Psychological Prozac

Psychological Prozac - Can we reliably attribute any improvements to CBT?

I was going to add some comments to the PACE trial thread but those comments grew too unwieldy to clutter up a thread with, so instead here's my debut (and given the length of this - possibly final) blog.

I don't know why I've been thinking about this. I've much better things to do with my time but the ubiquitousness of CBT as a proposed therapy for just about everything (CBT for AGW anyone?) jars with me. Don't get me wrong, outside of its misapplication to ME/CFS, I've no reason to believe that CBT isn't a reasonable and useful adjunctive therapy to help people cope with illness. I've also no reason to believe that its any better than mindfulness meditation in this respect which is why I distrust its apparent over-promotion.

Subjective outcome measures

We've bemoaned the lack of objective outcome measures in the PACE trial all throughout the PACE thread but I doubt the issue is even remotely exclusive to ME/CFS. In fact I've been hard pressed to find any objective measures used outside of the six minute walking test (6MWT) and grey matter brain volume (more on both later) regardless of the disorder being studied.

The problem, as discussed many times, is that much of CBT research has been carried out on psychiatric disorders; mood disorders such as major depression and disorders where 'maladaptive thinking' is assumed by the practitioners to play a part such as biopsychosocial model of ME/CFS. Even for the psychiatric disorders, there are few reliable or commonly used biomarkers meaning that 'disease progression' relies on self rating of subjective symptoms or alternatively observer ratings (presumably the therapist's subjective impression of, for example, a schizophrenia patient's observable symptoms).

Reality based thinking?

At its core, CBT is claimed to be a 'rational' therapy aimed at replacing a patient's maladaptive and unrealistic ways of thinking that may exacerbate their experience of symptoms with more realistic adaptive thinking. Where the only outcome measures are self reported fatigue; physical function; mood etc, there is an obvious difficulty in disaggregating any 'real' effects from a reporting bias bearing in mind that CBT is cognitive plus behavioural therapy. So for example, if part of the therapy was to maintain a diary, would the participant be encouraged not only to interpret their symptoms in a different way but also to describe them in less threatening terms (e.g. to describe themselves as 'pooped' rather than exhausted). Alternatively, if studying say major depression and part of the therapy included encouraging a higher level of activity, might changes in mood or fatigue might not be due to activity as a 'mediator' rather than CBT itself? Which would beg the question of why not just encourage activity rather than expensive therapy to address 'illness beliefs'?

Placebo, controls and standard medical care.

Another issue is the use of relevant control groups in clinical trials of CBT. Unlike pharmaceutical trials, there is no reliable way to use a placebo arm (unless you instruct therapists to agree with all the subject's 'maladaptive beliefs'?) which means that controls are usually dealt with in one of two ways. Either, as with PACE, patients are randomly assigned to either standard (specialist -sic) medical care only or SMC plus CBT. PACE of course added the further arms of the 'alternative pacing' straw man, plus the GET arm which embarrassingly seemed to outperform CBT on several measures. In pharmaceutical trials it is also not unusual for patients to maintain their current medication to give medication only and medication plus arms. The other alternative is to compare over time, patients undergoing CBT with healthy controls on certain parameters such as scores for anxiety, mood, fatigue or even brain volume. The rationale of course being that the patients have a deficit in one of these areas compared to controls and any progress towards 'normality' can be measured.

All of which seems perfectly logical as long as you can effectively disaggregate any additional effects CBT may have over and above standard medical care and that the 'deficit theory' is valid.

The PACE thread has already discussed the problems of disaggregating the additional effects of CBT above those of standard medical care. As above, other mediators such as increased activity may be contributing most of the therapeutic benefit rather than CBT per se and offer much more cost effective interventions. Another possibility, for example if treating major depression with antidepressants plus CBT, is that CBT may only serve to encourage patients to over-report the beneficial effects of the medication.

Correcting a deficit?

Is the deficit theory valid? On the face of it it would seem so. Patients can be clearly shown to have a deficit (whether scoring low on physical function or mood or high on fatigue) compared to healthy controls. Reduced brain volume would appear to be be a clear and measurable objective marker although one that may not necessarily be directly linked to the underlying pathology. In fact healthy controls are usually assumed to represent no deficit be it at or close to 100% on physical function; low or zero on fatigue; to represent optimal mood or representative healthy brain volume.

On the other hand, I'm not so sure that normal healthy controls always represent the 'gold standard' for comparison particularly when reporting on psychological variables in the chronically ill. I also wondered if, by any chance, there had been any research on the effects of CBT on healthy controls.

