Daily stress and worry are additional triggers of symptom fluctuations in individuals living with Long COVID... , 2025, O'Connor et al.

[Amw66]

Posting here prior to reading .

It's being circulated by an infectious diseases consultant based in Scotland who undertakes research into long COVID ( and as funding is being bunched together more often , likely will be involved with ME/ CFS )

It would be interesting to determine which scales were used for assessment given that many are flawed for chronic illness applications

ITs a behavioural science paper , so may have methodological issues .

Daily stress and worry are additional triggers of symptom fluctuations in individuals living with Long COVID: Results from an intensive longitudinal cohort study - White Rose Research Online

eprints.whiterose.ac.uk

 

[Trish]

Daily stress and worry are additional triggers of symptom fluctuations in individuals living with Long COVID: Results from an intensive longitudinal cohort study

O'Connor, D.B., Greenwood, D.C. , Mansoubi, M. et al. (18 more authors) (Accepted: 2025) Daily stress and worry are additional triggers of symptom fluctuations in individuals living with Long COVID: Results from an intensive longitudinal cohort study. Annals of Behavioral Medicine. ISSN: 0883-6612 (In Press)

Abstract​

Background: Recent research has shown that exertion in physical, cognitive, social and self-care activities trigger symptom severity in individuals with Long COVID.

Purpose: The current study aimed to investigate whether daily emotional exertions (stress, worry, rumination) were associated with symptom exacerbation, over and above influences of effortful daily activities, in individuals with Long COVID.

Methods: 376 participants were recruited from UK Long COVID clinics and community settings and completed daily assessments of activity and severity of 8 core symptoms every 3 hours for up to 24 days. 155 participants completed daily assessments of stress, worry and rumination for at least seven consecutive days.

Results: Days with higher stress scores were associated with increased severity of all symptoms on the same day, after adjusting for activities, demographic and medical factors (p-values ≤ 0.007). Days with higher stress scores also predicted more severe anxiety and depression symptoms 1-day later (p<0.001) and more severe anxiety (p<0.001) and dizziness symptoms (p=0.003) 2-days later. Days with higher worry scores were associated with increased fatigue (p<0.001), anxiety (p<0.001), depression (p<0.001) and cognitive dysfunction (p=0.002) on the same day, but decreased anxiety (p=0.003) and depression (p=0.002) symptoms 1-day later and less severe pain (p=0.002) symptoms 2-days later. Daily rumination was only associated with two symptoms.

Conclusions: Daily stress and worry are distinct factors linked to fluctuations in same-day and next day Long COVID symptoms, with daily stress showing the strongest association—consistent with patterns of post-exertional symptom exacerbation. These findings highlight the importance of considering stress and worry as potential therapeutic targets and integrating their management into self-care programmes.

https://eprints.whiterose.ac.uk/id/eprint/232563/

 

[Hoopoe]

Clearly the fact that long covid patients are normal human beings that suffer from stress and worries is an unusual phenomenon that must be studied further.

 

[Eleanor]

participants ... completed daily assessments of activity and severity of 8 core symptoms every 3 hours for up to 24 days.

One easy way to reduce your stress level is to avoid assessing the severity of 8 of your symptoms every 3 hours for 24 days.

 

[Yann04]

Stress is correlated with symptom burden.

That sounds wholly unsuprising.

I don’t see how they justify assuming causation, stress -> symptom burden, when the opposite seems more likely?

 

[Hoopoe]

I don't need a person that tries to assume the role of benevolent authority figure that explains my feelings, my levels of stress and its presumed causes, how to manage rest, etc. Because the advice is worse than my own, the interpretations are often very wrong, and I can ask myself when I can't figure out specific things. For this reason the benevolent authority figure does not come across as such. It often looks like a person that can barely hide their prejudices, that can't offer anything useful, and in the end is just using me for their own gain.

They could learn so much from patients if they had the humility to accept that they are not the experts.

 

[hotblack]

Conclusions: Daily stress and worry are distinct factors linked to fluctuations in same-day and next day symptoms of life, with daily stress showing the strongest association—consistent with patterns recognised throughout the history of humanity. These findings highlight the importance of considering stress and worry as potential things we should consider when building a society which benefits those who live in it.

 

[hotblack]

I would like to see a study on the number of cases investigating stress in Long Covid or ME/CFS compared to other biological diseases.

Or how the NIHR (who funded this) justify it and not other projects we badly need.

