Two age peaks in the incidence of chronic fatigue syndrome/myalgic encephalomyelitis: a population-based registry study from Norway…, 2014, Bakken+

[SNT Gatchaman]

Two age peaks in the incidence of chronic fatigue syndrome/myalgic encephalomyelitis: a population-based registry study from Norway 2008-2012

Spoiler: Authors

Bakken, Inger Johanne; Tveito, Kari; Gunnes, Nina; Ghaderi, Sara; Stoltenberg, Camilla; Trogstad, Lill; Håberg, Siri Eldevik; Magnus, Per

BACKGROUND
The aim of the current study was to estimate sex-and age-specific incidence rates of chronic fatigue syndrome/myalgic encephalomyelitis (CFS/ME) using population-based registry data. CFS/ME is a debilitating condition with large impact on patients and their families. The etiology is unknown, and the distribution of the disease in the general population has not been well described.

METHODS
Cases of CFS/ME were identified in the Norwegian Patient Register (NPR) for the years 2008 to 2012. The NPR is nationwide and contains diagnoses assigned by specialist health care services (hospitals and outpatient clinics). We estimated sex-and age-specific incidence rates by dividing the number of new cases of CFS/ME in each category by the number of person years at risk. Incidence rate ratios were estimated by Poisson regression with sex, age categories, and year of diagnosis as covariates.

RESULTS
A total of 5,809 patients were registered with CFS/ME during 2008 to 2012. The overall incidence rate was 25.8 per 100,000 person years (95% confidence interval (CI): 25.2 to 26.5). The female to male incidence rate ratio of CFS/ME was 3.2 (95% CI: 3.0 to 3.4). The incidence rate varied strongly with age for both sexes, with a first peak in the age group 10 to 19 years and a second peak in the age group 30 to 39 years.

CONCLUSIONS
Early etiological clues can sometimes be gained from examination of disease patterns. The strong female preponderance and the two age peaks suggest that sex-and age-specific factors may modulate the risk of CFS/ME.

Web | PDF | BMC Medicine | Open Access

 

[SNT Gatchaman]

Older paper (2014) predating S4ME, but relevant for review due to —

Audrey Ryback talked about her work on electronic health care records in Lothian. This is impressive. They appear to have a genuinely population based cohort. The prevalence was up at 0.8% - 7,000 cases out of 900,000 people. In other words everyone with ME/CFS in a defined area. This begins to look like solid population data. (And I suspect GPs in Lothian are relatively good about coding ME/CFS compared to some places.)

The data provide a variety of measures of disease dynamics. Audrey stayed with us Wednesday night and also showed me work she has been doing with Simon on age of incidence profile which are to me pretty mind-blowing. Clues from the old Norwegian study look as if they are replicable. […] To me it looks as if there is a totally novel aspect to ME/CFS disease dynamics worth a very deep look at.

I'm glad you think that. I am biased, but I've been stunned by the findings.

See general discussion thread The two age peaks in onset of ME/CFS.

 

[Hoopoe]

What is Ryback's work showing? The same two age peaks or something more?

What do these two age peaks suggest?

What are the hypotheses on the reason for the peak in adolescence? Brain maturation associated with becoming an adult? Sex hormones?

What other illnesses show two age peaks?

 

[SNT Gatchaman]

The related thread we have is Early-onset autoimmune vitiligo associated with an enhancer variant haplotype that upregulates class II HLA expression, 2019, Jin+ [two age groups].

Tag made for two age peaks to keep track.

 

[Sid]

I recall this study being discussed at great length. Perhaps it was on the previous forum.

 

[Jonathan Edwards]

What is Ryback's work showing? The same two age peaks or something more?

I think I can say that yes, two peaks come up again. Beyond that I don't want to open this up until Simon and Audrey are happy to do that. The best thing about it is that there is a replication across several different analyses.

 

[Evergreen]

Good to have a thread for this paper. I wonder should the two age peak thread @SNT Gatchaman helpfully linked to above be merged with this one?

Bakken et al.'s main finding was:

We observed a distinct age pattern in the incidence rate of CFS/ME, with the number of cases peaking in the age groups 10 to 19 years and 30 to 39 years.

