Digital physiological biomarkers predict within-person symptom changes in complex chronic illness, 2026, Aitken et al

Abstract changed as well:

Digital physiological biomarkers predict within-person symptom changes in complex chronic illness

Aitken, Annie; Sawyer, Abbey; Iwasaki, Akiko; Krumholz, Harlan M.; Preston, Rory; Calcraft, Paul; Leeming, Harry; Tosto-Mancuso, Jenna; Proal, Amy; Osborne, Michael A.; Putrino, David

Abstract
Altered heart‑rate variability (HRV) and resting heart rate (HR) are common in many complex chronic conditions. Mobile and wearable technologies now provide real-time, valid measurements of HRV and HR, advancing symptom monitoring and management.

The current study integrates a 60-s morning PPG assessment with evening symptom severity reports, yielding a high-density mobile health dataset (n = 4244) with an average of 125 biometric observations per participant. We examined whether within-person fluctuations in HR, HRV, and respiratory rate predicted daily changes in crash, fatigue, and brain fog symptoms and secondarily evaluated model predictive performance.

Model fit and variance explained were highest in models that included morning biometrics in addition to prior-day symptom reports and covariates. Within-person increases in HR and decreases in HRV in the morning were associated with worsening symptom reports in the evening. Walk-forward cross-validation showed a statistically significant improvement in model performance when morning biometrics were added to prior-day symptom reports (AUC = 0.82–0.85 vs. 0.73–0.83).

These findings represent the prospective utility of mobile health tools for precision monitoring and prediction of real-time symptom exacerbations in complex chronic illness.

Web | DOI | PDF | npj Digital Medicine | Open Access
 
They seem to be measuring raised heart rte and reduced HRV for a short period in the morning. That is presumably not going to reflect any episodes of exertion during the day. So it will not be a measure of how much you have overdone things, just a measure of maybe it being a bad day on which to do things. But to be sure that actually was the case I think you would need to show that exerting on those days really did produce more PEM than other days.

That seems to get a bit too complicated.

And presumably you cannot do much about the fact that your heart thinks it is a bad day to do stuff when it does.

I keep getting this sense that people have borrowed these measurements from the old deconditioning model and that a model of what exertion determines PEM needs to start afresh.
 
Utsikt said:
If it’s from Visible, you’re supposed to take a reading first time in the morning. I tried re-doing mine a few minutes later and they gave vastly different results.
Click to expand...

And wouldn't the result for that first minute depend on whether you were already feeling grumpy and trying to wash your face when the water was too cold or whether the sun was shining and someone had brought you a special cake for breakfast in bed? I can't quite get my head around how this is going to be any use.
 
I have often wondered when the day will come when computers rule the world and we are just their servants tending to their needs while foolishly thinking we are still in charge.

A little while ago I realised that that day already came.
 
Utsikt said:
What they are missing: the participants’s morning prediction of evening symptoms. We have no idea if this is better than what you can manage on your own.
Click to expand...
If I had a bad sleep, I can feel it upon waking up. On those nights I tend to have lower HRV.
If my HR is high upon waking, i.e. higher than usual, I can tell that it's going to be a rough day and my HR will remain higher than on a good day.
If I had a bad sleep, high HR (higher than usual) the following day is guaranteed.

The thing is, all of this is pretty clear to me within 30 mins upon waking, purely based on how it feels in my body and it goes beyond HR. It's about how wrecked I'm feeling.

While I was moderate, a good test for the day was making breakfast. I used to get up as soon as I'd wake up and make myself breakfast. The way I felt during it was a good predictor for the rest of the day.

The wearables are not telling me anything I don't know. They only put some numbers on it.

And because I feel worse than on a good day (although it's a blasphemy to call any of my days a good day), naturally I will try to do less.


Utsikt said:
If it’s from Visible, you’re supposed to take a reading first time in the morning. I tried re-doing mine a few minutes later and they gave vastly different results.
Click to expand...
I'm not an expert on HRV and things that I've read don't add up and don't necessarily correspond to some of my observations but I think ideally you'd be working with HRV during sleep. The old Visible band has to be charged during night, so the reading upon waking up is used as a surrogate. Their new band is able to give HRV reading during sleep.

