CDC Data Brief: Myalgic Encephalomyelitis/Chronic Fatigue Syndrome in Adults: United States, 2021–2022, 2023, Vahratian, Unger et al

I wouldn't agree that recovery is rare. I think that's heavily influenced by sampling bias (easy for us to think of lots of people who have had ME for a long time, not so much people who were sick for a year or three and then got better, perhaps not ever really knowing what they had). I think we've seen with LC as well that a significant portion of people do recover.

But leaving that aside, the CDC survey data here shows a 33% drop in ME/CFS prevalence from the 60-69 age group (2.1%) to the 70+ age group (1.4%). That's not really compatible with a model where recovery rates are low and all the people in this survey who answered yes do in fact have ME/CFS, so one of those two conditions must be false to some degree.

 

Not necessarily. It's also explainable by ME/CFS patients having a lower live expectancy than healthy people. I believe there is sufficient evidence and reasoning to make such a claim, albeit ME/CFS not being a directly deadly disease (lower income, less accessibility to health care and generally ill health leading to less physical activity, less sun exposure etc already reduces the life expectancy by quite a bit).

 

Good point. Not sure we have evidence to say this effect would be strong enough to explain the difference, but it could be part of the explanation.

I still have a hard time believing a full 2% of 50-59 year old Americans have a diagnosis of ME/CFS though (actual prevalence, maybe, but for people who have been diagnosed that seems high).

 

I think it's a combination of reduced life expectancy, older generations being less likely to receive a diagnosis, and perhaps older people no longer feeling like they have ME if the effects of old age "shadow" it. It's sensible that we have a reduced life expectancy due to the inability to be active, the possible effect of ME on immune function and cancer rates, perhaps the inability to make nutritious food, and receive poor healthcare in general. PwME may also be more likely to opt for comfort care over burdensome lifesaving interventions like chemo or major surgery, because of their pre-existing poor quality of life.

 

I'd take a skeptical eye at that prevalence. Even if I'm on the more severe end, I just find it very unlikely that 1.5% of the population have what I have. Like implausible. MS is .21%.

Almost everyone I've been told about who "has that [me/cfs]" doesn't have it, has something else, or recovers within 18 months. You know, the got better with Yoga kinda. Or is working a full time job and having an active social life. It's just not remotely compatabile with me/cfs and PEM as I know it. I'm guessing as hard as it for some to get dxed, others get it thrown at them when their cbc comes back normal.

I'm wondering if the age spike is somewhat related to age related fatigue and wearing being called me/cfs and those people becoming more accustomed/accepting of it by their 70s. It seems less likely to me more people are getting me/cfs and then having it go away at those times.

 

This statistic makes sense given that the substantial disability caused by ME/CFS, even at a “mild” stage, leads patients to not being able to work or only part time and thus to a lower income or none at all. Then, getting approved for disability benefits is often an arduous and long process due to the lack of medical recognition of the disease.

However, it is at odds with the “yuppie” aspect of the “yuppie flu” characterization, in the sense that “type A” young urban professionals tend to be better remunerated. It is also not consistent with the age distribution reported in the survey (nor with the geographical distribution as pointed out by @ME/CFS Skeptic).

Of course, there is the issue of how doctors diagnose ME/CFS which likely inflates the prevalence rate in this survey. Given the poor knowledge that mainstream MDs have on ME/CFS and the lack of (proper) teaching about the disease in medical schools, the diagnosis was probably most often made on the grounds that routine exams, if any, were normal in the presence of chronic fatigue. This is arguably looser than the loosest CFS diagnostic criteria, such as the 1991 Oxford criteria, which are known to significantly overestimate the prevalence of ME/CFS.

It is also possible that respondents answered that they were given a diagnosis of CFS when their doctor only evoked chronic fatigue — i.e. confusion and/or conflation between the terms CF and CFS —.

 

Massive, representative survey - but red flags too

This is a massive, nationally representative survey, which makes this very interesting:

There is a lot to go through here, but I am concerned by some mismatches with other data sources and evidence

1. 1.3% is way higher than the Lenny Jason 1999 community prevalence study (0.42%) and the CDC/Wichita community study (0.25%), both used random dialling (when people still answered surveys by phone) and actual diagnosis by specialists. Both of these community surveys spoke to tens of thousands of people in the in their initial screening.

