Preprint Peripheral metabolic signatures in patients with neuropsychiatric long COVID syndrome, 2026, Bergmann et al.

SNT Gatchaman

SNT Gatchaman

Senior Member (Voting Rights)
Staff member
Peripheral metabolic signatures in patients with neuropsychiatric long COVID syndrome
Dario Lucas Bergmann; Sophie Neugebauer; Tonia Rocktaeschl; Eva-Maria Dommaschk; Meng Li; Alexander Weuthen; Alexander Refisch; Nikolai Blekic; Michael Kiehntopf; Andre Scherag; Helgi B Schioeth; Chai K Lim; Nils Opel; Martin Walter; Bianca Besteher

Neuropsychiatric symptoms are considered the most common feature of long COVID disease. Recent studies have demonstrated structural brain changes and highlighted the importance of neuroinflammation in the development of cognitive deficits as seen in long COVID patients. In addition, peripheral studies have demonstrated heterogeneous molecular subtypes of long COVID pathology. However, it is unknown which peripheral metabolomic alterations occur in patients with neuropsychiatric long COVID syndrome and how these relate to symptom severity.

In the present study, we investigated differences in the peripheral serum metabolome profiles of healthy controls and patients with long COVID syndrome with neuropsychiatric symptoms. We found that patients with long COVID showed peripheral alterations in lipid species such as triacylglycerides and acylcarnitines. Furthermore, metabolites altered in patients with long COVID syndrome were also associated with depressive and fatigue symptom burden as well as with differences in cortical thickness in multiple brain regions.

Our results demonstrate a metabolic phenotype of long COVID patients that may reflect a dysregulation of lipid metabolism and deficits in mitochondrial energy production as potential contributors to symptom burden and brain structural alterations. These data may serve as a resource and basis for further studies aimed at investigating peripheral molecular alterations in patients with neuropsychiatric long COVID syndrome.

Web | DOI | PDF | Preprint: MedRxiv | Open Access
 
Plenty of members can provide comment on whether the study is sound, including you @mariovitali. What did you think of it?

The abstract is a bit unhelpful, it doesn't even tell us how many people were in the sample or if the lipids went up or down, so that raises a question mark for me.

I'd want to be checking out the characteristics of the people with the 'neuropsychiatric Long Covid' and the controls, to see if they were comparable, apart from the presence or absence of Long Covid.
 
Between April 2021 and June 2022, we included 54 long-COVID patients, 18 healthy survivors who had laboratory-confirmed COVD-19 but did not develop long-COVID syndrome, and 30 healthy controls without known or serologically confirmed prior COVD-19 infection from whom we measured metabolomic profiles. The groups were matched for age and sex.
Click to expand...
Yeah, except they were not matched for sex very well:
Healthy controls, no infection: 13 female, 17 male. 43% female
Healthy controls, infection: 10 female, 8 male. 56% female
Long Covid: 32 female, 22 male. 59% female

There were some differences on some other measures too, especially between the healthy controls who had an infection and the Long Covid group
BMI: Healthy infected: 23. 4 : Long Covid: 25.8
Age Healthy infected: 39.9 years : Long Covid: 45.2

So the Long Covid group tended to be older and more overweight that the infected controls.
And the Long Covid group had a bigger percentage of females than the non-infected control.
Both of those differences are likely to skew the comparisons, making the Long Covid group more likely to have certain lipids.

624 metabolites

Abundances were adjusted for sex, BMI, and age, which is potentially good, but not nearly as good as matching controls really well, and does create a possibility for some artifactual error.

There are q values (p values adjusted for False Discovery Rate). Metabolites with a q value of <0.1 were used in the analysis.

However, principal component analysis also revealed no major shifts between the three different cohorts along the first four principal components, accounting for ~45% of the total metabolome variance), indicating rather subtle differences in the total serum metabolome in
long COVID patients compared to healthy controls
Click to expand...
Overall, there were not major differences in metabolites
 
Last edited:
Just a note on the Long Covid cohort, they just had to have persisting symptoms for at least 12 weeks. Not everyone had fatigue, not everyone had cognitive problems. So, they have not been sorted well.

And that's a problem when we come to things like triglycerides in blood. Does the Long Covid cohort have people who are more unhealthy (besides having "Long Covid")? It's pretty hard to know what to make of the data when the cohort is so heterogeneous. There's so much that could be going on there - e.g. different percentages of people on different medications.

And there's potentially the same problem that Chris Armstong's team had - of choosing ultra healthy controls. The controls had less depression. Does that mean that they would be less likely on average to sit on the sofa and eat chocolate biscuits than jog around the block and then make a chicken salad? Possibly.

Sure, we can and should pick through the metabolites mentioned, but there's so much noise coming from so many directions in this. I suspect we really need paired blood tests - before and after the onset of post-Covid ME/CFS - to get data that is very useful
 
Hutan said:
Plenty of members can provide comment on whether the study is sound, including you @mariovitali. What did you think of it?
Click to expand...
I am sure they can, since you and @forestglip are reviewing the methodology of studies very frequently, I thought of mentioning you. If the way I wrote it appeared as overly demanding, this was not my intention and for this I apologise.
Hutan said:
The abstract is a bit unhelpful, it doesn't even tell us how many people were in the sample or if the lipids went up or down, so that raises a question mark for me.

I'd want to be checking out the characteristics of the people with the 'neuropsychiatric Long Covid' and the controls, to see if they were comparable, apart from the presence or absence of Long Covid.
Click to expand...
Thank you for your comments on this study and your time, as always
 
You must log in or register to reply here.
Back
Top Bottom