Gene expression alterations at baseline and following moderate exercise
- Thread starter Bob
- Start date
Good point, I hadn't noticed that. But even so, that leaves 3 of the 4 time points as not relevant (and they used AUC, the sum across all 4 time points, for most analyses).
Fundamentally, what I don't understand is why they didn't use an exercise test that is relevant to the real-world exertions that provoke PEM for people with CFS. 'Moderate exercise' is way more than most people with CFS need to provoke PEM, or indeed can manage. If you can mange moderate exercise (as 'limited a little') you probably score over 80 on the SF36 PF scale!
I certainly wouldn't be rushing to have the test (but then I'm severely affected). Saying that, hopefully it will give useful knowledge about the condition. I'm no expert but sometimes "challenge tests" both in research and/or even in clinical practice are a big unusual e.g. you get a big dose of ACTH or a buspirone challenge test (can't say I know too much about the latter).
I can certainly see the need to make the challenge big enough to ensure you are going to get a result (PEM in this case) but this still fails there own logic of trying to find something closer to real world situations (they deliberately avoided maximal exercise tests for this reason). And crucially, the level of challenge varied with severity: maybe it was moderate for the least affected but extreme for the most severely affected. This gives you unreliable data, particularly as the level of expression changes with level of exercise challenge in healthy controls. Maybe setting exercise based on a given level of RPE exertion would give a more appropriate test.
I still think what they are trying to do is very important, and their secondary analysis using matched RPE indicates they've found something real, but the methodolgy needs refining. The CAA (who funded this work) pointed out the need for replication by other groups; hopefully this will happen and other researchers will refine the exercise challenge.
I still think what they are trying to do is very important, and their secondary analysis using matched RPE indicates they've found something real, but the methodolgy needs refining. The CAA (who funded this work) pointed out the need for replication by other groups; hopefully this will happen and other researchers will refine the exercise challenge.
I think the RPE is influenced by neuromuscular channelopathy (Siemionow 2004, Jammes 2009 & 2005, Pietrangelo 2009, Whistler 2005), which might help. Really it's a confounding factor (unless any of those gene products disrupt these neuromuscular ion channels), but it confounds in the direction of presumably increasing RPE selectively in the ME/CFS sample.
Snow Leopard
Hibernating
- Messages
- 5,902
- Location
- South Australia
Is it way more than "moderate exercise" based on their fitness and day to day activity levels, or way more because of their disease?
The fact is that you have to do "way more" to provoke such a response when the study size is small. The bias with exercise studies is you are unlikely to have more severely affected patients, so "moderate" levels of exercise are higher. I really wish figures such as SF-36 PF scores and employment status were published though.
On the other hand, I agree the controls should be carefully matched based on fitness levels and sedentary activity levels (measured with actometers).
Sounds good to me. With lots of objective post-exertional testing, from start to finish.
Way more because of activity levels: 25 mins @ 70% max heart rate is not appropriate for anyone who has been rated housebound/bedbound (as applied to some in this study). I'm arguing that the challenge needs to be strong enough to provoke the PEM that this study focuses on, but not so strong that it confounds the whole study by pushing people to extreme exercise. Matching controls on fitness levels as you suggest would help.
is there a way we can use other disease groups as controls, in order to match sedentary-ness, without introducing inappropriate confounding? actually, we could then know whether the test was diagnostic and whether the effect was ME/CFS-specific. this would be better than comparing to healthy sedentary population in this regard, wouldn't it?
It may not be perfect but this group, to their credit, is looking at other disease groups:
As Dolphin points out, the authors have submitted a paper on this and it will be interesting to see the detailed results. My concern would be on how sedentary they are: MS can cover a very wide range of abilities. But if they are appropriately matched then these MS results would be compelling evidence of a unique process to CFS/ME. However, the cytokine gene expression profile may not be comparable as this would be expected to be disturbed in MS anyway.