2. Including long covid could make a big difference (less so in the 2021 edition) but in the UK at least, doctors don't seem keen to hand out ME/CFS diagnoses for LC so far.

3. The sex ratio (<2:1 female:male) looks very low, especially vs DecodeME (ratio c. 5:1), which also selected primarily on patient-report of diagnosis by a health professional .

4. The age prevalence figures indicate peak incidence (new cases) age 50-59, which rewrites the book:

age band, prevalence, implied incidence:
18-39 = 0.7%, +0.7%
40-49 = 1.2%, +0.5%
50-59 = 2.0%, +0.8%
60-69 = 2.1%, +0.1%

Other data would support highest incidence for the 18-39 age group, which also spans 21 year vs 10 year bands for the others.

My concern is that they could be picking up lots of cases of mis-remembered chronic fatigue, a very common diagnosis (I think I've seen c.5% of the pop) that does get more common in middle age. But we don't know.

OK, I'm cooked now. There's clearly a lot to discuss about this survey.

 

Not a recommendation
CDC reports 3.3 million Americans experience chronic fatigue syndrome

5NEWS

https://www.youtube.com/watch?v=qvXVQ7JCjYo

 

It is kinda of strange that the true prevalence studies found low prevalence rates (0.2-0.4%) where the majority of patients (80-90%) are undiagnosed and that these surveys find that 1.4% of the population says they have been diagnosed with CFS. That is quite a large mismatch.

if I were to guess, I think it's due to ambiguity surrounding the term 'chronic fatigue syndrome'. I suspect that a lot of healthcare professionals use it for idiopathic fatigue, burnout, sleep problems and similar fatiguing conditions that have no good explanation. In my own country, doctors often speak of 'a fatigue syndrome' in such cases, even though patients would likely not meet official ME/CFS criteria.

 

A 1.3% prevalence is 1 in 80 people having a diagnosis. I had always been told that diagnosis was very hard to get in the US. Excluding those we met through this illness, do 1 in 80 people we knew have a diagnosis of ME - maybe 1 in 40 actually having it?

I know Charles Shepherd of the MEA often casts doubts on high prevalence rates because they don't match with the number of patients on GP lists. 1.3% in the UK would be about 25 diagnosed patients per GP. The GPs I've had have implied they each have only a small number. (Maybe I should ask my GP for a view for her practice.)

 

Both can be true, the prevalence numbers are for whatever reason too high and the new prevalence numbers can be used effectively by advocacy to push for higher funding of ME/CFS. Of course, one can also 'try to get to the bottom of this' but it's most likely a waste of energy. There will be no lack of people trying to dismiss this numbers, but, there will be a lack of people putting this numbers to good use for us.

 

A lot of undiagnosed are probably told at some point, it's just never coded because clinicians are discouraged, if not told directly, never to do that. It's the refusal to code that distorts reality, on a BS excuse no less. There are plenty of conditions without any treatment, they still get diagnosed. How else would treatments ever be developed and patients get support?

No one can claim to be surprised by this. The model dealing with chronic illness is mass lying. This is what mass lying does. No one can pretend they don't understand what it leads to, but until there are consequences it won't change, especially as right now there are consequences for doing the right thing. Which, also, no one can pretend they don't understand is wrong and can only lead to terrible outcomes for patients.

 

And quite obviously, it massively skews early on. Most recoveries happen within weeks, then fewer still within months, even fewer at 6 months and by that point it seems to be where we hit the spot where recovery is as we know it, around 5%.

Given everything I've seen, I also see no reason to think that there's anything different, that a change occurs over time between one illness and another, i.e. this old distinction between "self-limiting" PVFS and ME/CFS. As we've seen from a lot of data, the conditions seem to be set early on, and whatever difference is there from the start is what makes the recovery process occur over weeks and months, or not.

And it's probably down to chance, some combination of genetics, epigenetics, past exposure and probably local conditions having to do with which cells get infected and their location and immune environment. Maybe it's up to the speed at which the immune response, the training in the thymus, occurs. Likely it's not something that medicine will have the technology to look for until many years. For similar reasons that the most powerful computers can only predict the weather so accurately, since the atmosphere is a chaotic system.

