Brain function characteristics of chronic fatigue syndrome: A task fMRI study, Shan et al, 2018

The idea of having to put in more effort to do even simple tasks sounds true to me though.

It's probably just an indication of mental fatigue. Without a group with some other illness we are not really learning anything new. We already know that ME/CFS patients are ill.

 

@Adrian's right.

Luckily, in an event-related fMRI (which this was), all these things get factored out, because you’re only interested in the regions that differed significantly between your control trials and your experimental trials.

Since you’re likely to be having similar thoughts throughout the entire block of trials (its pretty unlikely particular thoughts spring into your mind only during the experimental trials and then completely disappear during the control trials), that activity does not contribute to the difference score.

 

I don't follow that, @Woolie, what are the experimental and control trials? Surely if the experimental trial is asking someone to do a difficult task that will raise different thoughts for people with ME than healthy people just because they know they are being challenged in relation to something that might be about their illness.

 

All the trials involve the same task - name the colour a word is written in. What differs is whether the word itself provides distracting information or not (e.g., the word blue written in red). it takes around 500ms to complete a control trial, 900ms for the distracting info trials.

In your post you talked about thoughts, that is, cognitions open to awareness. Its important to remember that thoughts are resource-intensive. We cannot think of more than one or two things at a time (we can perform lots of different cognitive operations at the same time, but we can entertain only one or two thoughts).

During the difficult trials, you're unlikely to be doing any thinking outside the task at all. For that 900ms (or at least the first 700ms of it), your entire cognitive focus on attention will most likely be on the task. Those trials are very resource intensive, so there won't be much space for mental explorations about how difficult you're finding it, or stuff like that. You can probably squeeze in some thinking during the easier (control) trials, they're pretty routine. And you'll be thinking stuff in between trials, but that won't get measured.

You'll still be experiencing emotions throughout. They are essential for success. Our desire to do well on a task ensures a strong pattern of action in the ACC, and without that emotional aspect, we won't do well.

So if PwMEs are especial keen to do well on the task, more so than controls, that might actually heighten activation in some key areas like the ACC.

 

I am still sceptical @Woolie. When I was talking of thoughts I was thinking of what one might call 'mindsets' that will be hanging around in the brain throughout the proceedings, in the way that priming stimuli do. These may set circuits to spread more widely when tasks are difficult, not because of some abnormal wiring but because of the normal effect of having a different background mindset. The effect could easily play out in 500msec. Whether or not overt thoughts occurred at the time I don't think matters. The overt thoughts might well pitch up after a bank of test questions had been dealt with.

As perhaps an example, I once heard a videotape of a teenager playing a Chopin piece at a school concert. The first exposition was technically fine but lifeless. Then the repeat exposition suddenly came to life, full of expression, because the girl had become comfortable with playing. Her brain actions were pouring out every every 200msec or so but they were informed by a different underlying mindset. I suspect that she was aware that she had changed mindset over a timescale that spread across the timing of the immediate responses. I bet her fMRI looked quite different the second time around.

What I think really would be helpful in this sort of situation would be if patients with ME turned out to show two quite distinct patterns on BOLD. Maybe comparing Gulf War Syndrome with ME would be good - very much what James Baraniuk has been doing. He thinks there are two quite different patterns of brain action even within Gulf War Syndrome.

 

@Jonathan Edwards, I'm not sure about the concept of a "background mindset", it sounds very vague to me. What you're describing seems to be the myriad ways our cognitions are shaped by context. That context can include current health, recent activities, current emotions.

But in any case, few of these things are likely to be event-specific so will get factored out during the comparison between the experimental and control trials. The only one that might be event-specific is priming (very short lived changes in cognition as a result of the previous 1-2 trials), but that would not be expected to differ between patients and controls.

Yes, a second group of controls with some other illness would be good.

But distinct patterns on BOLD? That's an impossible ask, unless you study a sample that have no brain tissue in the regions normally associated with performing the task (stroke, surgery, etc.). Generally speaking, there's not huge variation in the regions and networks people engage to do these tasks.

 

Nothing very vague, I would say.
Think of the mindset of a jazz player in a jam session whose brain is set to respond to each phrase from other musicians with a further interpretation of "Ain't Misbehaving' " rather than "Honeysuckle Rose". All his responses are conditioned by something he may not be overtly thinking about at all. And certain BOLD patterns may result only when he is responding to a new member of the group playing something a bit unexpected.

To my mind, our understanding of how 'cognitive processes' evolve is so minimal that we cannot exclude all sorts of possibilities like this!

And I do not quite follow that. Unless we are hoping to find different BOLD patterns we have no result. By distinct I mean discriminable in some way, though, not mutually exclusive. More extensive would do for that - which seems to be the claim here.

 

Here is the bit that you still need to attend to (from my last post):

Remember, trials last for around 1-2 seconds.

 

But that is exactly my point. A person who knows that they are being tested as a patient may specifically respond to the tricky stroop tests in a different way just as the jazz musician may specifically respond to the riffs from the new member in a different way if they are playing a tune they have not tried before, or whatever. Each riff may last from 1-2 seconds, or not far off.

