Discussion in 'Latest ME/CFS Research' started by Firestormm, Feb 25, 2014.
7 February 2014
When they speak of significant decreases in NK cytotoxic activity, what is meant exactly? Reduced number of the relevant cells, or the cells not working as well as they should?
It means the cells are not working as well as they should.
So I guess these are the initial results from their new flow cytometer that they have been excited about at NCNED* at Griffith University.
* NCNED = The National Centre for Neuroimmunology & Emerging Diseases.
Previously known as 'PHANU' (Population Health and Neuroimmunology Research Unit.)
These findings need replication
I hate to say this, but while I think the findings look interesting - and they used some very fancy kit, the findings may not be that reliable because of methodological issues, and really need replicating. Specific issues:
1. Patients not properly diagnosed
Patients were not properly diagnosed according to Fukuda (the definition they cited) as they used a questionnaire but did not do (or did not report doing) the physical examination, blood tests and psychiatric interview that are required for a Fukuda diagnosis.
2. Controls not matched for deconditioning/activity
They used a "non-fatigued control group (n=18)". However, deconditioning is known to affect the immune system so it's possible that the changes seen are non-specific ones related to deconditoning, rather than being specific for CFS.
3. Statistics: false positive risk
The study set a p value for a significant finding of p=0.05 or lower. This is pretty generous given the large number of factors examined, with each factor comparing controls v moderate, controls v severe and severe v moderate patients. With so many comparisons there are likely to be a substantial number of false positives even with no real difference between the groups. Some of the p values reported were very low eg p=0.001, which suggests a real difference, but quite a few were only a little below 0.05 and these may well not replicate.
3.1 They also excluded outliers (data points that are a long way from where other data points cluster), but didn't say how many times they did this - or whether excluding outliers changed the results, eg by chaning a non-significant finding into a significant. Maybe the outliers were real data points that should be in the data set - another reason replication is needed.
It would be interesting to know how the results would look using a more realistic p value (at least p<0.01) and with outliers included.
I would be surprised if some of these results didn't replicate, but I would be very surprised if they all did, given the concerns above. What's really needed is a replication using robustly-defined patients and activity-matched controls (at least matched to moderate patients, prob not feasible for severe patients). Lets hope this happens.
edit: I did read somewhere that Sonya Marshall-Gradisnik's group have used pilot studies to win funding for much larger studies, though can't find the details (anyone?) - would be great if this happens here.
I think they are a young group still learning. Most of the researchers doing this work are Phd students and I believe their knowledge is growing.
With regard to your 1st point about proper diagnosis - as well as the questionnaire they do require that you have been diagnosed by a doctor. They also ask on what basis you were given the diagnosis, ie what tests or clinical judgement was used to diagnose. We all know that's not foolproof but it is better than just a symptom questionnaire.
I think they are more like pilot studies...I agree with @Simon, but I put a like on it because I personally like what they are trying to look at in the immune system. I could be wrong in my thinking though
Only if you regard fatigue as the defining element of me/cfs. The fact that they used a control group at all already puts them ahead of the crowd in terms of quality of study. To mark them down or the study down because they did not also use deconditioning as a control is very miserly.
The other point is that these were disabled according to several scales noted in the study, not "deconditioned".
Then let's focus on those areas where there was a real difference. The fact that some areas may or may not show a large difference is not really relevant.
My point is that the strengths of this study outweigh the negatives by a country mile.
Sonya Marshall-Gradisnik has a very good reputation and is not a new-comer. A lot of research is conducted using PhD students, under the direction of a more qualified researcher, to keep costs down. This makes sense, as most of the work is grunt work.
The link is no longer available to me, can anybody else check? I was about to share this with health care providers.
Appears that their site is playing up. Can't access from anywhere.
The link and abstract above makes no mention on which journal it's been published in, and in a very recent communication on FB Dr marshall-Gradisnik made no mention of that paper, referring to her latest publication being from summer 2013
it makes me wonder if it was published too early by mistake and removed
I didn't download the PDF, but @Firestormm's links in the opening post are for a third party website, which isn't an actual journal, so perhaps they erroneously published it before it was published in the journal itself.
Edit: Actually, it's not a third party website. It's the journal's website. But it's been removed anyway for whatever reason.
Doing a search for the paper on the OMICS Group website gives this result:
Thanks Bob. I am sure that is what happened. I couldn't find the article and gave up looking after a while. Figured it would come out sooner rather than later
Journal of Clinical & Cellular Immunology
J Clin Cell Immunol 2014, 5: 190
Analysis of the Relationship between Immune Dysfunction and Symptom Severity in Patients with Chronic Fatigue Syndrome/Myalgic Encephalomyelitis (CFS/ME)
Sharni Lee Hardcastle, EkuaWeba Brenu, Samantha Johnston, Thao Nguyen, Teilah Huth, Manprit Kaur, Sandra Ramos, Ali Salajegheh, Don Staines and Sonya Marshall-Gradisnik
2014, 5. Would that mean May?
2014, 5: 190
It usually means Volume 5 (of the journal), page 190.
Volume 5 is yet to be released, according to their website:
This is now available at
I was part of this study, in the severe group, so it's especially pertinent to me. The interesting findings for me were the different immune profiles between moderate and severe CFS groups.
Griffith are doing a follow-up study in May with the hope to expand.
From Sharni Hardcastle (firstname.lastname@example.org):
You can also try a Google Site Search
Separate names with a comma.