I am relying mainly on the comments above (thank you) and the authors’ conclusion, but the paper seems to be saying that looking at the blood metabolomics in the way that has been done to date, by a variety of researchers, has basically drawn a blank. That also means, if I've understood this right, that the early promise of the Bob Naviaux study has not been fulfilled.
5. Conclusions
We are reporting on the largest number of metabolites in the ME/CFS field to date, about 1750
plasma compounds, encompassing 20 super‐pathways and 113 sub‐pathways. However, we have not unequivocally identified a plasma biomarker or set of biomarkers with abundances drastically
different between controls and ME/CFS patients, despite the fact that our clinical data indicates our patient cohort had a substantial level of disability (Table 1). The same conclusion can be drawn from examining various groups’ prior reports, including ours, where distinct populations were recruited, alternate instrumentation was used, or even serum was analyzed instead of plasma [9 refs]...
As for the significant findings:
Conclusion:...Nevertheless, the metabolites emphasized as a result of our analysis, most specifically acyl cholines and steroids, should be considered in light of the metabolic impact even modest changes can have along with the complexity of sources that drain metabolites into the circulatory system. Indeed, the changes observed might result from a disturbance occurring in another part of the body (e.g., brain, muscles), one whose impact may be diluted in plasma...
That looks to me to be more speculative, particularly given the point above about the significant age difference between patients and controls and how that could affect results. Subject also were not matched on exercise, inevitably, and perhaps that's could affect the results as well. And several people inc
@Trish highlighted potential dietary difference. These ‘confounding’ factors make it harder to be certain the positive findings are important. Or if they are cause or effect (the authors seem to be arguing for cause).
This looks more significant, where the authors plot a way forward for the field:
... Acquisition of samples in conditions that increase patient symptoms (e.g., PEM) could further widen the observed gap between the plasma metabolome of patients and controls. Because liquid biopsies such as plasma and serum remain the most accessible resource from human subjects, our results combined with more explorative metabolomics will help stimulate our efforts before we move toward different, yet more invasive sample collection techniques.
Furthermore, set‐enrichment statistics with a control for false discovery rates can be used to identify chemical groups instead of individual metabolites that can be associated with a disease or phenotype. Follow‐up studies can develop new targeted analytical methods to measure those chemical groups in studies with power calculations from this discovery stage data.
Overall, this appears to be an excellent study that helps move the field forward substantially. It says, we tried this, and were not really getting anywhere. Perhaps we can make progress if we try a different tack. This is so much better than the situation we’ve often seen in the past, where one promising study gets a lot of attention and quoted for years afterwards - without any follow-up research to see if the initial finding is reliable/meaningful.
Hats off to Maureen Hanson, Susan Levine and all the authors for a thorough study.
And please point out if my analysis is flawed. Migraines prevent me taking a thorough look at this.