Different wearables calculate HRV differently. My morning Visible HRV is always within a narrow interval and I don't trust it. I get more variation in HRV readings from another wearable which calculates it during sleep and it correlates better to how I'm feeling upon waking up. That being said, I do not want to read into HRV too much as indicated in the previous paragraph.
 
Felis Catus said:
What do you mean here (in bold)?
Click to expand...

The literature that gets referred to focuses on reduced HRV with deconditioning as opposed to fitness in athletes where HRV is highest. I am not sure why a short time sample of HRV should tell us anything useful about the risks of PEM on a particular day. I would have thought PEM was likely to coccur after days when a big difference in heart rate occurred because of exertion.
 
Jonathan Edwards said:
And wouldn't the result for that first minute depend on whether you were already feeling grumpy and trying to wash your face when the water was too cold or whether the sun was shining and someone had brought you a special cake for breakfast in bed? I can't quite get my head around how this is going to be any use.
Click to expand...

You always have to take the reading when you’re in the same state – that is, at rest, before washing your face, eating cake or soaking up the sun ;)
If I get up, my heart rate shoots up, and only then do I remember I need to take a reading, the result ends up all over the place.
Being in a bad mood doesn’t seem to have a direct impact.

I’ve been using Visible to take readings for a year now.
Looking at the data now, it’s clear that the times when I was feeling worse were accompanied by a lower score on Visible.
Now that I’m a bit fitter, I’m consistently getting a better score.

Is that really useful to me? No. But it is interesting...
 
Jonathan Edwards said:
The literature that gets referred to focuses on reduced HRV with deconditioning as opposed to fitness in athletes where HRV is highest. I am not sure why a short time sample of HRV should tell us anything useful about the risks of PEM on a particular day. I would have thought PEM was likely to coccur after days when a big difference in heart rate occurred because of exertion.
Click to expand...
Ok. My understanding from the literature I've read is that decrease in HRV can indicate overtraining in athletes. So translated to ME/CFS world, it could indicate overexertion from which one hasn't recovered. I say "can" and "could" because other factors can influence HRV.

Visible takes into account the data from previous days. I don't know how many.
HR, HRV, reported sleep quality and reported symptom scores.
 
Jonathan Edwards said:
Isn't that the exception though?
Click to expand...
I'm not sure what do you mean by exception. That overtraining is an exception which happens rarely or that decrease in HRV (following overtraining) is rare?

HRV seems to be used, among other things, to guide optimal training load.
Jonathan Edwards said:
Training increases HRV 95% of the time surely?
Click to expand...
Apparently high intensity training decreases HRV for up to 72h.
You'd expect to see an increase in HRV over time, assuming adequate rest and everything else being ok in their life.


My impression is that there is a lot more nuance in interpreting HRV than can be summarised in a couple of sentences. I don't know if that's because the whole thing is weak and people are pretending to be smart OR simply because you can't use HRV on its own out of context to guide anything.
 
Last edited:
Jonathan Edwards said:
I keep getting this sense that people have borrowed these measurements from the old deconditioning model and that a model of what exertion determines PEM needs to start afresh.
Click to expand...
It's especially annoying that the data on this have thoroughly debunked the deconditioning model, and it makes no difference. Here we are talking about variations over days, minor fluctuations that are more likely to be consequence than cause. But the entire basis of pinning everything on deconditioning, stemming from the original made-up lies by Wessely and his gang where they imagined scenarios of severe acute illness leading people to be bedbound for days, weeks even, is supposed to find significantly lower HRV, as it's claimed to be a marker of physical fitness.

We are obviously not physically fit. I have not walked more than 25 minutes in a decade, at a slow pace. I cannot do more weight lifting than 16 lbs in each hand for my shoulders, to relieve pain in my back. And yet for my age I have an excellent HRV, about 55 on average.

But that's not what's found here. There are variations, fluctuations, that don't seem common, but we don't see things like sub-20s HRV as we should expect with severe deconditioning. Either HRV isn't actually a useful measure of fitness, or there's something odd about us that should be a clue about how we are different.

But with the recent rehab obsession in Long Covid, no one has even bothered to reason beyond "well, they don't function, so we'll have them function, with coaching and music and a petting zoo", with zero regard for any of the old myths that made them go for this debunked myth in the first place.
 
