Just a couple more thoughts from me. Running it through my normal mental checklist for such studies I'm reassured by:
- It is published in a high quality, non-predatory journal;
- The selection criteria seem ok (1994 CDC criteria and 96%-97% of patients have PEM and/or cognitive impairment);
- Decent sample size;
- Reasonable match (race aside) between patients and healthy controls;
- They've corrected p values for multiple comparisons.
But some worries around:
- Main finding is due to a subgroup analysis, which appears to be post-hoc;
- No validation cohort / attempt at replication;
- Press release doesn't quite match the study (often not the fault of a study's author, it should be noted)
Whilst I agree with some of the criticism above that this doesn't look like it'd provide a useful blood test, given that the patient group cannot be distinguished from healthy controls. That would be enough for me to normally write this off as a decent biomarker study...
...but there is an interesting parallel to be drawn between the results of this cross-sectional study and the (initial, unpublished) results coming from the longitudinal study currently underway at the Younger Lab. They too are finding a correlation between certain cytokines and severity of symptoms, and again finding that whilst this correlation exists the actual cytokine levels on average are within normal tolerances. This seems consistent to me and if the case would indicate that cytokines drive the disease but it would appear that there is some kind of abnormal response to these cytokines or that the cytokines are correlated with another disease process that is doing the actual damage.
So, overall, slightly over-optimistic press release aside, this is an interesting study and I'm interested as to whether this cross-sectional result will be replicated in a longitudinal study.
What I was pointing out was that there may be some similar processes going on with respect to cytokines in depression and CFS, and that by itself is - as I think - an interesting finding.
Which is entirely plausible given the increasing interest in using anti-inflammatory drugs to treat severe refactory depression. There was a BBC report recently on a new trial using auto-immune treatments (I don't think it was Rituximab, it was another monoclonal antibody, IIRC) - I'll see if I can find a link if I don't get beasted at work today.