2013: Attention processes in chronic fatigue syndrome: Attentional bias for health-related threat an

[Esther12]

http://www.sciencedirect.com/science/article/pii/S0005796713001800

I've had this abstract open on my PC for ages because it smelt like a spun null result and I've only just got a copy, so am going to post some stuff up. I quote quite a lot from it as I've really not gone through it that carefully, but thought others might be interested. Also, I'm feeling a bit brain dead, and am just knocking this out: low expectations please!

For a quick summary, skip to the very end where I quote bits from their abstracts and show their results... that's the only really interesting bit imo.

Attention processes in chronic fatigue syndrome: Attentional bias for health-related threat and the role of attentional control

• Ruihua Houa,
• Rona Moss-Morrisb, , ,
• Anna Risdalec,
• Jeannette Lynchd,
• Preshan Jeevaratnama,
• Brendan P. Bradleyc,
• Karin Moggc

Highlights
•
Attention bias for health-threat, and executive attention, were studied in CFS.

•
Attention bias to health-threat is associated with poor executive attention in CFS.

•
Results show heterogeneity in CFS in attention bias and executive attention.

•
Heterogeneity in attention functions in CFS has implications for treatment.

Abstract
Cognitive behavioural models of chronic fatigue syndrome (CFS) propose that attention processes, specifically, enhanced selective attention to health-threat related cues, may play an important role in symptom maintenance. The current study investigated attentional bias towards health-threat stimuli in CFS. It also examined whether individuals with CFS have impaired executive attention, and whether this was related to attentional bias. 27 participants with CFS and 35 healthy controls completed a Visual Probe Task measuring attentional bias, and an Attention Network Test measuring executive attention, alerting and orienting. Participants also completed self-report measures of CFS and mood symptoms. Compared to the control group, the CFS group showed greater attentional bias for health-threat words than pictures; and the CFS group was significantly impaired in executive attention. Furthermore, CFS individuals with poor executive attention showed greater attentional bias to health-threat related words, compared not only to controls but also to CFS individuals with good executive attention. Thus, this study revealed a significant relationship between attentional bias and executive attention in CFS: attentional bias to threat was primarily evident in those with impaired executive attention control. Taking account of individual differences in executive attention control in current intervention models may be beneficial for CFS.

Keywords

• Chronic fatigue syndrome;
• Attentional bias;
• Executive attention control;
• Visual Probe Task;
• Attention Network Test

Also, there was this earlier abstract from a conference which seemed stronger (and has some different authors):

P-229 - An investigation of attention processes in chronic fatigue syndrome: health-threat related attentional biases and attentional control

• R. Hou
  ,
• R. Moss-Morris
  ,
• A. Risdale
  ,
• P. Jeevaratnam
  ,
• K. Mogg
  ,
• B. Bradley
  ,
• R. Peveler

• Abstract
• PDF

Background
Attentional bias is an important psychological mechanism that has been extensively explored within the anxiety literature and more recently in chronic pain. Cognitive behavioural models of chronic fatigue syndrome (CFS) and chronic pain suggest an overlap in the mechanisms of these two conditions. The current study investigated attentional bias towards health-threat stimuli in individuals with CFS, compared to healthy controls. The study also examined whether individuals with CFS have impaired executive attention, and how it was related to attentional bias.

Methods
Two participant groups, CFS (n

=

27) and healthy control (n

=

35), completed a Visual Probe Task measuring attentional bias towards health-threat stimuli (words and pictures) presented at 500ms and 1250ms, and an Attention Network Test measuring alerting, orienting and executive attention. Participants also completed a series of standard self-report measures.

Results
When compared to the control group, the CFS group showed greater attentional bias towards threat-words, but not pictures, regardless of stimulus duration. This was not related to anxiety or depression. The CFS group was also significantly impaired on executive attention compared to the controls. Post-hoc analyses indicated that only CFS individuals with poor executive attention showed a threat-word bias when compared to controls and CFS individuals with good executive attention.

Conclusions
The findings from this study suggest that CFS individuals show enhanced attentional biases for health-threat stimuli, which may contribute to the perpetuation of the condition. Moreover, the attentional biases in CFS are dependent on an individual's capacity to voluntarily control their attention.

I'm not entirely clear what analyses were planned and which were post-hoc.


