CFS/ME starts most often age 10-19 & 30-39: Norwegian population study

[Bob]

Thanks for the clarification, Woolie.

 

[wastwater]

I find the graph on schizophrenia interesting,with a lot of neuropsychatric illnesses having a typical age of onset in late adolescence

 

[wastwater]

I wonder if there is a different presentation in the two groups ie more neurological in the 10-19 and more muscular in the 30-39 or vice versa or any other stand out differences.

 

[Jonathan Edwards]

@Bob, I understand that both graphs refer to formal diagnosis: the first to the age at which this diagnosis was first made, and the second the percentage of the population in the different groups currently carrying that diagnosis.

Neither records actual symptom onset, which could be many years before formal diagnosis. This period is a difficult one for people with MS (one blogger referred to it as "limboland"). I saw one statistic suggesting that a third of MS patients get a diagnosis of ME/CFS during this period.

I am a bit confused. Are people comparing the two graphs from the same paper or the incidence graphs from the two different studies. It was the difference in incidence at age 10-20 that seemed to me might need explaining more.

 

[Simon]

Not sure if this is a point worth making but if ME is, say, three times more prevalent in women and you just multiply up the men's figures by three, you'd get a more prominent peak in women just by virtue of that multiplication.

I tried to do that by eyeball and that women's graph looks to me like three of the men's graphs stacked on top of each other.

Our eyeballs are calibarated differently! I take the point about multiplication, but compare the two clear peaks in the female graph vs one clear early peak for men - the shape of the graph is quite different, and multiplication won't change that. Eg the female second peak is sightly bigger than the first peak, and it has more consistent highs than the male one, giving it a more credible peak (looks like a lot of noise in the male data rather than a clear second peak).

Figure One: twin age peaks:


(this image may be copyright: it's from the linked open source paper above, but please don't reproduce without linking to the original paper and explaining it may be copyright).

 

[Bob]

Something significant to note in relation to the MS graphs: One illustrates symptom onset, while the other indicates age of diagnosis, so they may not represent the same type of data. This would probably explain the huge difference in the 11-20 age bracket, perhaps indicating that people get a diagnosis years after the initial (often gradual) onset of symptoms.

@Bob, I understand that both graphs refer to formal diagnosis: the first to the age at which this diagnosis was first made, and the second the percentage of the population in the different groups currently carrying that diagnosis.

Neither records actual symptom onset, which could be many years before formal diagnosis. This period is a difficult one for people with MS (one blogger referred to it as "limboland"). I saw one statistic suggesting that a third of MS patients get a diagnosis of ME/CFS during this period.

I am a bit confused. Are people comparing the two graphs from the same paper or the incidence graphs from the two different studies. It was the difference in incidence at age 10-20 that seemed to me might need explaining more.

Ah, well spotted Jonathan. I was commenting on the two MS graphs that have been posted in separate posts re (1) age at symptom onset and (2) age at diagnosis. I had assumed that @Woolie was looking at the same but it's seems we're probably at cross-purposes and woolie was commenting on the two graphs posted side-by-side in Simon's post.

Going back to my comment, I think that would explain the differences between the graphs, but it's not explicitly explained in the research paper what they are actually referring to by "symptom onset", and the study is only n=121. It's not a large scale epidemiology study.

MS graph 1: (click here)
MS graph 2: (click here)

 

[Sasha]

Our eyeballs are calibarated differently! I take the point about multiplication, but compare the two clear peaks in the female graph vs one clear early peak for men - the shape of the graph is quite different, and multiplication won't change that. Eg the female second peak is sightly bigger than the first peak, and it has more consistent highs than the male one, giving it a more credible peak (looks like a lot of noise in the male data rather than a clear second peak).

I've got to say, they look pretty damn similar in shape to me! But as you say, maybe that's our eyeballs. The micro-spiking for the female data looks odd and gappy on that peak, but I take your point. I just had a look to see if the numerical data are given for that chart so we could just multiply up the male data and see what it looks like, but it's not given.

 

[Marco]

Just flagging this up before I forget.

Digging around on how you might explain bimodal distributions in disease there are various possibilities - but for Hodgkin's and other cancers with a bimodal distributions, psoriasis and possibly schizophrenia where early incidence is distinguished by much greater severity, the most likely explanation as per @Jonathan Edwards is that you are dealing with two different diseases with overlapping distributions.

