Typing myalgic encephalomyelitis by infection at onset: A DecodeME study, 2023, Bretherick et al

[forestglip]

They are just testing individual answers on the questionnaire against the likelihood of a self-report of 'moderate, severe or very severe ME/CFS' (as opposed to a self report of 'mild' ME/CFS. [...] Is someone who says that they are male more likely to report that their ME/CFS is not mild? Also yes.

Are you sure? Isn't that the simple formula they used for females being more severe? I think figure 5 is a model with every single symptom and variable at once:

To address “What symptoms are associated with severity?” for 80 symptoms we used the model: Severity ~ age + sex + symptoms + intercept – Figure 5.

 

[Hutan]

Are you sure?

No, I'm not sure about most of this. I don't think the paper is clear.

 

[forestglip]

No, I'm not sure about most of this. I don't think the paper is clear.

I think it probably is females are more severe. I wish they would have done sex adjusted for non-symptom variables like age and duration of illness. But in the case of age, it would probably make the association even stronger since the males were older. Not sure about duration, not sure if they say which sex was ill longer.

I'm not sure how useful this association is anyways. It shows that females diagnosed with ME/CFS and who enrolled in DecodeME are more likely to self report as more severe than males.

Maybe females aren't as likely to realize they have ME/CFS until their illness is more severe, possibly because of similarities with common comorbidities. Or they may be less likely than males to be diagnosed by a doctor if mild.

Edit: But I think it still might go to that Wikipedia editor's point:

When you have a large bias in diagnosis rates between genders, often the threshold is lower for the more-diagnosed gender, and they are affected less severely.

I don't know if that assumption is valid, but if it is, then this does seem to agree with their point. The threshold for diagnosis appears to be higher for females since those who are diagnosed and enter a study are more severe. If there was no real sex bias, you would thus expect the opposite observation for sex differences: more males getting diagnosed.

 

[Hutan]

I think it probably is females are more severe.

I can't square that conclusion with Figure 5.
@Chris Ponting, sorry to bother you, but can you or another author please explain what Figure 5 is showing with respect to male participants in the study and reported severity?

 

[Trish]

Given that there are males and females across the full range of severity, and any individual can experience the full range of severities over time, I don’t know how it can be claimed ME is more severe in one sex or another.

 

[forestglip]

I don’t know how it can be claimed ME is more severe in one sex or another.

In this study, it's more like just, how severe are males vs females at the specific time they complete the study. It might point to selection bias. If females are more severe, maybe that means mild females aren't as likely to enroll or be diagnosed.

Or maybe it really could be females are more severe for some reason. Maybe if 100 males and 100 females developed ME/CFS at the same time, and you asked them all a year later about severity, maybe 80 males are mild vs 20 severe, while 20 females are mild vs 80 severe.

 

[Chris Ponting]

I can't square that conclusion with Figure 5.
@Chris Ponting, sorry to bother you, but can you or another author please explain what Figure 5 is showing with respect to male participants in the study and reported severity?

Sorry to be late in responding. The analysis for Figure 5 considered the association of 80 symptoms as well as age and sex in the same model. The important thing to say is that 62 of these symptoms are significantly female-biased, so these tend to favour greater severity in the model; being male will tilt this bias back somewhat but not completely. The key analysis that we used to infer greater severity when female is different: the comparison of those with mild ME/CFS against the two-thirds with moderate, severe or very severe illness (sex: p=4.5×10-4). This analysis did not account for the 80 symptoms - including the 62 that are female-biased. Hope this helps.

 

[Hutan]

Thanks @Chris Ponting.

So 'being male' was an input in your severity prediction model; and 'being male' tended to predict reporting severity as greater than mild?

It looks as though, if you took the top ten factors to predict severity being worse than mild in the DecodeME sample, being male is one of the factors. Other factors include reporting 'fatigue is disabling', increasing age, reporting fatigue greater than 50% of the time, reporting difficulty remaining standing.

I didn't see simple percentages of males and females for each severity rating in the DecodeME sample. Is the data available so I could calculate those?

 

[Fero]

There's some people (usually AFAB, I think) who report some improvement when taking progesterone, but I don't know if that's been studied.

I was so much worse with too much progesterone.

My avatar shows my resting HR over 6 weeks when I was accidently taking twice the amount of progesterone. My HR shot up over the 2 progesterone weeks to reveal this sine wave pattern.

That was a terrible time. Starting improving the month I was on the correct dose of HRT - ie I was able to speak to my carers for the first time.

(I had early onset menopause because of ME)