Thought it would be useful to have Bakken et al.'s figures here, for people who cannot look at the paper. I've prefaced each figure by what Bakken et al. write about each figure in the results section:

Figure 1 shows the number of cases in one-year age intervals for women and men separately. For both sexes, two distinct peaks in the number of cases were observed, the first in the age group 10 to 19 years and the second in the age group 30 to 39 years.

Figure 2 shows the number of cases per year by age category for men and women separately, estimated from the Poisson regression model. The non-overlapping CIs support the existence of two age peaks in the distribution indicated by the raw data (Figure 1).

Figure 3 shows the estimated incidence rate per 100,000 person years for men and women separately. After adjustment for population figures, the pattern with two age peaks is clear for women, whereas a second peak is not evident for men.

 

[Evergreen]

On the general thread, @Russell Fleming highlighted a study Bakken co-authored - Collin et al. 2017 - which found increased healthcare use long before diagnosis. While there are a number of possible reasons for that, I think this gets more interesting with increased awareness of a potentially similar pattern in MS.

This is from Collin et al. 2017. Females are green, males are orange. Solid lines are pwME/CFS. Dotted lines are controls:

 

[Hoopoe]

I remember a turning point where I began experiencing episodes of exhaustion, malaise, poor sleep. I consider that the onset of clinically significant illness, but I believe that 1-2 years before that, I was already showing subtle symptoms. The most important symptom was, occasionally and without any reason, waking up with malaise and as if I had barely slept. Since my onset was at the age of 15 this would fit with the Norwegian data showing a sudden increase in incidence in the age 10-14 group.

Thinking about it my onset was probably actually at the age of 14, one month and a week before my 15th birthday.

 

[Jonathan Edwards]

Thought it would be useful to have Bakken et al.'s figures here, for people who cannot look at the paper. I've prefaced each figure by what Bakken et al. write about each figure in the results section:

One thing I think can be said at this point is that although it looks as if the second peak for men was partly an artefact there are also good reasons to think it is still there. It would be easy enough to link two peaks to changes in female hormone levels but if the male peak is real that doesn't work.

 

[Midnattsol]

Another Norwegian study with age peaks were Tjenesten and MEg, but I don’t remember which part of the project it was mentioned.

 

[Evergreen]

One thing I think can be said at this point is that although it looks as if the second peak for men was partly an artefact there are also good reasons to think it is still there. It would be easy enough to link two peaks to changes in female hormone levels but if the male peak is real that doesn't work.

So if the second male peak is real, would we have to argue that the second peak being stronger for females is because of differences between the sexes' immune systems as they encounter infections, perhaps from children?

Or are there other explanations?

 

[Kitty]

Or are there other explanations?

We'd have to rule out sex hormones as a critical driver, wouldn't we, if it is real? Both female peaks align roughly with changes in hormone levels, but they don't in males. And the second female peak would be more persuasive if it occurred between 40 and 50 instead of 30 and 40.

Presumably it would be quite helpful to be able to rule that out, though I suppose it's difficult to do so entirely.

 

[Peter T]

One thing I think can be said at this point is that although it looks as if the second peak for men was partly an artefact there are also good reasons to think it is still there. It would be easy enough to link two peaks to changes in female hormone levels but if the male peak is real that doesn't work.

I was wondering that, as a male with a sudden onset in my mid thirties associated with an acute EBV infection I had no obvious reasons to suspect any hormonal issues at that point in my life. Also anecdotally the men I know about with adult onset also had their onset in their thirties. I wonder, given the smaller number of males in the various studies, if larger sample sizes would make a second male peak a more robust phenomenon.

I did have lots of viral infections in my twenties, but that was when I was working on a group of islands into some twenty schools and three hospitals. However they were usually mild and short lived and after several years I suspect I adapted to the local virus population, then having brief mild infections only when I visited the mainland. However my health was then pretty stable once I had settled in back into living in England, that is until I got the triggering glandular fever (mono).

Has any one compared levels of sex hormones in both male and female patients with appropriate controls?

