Neuropsychological dysfunction in chronic fatigue syndrome and the relation between objective and subjective findings, 2019, Rasouli et al

Paywall, https://psycnet.apa.org/doiLanding?doi=10.1037/neu0000550
Not available via Sci Hub at time of posting

 

Unsurprising. You find similarly weak relationship between objective and subjective cognitive dysfunction in other brain conditions like Parkinson's disease and major depression. To my mind this is another reason why subjective outcome measures in trials are so dangerously misleading.

 

Egil Fors as co author is a red flag. He is a proponent of the bio psychosocial approach to ME, defining it as a functional disorder and recommends CBT/GET as treatments. He's also written at least one article together with the Lightning Process coach Live Landmark (2011).

 

At least the abstract makes the classic error of presenting p values as if they are measures of biological relevance. They are simply artefacts of sample size. Biological relevance depends on the actual data. I had an immediate sense of an axe grinding somewhere.

 

I'm getting the impression these cognitive tests are inadequate.

 

My brain is a complete mush. If objective tests of cognitive difficulties do not show significant deficits than the tests used are inadequate. We would all pass the severe dementia test as well, where you have to circle the camel and stuff. This only means the proper tests aren't being used.

You need genuine expertise to get this right so it's not surprising that this would fail coming from someone who wants to prove a point, rather than do actual research.

 

My feeling from the abstract is that they had not got a handle on the cognitive impairments in
ME or how they might vary within and between individuals. We need research to know which which cognitive variables are relevant and also when/how to reliably test people with ME. We do not have good data on what cognitive factors are relevant to people with ME. We have no idea that the assessments used are relevant to the subjective cognitive difficulties experienced by patients in real life situations.

A big problem with such research is not having data on premorbid cognitive levels. For example I suspect I still have better than average reading skills though it has deteriorated significantly since the onset of my ME. So a normal control for me would need to be someone with comparable academic levels, as well as the obvious age, gender etc. Assessing my reading skills further needs to involve a complex enough task to avoid any ceiling effects. Also with complex written material my comprehension will vary enormously in relation to my familiarity with the content, so I can understand a psychological text, but would fail completely with for example an economics text. With the former I suspect I would have above average reading comprehension where as with the later I might have bellow average results.

The only way to get round not knowing individuals’ premorbid abilities is to use very large sample sizes so that you can talk about population averages. I am not sure that their sample although over two hundred would be big enough to get over the problem of not having appropriately matched controls.

Also when I am assessed will significantly impact my result, so example for reading for comprehension, I may be able to cope with a complex text when rested on a good day, but really struggle to make sense of comparable material when experiencing brain fog. A good example of this variation is my immediate recall of a string of digits or digit span can vary from 8 to 3 depending on my ME and the circumstances, though the normal is 7 plus or minus 2. A practical example of this is that on a good day I may have the ability to hold a phone number in my head long enough to dial it, but on a bad day I have to put my finger on each written digit and dial the number very slowly one digit at a time.

Further there is potentially the issue of overall demands of a task. I may be able on a good day to perform a cognitive task within normal limits, whether or not at my premorbid level is another question, but introduce other factors such as time constraints, distractions or novelty of materials and
I may struggle. Recently my goddaughter had organised a birthday dinner and asked us all to pick on line what we wanted to eat in advance. I managed easily at home online to pick something I thought I would like and fitted my dietary constraints, however when we got to the restaurant I had had to walk further than was comfortable from the car as the restaurant was in a pedestrian only area, there was a crowd and noise. At the restaurant the idea of our preordered choices was abandoned and
I was presented with a menu to make my choice again. I was completely unable to undertake the task and my goddaughter had to make the choice for me.

The ME sample [in this study] would be fairly heterogenous because of how they selected subjects making matching controls difficult, but further I suspect it would not include people with severe ME meaning the cognitive impairments sampled would be relatively mild and so much harder to achieve significant differences.

My feeling is we need more work on profiling the cognitive deficits associated with ME and how they vary within and between individuals before such studies as this can be meaningfully pursued.

 

Quite right, these studies are hard to control properly and many of them have shown both patients and controls having well above average intelligence - or abnormally distributed intelligence, suggesting substantial biases.