This article discusses bimodal cancers :

Here are the causes for cancer multimodality (multiple peaks in a graph of cancer occurrences by age)

"1. Multiple environmental causes targeting different ages
2. Multiple genetic causes with different latencies
3. Multiple diseases classified under one name
4. Faulty or insufficient data
5. Combinations of 1,2,3 and 4"

http://julesberman.blogspot.fr/2009/01/medical-importance-of-bimodal-cancers.html

and this is a very interesting paper from 1966 discussing the epidemiology of Hodgkin's :

http://cancerres.aacrjournals.org/content/26/6_Part_1/1189.full.pdf

 

[Sasha]

Just wondering if this could tie in with Ian Lipkin's finding that the disease changes its nature three years in, but I can't see how. Just mentioning it in case anyone else can!

http://phoenixrising.me/archives/19083

Simon said:

Two new patient types revealed
Ian Lipkin said they found something very surprising in their data: there appear to be substantial differences in profiles between those people who have had the disease for 3 years or less and those who have had the disease for more than 3 years. And that this is important because, “it could have implications for therapy as well as for diagnosis”.

In the ‘early group’, who have been ill for less than 3 years, there seem to be a number of markers suggestive of some sort of allergy aspect. For example there are often increased numbers of eosinophils in blood, with more differences in cytokines. But while this is tentative, he was more confident in the finding the IL-17 was elevated in these ‘early’ patients, compared with reduced levels in patients ill for over three years.

Lipkin said that he thought this 3-year division could be very important and that it hadn’t been something they were looking for; but would be looked at in any future work they did.

Cerebrospinal fluid: different differences
Lipkin’s team also found differences in cerebrospinal fluid biomarkers between patients and controls. I’m not sure from what he said if there were significant differences between the ‘early’ group and patients who had been ill for more than three years. However, they did find that patients had elevated levels of the TH-2 type cytokines IL-10 and IL-13 and elevated levels of four TH-1 cytokine: IL-1 beta, TNF-alpha, IL-5, and IL-17. Lipkin said this is compatible with a profile of some particular types of response that may provide insights into immunological dysregulation in Chronic Fatigue Syndrome.​

 

[Jonathan Edwards]

Just wondering if this could tie in with Ian Lipkin's finding that the disease changes its nature three years in, but I can't see how. Just mentioning it in case anyone else can!

I suspect they will have studied PWME over 18, if their ethics committee restrictions are anything like the UK ones. And like you I cannot see how it would relate to the histograms in relation to 3 years, although some of the 'more than 3 years' might be childhood onset disease of course.

 

[MeSci]

Just wondering if this could tie in with Ian Lipkin's finding that the disease changes its nature three years in, but I can't see how. Just mentioning it in case anyone else can!

http://phoenixrising.me/archives/19083

Simon said:

Two new patient types revealed
Ian Lipkin said they found something very surprising in their data: there appear to be substantial differences in profiles between those people who have had the disease for 3 years or less and those who have had the disease for more than 3 years. And that this is important because, “it could have implications for therapy as well as for diagnosis”.

In the ‘early group’, who have been ill for less than 3 years, there seem to be a number of markers suggestive of some sort of allergy aspect. For example there are often increased numbers of eosinophils in blood, with more differences in cytokines. But while this is tentative, he was more confident in the finding the IL-17 was elevated in these ‘early’ patients, compared with reduced levels in patients ill for over three years.

Lipkin said that he thought this 3-year division could be very important and that it hadn’t been something they were looking for; but would be looked at in any future work they did.

Cerebrospinal fluid: different differences
Lipkin’s team also found differences in cerebrospinal fluid biomarkers between patients and controls. I’m not sure from what he said if there were significant differences between the ‘early’ group and patients who had been ill for more than three years. However, they did find that patients had elevated levels of the TH-2 type cytokines IL-10 and IL-13 and elevated levels of four TH-1 cytokine: IL-1 beta, TNF-alpha, IL-5, and IL-17. Lipkin said this is compatible with a profile of some particular types of response that may provide insights into immunological dysregulation in Chronic Fatigue Syndrome.​

Mady Hornig found much higher levels of interferon-gamma in patients ill for less than 3 years, according to this post.