 

[Amw66]

We'd have to rule out sex hormones as a critical driver, wouldn't we, if it is real? Both female peaks align roughly with changes in hormone levels, but they don't in males. And the second female peak would be more persuasive if it occurred between 40 and 50 instead of 30 and 40.

Presumably it would be quite helpful to be able to rule that out, though I suppose it's difficult to do so entirely.

Stopped for lunch before a meeting and reading this thread . Random thoughts which may not be relevant.

Hormones . Ratios rather than absolutes may be important.
Whilst focus tends to be on estrogen , it could equally be testosterone.
Taking it back further the basic block is cholesterol which factors in lipids

 

[Jonathan Edwards]

So if the second male peak is real, would we have to argue that the second peak being stronger for females is because of differences between the sexes' immune systems as they encounter infections, perhaps from children?

I don't think it need have much to do with encountering infections. We have to assume that some pathway in women is easier to subvert. It is likely to be "immune" but if involves an innate immune signals like TLR7 it might impact on susceptibility of brain cells without involving immune responses elsewhere to specific antigens. There may be an infection trigger step but it is what happens after that we need to understand.

The second male peak may get lost for technical reasons but it is relatively small cmpared to the first male peak in comparison to women. That raises an interesting issue. If female susceptibility is due to incomplete X chromosome inactivation it ought to operate at age 15 as well as 35. Unless of course the incomplete inactivation is oestrogen dependent.

 

[BrightCandle]

There is also a possibility that the second male peak is being impacted by diagnosis rates. We know that men are less likely to get a diagnosis, its possible the older men are dismissed more than the male children. Because this is gathered from a public health registry its quite likely its impacted by the medical systems own biases and there is a lot of that and Norway has been one of the more problematic places for this over the years and this study is from well before they started to accept ME/CFS wasn't psychosomatic.

 

[Evergreen]

We'd have to rule out sex hormones as a critical driver, wouldn't we, if it is real? Both female peaks align roughly with changes in hormone levels, but they don't in males. And the second female peak would be more persuasive if it occurred between 40 and 50 instead of 30 and 40.

Presumably it would be quite helpful to be able to rule that out, though I suppose it's difficult to do so entirely.

My post followed on from Jonathan's:

It would be easy enough to link two peaks to changes in female hormone levels but if the male peak is real that doesn't work.

My understanding is that if the second peak in males is real rather than an artefact, then we can rule out changes in hormones as a critical driver, because males don't have dramatic changes in hormones at that age in the way that women do.

So if the second male peak is real, we need another explanation. Or one explanation for females and a different one for males.

 

[bobbler]

We'd have to rule out sex hormones as a critical driver, wouldn't we, if it is real? Both female peaks align roughly with changes in hormone levels, but they don't in males. And the second female peak would be more persuasive if it occurred between 40 and 50 instead of 30 and 40.

Presumably it would be quite helpful to be able to rule that out, though I suppose it's difficult to do so entirely.

The age of menarche has got younger over the decades for the general population in industrialized countries (so if it’s linked to industrialization then countries who have become or parts of it have become affected in this way ie have become more industrialized later over the years would show the same trend with a time lag)

No idea re menopause (and I guess that’s been complicated by record keeping and ‘diagnosis’ changes related to medical developments and attitudes) and as you say the peak is 30-40 which could fit to other things

It would be a fearsome calculation fraught with other factors , and would require data on me/cfs that probably doesn’t exist across countries over long time periods

But theoretically I guess it could give clues if it did existed as to how much is ‘explained by directly/indirectly linked with’ those points of hormone change if there were any shifts that correlate with those same trends of menarche age in a country/population ?

 

[Mij]

I was wondering that, as a male with a sudden onset in my mid thirties associated with an acute EBV infection I had no obvious reasons to suspect any hormonal issues at that point in my life. Also anecdotally the men I know about with adult onset also had their onset in their thirties. I wonder, given the smaller number of males in the various studies, if larger sample sizes would make a second male peak a more robust phenomenon.

I had no obvious reason to suspect hormones in my case. I had sudden viral onset at age 29 PVFS that developed into ME in my mid 30s.. The ME specialist I saw in the 90s told me that my age group for ME was the most common he had seen in his 20 years experience.