The maintained attention assessment in patients affected by [ME/CFS]: a reliable biomarker?, Murga et al, 2021

Conclusion
General cognition remains preserved in most patients, only a small group of them shows a significant mild cognitive impairment. Maintained attention is clearly deficient, showing a marked fatigability after the Toulouse-Piéron test. The effort was perceived as very hard by both gender, but higher by women.

This study proposes a simple clinical way to assess maintained attention. Present results support the reliability of maintained attention as biomarker of CFS/ME. Attention deficit is a significant disability in patients affected of central fatigue.


https://translational-medicine.biomedcentral.com/articles/10.1186/s12967-021-03153-1

 

Full title: The maintained attention assessment in patients affected by Myalgic encephalomyelitis/chronic fatigue syndrome: a reliable biomarker?

Abstract

The maintained attention is the cause of great functional limitations in CFS/ME, a disease that mainly affects women in the central period of life. Cognitive function is explored using the Montreal Cognitive Assessment, the maintained attention using the Toulouse-Piéron test with which the Global Index of Attention and Perception (GIAP) is obtained, the fatigue using the visual analog scale and the perception of effort using the modified Borg scale. The final sample were 84 patients (66 women/18 men) who met diagnostic criteria (Fukuda-1994, Carruthers-2011) and 22 healthy controls (14 women/8 men). Most of patients maintain normal cognitive function, showing low or very low attention score in the 70% of patients with a marked cognitive fatigue compared to the control group (p < 0.05). There were no significant differences between genders in GIAP or fatigue for CFS/ME; however, sick women perceive cognitive effort higher than men. Deficits in sustained attention and the perception of fatigue, so effort after performing the proposed test are a sensitive and reliable indicator that allows us to substantiate a clinical suspicion and refer patients for further studies in order to confirm or rule out CFS/ME.

 

I haven't looked at this yet, but in the last 5 years or so my ability to think is very impaired most mornings and early afternoons, so it will make a large difference what time I am tested.

 

i would hazard an opinion that many of the women in this assessment had far more responsibilities ie child care managing of the house etc . sorry if this comes across as sexist but it is what happens in many families sick or not the women of most households are still carrying most of the burden .

 

Interesting paper, one of the hardest parts of my illness when in relapse is the distractibility (inability to maintain attention) and cognitive fatigue which has major occupational impacts. They discuss the literature around oxidative stress.

They particularly note no differences between males and females with diagnosis.

 

If replicated, these findings are of huge importance, IMO.

Over 80% of people with ME report cognitive issues yet research studies have never found very much (including the Cognitive Function Montreal test reported here). The biggest effect has been for reaction speed — and I've never heard anybody complaining that the problem with ME is that their reactions are slow . (We complain about brain fog all the time.)

As the graph posted by @Hutan shows, there is a huge difference in the Global Index of Attention and Perception between patients and controls. Even though the study didn't correct for multiple comparisons, the p-value is so small it doesn't matter.

As several people noted, some people might not have a problem depending on the time of day or the kind of day they are having. Which makes the results all the more impressive.

A few concerns:
— I'm always concerned when data seems almost too good to be true (not because I distrust the researchers but because I've just been disappointed too many times in the past).
— almost all the patients are on medication and this could affect the results, as the researchers point out. But I do find it hard to believe that the drugs would make such a big difference. Do antidepressants really reduce attention? I would have thought they would help.
— As, again, Hutan points out, the reported fatigue scores don't really make sense (if they were right, with some controls showing reporting more fatigued than some patients, the results are even more extraordinary).

I was thinking of writing to the authors to ask if they are planning a replication and also to query the fatigue scores. Does anyone else have any questions they'd like to put to the researchers?

 

Yes @Hutan, depending on the antidepressant (and other medications, psych or non psych) they can affect attention and concentration. Many have anticholinergic effects known to disrupt cognition.

I had similar concerns about medication that people were on in this study, which can significantly alter attention and concentration. This study needs to be replicated controlling for this.

Also may not be useful to use non-ME people as our pathophysiology is different from “normal” people

 

I find it hard to imagine that they had done this but failed to mention it in the paper ;-). Perhaps I will simply ask if they plan to do this (or plan to test a significant number of ME patients who are not on medication.)

good point.

 

Nice to see a study looking into some of our cognitive issues in a more precise way than just at general, hard-to-define brainfog. Brainfog really needs to be broken down into its constituents to better understand what's going on.

Wondering about best experimental set up for pwME. In this study they made some effort to ensure testing conditions were the same for all, e.g. temperature and time of day for testing. Which makes sense in most cases but with ME, could this introduce more confounders than it eliminates? Some of us operate better when cold, others when warm, some better in the morning, others in the afternoon. Would it be better to test pwME under conditions they feel their best or their worst but not a mix of both?

I actually did a similar test to the Toulouse-Piéron test used in the study for sustained attention (d2 Test of Attention) about 35 years ago during one of my much better periods and scored a clean 100% perfect. I may be deluding myself but I'm pretty sure if I did it now when well rested I'd still score very well. For me the sustained attention test would likely only be a marker of anything if taken twice, before and after exertion.

There are other parts of my cognition that are more 'off' even when well rested, notably working memory and processing speed. For example, if you spell a word to me I understand the letters but am unable to process them into a word. Or if you read me a phone number (or even a shorter number) I'm unable to remember it long enough to write it down. So for me those aspects of cognition would be closer to a biomarker than sustained attention (since 'sustained' in those test is still only a matter of minutes).

Then there's always the issue of not having a baseline of cognitive function before illness. This may not matter that much for research if you have large enough cohorts but it could matter if it's used as a biomarker on the individual level. How do they overcome this lack of baseline info in other diseases?