From the paper:

The psychological versus biological
pathogenesis of the condition has been hotly debated. However,
there is increasing consensus that CFS is multifaceted and heterogeneous
in nature. The cognitive behavioural model of CFS attempts
to incorporate this heterogeneity through describing the
interaction between biological and psychosocial factors. The model
suggests that factors such as genetics, high perfectionist tendencies,
and/or a history of psychological distress predispose
individuals to CFS. The initial symptoms are then precipitated
through events such as an acute infection and/or stress. Cognitive
and behavioural factors such as negative illness beliefs and all-ornothing
behaviour perpetuate the symptoms and associated
disability (Moss-Morris, 1997; Moss-Morris & Petrie, 2003; Surawy,
Hackman, Hawton, & Sharpe, 1995; Wessely, Butler, Chalder, &
David, 1991).
The perpetuating factors are seen as key to the model and are
the focus of successful treatments for CFS (Castell, Kazantzis, &
Moss-Morris, 2011). There has also been considerable empirical
support for the content of the illness cognitions hypothesised to be
important in the maintenance of CFS, including negative beliefs
about the consequences, timeline and controllability of the condition
(Moss-Morris, 2005; Moss-Morris, Spence, & Hou, 2011). Little
research has, however, been conducted into the cognitive processes
that may develop or maintain particular cognitions or beliefs. It is
believed that distortions or biases in cognitive processing (e.g.,
selective attention to health-threat information, negative bias in
the interpretation of somatic information) may influence the
development of negative illness representations, thus maintaining
the severity and duration of the symptoms (Moss-Morris & Petrie,
2003). Furthermore, there may be a reciprocal relationship between
illness beliefs and cognitive biases, as illness beliefs may
promote increased attentional focus on somatic symptoms (Deary,
Chalder, & Sharpe, 2007).

They really need some evidence of bias and cognitive distortions to support the way they've treated patients in order to avoid being utter quacks.

How they looked for attentional bias:

It involves presenting a series
of pairs of stimuli (e.g. a threat-related picture paired with a neutral
picture; or a threat word paired with a neutral word) on a computer
screen. Each stimulus pair is presented briefly (e.g. 500 ms), followed
by a probe (e.g. arrow or dot). Participants are asked to
respond to the probe as quickly as possible by pressing a response
button. An attentional bias towards threat is reflected by faster
response times to probes replacing threat cues than neutral cues.
The exposure duration of the stimuli can be varied to examine the
time-course and component processes of attentional biases.

You would think that this means that they should use controls suffering from ill health?:

A bias in initial orienting may reflect rapid automatic capture
of attention by salient stimuli, in the absence of detailed elaborative
processing; whereas a bias in maintained attention may reflect
more prolonged dwelling and rumination on personally relevant
information. Anxiety is commonly associated with bias in initial
orienting to threat; and depression with bias in maintained attention
on information related to loss and sadness, although overlap in
these patterns of bias is sometimes noted (e.g., reviews by
Armstrong & Olatunji, 2012; Bar-Haim et al., 2007; Gotlib & Joormann,
2010). Chronic pain has been associated with bias for painrelated
cues in both early and later aspects of attentional processes
(Schoth et al., 2012); moreover, this attentional bias is not
accounted for by anxiety or depression, and may operate primarily
for stimuli related to the individual’s pain-related concerns
(Crombez et al., 2013). Across pathologies, attentional biases are
likely to be found for information which has high personal salience,
or relevance to the individual’s disorder (e.g., Crombez et al., 2013;
Hankin, Gibb, Abela, & Flory, 2010; Williams, Mathews, & MacLeod,
1996).

Few studies have investigated attentional bias in CFS. Using the
modified Stroop task, Moss-Morris and Petrie (2003) failed to find
an attentional bias for somatic words (e.g., sick, dizzy) in CFS individuals.
In contrast, Hou, Moss-Morris, Bradley, Peveler, and
Mogg (2008), using the VPT, demonstrated an attentional bias towards
health-threat stimuli (words and pictures) in individuals
with CFS compared to healthy controls. However, this study used a
relatively small sample size (14 participants with CFS) and attentional
bias was only assessed at a single duration (500 ms). More
recently, Martin and Alexeeva (2010) used a modified exogenous
cueing task and found no attentional bias to illness-related information
in individuals with CFS, compared to healthy controls. They
presented their stimuli for 100 ms, compared to 500 ms in the Hou
et al. (2008) study, which led Martin and Alexeeva to suggest that
the attentional bias in CFS may primarily occur at a later, more
strategic stage of processing (which was not assessed with their
100 ms stimulus duration). However, the modified cueing task
(which presents a single cue on each trial, rather than a pair of
stimuli) has methodological complications which can make the
interpretation of results unclear (see Mogg, Holmes, Garner, &
Bradley, 2008, for details). Further research is needed to examine
the time-course of attentional bias towards health-threat stimuli in
a larger and well-defined CFS sample. Such research may help
identify specific cognitive anomalies in CFS; e.g., whether or not
attention is automatically captured by illness-related information,
and whether individuals with CFS maintain their attention on such
information (i.e., similar to the pattern of bias found in chronic pain,
Schoth et al., 2012). In the longer term, it may also prove useful to
assess the effects of treatment on such specific attentional biases,
and to refine cognitive interventions to alter biases that may
maintain or exacerbate CFS.

