK Biobank, there are four ways to categorise people with ME/CFS.

This study’s results show that just over 1% of the UK Biobank participants could be categorised as having a ME/CFS medical diagnosis. However, not all evidence for their ME/CFS diagnosis is consistent.

By cross-referencing different data in UK Biobank, we show that a participant’s ME/CFS diagnosis is best supported by two or more lines of evidence. Using the most consistent evidence, and accounting for those who – due to their illness – could not participate in the UK Biobank, we estimate that the UK’s prevalence of ME/CFS is 410,000.


Keywords
Myalgic Encephalomyelitis/Chronic Fatigue Syndrome prevalence; Population bioresource; Electronic health records; Post-viral fatigue syndrome

https://bsky.app/profile/cgatist.bsky.social/post/3lnuillx7us26

 

I’m not sure what they mean by multi-system disorder. I prefer how the factsheet presented it by just focusing on the impact on life:

————

Good thing they used report.

No mention of the delay, which seems to be a key element for the definition of PEM. I hope this is corrected.

There should be a mention of how this list isn’t exhaustive.

I didn’t know it was that high, but it’s from a DecodeME paper so I guess it’s a good as anything.

I think this is an important piece or information mostly because money seems to be the only thing that might make the government act (even though it’s cheaper short term to just ignore the problem).

It feels like they are throwing some shade on the effort preference study, but that might just be my bias. I’m glad they show how funding their kind of work is relatively cheap.

For the sake of countering the BPS narrative of «the lack of evidence so far is evidence for the lack of a biological basis», it’s useful with an alternativ explanation.

 

A very useful paper. A few initial comments after a skim through:

The authors use 3 UKB PQ questions as proxies for PEM:

"Do you get tired after minimal physical or mental exertion?"
"Does this tiredness, weariness or fatigue go away when you rest?"
"Is this tiredness, weariness or fatigue happening only because you have been exercising and/or working too much?"

Of course, as they acknowledge, these are limited: they ask about "tiredness" rather than malaise, cognitive dysfunction, pain, etc, and don't assess delayed onset or a prolonged recovery time. The third question would include ordinary fatigue from overexertion. 76.1% of C2 reported tiredness after minimal exertion vs 48.2% of controls, but about 24% of self-declared pwME deny tiredness after minimal exertion. Completion of the PQ was itself biased (41% vs 33% C1 vs not-C1). Requiring not only a positive answer to the persistent fatigue question but also all three PEM-congruent questions - only ~57% of C2 would meet that pattern.

There is an unusual level of instability across UKB visits (44-69% of C1 not consistently reporting CFS+ve). It's possible that some participants might misinterpret questions, confuse "chronic fatigue", the symptom, with CFS, or self-diagnose, or perhaps even interpret the question differently on initial screening vs follow-up (lifetime vs. current diagnosis?) but this nonetheless seems high.

Does the MCAR assumption really hold? There is considerable "missingness", with only 36% of 5,354 participants having >=2 pieces of evidence. But people with more severe ME/CFS may be less likely to complete questionnaires or have more hospital admissions. Sensitivity analyses under different assumptions (MAR, MNAR) might have been useful here? The statistical approach is straightforward but might logistic regression also have been useful to consider confounders (sex, age, socioeconomic status).

The reported female-to-male ratios (lowest 1.85:1 for C2; highest 2.38:1 for C3) are notably lower than estimates from clinic studies. Perhaps recruitment/ascertainment bias (UKB's older, healthier cohort may under-ascertain female cases) but this may be worthy of deeper investigation: e.g. are men more likely to be underdiagnosed in clinics or report their symptoms differently, or are women more likely to be severe, etc? It might also be useful to see if there were systematic differences - demographics, severity, comorbidities - between those in multiple cohorts vs those in just one.

And on the temporal factors: use of different diagnoses & codes & different understanding of the features of the diagnosis are likely to result in significant differences over time. Some G93.3s may also have been due to self-limiting PVFS. The use of G93.3 has probably increased significantly over the last few decades (wasn't there even a patient campaign about ME/CFS being coded correctly?); previously there may have been many more F4x.x or CF symptom codings.

The prevalence estimate is derived from 0.49% of UKB - 330,000 adjusted upwards to 410,000 (0.6%) on the assumption that there are 25% of pwME who are house or bedbound and therefore absent from UKB. But about the 25% adjustment: if the housebound/bedbound 25% are completely absent from UKB then 330k/(1-0.25) = 440k - whereas they give 410k?