Felis Catus said:
My impression is that there is a lot more nuance in interpreting HRV than can be summarised in a couple of sentences. I don't know if that's because the whole thing is weak and people are pretending to be smart OR simply because you can't use HRV on its own out of context to guide anything.
Click to expand...

Yes, it seems likely that the relationship between heart rate at various times and symptom patterns in ME/CFS will be very complicated. I just get the impression that a short window measurement of HRV has been lifted from other contexts and applied to fatigue situations without any real thinking about why. My own thought about actometry was that it should be done 24hrs a day for weeks and patterns that might be specific for ME/CFS extracted using statistical methods. I doubt those patterns would have anything to do with a daily HRV measure.
 
Jonathan Edwards said:
And wouldn't the result for that first minute depend on whether you were already feeling grumpy and trying to wash your face when the water was too cold or whether the sun was shining and someone had brought you a special cake for breakfast in bed? I can't quite get my head around how this is going to be any use.
Click to expand...
I don't know how well most people follow the instructions, but it's supposed to be first thing in the morning, after waking up, while still in bed, to avoid some of that exertion. But that can't be verified, so the data are only marginally better than what a questionnaire allows.

I don't personally see significant differences from repeated measurements. About the only one I can observe is that my resting heart rate can shoot up by 10-15 from simply moving my body into a resting position, but waiting for 1-2 minutes for it to go down usually gives me reliable repeated measurements. It is supposed to measure 'resting' rate, after all, and while it's not supposed to shoot up by 20-30 from simply rolling around in bed, it is the case for many of us, and that's interesting data, but clearly no one knows what to do with this.

Also measuring this in the morning should be seen as a function of yesterday's (and 1-2 days before) exertion, not today's, which is consistent with trying to relating this to PEM. I do personally observe this. When I am overdoing it, my HRV will tend to lower by 5-10, and my RR to increase by 10-25.

But it's all after the fact, this tells me that I have overdone it in recent days, and on that I can say that it's rather good. Not any better than how I feel, but I have ignored increasingly feeling unwell far too many times not to trust the morning readings, or rather whether they are elevated compared to a time when I am better. It takes years of data to see those trends.

The biggest differences by far are when I am sick, then my HRV can drop in the 30s and my RR shoot up to 100+. Which then that clearly has nothing to do with fitness. So then I don't know what the hell anyone is supposed to do with this, if it doesn't even measure what it's supposed to measure, and those measures don't even support the working model anyway, and the model still completely dominates because nothing matters.
 
Felis Catus said:
I'm not an expert on HRV and things that I've read don't add up and don't necessarily correspond to some of my observations but I think ideally you'd be working with HRV during sleep. The old Visible band has to be charged during night, so the reading upon waking up is used as a surrogate. Their new band is able to give HRV reading during sleep.
Click to expand...
On that, I have worn a device 24/7 for a while and it mostly compares to what Visible gave me. It's not perfect but the differences are roughly the same, basically similar data in a shifting base. It's probably still better to have it during sleep, but it only adds little accuracy overall.
 
Felis Catus said:
I'm not sure what do you mean by exception. That overtraining is an exception which happens rarely or that decrease in HRV (following overtraining) is rare?
Click to expand...
Which actually tracks with most research converging on medium intensity exercise providing the best benefits, and low-intensity being good enough anyway, with the best return on investment. High-intensity quickly starts causing problems that cancel out many of the benefits of simple activity, and it does seem to more or less track similarly to overtraining. In a sense maybe this is what PEM is, overtraining in a body where the line for overtraining is below activities of daily living, or something like that.
 
What HRV metric is being used here, I can't see it skimming through the paper? Is it some wearable derived score where we don't even know what it's supposed to be measuring or something more standard like SDNN? I agree with @Jonathan Edwards . I think we need someone sensible to start these things back from the beginning. I don't see that happening in the "wearable field" and with app providers. You need someone that genuinely wants to understand something not someone who believes they already do and want to sell you that idea. Gather genuine data over many months, don't feed the output to the patients and then let statisticians analyse the data carefully to see if anything pops out.
 
You must log in or register to reply here.
Back
Top Bottom