Unsurprisingly, those with CFS tend to be a bit rubbish at testing:

Although evidence of attentional bias in CFS is mixed, it has
been established that CFS is associated with attentional deficits
(Cockshell & Mathias, 2010; Dickson, Toft, & O’Carroll, 2009).

CFS patients also took nearly twice as long on average to respond to this test.

'Recovered' PACE participants could have been included [can't resist the dig]:

To be included in the study CFS participants needed to have
received a diagnosis of CFS in the past 6 months and score more
than 4 on the Chalder Fatigue Scale (Chalder et al., 1993), which is a
standardised cut-off for caseness of fatigue.

Here are their results:

They don't seem that interested in the fact that CFS patients classed as having good executive function seem to have a bias against threat words. To me, those results look like a mess indicative of nothing. It looks like there would not be overall evidence of bias from pictures and words, as the 'bias' is going in opposite directions.

Visual Probe Task: attentional bias in CFS versus control groups
The ANOVA of bias scores indicated that the CFS group showed
greater attentional bias towards health-threat words than pictures,
compared to the control group; i.e., the 2 (group) 2 (stimulus
modality) 2 (stimulus exposure duration) ANOVA demonstrated
a significant interaction between group and stimulus modality,
F(1,59) ¼ 3.98, p ¼ 0.05. Mean bias scores in CFS and control groups
were 10 ms and 3 ms, respectively, for threat words; and 10 and
0 ms, respectively, for threat pictures. Post hoc between-group
contrasts were not significant. There were no other significant
ANOVA results; e.g. the main effect of group, predicted by
Hypothesis 1, was not significant, F < 1. See Table 2 for means.
Attention Network Task: executive attention in CFS versus control
groups
The CFS group had significantly higher ANT conflict scores than
the control group, t(59) ¼ 2.65, p ¼ 0.01, indicating that the CFS
group had impaired executive attention; see Table 2. There was no
significant difference between the two groups in alerting,
t(59) ¼ 1.31, p ¼ 0.20, or orienting, t(59) ¼ 0.56, p ¼ 0.58, scores.
Relationship between attentional bias and executive attention
There was a significant positive correlation between ANT conflict
scores and attentional bias for threat words in the CFS group,
r(23) ¼ 0.60, p ¼ 0.001, but not in the control group, r(33)¼0.23,
p ¼ 0.19. That is, poorer executive attention in the CFS group (i.e.,
higher conflict score) was associated with increased attentional
bias for threat words. To assess whether the relationship between
CFS and attentional bias is clarified by taking account of attention
control (as predicted by Hypothesis 3), the CFS group was split into
two groups with good executive attention versus poor executive
attention, based on the median ANT conflict score; M (SD) of conflict
scores were 38 (19) and 148 (82) ms, respectively
(median ¼ 66; n ¼ 13 in each subgroup). The high and low executive
attention CFS subgroups did not differ significantly in missing
RT data (9 versus 11% of trials, respectively), t(24) ¼ 1.50, p ¼ 0.15,
or overall mean RT (898 versus 1004 ms, respectively), t(24) ¼ 1.10,
p ¼ 0.28. The CFS subgroups also did not differ significantly in
gender, marital status, level of education (ps > 0.2), or in age,
duration of CFS, fatigue severity, HADS anxiety or HADS depression
(all ts(24) < 1, ps 0.4).
A 3 2 2 mixed design ANOVA of attentional bias scores was
carried out with three levels of group (CFS with good executive
attention; CFS with poor executive attention; healthy control) as
the between-subjects factor, and stimulus modality and exposure
duration as within-subject factors. This demonstrated a significant
main effect of stimulus modality, F(1,57) ¼ 5.24, p ¼ 0.03, and, more
importantly, a significant interaction between group and stimulus
modality, F(2,57) ¼ 6.66, p ¼ 0.003. See Fig. 1.
The latter interactionwas clarified using one-way ANOVA of bias
scores for words and pictures separately. There was a significant
main effect of group on attentional bias scores for threat words,
F(2,57) ¼ 8.44, p ¼ 0.001, but not for threat pictures, F < 1. Post hoc
contrasts showed that the CFS group with poor executive attention
had significantly greater attentional bias for threat words (M ¼ 31,
SD ¼ 41), in comparison with the CFS group with good executive
attention (M ¼ 8, SD ¼ 22, p < 0.01) and also with the control
group (M ¼ 3, SD ¼ 22, p < 0.01). There was no significant difference
between the CFS group with good executive attention and
the control group in threat-word bias scores (p ¼ 0.56). Comparison
of bias scores against a value of zero showed that only the CFS
group with poor executive attention showed a significant bias for
threat words, t(12) ¼ 2.68, p ¼ 0.02.
Relationships between attentional bias, executive attention, and
symptom measures in CFS group
Because the CFS group had higher anxiety and depression than
the control group, we further explored using correlations whether
anxiety and depression could account for the poor executive
attention and the attentional bias within the CFS group. Pearson
correlations showed that there were no significant relationships
between the self-reported symptom measures (anxiety, depression,
and fatigue severity) and attentional measures (attentional
bias for threat words and pictures, and ANT conflict scores
reflecting executive attention); r values ranged between 0.07 and
0.17, ps > 0.39.