Which is exactly where the diagnostic criteria of waiting for 6 months screwed up all the data. Ironically, it's literally people who don't believe in the illness who caused this real perception that recovery is rare, when actually it's that it's rare after 3 months, and around the 5% mark at 6 months. And it is actually rare at the 6 months mark, it's just that this arbitrary timepoint wrecked all the data.

None of this is surprising, any partial and biased count of anything will distort the picture. This is why patients have been begging for decades to get accurate counts. Medicine only has themselves to blame for refusing, leading to not only mass suffering, and the early death of far too many, but the complete distortion of reality.

 

Probably a bit of that, but also mostly having to do with the "not otherwise explained by a known diagnosis" qualifier. Reaching their 70's, people usually have multiple diagnoses, and general fatigue can and will always be put down to ageing, regardless of the age of the patient.

 

Feeds into chemical soup theories .
Potatoes here are generally " treated" with average 32 chemical compounds before harvesting .
There ' s a lack of research into the effects of the cumulative chemical load.

 

What is the true prevalence rate for MEcfs?

I agree this prevalence figure will help advocacy, especially in the US. It's a very large (57k), nationally representative study from the CDC and the CDC has put its name to a high prevalence rate for ME/CFS. This will be a real asset for advocacy.

At the same time, a good handle on prevalence rates is a fundamental for understanding any illness. We still aren't clear on this for ME/CFS.

The 0.24% CDC/Wichita study has issues: long-term follow-up showed many diagnosed with CFS dipped in and out of meeting criteria in later years. I think it's based on under a hundred diagnosed cases - the Lenny Jason one too. The new CDC study 1.3% figure implies they had around 750 cases in their national survey.

If we can understand why the figures look wrong, it will help us understand what methodology might work better. I think UK Biobank cases are diagnosed with a very similar question to the new CDC one (not sure if the sex ratio and age prevalence have similar issues to the new study).

I'd love to see research/data that would pin down robustly prevalence rates and also onset age and sex ratio. These are epidemiological basics and I think it's time they were addressed properly It's not as if we understand what's going on in ME. We shouldn't still be relying on a couple of small studies from the 20th century.

 

Sure!

I am not against having better data.

However, a good way to obtain the best/correct epidemiological data is to work with the CDC numbers, aiming to increase funding. This funding can be used, among other things, to support epidemiological studies.

Additionally, I have doubts that we will see significant improvements in the next decade or so, primarily due to considerable ambiguities in terms of diagnosis (including criteria, the absence of biomarkers, untrained MDs, etc.).

 

It seems like this question would be an obvious candidate for a health records based study. We've seen all kinds of EHR studies on LC, why not ME? I suspect the coding is poor and there would be some work needed to clean things up, but getting a lower bound for diagnosed ME doesn't seem like it should be challenging.

 

It seems to me that it would be fairly easy and useful to do some statistically significant sampling of health records within the CDC study or a similar one (sampling a subset of the patients in the survey to have doctors appointments would also be great, but probably too much effort).

The diagnosis might not be accurate, but at least one then knows, especially for future studies, whether the method of asking someone if they have a diagnosis is somewhat accurate and would also have some information on which diagnosis is being referred to (Fukuda vs IOM vs CCC or even just chronic fatigue, and you could also find out why patients answered incorrectly leading to better questionnaires next time around). Perhaps that is something that could still be done retrospectively?

This of course involves issues of data privacy and naturally involves extra effort, in particular funding and of course the statisticians at the CDC are more than aware of the use of such an extra step and also far better methodologies, but perhaps the specific relevance to ME/CFS isn’t clear enough or it isn’t clear why the extra effort involved could be worth the effort in ME/CFS.

After all as @Simon M points out thus far the epidemiological basics haven't been covered yet. Hopefully the CDC survey can be used for advocacy so that at least that can change. I'm sure even some people at the CDC want to know how accurate these horrendous numbers they've just gotten are.

 

I agree that it would be great to cross check the answer to the question about being diagnosed with ME/CFS against other data. I don’t know if they have links to respondents‘ health records of permission to access them. As you say, that could improve the question used and the interpretation of it – The findings could be used to inform other studies as well.

Maybe there is other data within survey that we could use, such as quality of life data, particularly on physical function.

if neither of those are possible, I wonder if a future survey could have a follow-up question or questions to a subset of respondents.