We need a diagnostic test for fatigue induced by sustained activity

Not easy at all. There are just too many factors that change during exertion ... and time of day, and day to day, and what they ate and what microbes they encountered, and different responses to exercise for each individual and time since the last exercise. All that just to find one way of triggering PEM. "Who pulled the trigger?" doesn't reveal much about the damage to the body or how to fix it.

 

With the PEM/non-PEM testing, you can add individuals who have reliable cognitive triggers, as another comparison. I think cognitive-triggered PEM is an even better subject for testing, since it should change fewer factors elsewhere in the body.

 

I'm sorry there aren't more people who have found reliable PEM-blockers. That would add another comparison: trigger activated, but with or without the symptoms.

 

I've started another thread about COX-1 inhibitors since they help to minimize PEM (though not completely block it) for me. A couple others have responded that they help lessen PEM though not everyone reported the same effect.

The only thing that has completely blocked PEM for me is malic acid, though I seem to get more benefit from it than others potentially because I'm on a stimulant. It also doesn't seem to help universally.

I think you run the risk of picking up on underlying physiological differences that make something work to block PEM in some but not others. Though that's not a guarantee, still might be a useful comparison.

Thanks for the suggestions!

 

I'd definitely start doing measurements during activity for that reason, though there's a distinct possibility that there's a whole cascade that happens leading to PEM.

I think there's two possible markers to think about here:

1) What can you detect during activity that is predictive of PEM at a later time?

2) What marker distinguishes currently being in PEM vs. not being in PEM?

During activity you might find some "trigger" that's correlated with PEM at a later time, but you don't know if that same signal will still be around when PEM finally happens. There might be a couple degrees of remove in between the activity and what's going on during PEM, so a timecourse study would probably be most appropriate.

 

What does ‘timecourse’ mean in practice in this context? Multiple blood draws over time?

 

Yes exactly. It would add logistical complexity but I think it would be necessary.

What I'm thinking is:

1) multiple pre-activity draws to establish a baseline distribution.
Since any metabolite can feasibly fluctuate over time, we need a sense of how much each metabolite typically fluctuates outside of PEM.

2) multiple draws at regular intervals during activity.
The exact time interval would probably be determined by the activity tolerance of the participants. I'd do at least 3-4 draws over the course of the activity.

3) multiple post-activity draws, starting at least a few hours after activity and then once every 24 hours for 3 days.
Ideally you'd be able to get a post-PEM sample to compare to your baseline measurements.

Biggest problem with this is it just becomes extremely financially and logistically difficult the more samples you need to process.

I'm getting a headache just thinking about how I could plan something like an LC-MS panel for each sample without introducing lots of confounders re: sample processing.

You'd almost need to pay for a participant to be posted up in a hotel for an extended period of time and have multiple lab techs on deck and reserved machine time to process samples instantly.

Though there might be more wiggle room than I'm anticipating here. I'd definitely be consulting with a lot of other researchers before planning something like this (if funding ever made it possible)

 

Røysumtunet in Norway might be the only viable option:
https://www.s4me.info/threads/norway-røysumtunet-a-centre-for-severe-me-patients.13841/

It would be a small sample size, though.

 

Probably would have to be a small sample size anyway to afford to process all those samples.

 

Yes that’s definitely on my mind as well. I think that’s what would inevitably happen in a signaling cascade.

What you would need for a regression type-analysis to zero in on a marker is at least one point in time where the participant can clearly say “yes, this is PEM” and another where they can clearly say “I’m not in PEM right now.”

That’s just a necessity so your model can exclude whatever is not changing between those two states. It might be harder for some pwME to be sure about those judgements, so you’d probably have to start with people who are more confident about it.

Another option would be to do a linear regression to see which signals correlate with degree of PEM-ness as reported by the participant, but that might be trickier due to the subjectivity you note. It’s possible to add splines and such to account for non-linearity, though it gets trickier the more complex the system.

My hope would be that a biological PEM signal is strong enough to show a correlation despite some of the variation inherent in subjective judgements.

 

How about just having participants track their symptoms every day at home over several weeks? I'm not sure there's any good way to do blood samples, but maybe everyone could store urine samples in a fridge every day on their own, to bring to a lab later. (Ideally the researchers would provide a mini-fridge for this.)

And then look to see if any chemicals in the urine are correlated to reported symptom level within individuals.

 

It’s a possibility, though many signaling molecules of interest may have a very short half-life (on the scale of minutes/hours), which is why I’m leaning towards some arrangement of instant processing.