 

Those do not seem to me to be likely to pick out PEM.
Being tired after exertion is very nonspecific and if someone is feeling generally tired they might easily say yes. And tiredness tends not to go away with rest. Exhaustion or exercise fatigue does, but not tiredness I think.
The last question I find hard to understand - how is the person to know what caused the tiredness? I would guess that someone with PEM could very well say no because it isn't the exercising that caused it so much as the ME/CFS being there that caused trivial exercise to be followed by tiredness.

But the study is useful in demonstrating the problem of diagnostic ascertainment. The limited overlap between groups may go a long way to explaining why nothing came up on the genetics screen for this group.

 

Yes that’s been my same concern for a while. In terms of a statistical association, it would be like trying to find genetic predictors of “happiness” where some of the participants were labeled as “happy” because they weren’t severely depressed, others because they reported feeling content with their lot in life, and others because they smiled more than average throughout the day.

Even if you do find a strong association, it becomes uninterpretable biologically.

 

I have visions of knock-in mice with a super active happiness gene allele. Whiskers twitching all over the place.

 

I'm worried to see this too. When I think about the times I was moderate but younger and didn't know about PEM specifically I absolutely knew that the pattern was waking up some days unable to function vs other days being or seeming as good as anyone else. I didn't think I had 'CFS' or 'ME' because of the misnomer that it was just a 'being exhausted all the time' or 'getting tired after small amounts' type thing.

In fact the defining feature isn't until you get more severe the getting tired sooner than others at the time but sleep and pain and waking up to the wonderful surprise of not being able to wake up... except not in a way you expect and think 'well that will be because I did a marathon yesterday' or even 'that will be because I did more than I would normally yesterday', given the delay and the cumulative issue.

Strangely the consultant years ago who diagnosed me in a clinic was biomedical (although still limited by all sorts) and actually said the illness was characterised by having inappropriately timed energy. ie that it wasn't just about tired all the time, but they were interested as much in the times when others who did a busy day at work would get home and get some sleep and I ended up not because that is how ME/CFS works - in a much more extreme scenario than the boss saying 'you'll need some time to wind down before you go to sleep if you work into the evening'. Of course that didn't mean that I didn't feel pains in my legs due to over-use and dog exhausted in particular every single morning, most evenings as the end time for work couldn't come soon enough and definitely when thinking about whether to book in any social life.

Of course all that relates to the pushing through I assume, but it is more important therefore that we acknowledge that exists to new people rather than the vice versa of pretending it doesn't then surely - because most newbies therefore will be puzzled by this strange pattern. So the logic of cutting it out, rather than making sure it is in 'because new people' seems the wrong way around to me?

And I mention this because I think the cumulative doesn't get enough mention and how it builds up and then surprises you in its scale/order of magnitude (even though if you are in rolling PEM you almost certainly are feeling it as it builds because it is like trying to function in a situation that feels like for healthy people they would feel if they were trying to get up and go to work when they'd spent 5 days in transit flying to somewhere sitting in uncomfortable seats and without any proper sleep - yes you can keep reaching for the caffeine and other cajoling and setting 25 alarms to make your body work, but it hurts)

I thought we'd got to the point where people agreed that the ME/CFS that everyone is now supposed to be working towards and trying to get provision actually provided for is the counter-intuitive thing that screws up your sleep and involves huge amounts of exhaustion but, even to those who are new to it and not realising what it is or in rolling PEM, can spot it isn't as straightforward as the 'post-viral gets tired easier than others' thing.

I'm a bit worried that all over the place in the name of good intentions (like inclusivity supposedly or 'new people' - except there are other ways to say 'this might feel like' or 'x pattern of do--> exhaustion or how it takes a week to recover and you mightn't sleep for the first few days of it' or importantly 'you might be exhausted and needing sleep to feel like you did before that, but it screws up your good sleep' or 'you get better rest/sleep only after you've rested/slept enough to get that good sleep' to help people relate, or give them thought or real experiments) we are taking whether the illness exists backwards and writing descriptions that will just be read by most as fatigue/chronic fatigue.

We are kidding ourselves that when people who think they know read something that sounds approx like what they thought they knew that they will note that little caveat of 'and there are also those who are severe' or 'sometimes it is delayed' when these things, particularly describing it as a spectrum, need to be the meat of the sandwich and not the and also.

 

Same.
Saw the name “chronic fatigue syndrome” a coupe times.
Was like — nope, that’s not my problem. My proboem is whenever I exercise I get a migraine, feel sick, and my stomach hurts for the next couple days.

This is actually how I mostly experience PEM. More this than the sort of small activity triggers PEM small crash passes. So my PEM could be delayed by months in that I’ll push for months and then crash.

This is a major problem with “chronic fatigue syndrome”. It redefines the condition. I’m willing to bet 50% of people who were convinced they had “CFS” or diagnosed as “CFS” at some point don’t have PEM.