 

[Simon]

At the risk of over-analysing a single set of data, I wanted to make a few more comments on the 'twin peaks' graphs and what it might mean

I tried to do that by eyeball and that women's graph looks to me like three of the men's graphs stacked on top of each other.

Here's what it actually looks with men scaled up by 3:

My interpretation of this:
1. There is a definite male adult peak (like Sasha said), though it doesn't appear to be as strong or as clear as the female one
2. So any explanation needs to account for a male adult peak, as well as a stronger female one
3. The relatively bigger 'young' male peak may be an artefact: the 3x factor seems to cut in at puberty (seen in this study as well as one by Esther Crawley) so some of the early male data has probably been scaled up inappropriately as I scaled all male data x3 (hey, even that was pushing my photoshop skills)

• On that 3x factor, a separate Norwegian study found higher levels of fatigue in women than men, though the difference was less pronnounced " More women (26.2%) than men (19.8%) experienced high fatigue (p = 0.004)= 1.3x, though I have seen higher ratios in other (non-Norwegian) studies. But that does suggest the very high female ratio isn't a generic fatigue thing.

• Going back to the idea that childbirth cold be driving the female adult peak, Norwegian women had their first child at an average age of 28 (in 2000, approx midpoint of this study), which seems a bit young to account for the female peak, unless the factor is through having young children or a 'doseage' effect of multiple births. I have heard there was a link between autoimmunity and pregnancy, but wasn't sure if there was any data to back that up.
• The authors of this Norwegian study themselves suggested childbirth and resullting hormonal changes were the most likely explanation for the adult female peak in the graphs
  
• We still need to account for the male adult peak

 

[Marco]

We still need to account for the male adult peak

Sympathetic pregnancy?

 

[MeSci]

Sympathetic pregnancy?

But how many women with ME haven't had children? There seem to be several here...but maybe that's because many of the others are busy looking after children?

 

[Jonathan Edwards]

I think the timing is wrong for pregnancy and kids and even if fathers undergo physiological changes during pregnancy I very much doubt they bear any relation to the ones mothers do. And anyway if all the peak was due to pregnancy somebody would have twigged to the association by now.

 

[Sasha]

A
Here's what it actually looks with men scaled up by 3

Photoshop! That was clever. I was going to measure each bar in mm and put it in a spreadsheet and.... So let's just say I'm really glad you posted this sooner rather than later.

• Going back to the idea that childbirth cold be driving the female adult peak, Norwegian women had their first child at an average age of 28 (in 2000, approx midpoint of this study), which seems a bit young to account for the female peak, unless the factor is through having young children or a 'doseage' effect of multiple births. I have heard there was a link between autoimmunity and pregnancy, but wasn't sure if there was any data to back that up.
• The authors of this Norwegian study themselves suggested childbirth and resullting hormonal changes were the most likely explanation for the adult female peak in the graphs
  
• We still need to account for the male adult peak

It's interesting that a lot of women WME say that their symptoms improve while they're pregnant.

As you say, doesn't account for a male peak.

 

[Marco]

But how many women with ME haven't had children? There seem to be several here...but maybe that's because many of the others are busy looking after children?

Just kidding. I don't have kids. Didn't seem sensible with no guarantee of being able to look after them.

 

[MeSci]

Just kidding. I don't have kids. Didn't seem sensible with no guarantee of being able to look after them.

But the suggested association is that something about having kids can cause ME, isn't it?

Anyway, after several fruitless generic internet searches for reference ranges for hormones and variance with age, I settled for a Wikipedia entry, which has a graph for bone density showing male and female peaks around our second ME/CFS peak. There is also stuff about various hormones varying with age. It says, for example, that DHEA declines steadily with age. I have read something about us having low DHEA - not sure if it's anecdotal.

Even with a steady decline, it's possible that going below a threshold could increase risk, isn't it?

 

[Marco]

But the suggested association is that something about having kids can cause ME, isn't it?

Even with a steady decline, it's possible that going below a threshold could increase risk, isn't there?

I personally don't see any strong reason to associate the two. I do agree though that a threshold or rate of change of 'something' could be important.

 

[Kati]

Perhaps the 2 peaks represent hormonal vulnerability.