I do remember now that you posted this. Also there's a group that compares to liver disease, but I can't think of any citations right now or what they study.
wouldn't that help us know what was diagnostic? and also help demonstrate similarity of ME/CFS to MS (rather than have it thought to be more appropriately compared to anorexia nervosa)?
With regard to liver disease, I think you're thinking of Julia Newton's research team who have compared the two e.g.:
e.g. free full text: http://qjmed.oxfordjournals.org/content/100/8/519.long
The aim is to find out if the gene expression changes after exercise are part of ME/CFS. MS and other illnesses such as RA could provide suitable sedentary controls but with many diseases you would expect a non-normal cytokine profile because of infection/autoimmunity etc, which would confound the results. It would make it very hard to intepret any differences/similarities in cytokine profile between MS, say, and ME/CFS.
Need for replication of these findings
I think that, along with XMRV and Riuximab, this report of increased gene expression in response to exercise is amongst the most interesting findings to emerge about CFS. However, as with XMRV and Rituximab, replication is crucial.
The biggest concern over this study relates to the higher exertion levels of CFS patients than controls, as has been extensively discussed on this thread. The authors had indicated that submitted research would demonstrate that in response to exercise MS patients showed no change in the expression of the genes that increased post-exercise in CFS patients. This would go some way to demonstrate the changes were unique to CFS and not related to deconditioning, particularly if the MS patients were well-matched with the CFS ones in terms of activity levels.
This submitted paper on MS patients has now been published and while it shows that metabolite receptor gene expression is not increased in response to gene expression, it also shows that the adrenergic recepotor genes are similarly increased in MS and CFS patients. Which makes it far less conclusive. It's not clear how well matched the MS patients were to the CFS patients.
Furthermore, the moderate-to-strong link found between gene expression and fatgiue and pain scores across all the genes implicated in this (2011) paper was not replicated in the new MS/CFS paper. The only correlation found was for a single adrenergic gene, P2X4, and then only for pain and not for fatigue. It was this correlation between gene expression and fatigue and pain that made the findings of this 2011 paper so compelling.
Replication studies have not been kind to the XMRV findings to date, though the largest and most rigorous (Lipkin) study has yet to report. Replication of the Rituximab findings are planned by its authors and at least one other group. The authors of these gene expression findings have made it clear they believe their work to date is sufficient:
I think that, along with XMRV and Riuximab, this report of increased gene expression in response to exercise is amongst the most interesting findings to emerge about CFS. However, as with XMRV and Rituximab, replication is crucial.
The biggest concern over this study relates to the higher exertion levels of CFS patients than controls, as has been extensively discussed on this thread. The authors had indicated that submitted research would demonstrate that in response to exercise MS patients showed no change in the expression of the genes that increased post-exercise in CFS patients. This would go some way to demonstrate the changes were unique to CFS and not related to deconditioning, particularly if the MS patients were well-matched with the CFS ones in terms of activity levels.
This submitted paper on MS patients has now been published and while it shows that metabolite receptor gene expression is not increased in response to gene expression, it also shows that the adrenergic recepotor genes are similarly increased in MS and CFS patients. Which makes it far less conclusive. It's not clear how well matched the MS patients were to the CFS patients.
Furthermore, the moderate-to-strong link found between gene expression and fatgiue and pain scores across all the genes implicated in this (2011) paper was not replicated in the new MS/CFS paper. The only correlation found was for a single adrenergic gene, P2X4, and then only for pain and not for fatigue. It was this correlation between gene expression and fatigue and pain that made the findings of this 2011 paper so compelling.
Replication studies have not been kind to the XMRV findings to date, though the largest and most rigorous (Lipkin) study has yet to report. Replication of the Rituximab findings are planned by its authors and at least one other group. The authors of these gene expression findings have made it clear they believe their work to date is sufficient:
As the CAA, who funded this study say "As with any research, these findings will need to be replicated by another research group in another group of CFS (and FM) patients", to establish that these gene expression are a real and important link to fatigue and pain in CFS. Hopefully other groups will do exactly this.