And I think this was kind of on purpose by the psychiatrists who redefined ME as CFS. Much easier to dismiss an illness as a social contaigon or maligners and depression who just need to pull themseleves up by their bootstraps if you define it as a vague state of fatigue.

 

The minimal qualifier is the important part and accounts for that, and should probably be emphasized a bit more, because no one feels exhausted after minimal exertion unless there is something medically wrong (or, I dunno, if they're wearing a very heavy armor, or whatever).

Although of course it's not always minimal. When we get to that point, it's probably too late. Many of us, as commented above, have gone through the PEM cycle where a moderate level of effort wiped us out in unpredictable ways, with the mind-messing delays that make it impossible to attribute cause.

A simple but very hard question to get right.

 

Interesting study. Sad that the concordance between the different ME/CFS pieces information in the UK biobank is so low.

For example: 2,312 (0.46% of the 0.5 million biobank participants) self-reported a clinical diagnosis of CFS (C1). But of these, 28% also reported 'good' or 'excellent' overall health when they first volunteered their CFS diagnosis. This seems like a contradiction.

Next to the verbal interview there was also another data source (C2), a pain questionnaire that explicitly asked: “Have you ever been told by a doctor that you have ME/CFS?” 2,720 people (1.63%) said yes. But in this group, we also see discordance with other questions. For example: of these self-reported ME/CFS cases, 31.0% did not report ‘persistent or recurrent tiredness, weariness or fatigue that has lasted over 6 months. 24% of this cohort did not report ‘feeling tired after minimal physical or mental exertion’.

Combined with diagnostic records in hospitals (C3) and primary care (C4), there was very little overlap.

 

I think I've done something similar on study surveys. I'm not totally sure why. I guess it feels like it's asking "apart from the main thing in question that we've already asked about, how would you say your health is? Got autoimmune disease, cancer, diabetes, poor dental health?" And I'm mostly good in other departments. Could be a similar thing going on with other people.

Or maybe I'm imagining how I might be described by someone who isn't convinced fatigue and mental health are part of Health™: "You're tired all the time, and have anxiety and trouble thinking, but at least you're healthy."

 

That's a good point, it might need a look at the whole survey and how it was presented at the time to get a better picture. Isolated responses can be misleading if they're even slightly out of context.

 

This may seem pedantic but there is a subtle difference between what the paper reports the questionnaire says ("that you have") and what it actually says ("that you have had"):

BLOCKB1 Have you ever been told by a doctor that you have had any of the following conditions?
B1j Chronic Fatigue Syndrome or Myalgic Encephalomyelitis (M.E.)

I suspect there are rather a lot of people who, perhaps suffering from a self-limiting post-viral fatigue state or something similar, have been told by some doctor at some point that they have "chronic fatigue syndrome". The "feeling tired or having no energy" question is limited to recent feelings ("We next want to ask a few questions about your mood and feelings recently").

 

I'm sure this happens. I suspect a bigger risk is that people have been told they have chronic fatigue (around 10-fold more common than ME/CFS) and misremember it as or confuse it with CFS, as the paper suggests.

 

they don’t even need to misremember or confuse it.

If my doctor diagnoses me with migraines and a survey asks me if I’ve been diagnosed with “Migraine syndrome”, the reflex of pretty much anyone who isn’t super informed or pedantic will be to say yes. I mean you’ll just assume their the same thing and “syndrome” is just the official medical name. Same for “anxiety” and “anxiety syndrome” or something. So someone diagnosed with “chronic fatigue” will say they have “chronic fatigue syndrome” if asked, unless they specifically know about CFS as a seperate thing.

Here’s it’s just more of a case of “CFS” being a terrible name.

you can’t expect laymen to know the difference between CFS and CF because the naming is made in such a way people are to automatically assume they are equivalent. I mean most clinicians can’t even tell the difference.

 

Given that the NHS has been doing its best to gaslight people that life with ME is “good health”, I don’t see any contradiction in some patients with ME reporting that.

On the rare occasions when I’ve been able to access healthcare, the conversation has gone like this:
- Do you have any health problems?
- Yes, I have severe ME. I am largely bedbound, and need help in every aspect of daily life and can’t live independently. With someone to help and a wheelchair, I can occasionally get out of the house if it’s essential, although for the next week, because of this appointment, I will have even worse symptoms and may for example be too weak to eat, drink or use the toilet without help.
- Right, but apart from that, you’re in good health?

I find it hard to imagine such a conversation around MS or any other potentially debilitating chronic disease.

Doctors teach patients that the NHS does not consider ME symptoms to constitute poor health. Some people asked to fill in a UK medical form will respond accordingly, even if they are very unwell.