This bit would seem to be more of a problem for them than their abstract acknowledges:

The CFS and control groups did not significantly differ in overall
attentional bias (i.e. averaged across both pictorial and linguistic
stimulus modalities and both stimulus exposure durations). Thus,
the first hypothesis was not supported. In a previous study by our
research team, individuals with CFS showed enhanced attentional
bias, averaged across both health-threat pictures and words (Hou
et al., 2008).
Methodological differences between studies may contribute to
the different pattern of results. For example, the studies used
different inclusion criteria and sources of CFS participants; the
former study used a smaller CFS sample recruited from a support
group using self-reported CDC criteria, whereas the present study
used a larger well-defined group diagnosed by a specialist medical
practitioner at a CFS outpatient clinic. It may be that patients
recruited from the support group were more homogeneous in
nature than those recruited from a hospital clinic. There is some
evidence that belonging to a CFS support group is associated with
poorer prognosis and higher level of functional impairment in
everyday activities (Sharpe, Hawton, Seagroatt, & Pasvol, 1992).
Perhaps health-threat images (e.g., person in wheelchair) may be
perceived as subjectively more aversive by CFS sufferers who have
higher levels of functional impairment, and thus more likely to
capture their attention. While this is speculative, it could be
explored in future research by asking participants to rate the subjective
unpleasantness, or concern-relatedness, of the stimuli.

Neither attentional bias nor executive attention was associated
with self-reported measures of anxiety, depression or fatigue
severity in individuals with CFS. This is consistent with a recent
meta-analysis in pain reporting that attentional bias is not associated
with anxiety, depression, or pain severity (Crombez et al.,
2013), and also with our previous finding that attentional bias in
CFS was independent of anxiety and depression (Hou et al., 2008).
The lack of relationship between attentional bias and negative
mood might perhaps seem surprising given that anxiety and
depression are associated with attentional biases for negative
stimuli (Bar-Haim et al., 2007; Gotlib & Joormann, 2010). However,
other research has shown an attentional bias for health-threat cues
(across the two stimulus exposure durations of 500 and 1250 ms)
which was associated specifically with physical-health worries,
rather than anxiety or depression (Mogg, Wilson, Hayward,
Cunning, & Bradley, 2012). Thus, attentional biases in healthrelated
conditions, such as CFS or chronic pain, may be largely a
function of the personal salience or subjective aversiveness of the
health-threat cues (which is likely to depend more on the relevance
of the stimuli to the individual’s illness concerns and beliefs, rather
than anxiety or depression per se), as well as the person’s ability to
control attention and inhibit processing of task-irrelevant information.
Therefore, attentional bias for health-threat related information
may be a discrete cognitive feature, which may contribute
to specific symptoms in a subgroup of patients, rather than being a
maintaining factor in all patients with CFS.

What they see as the implications (not 'fire the quacks'):