So, they use “Ecological Momentary Assessment” that is asking people for ldata on stress, worry and rumination via an app 5 times a day. Interestingly it is often talked about in other papers that constant watching and monitoring causes stress and worry.

Despite this granularity they don’t seem to have used it?

Scores were aggregated over the day for ease of computation and to facilitate estimation of delayed responses to triggers over the following days.

There was high dropout.

Out of 514 participants who were approached (351 from clinics, 163 community), 420 (82%) consented to participate in the study (301 from clinics, 119 community). A total of 376 (73%) provided symptom data (273 from clinics, 103 community), and 155 (41%) of these completed questionnaires on stress, worry and rumination for at least seven consecutive days’

It’s all very subjective, but tbh I don’t know much about the scales they used

The core symptom list was adapted from the COVID-19 Yorkshire Rehabilitation Scale (C19-YRS) and included fatigue, pain or discomfort, dizziness, palpitations, cognitive dysfunction, anxiety, and depression [22].

But the 3 main factors were just 0-10, how do you feel stuff

Additionally, each EMA asked participants, “Thinking about the last hour, to what extent have you? 1. Felt stressed, 2. Worried about your illness in the future, and 3. Thought about your illness in the past " (rated on a continuous scale from 0 “Not at all” to 10 “a great deal”). These latter items were based on the UK COVID-19 Mental Health and Well-being Study [18, 23] which demonstrated good face, current and predictive validity.

Someone who is better at the stats than me should comment on this but I have concerns.

The epidemiological exposures were feeling stressed, worried about their illness in the future, and thoughts about their illness in the past (rumination). These exposures were used to predict the severity of eight self-reported symptoms which were modelled as joint multivariate outcomes.

That sounds like a lot of statistical tests but..

Second, we also acknowledge that we elected to not correct for multiple comparisons.

They explain why here

The primary reasons for this decision were because: i) this is one of the first intensive longitudinal, EMA studies in this patient group and therefore, we wanted the study to be hypothesis-generating with a focus on identifying possible signals, ii) within this context, applying strict multiple-comparison corrections are usually overly conservative and potentially obscure meaningful patterns between the daily triggers and the symptoms, and iii) we were keen to present the unadjusted p-values to ensure full transparency.

This was also very short timescale it seems. Or maybe fragmented data, I’m unclear

We recognise there are some limitations with the current study. First, the main analysis is based on 155 of the 376 participants who provided symptom data. This was because we only included participants who had completed at least seven consecutive days of daily stress, worry and rumination measures. This ensured we had a sufficient number of days to model fluctuations in daily stress, worry and rumination and to test delayed effects of the predictor variables on subsequent consecutive days.

Is it really enough to model something like this with confidence?

 

[rvallee]

Background: Recent research has shown that exertion in physical, cognitive, social and self-care activities trigger symptom severity in individuals with Long COVID.

Swing

Purpose: The current study aimed to investigate whether daily emotional exertions (stress, worry, rumination) were associated with symptom exacerbation, over and above influences of effortful daily activities, in individuals with Long COVID.

and a miss.

So close to getting it, yet as far as anyone could ever be. They can listen to thousands of people explaining things to them, and they will still completely miss the meaning of those words. This is an actual talent. A negative one, but still a remarkable talent. Especially when you consider that the eternal solution, which has never once worked, is to add more exertion.

Daily stress and worry are distinct factors linked to fluctuations in same-day and next day Long COVID symptoms, with daily stress showing the strongest association—consistent with patterns of post-exertional symptom exacerbation. These findings highlight the importance of considering stress and worry as potential therapeutic targets and integrating their management into self-care programmes.

And again the swing and a miss, re-stated, yet still unable to actually grasp it. They understand the association with PEM, but not the fact that what they describe as 'stress' and 'worry' are themselves exertion, and not 'above' or 'beyond' or anything unlike the exertion of daily activities. All of which is besides the point that the total uncertainty, chaos and misery of living with a chronic illness would be too much even for healthy people, while for us it's on top of everything else.

I guess it must not be exertion if someone isn't grunting with effort? Because normal activities of daily living can't possibly count as exertion since to a healthy person, which everyone who works in health care must be, they are easy to do? I have no idea how to deal with these people and how they think.

 

[Utsikt]

Participants received push notifications on their phones to complete an EMA every three hours between 09:00 and 21:00 each day (Supplementary Figure 1), with responses allowed within a 45-minute window. The EMA was co-designed with Long COVID patients, building on prior research[9].