 

I think you could reveal how ME affects my cognitive abilities pretty well by giving me an easy crossword (like the Guardian Quick) when I was well rested, and after making me stand up for a few minutes.

I know this because I've been doing that crossword most days for donkey's years. I can finish it in a few minutes when rested, but am spectacularly incompetent when tired. There are four main blocks: recalling a specific word (I 'know' the answer, but can't call it to mind); thinking of alternatives for common words; following really obvious clues and puns; and spotting that a word has more than one sense or pronunciation (e.g., rowing as in propelling a boat and rowing as in arguing).

That kind of approach might to some extent get around problems like not having a pre-illness baseline, because it would measure the impact of an activity rather than trying to get some kind of broad picture of a person's cognitive abilities. It would be easy to compare it with other conditions, and there are lots of existing cognitive tests that could be used.

 

There are known deficits for pwME on formal neuropsychological testing. Sustained attention is concentration, then manipulating the information requires intact working memory and processing speed is reduced in ME/CFS

Many GP’s have some standard questions they may ask if they think someone is not concentrating (distractable) and/or showing memory deficits The MOCA is a common screening tool for GP’s (or the MMSE) for cognitive dysfunction to start diagnositic work ups e.g. medication review, cardiovascular and endocrine examination etc. and considering a referral to a neuropsychologist or psychiatrist or psychogeriatrician (who specialises in cognitive disorder across adulthood).

It is a bit of a leap to say this is a bio-marker - There is no biological marker as yet but imaging is pointing to the limbic area where attention is managed (I am sure others can give more detailed explanations). MOCA is a screening test not a diagnostic test (it indicates dysfunction but does not specify where it is).

But I do think this study does present information that a GP could consider when seeing their pwME, but also needs to be longitudinal so done at different testing times and I believe it is is also related to the severity of the ME/CFS illness and where the pwME sees themselves, eg their baseline and how cognitively or physically challenged they are at that time.

 

One potential way is to seek to match the patient and normal groups on a factor that is presumed to relate to premorbid cognitive ability, such as level of education or previous work status. So you would seek to find a match pair for each patient to make up the control group, and you would measure the difference between the patient and their matched control rather than between the group averages. However, it is not clear how doable this is without some independent confirmation that we could reasonably expect that say solicitors would have similar premorbid baselines as other solicitors and say postmen would have similar premorbid baselines as other postmen.

The only sure fired way to get round the lack of a premorbid baseline, is to to have samples that can be unambiguously said to represent what is normal for the target populations. Usually this can only be achieved where the sample size is large enough to eliminate any possibility that between group differences are not due to normal individual differences within the sample, or if the between group differences are so big as to make normal variation an implausible explanation, usually an all or nothing difference.

Personally, my cognitive difficulties, other than my specific issues with such as spelling or navigation, depend on processing load, any distractions and novelty of the task.

I can read most single sentences, but my reading comprehension brakes down on longer passages requiring integration of information between sentences containing unfamiliar vocabulary items or under time pressure, so premorbidly I would read three or four books a week including non fiction, whereas now I have only managed to read two books in the last five years, each taking over six months with frequent re reading of chapters, whereas I probably process the same volume of text as premorbidly but in the form of Facebook posts or forum comments.

My ability to place an on line order depends on my familiarity with the site and/or the items being purchased, so I can, having had supported practice of making on line grocery orders, now reasonably order my groceries unaided from my saved list of favourites, but then fail at the superficially simpler task of a one-off turf order on a site that only required me to fill in a total area and a delivery date, where failing to accurately add three numbers together I ordered 1,500 square metres, rather than the required 500 hundred square metres (in retrospect I now know I should have got the person doing the measuring to also total the three component areas, and have learnt to only use unfamiliar sites when there is someone present to check my order before a final purchase.)

 

I often have severe difficulty in maintaining attention and it's very much something that appears over time with mental exertion. In the morning this is less noticable because there is less fatigue.

 

What is interesting is that none of the me/cfs patients were above the average score for the test for controls even though the average difference was only 2/5 vs 3/5. You can try the test online, although it does indeed look nightmarish for a person with me/cfs. The problem with these test is I think what it "means" is a little bit less well established - fatigue, disinterest, attention and many other factors come up when I look around at it. Btw, the Carruthers 2011 criteria is just the ICC if anyone else is also wondering what that is.

 

Interesting study, hope this will replicate.

 

This study found a huge difference between patients and controls on a relevant measure
People have made a lot of valid points about how difficult it is to produce a "fair" test of cognitive function in people with ME, including issues around time-of-day and state of health, and the best way best kind of test to reveal a deficit (which depends on the person being tested). @Ravn @Kitty @Peter Trewhitt @Trish

It's worth bearing in mind that these factors apply to all studies of cognitive function and apply to most cognitive tests. What makes these findings so interesting – if they replicate — is the effect is unusually big.

The way to compare results is using effect sizes. Crudely, an effect size of 0.25 is trivial, 0.5 is modest and 0.8 and above is large. You need the raw data to calculate this properly but you can get a pretty good estimate by dividing the average difference between groups by the average standard deviation of the groups. Here that is 3.04-1.98=1.1/((0.8/0.76)/2=0.78)=1.44

Effect size= 1.44

Or, in lay terms, "bloody enormous".

Previous studies have found only small to modest effect sizes, with the exception of reaction time (effect size=0.9) and that hardly seems central to mecfs.

It's still not a biomarker because of the level of overlap between patients and controls.

But, just maybe, this is a test that, for the first time, provides a useful way to identify meaningful cognitive problems in mecfs, even if only at the level of groups of patients.

 

Good point, I will ask the researchers about this too. Though people who wind up in research studies tend to be hyper-educated (lest we ever think they are really random samples).