The present findings have potential clinical implications. First, as
noted by Solberg Nes, Roach, and Segerstrom (2009), executive
attention may play an important role in the development and
maintenance of chronic multisymptom conditions, such as CFS and
chronic pain, because executive functions contribute to selfregulatory
capacity (i.e., ability to exert control over cognition,
emotion, behaviour and physiology), including the ability to resist
attention capture by health-threat cues. As attention training
methods can improve executive attention (e.g. using computerbased
exercises, Tang & Posner, 2009), individuals who have a
combination of CFS and poor attention control may benefit from
such training to reduce executive attention deficits (e.g. improve
concentration, reduce distractibility) and reduce attentional capture
by health-threat cues. Second, pre-treatment assessment of
attentional bias to threat cues may help predict responders to
cognitive behaviour therapy (CBT) in CFS, given that recent studies
have shown that attentional bias for threat information predicts
treatment response to CBT in other clinical conditions, including
anxiety (Price, Tone, & Anderson, 2011; Waters, Mogg, & Bradley,
2012). Third, a novel computer-based intervention has been
developed, namely, attention bias modification training, which is
specifically aimed at reducing attentional bias to threat stimuli
(Hallion & Ruscio, 2011; MacLeod & Mathews, 2012). Combining
such training with CBT may also benefit this subgroup of CFS patients.
Thus, it would seem useful for future research to examine
the extent to which attentional variables, such as attentional bias
towards threat and poor executive attention control, predict
treatment response to CBT in CFS sufferers, given that their
response to CBT is variable; and also to design attentionmodification
tasks aimed at improving executive attention and
reducing attentional bias to threat, which may help enhance
treatment effects of CBT for CFS. It would also be helpful to clarify
the inter-relationships between attentional bias, executive attention,
and health-related cognitions, including illness concerns and
beliefs, in CFS. For example, while poor executive attention may
increase susceptibility to attentional bias and intrusive illnessrelated
cognitions, a high level of illness concerns and preoccupations
may also impair executive attention control and
contribute to the attentional bias for health-threat cues. Thus,
reduction of illness concerns and beliefs, e.g. using CBT, may also
help reduce capture of attention by health-threat cues in CFS.

This is the only bit really worth reading:

Back to their abstracts:

Compared to the control group, the CFS group showed greater attentional bias for health-threat words than pictures; ... attentional bias to threat was primarily evident in those with impaired executive attention control.

Results
When compared to the control group, the CFS group showed greater attentional bias towards threat-words, but not pictures, regardless of stimulus duration. This was not related to anxiety or depression. The CFS group was also significantly impaired on executive attention compared to the controls. Post-hoc analyses indicated that only CFS individuals with poor executive attention showed a threat-word bias when compared to controls and CFS individuals with good executive attention.

Conclusions
The findings from this study suggest that CFS individuals show enhanced attentional biases for health-threat stimuli, which may contribute to the perpetuation of the condition. Moreover, the attentional biases in CFS are dependent on an individual's capacity to voluntarily control their attention.

And their results:

As I said:

They don't seem that interested in the fact that CFS patients classed as having good executive function seem to have a bias against threat words. To me, those results look like a mess indicative of nothing. It looks like there would not be overall evidence of bias from pictures and words, as the 'bias' is going in opposite directions.​

Also, I thought I'd post the conclusion from the smaller Moss-Morris study which was positive, and this is a larger follow-up for:

Conclusion:
The finding of an AB towards
health-threat information in individuals with CFS is supportive of
models of CFS which underlie cognitive behavior therapy.

http://www.jpsychores.com/article/S0022-3999(08)00113-X/abstract

Shouldn't the conclusion for their larger study have been:

"The absence of evidence of an AB towards health-threat information in individuals with CFS challenges the models of CFS which underlie cognitive behavior therapy."

PS: That earlier study is cited by them in this 2013 chapter on CFS I was sure I'd read:

http://www.sciencedirect.com/science/article/pii/B9780444529015000253

You can get a lot of it on google: http://books.google.co.uk/books?hl=...aC50M2v2MG1ANXNY32T99djpM#v=onepage&q&f=false

They cite the article to say "CFS patients also tend to be hypervigilant to illness and symptom information" - the whole chapter is pretty irritating and full of claims built on sand. Okay - night-night!

 

[Cheshire]

Thanks a lot Esther 12!
I was wondering yesterday if there were any study proving the illness belief distortions that we are supposed to have. This study is not the one that is going to prove anything. It is just money waste...

How can they pretend to have a scientific approach when there is no control group with a disease similar to ME/CFS???

Do someone know about other studies aiming at proving our cognitive bias?

 

[WillowJ]

I agree that they should have used a disease control group.

I also couldn't find a part where they compared "threat" with "neutral" reaction times. From the data given (unless I missed something), isn't it plausible that people with ME/CFS have slightly more difficulty interpreting pictures than healthy people?

 

[A.B.]

How can they pretend to have a scientific approach when there is no control group with a disease similar to ME/CFS???

There is no intention to investigate reality. It's about pretending to be scientific just enough to justify intervention. Psychology is pseudoscience. Notice how even with a control group, there would be no way to know whether an allegedly increased attention towards illness has any clinical significance. It is just assumed to have significance. This study is a complete waste of money. I wonder when the house of cards will come crashing down.