Each EMA collected information on the primary activity performed in the past 30 minutes (categorized as physical, cognitive, social, self-care, rest, or sleep), the level of effort required (rated from 0 “no effort” to 10 “most effortful”), the presence of symptoms, and their severity (rated from 0 “no problem” to 10 “severe problem”).

The core symptom list was adapted from the COVID-19 Yorkshire Rehabilitation Scale (C19-YRS) and included fatigue, pain or discomfort, dizziness, palpitations, cognitive dysfunction, anxiety, and depression [22].

Additionally, each EMA asked participants, “Thinking about the last hour, to what extent have you? 1. Felt stressed, 2. Worried about your illness in the future, and 3. Thought about your illness in the past " (rated on a continuous scale from 0 “Not at all” to 10 “a great deal”).

These latter items were based on the UK COVID-19 Mental Health and Well-being Study [18, 23] which demonstrated good face, current and predictive validity.

It’s concerning that for physical activity, they asked about the last 30 minutes, but for stress, worry and rumination they asked about the last 60 minutes?

With double the time period you’re bound to pick up more signals.

Models showed that days with higher stress scores were significantly associated with increased severity of all symptoms on the same day, after adjusting for effortful activities, demographic and medical factors (all p-values ≤0.007).

This could just as well mean that on days with increased symptoms, the increased overall symptoms leads to more perceived stress.

Days with higher stress scores also predicted more severe anxiety (1.2 points higher; 95% CI: 0.8 to 1.6; p<0.001) and depression symptoms 1-day later (1.2 points higher; 95% CI: 0.8 to 1.6; p<0.001). Participants also reported more severe anxiety (1.0 points higher; 0.6 to 1.4; p<0.001) and dizziness symptoms (0.7 points higher; 0.3 to 1.2; p=0.003) 2-days later (See Figure 2 & 3, Supplemental Tables 2, 4 and 5).

Again, if higher anxiety coincided with more symptom in general, it makes sense that there would be more symptoms the following days as well if you have PEM.

The authors go on to only talk about one-way causality based on observed correlations:

Here we identify stress and worry about one’s illness as further triggers or emotional exertions that lead to changes in symptom severity.

Daily stress consistently has a negative impact on symptoms on the same day and on subsequent days.

This causality has of course not been established.

However, in contrast, there is evidence that engaging in higher levels of worry yielded beneficial effects on the following days. A likely explanation for this is that worrying about one’s illness may trigger engagement in protective coping behaviours and changes in emotional regulation that lead to symptom reduction.

For example, worrying about one’s illness may prompt individuals to rest and to disengage in activities that lead to symptom exacerbation. Moreover, this is consistent with a body of work that draws a distinction between excessive general worry and disease-specific worry [29-31].

I’m pleasantly surprised by this analysis. The changes is emotional regulation seems a bit BPSy, but adapting behaviours is very much in line with pacing. I.e., the finding suggests that being aware of your limitations and symptoms, and the consequences of doing things might be beneficial.

It was notable that daily rumination about one’s illness had a limited impact on same day and following day symptoms. This was surprising as rumination, as well as worry, has been reliably found to influence somatic symptoms and health-related physiological processes [32, 12, 14].

If they had bothered actually talking to patients instead of obsessing over their own models, this would not be very surprising at all.

They go on to suggest that using CBT and/or apps might be beneficial because it reduces stress, and that will in turn reduce symptoms. What a nice fairytale..

 

[Utsikt]

That sounds like a lot of statistical tests but..

They explain why here

Not correcting for multiple testing for transparency reasons?!

The obscuring meaningful patterns is also a weak argument. If the pattern doesn’t show through multiple testing, it’s probably not strong enough.

It sounds more and more like a fishing trip..

 

[hotblack]

It sounds more and more like a fishing trip..

That sounds relaxing. Maybe they could prescribe them?

Thanks for highlighting the stress vs worry bit, it confused me and I was going to add something, but you have both covered it.

 

[forestglip]

Not correcting for multiple testing for transparency reasons?!

That's a very funny (and nonsensical imo) reason.

Second, we also acknowledge that we elected to not correct for multiple comparisons. The primary reasons for this decision were because [...] and iii) we were keen to present the unadjusted p-values to ensure full transparency.

Odd that loads of other studies that adjust for multiple comparisons have no trouble displaying unadjusted p-values if they so choose.

 

[MrMagoo]

Breaking news : Study finds that prompting people to ruminate every 3 hours causes rumination.
Aren't there supposed to be ethics for these kind of psychological experiments?