 

Relevant background about the UK Biobank cohort

The UK Biobank was set up as a prospective study of mid- to late-life illnesses. Via the NHS, it invited over 9 million people aged 40 to 69 who lived close to a string of assessment centres to take part. Over 502,000 people were recruited, over 90% between 2008 and 2010, when recruitment closed.

As usual for such studies, the cohort was biased toward healthier, female (54.4% versus 50.7% nationally), wealthier/better educated, and White.

The average age is 56.4 at recruitment, and older people were overrepresented, as you can see from the participation rate below (note they note there are fewer and fewer people in each increasing age band).


Each participant was assessed at baseline with blood draws et cetera, measurements and questions.


Most participants made visited UKB centres just once, but between 80,000 and 1000,000 made up to four:

Visit 1: baseline assessment undertaken by everyone (2006-2010).

Visit 2: about 20,000 people had a repeat of the complete baseline assessment (2012-13).

Visit 3: Full body imaging, target 100,000, currently 80,000. (2014+).

Visit 4: Repeat imaging visit, undertaken by 60,000 people to date (2019+).

 

Data defining the four cohorts of this study.

1. Self-report of chronic fatigue syndrome (defining cohort C1)

This comes from the baseline assessment (and visits 2-4 where made), and it looks like participants were not prompted with a question about CFS.

Participants completed a touchscreen questionnaire that asked about a small number of specific illnesses and asked if they had "been told by a doctor that you have other serious illnesses or disabilities" (my italics).

Then, during a 'verbal’ (presumably face-to-face) interview with a nurse, those who said they had another illness or disability were asked what they are. Crucially, there appears to be no prompt for CFS:

From UK Biobank (ref 23)

What isn’t clear is how nurses and the system handled a 'chronic fatigue' response. I guess it could have been recorded as CFS or entered as Chronic fatigue in a text box. Potentially, this test response could have been recorded as CFS or ignored.

So first, the participant must declare the illness as serious—would everyone consider chronic fatigue serious? Then, the nurse must record the response as CFS, and it seems unlikely that all "chronic fatigue" responses are recorded and registered as CFS.

The UKB page for this question lists CFS under immunological/systemic disorders, and the CFS response is coded 20002.1481_ix [i0, i1, i2, and i3 designate the four potential visits].

Have I missed something? If CFS is not prompted, that presumably reduces the risk of inaccurate reporting.

 

Data defining the cohorts:
2. Self-report of ME/CFS (Cohort 2, C2)

Experience of Pain Questionnaire (optional and online, 2019-2020)

This questionnaire directly asks if respondents have ME/CFS:
Have you ever been told by a doctor that you have had any of the following conditions? There are a list of 15 diseases, including:
“Chronic Fatigue Syndrome or Myalgic Encephalomyelitis (M.E.)”. [Field ID 120010]

This is different from C1, above, where people are asked if they have an 'other' illness and then asked to specify it.

Studies asking general populations, such as this one, if they have chronic fatigue syndrome (or ME/CFS) have a poor record. For example, Louis Nacul’s 2024 Preprint shows that only 37% of the British Columbia Generations Project who self-reported having been diagnosed with Chronic Fatigue Syndrome met diagnostic criteria (IOM or CCC) assessed by questionnaire. Probably fewer still would have met the criteria after the necessary clinical workup.

So, despite asking about ME/CFS, rather than CFS (which should be helpful), I'm not sure the pain questionnaire question is very helpful.

 

This gets even more exciting. Like the posts on self-reports, tries to understand how the cohorts were selected rather than the data itself.

G93.3 in hospital records (1997-203), Cohort 3, C3
Hospital records covers those 'occupying a bed', so inpatients and day patients, but not outpatients or A&E attendees. 4767k/88% have hospital records, which might be so high because 90% of participants were aged 55-84 when the hospital records analysed here were last updated in 2023. Few outside Soctland would have been aged below 30 when English & Welsh records began in the mid 1990s. The F:M ratio is 1.2:1, so maternity can't be contributing that much.

9% of these had a G93.3 diagnoses (PVFS but seems widely used for ME/CFS) as the main reason for hosptial admission, while the rest had it as a secondary or external diagnosis.

Primary care "ME/CFS" diagnoses (pre-1987-2016/17), Cohort 4, C4
Although all participants consented to sharing their primary care records, UKB only has access to 45% to date as it had to arrange access through various companies that hold the data. Data runs from the mid-80s to 2016/17, though UKB has mapped older paper records information into current codes for some, going back to the 40s.

The one issue with this data is that 6/23 of the codes used to code ME/CFS also include PVFS, and it's not clear what proportion of PVFS diagnoses this includes. Historically, PVFS diagnoses far outnumbered ME/CFS diagnosis.