 

[Valentijn]

A good indication of crap results is when the patients are divided into multiple groups based upon a largely arbitrary measurement. In this case, there's only a "significant" difference between a subgroup of patients and the controls, which means they found no difference between controls and the patients as a single group.

In short, they found nothing, and messed with the data until they could find a correlation. And even then, the correlation only applies to so-called "threat words" but not photos. Their study was a huge failure, and they are engaging in extensive spin to present results which look positive to people who don't know any better.

 

[alex3619]

Lets presume there really is cognitive bias toward threat-related words, their best-case scenario. What does this tell us?

If you presume the behavioural models, it tells us this supports their view. If you presume a biomedical model, then being more cogntively impaired means you might be sicker, and so pay more attention. So it supports a fully biomedical model too. When data supports whatever model you choose, using the model as a presumption, then the fallacy of Begging The Question is in play. Its the model driving the interpretation, not the data. Its a methodologically flawed study, with flawed analysis.

The sad fact is that, yet again, they are conflating association with causation, by implication. This is not merely bad science, its evidence that these people should never get through peer review. In a journal with strict and enforced scientific standards this should not happen.

Not all psychology is pseudoscience. They just have a slightly larger share of it. Psychiatry has a big share too. Psychogenic psychiatry on the other hand is mostly pseudoscience by every definition of pseudoscience I have so far looked at.

One current estimate is that half of all psychological findings are wrong. Thats too much for the field to be considered strong science, but then it never has been ... its soft science, an euphemism for not really scientific but trying to be.

 

[alex3619]

A good indication of crap results is when the patients are divided into multiple groups based upon a largely arbitrary measurement. In this case, there's only a "significant" difference between a subgroup of patients and the controls, which means they found no difference between controls and the patients as a single group.

In short, they found nothing, and messed with the data until they could find a correlation. And even then, the correlation only applies to so-called "threat words" but not photos. Their study was a huge failure, and they are engaging in extensive spin to present results which look positive to people who don't know any better.

This is just one way results can be spun and twisted to support a view. It has a formal name, but it escapes me at the moment. Its a deeply entrenched and widely used methodological bias that generates oodles of spurious results in the literature.

 

[Bob]

Interesting comments. There's enough material in these comments for a letter to the journal.

 

[biophile]

There is an understandable acknowledgment that poor executive function could contribute to attentional bias, but there seems to be an unjustified assumption that attentional bias is contributing to symptoms via hypervigilance. The results are weak and contradictory, so for all we know, they cannot be confidently extrapolated to paying attention to symptoms.

Dare I suggest that illness puts the body under chronic biological stress, and as part of legitimate sickness behaviour, the brain is primed to detect possible further threats to homeostasis, including on some level, threatening words and pictures.

An adequate disease control group would need to experience similar symptom categories and similar levels of stigma. Compared to other chronic conditions, ME/CFS basically has a rare ratio of symptom burden to societal ridicule/dismissal, so finding such controls in a wealthy country where the research is conducted may be rather difficult.

I wonder if CBT/GET proponents experience attentional bias towards scary threat words such as "actometers", "exercise testing", "letters to the editor", "PubMed Commons", and "patients researching their condition on the internet"?

Would they click faster on the buzzer in response to the pictures and words in these two videos?
http://forums.phoenixrising.me/index.php?threads/the-fable-known-as-the-pace-trial.29991/

 

[Wildcat]

.
The psychosocial research funding gravy train rolls ever onwards; there seems to be some kind of bottomless pot of research funding gold for it.

Who on earth keeps funding this dross? And how come they have so much money to burn in poverty struck Britain (poverty struck when it comes to healthcare, and useful ME bioresearch, that is)

.
When concerned patients and carers wanted to submit evidence to the West of England Research Ethics Service to explain why doing a Lightning Process study on youngsters was not a good idea (Dr Esther Crawleys SMILE Trial)..... it transpired that local Research Ethics Committees only accept evidence from the researcher and the research funders.

The National Research Ethics Service (NRES) made an exception to receive evidence from patients and carers about Lightning Process, which the NRES passed on to the West of England RES.

.
I am not suggesting that this 'An investigation of attention processes in chronic fatigue syndrome: health-threat related attentional biases and attentional control' research is unethical, but that funders and LRES's seem to rubber stamp the same kinds of research by the same researchers decade after decade.
.

.
.

 

[peggy-sue]

I'm going to have to have a big think about this - but I reckon they're simply tapping in to areas of cognitive disability and not taking into consideration the actual time taken for the brain to achieve different processes.

I've only read the first bit and the stuff about health-treat words being recognised before health threat images - but we are slow at getting info. from an image and putting it together and interpreting it.