 

I think the problem is that none of this can be factored out because we simply do not know how these processes work. I raised this with Neil Harrison when he showed a signal in the basal ganglia on one side after giving something like interferon or typhoid vaccine to a group of patients. He agreed that all sorts of interpretations are possible.

 

Could they instead measure ME/CFS patients against other ME/CFS patients? I'm thinking in terms of stratifying patients by some other test(s) of cognitive impairment and then seeing if their results on this type of fMRI test correlated to their degree of cognitive impairment established by other means. If there was a correlation to cognitive impairment, it seems like that would at least show that the results were not solely due to the different background thoughts of ME patients under stress.

 

Yes, I think some sort of internal comparison would be best. But if you sort people by brain abilities you may drag along another confounding correlation. What might be ideal would be showing a difference between PWME of viral onset and those with gradual onset or something like that.

 

A non-technical perspective:

When I tried the Stroop test, I found it incredibly hard. My brain was totally focused on that and nothing else.

Sure, I wasn’t in an MRI scanner, but I don’t think that would’ve made any difference. My brain had no spare capacity for anything beyond the test itself. Maybe I had the vague thought “wow, this is hard!". I doubt that any subjects would be familiar with the Stroop test, so its novelty is another reason it would demand complete attention. Maybe if I played jazz I would respond differently.

 

Hmm. I don't mean to criticise Neil Harrison specifically. A lot of otherwise good researchers who just lose it when it comes to making causal inferences in this field.

The study's results are pretty clean, imo. Jonathan's idea that patients might have some disproportionate behavioural/emotional reaction to the incongruent trials specifically, not fully removed by the subtraction, something that should be considered a nuisance variable, seems to me a stretch. Without a proper hypothesis as to what these highly trial-specific and very short-lived reactions or behaviours are, I don't think it can take us anywhere (I could formulate a hypothesis based on abnormally heightened arousal levels in response to the difficult trials, but it would predict the opposite outcome to the one observed).

What's not clear is what the finding means. Adding another disease group would be useful here. Even activity-matched controls. If you compare sedentary people to otherwise comparable people that exercise regularly, you'll most likely get a similar effect to the one observed here. So is it just an artefact of cardiovascular differences between the active controls and the less active MECFS patients? Or something more?

I'd also like to see correlations between specific measures, things like overall activity levels, rating of severity of specific symptoms - including self-rated fog. It would also be good to see if the differences persist within the same group of individuals before and after PEM has been induced.

 

I've done the stroop test too as part of an undegraduate student project and I was quite nervous too. The fact is we have to work harder to concentrate and that may well also stimulate more areas of the brain (some of that extra stimulation may aid performance - should not be ruled out)

 

Nervous is good. Its associated with better performance.

That's exactly the point. That's what the researchers hypothesised.

As to whether the "abnormal" activity is doing anything useful, the results of this study would seem to suggest not. Much of that "abnormal" activity is not very well coordinated, so probably not doing much that's very useful.

 

A few comments if I may.

The researchers disappointingly did not assess patients (or controls) mood states using a standardised structured clinical interview (SCID, e.g. First et al) by clinicians uninvolved in the study during the patient selection process. This leaves the door open for the psycho-social brigade to claim that the differences seem are due to differing emotional states and this can/could explain the differences seen. Disappointing to see this lack of process/procedure as fMRI studies are not cheap equipment wise or time wise.

Stroop has quite a few forms. I use pen and pencil version - 3 tests in one - in my clinical practice with patients with neuro disorders including pwME. It often shows up differences in reaction time. And it is fair to say that pwME can do the test as well as HCs but a lot slower. Also, it is quite evident from observation that patients use a lot more energy and need to concentrate to achieve a slower result. If they'd asked me I would have suggested using the PASAT test which would probably have been more effective. And there might be HC databases of this available for this test too. Frustrating. PASAT is used in MS and TBI research frequently.

My interpretation of this paper would be that this type of research shows up the increased effort pwME have to put in to achieve a slower result. If the low blood flow to the brain issue is also investigated more (e.g. via SPECT scans) then this really works together to highlight the multiple neurocognitive problems pwME have. I guess all scan results will be carried out with patients lying down which could improve many pwME's performance - but not all.

Testing like this demonstrates patients performing at their very best. It is not necessarily transferable to everyday tasks / functioning as this test takes 10 mins or so to do - and it's done once. The challenge for researchers is to assess before and after physically and mentally demanding tasks - that really shows up differences. I did this in my doctoral work. I can post summary if folks would like to see?

Bw Joan Crawford
Chartered Counselling Psychologist
Chester, UK

 

Please!

 

That's interesting to hear you found the PASAT to be good at distinguishing PwMEs from controls. We've heard similar findings before, but I don't think your own study was mentioned at the time.

The authors of this study don't give any rationale for their test choice - neither theoretical nor practical. But based on fMRI studies using similar tasks to the PASAT (e.g., the N-back task), I would expect fMRI results from that task would be even harder to interpret than those from Stroop as there is such widespread neural activity even in healthy people. The Stroop is cognitively "cleaner".

Would love to see a summary of your doctoral work, @Joan Crawford!