Individual common words are instantly recognised by fluent readers, much as an advertising logo can be recognised.
It's a much shorter processing time.

(Attentional processes were someting I specialised in a bit at uni.)

 

[peggy-sue]

I wrote that before reading the rest of the thread, before I forgot.
I am confirming Willow's suspicion - the reality of how the brain works backs it up.

Normal/control subjects would have paid more attention to health threat words/images too.

 

[Esther12]

People seem more interested in this than I expected.

To me, the fact that the CFS group 'with good executive function' had a significant bias against threat words (as opposed to what was hypothesised), and their disinterest in commenting upon this, seems particularly amusing.

(There was a recent prospective study which had predicted children with lower IQ would be more likely to develop CFS, but...: "We found an association between higher cognitive ability and self-reported CFS/ME, which was in contrast to our hypothesis. Since the association was only found when additionally adjusting for psychological symptoms and since it was rather small, we might consider this as a chance finding." http://forums.phoenixrising.me/inde...ldhood-cognitive-ability-and-somatic-s.26753/ )

I also couldn't find a part where they compared "threat" with "neutral" reaction times. From the data given (unless I missed something), isn't it plausible that people with ME/CFS have slightly more difficulty interpreting pictures than healthy people?

I think that is what the main results are - comparing responses to threat words/images with responses to non-threat words/images.

Just opened the paper up to get some quotes:

The CFS group had
slower overall mean RT than the control group (960 vs. 577 ms),
t(59) ¼ 8.09, p < 0.001.

Attentional bias scores were calculated for trials with threat neutral
stimulus pairs by subtracting the mean RT when the
threat stimulus and probe were in the same position (congruent
trials) from the mean RT when the threat stimulus and probe were
in different positions (incongruent trials). Positive values of bias
scores reflect an attentional bias towards threat. Negative values
indicate avoidance of threat. Bias scores were calculated separately
for each participant, stimulus modality (picture vs. word) and
exposure duration condition (500 ms and 1250 ms) (Lees et al.,
2005; MacLeod et al., 1986).

Also thought I'd post results for the combined scores:

Table 2

Attentional bias scores for health-threat from the VPT, and alerting, orienting and
conflict scores from the ANT in the CFS and control groups.

VPT
Picture bias (500 ms) CFS: -10.3 (71.8) Control: 3.9 (35.1)
Picture bias (1250 ms) CFS: -9.0 (51.4) Control: -4.8 (27.5)
Word bias (500 ms) CFS: 7.4 (65.4) Control: -4.7 (39.4)
Word bias (1250 ms) CFS: 13.5 (42.8) Control: -0.5 (31.5)
Overall picture bias CFS: -9.7 (43.6) Control: -0.4 (23.6)
Overall word bias CFS: 10.4 (38.3) Control: -2.6 (21.8)

Also - here are their hypotheses.

The main predictions were as follows:
Hypothesis 1. Individuals with CFS, relative to controls, will show
an attentional bias for health-threat words and pictures across both
stimulus durations (500 and 1250 ms).

Hypothesis 2. They will also show impairment in executive
attention on the ANT.

Hypothesis 3. Individuals who have a combination of CFS and
poor executive attention control will show greater attentional bias
to health-threat cues, compared with individuals who have CFS and
good attention control, as well as in comparison with healthy
controls.

1:

There were no other significant
ANOVA results; e.g. the main effect of group, predicted by
Hypothesis 1, was not significant, F < 1. See Table 2 for means.

Hypothesis 2, that patients with CFS would perform less well in testing, was just right.

3:

They seem to avoid deciding whether they think that their data supports hypothesis 3 or not:

To assess whether the relationship between
CFS and attentional bias is clarified by taking account of attention
control (as predicted by Hypothesis 3), the CFS group was split into
two groups with good executive attention versus poor executive
attention, based on the median ANT conflict score; M (SD) of conflict
scores were 38 (19) and 148 (82) ms, respectively
(median ¼ 66; n ¼ 13 in each subgroup). The high and low executive
attention CFS subgroups did not differ significantly in missing
RT data (9 versus 11% of trials, respectively), t(24) ¼ 1.50, p ¼ 0.15,
or overall mean RT (898 versus 1004 ms, respectively), t(24) ¼ 1.10,
p ¼ 0.28. The CFS subgroups also did not differ significantly in
gender, marital status, level of education (ps > 0.2), or in age,
duration of CFS, fatigue severity, HADS anxiety or HADS depression
(all ts(24) < 1, ps  0.4).
A 3  2  2 mixed design ANOVA of attentional bias scores was
carried out with three levels of group (CFS with good executive
attention; CFS with poor executive attention; healthy control) as
the between-subjects factor, and stimulus modality and exposure
duration as within-subject factors. This demonstrated a significant
main effect of stimulus modality, F(1,57) ¼ 5.24, p ¼ 0.03, and, more
importantly, a significant interaction between group and stimulus
modality, F(2,57) ¼ 6.66, p ¼ 0.003. See Fig. 1.
The latter interaction was clarified using one-way ANOVA of bias
scores for words and pictures separately. There was a significant
main effect of group on attentional bias scores for threat words,
F(2,57) ¼ 8.44, p ¼ 0.001, but not for threat pictures, F < 1. Post hoc
contrasts showed that the CFS group with poor executive attention
had significantly greater attentional bias for threat words (M ¼ 31,
SD ¼ 41), in comparison with the CFS group with good executive
attention (M ¼ 8, SD ¼ 22, p < 0.01) and also with the control
group (M ¼ 3, SD ¼ 22, p < 0.01). There was no significant difference
between the CFS group with good executive attention and
the control group in threat-word bias scores (p ¼ 0.56). Comparison
of bias scores against a value of zero showed that only the CFS
group with poor executive attention showed a significant bias for
threat words, t(12) ¼ 2.68, p ¼ 0.02.

 

[peggy-sue]

I'm never keen on using ANOVA - it's only use is for ascertaining you actually do have a realistic graph drawn and that the points do fit on it, rather than it being a plot on top of a spread of points.

Paired t-tests would have been far more suitable and informative.

BTW - have they got any evidence for their measuring "executive function"?
Are they using this term from a "cognitive" point of view - ie, it is an assumed bit of "software of the mind",

or have they got realistic biological underpinnings of what they consider it to be in terms of real brain function?

 

[peggy-sue]

PWME have slower reaction times to pressing buttons than normals, for a start.
It's not just our brains that are slower and require more areas to perform oordinary things - we are physically slower to move our arms and fingers too.
I'm even sure it's been proven in a study, but do not ask me where.

This is just typical experimental psychology.
Find a couple of groups of folk, shove them in experimental conditions, make a few measuremennts, change the conditions, make a few more measurements...
(conditions changing should generate different numbers)

Apply the great god of ANOVA...

then off to the ivory towers to fiddle with the all important numbers (this is what makes them a science, you see - they have numbers to fiddle with) and speculate a load of philosphical garbage.

 

[alex3619]

There are so many ways to stack studies its causing a crisis in psychology. Yet the practices are endemic, and journals and reviewers do not typically look for these issues. This is before you get to the serious biases in psychogenic CFS research.

 

[Esther12]

I'm never keen on using ANOVA - it's only use is for ascertaining you actually do have a realistic graph drawn and that the points do fit on it, rather than it being a plot on top of a spread of points.

Paired t-tests would have been far more suitable and informative.

BTW - have they got any evidence for their measuring "executive function"?
Are they using this term from a "cognitive" point of view - ie, it is an assumed bit of "software of the mind",

or have they got realistic biological underpinnings of what they consider it to be in terms of real brain function?

re executive function: I think that this was just based on average speed of response. Something like that. If important I could go back and check.

re ANOVA: Do you fancy doing a 'for dummies' explanation of that? I didn't really follow that section of the paper - don't worry if it's tricky though.

 

[peggy-sue]

I seem to remember "executive function" was the big controller of all other "mind" functions in psychology; rather a vague and very assuming sort of thing.

Oh dear, Esther, I don't think I could do a "for dummies" now.

I did a big sort of exposition on it at uni, for myself, because I didn't trust computers or the SPSS package I was being made to use.
I like to work a few examples of all stat tests through on my own - to check that the computer does give the same results and to be certain I'd entered all the numbers into the right places on the spread sheets.
And to find out exactly what the test is doing with my numbers.

I did a few ANOVAs on paper, but while doing it, was not at all happy about what it was measuring - it does means or averages of differences, and it can't tell you when there are interactions although it claims to, because of the problem of the number of tails. Psychology does not seem to understand that in a living system, a response of "no response" is still a real response.
.

 

[Esther12]

No worries peggy-sue. Whenever I try to learn stats stuff I end up forgetting it the following month anyway!

 

[Dolphin]

I was just coming here to make some points on this but see Esther12 has said what I was going to point out:

You would think that this means that they should use controls suffering from ill health?

Yes

View attachment 7221

They don't seem that interested in the fact that CFS patients classed as having good executive function seem to have a significant bias against threat words. To me, those results look like a mess indicative of nothing. It looks like there would not be overall evidence of bias from pictures and words, as the 'bias' is going in opposite directions.

(see attachment above) Yes, exactly.