A stupid idea you might think as there is no deficit to correct and you would hardly use healthy controls as the comparison arm in a pharmaceutical trial?

The eternal optimist?

Given the above, rather than CBT's claimed impacts on specific symptoms, I wondered if there might be any literature on how CBT might affect some broader domains of how normal healthy people experience life such as global measures of well being; happiness; life satisfaction etc? Not surprisingly I was disappointed.

But I did find this interesting discussion of cognitive therapies (presumably rational emotive therapy and the like) and CBT in the context of optimism and pessimism or positive and negative thought patterns (I wouldn't bother reading it unless really interested in these things). Again, it discusses how cognitive (behavioural) therapies aim to replace unrealistic pessimistic patterns of thinking with more realistic optimistic patterns :

Riskind and his colleagues (1996) compared the effects of three group treatments, (standard cognitive therapy, optimism training, and cognitive priming) with progressive relaxation training (as a control group) in a sample of 83 college students. This studys results yielded promising initial support for the effectiveness of the optimism training condition. The optimism training group was superior to the control group on four out of five measures. This included a significantly higher level of optimistic interpretations for negative events and a higher level of positive self-statements.

A (fictional?) case study is discussed of Art, a client who suffers from low self esteem and social insecurity and finds it hard to form relationships with the opposite sex. He expects every attempted romantic encounter to fail. Should Art's unrealistic pessimism be replaced with a similarly unrealistic optimism that encourages Art to believe that he will always be successful when attempting to start a new relationship and what psychological impact would it have if this inevitably turns out to false?.

Thankfully the therapist takes a more sensible approach that encourages Art to be confident enough to attempt to start relationships while being realistic enough to expect to get the odd knock-back.

I wonder whether the PACE participants were encouraged to expect a full recovery and what 'cognitive dissonance' they might experience if in the longer term this isn't the case?

Is normal subnormal?

If cognitive therapies can increase levels of optimism (and how people report their experiences including health related experiences) in healthy people, it does suggest that rather than representing a 'gold standard' comparison group for measuring 'deficits' (read as representing 100%) there is scope on certain measures to score above 'normal' and that CBT can encourage a positive response bias. Which reminds me of the often reported finding that the perceptions of patients with depression are often more realistic than those of the non-depressed and to be 'normal' involves a certain degree of thinking through rose-tinted glasses. Is it at all likely that a similar tendency following CBT to respond positively also affects health related measures such as fatigue and physical function regardless of the 'reality'?

Unfortunately, I can't track it down at this point in time but a paper was discussed on the PACE thread (perhaps relating to Fibro or MS?) that found, to the researchers' puzzlement, that following treatment patients reported less fatigue and better function than healthy controls while no-one was claiming that they were in any way cured?

Hard evidence?

What about the rare objective measures for which CBT has claimed some success such as brain grey matter volume and the 6 minute walk test. Grey matter reductions have been shown in ME/CFS and schizophrenia for example and papers published that appear to show that CBT can reverse these reductions in both cases (Tom Kindlon has previously suggested alternative explanations regarding the ME/CFS paper and similar concerns were raised about the schizophrenia paper). Other technical objections aside, again mediating factors such as increased activity may be responsible for any reversal in brain volume as exercise and learning to play a musical instrument for example have both been shown to increase grey matter volume in healthy people. Its at least possible also that any regular cognitive activity would have similar effects.

Again, not only do 'normal healthy people' not represent a 'gold standard' comparison for brain volume against which to compare a 'deficit' but they also don't represent an 'upper bound'. Normal healthy subjects may be considered to have a grey matter volume deficit compared with a professional musician.

Finally, we've discussed at length how moderate the affects of CBT and GET were on the only remaining objective measure left in the PACE trial the 6 minute walking test.

Spare a thought for the 'surprise' awaiting researchers using CBT to encourage exercise in elderly patients :

Unexpected effects of cognitive-behavioural therapy on self-reported exercise behaviour and functional outcomes in older adults

I'll put you out of your misery. While patients reported taking more 'strengthening' exercise than controls, they found that on the objective measure, their 6 minute walking distance was reduced.


Are we then to assume that normal healthy individuals lead their lives with a sub-optimal brain volume and a less than optimal mood due to a tendency to think in an unrealistic and pessimistic fashion? Do normal healthy people perceive and report on the world in an unrealistic manner? It would seem not as they already appear to take a 'rose-tinted' view of the world. Or is it the case that CBT encourages a (temporary?) state of 'hyper-optimism' akin to a psychological form of Prozac that is even less rooted in reality than 'normal' thinking.

Should it then be the case that any clinical trial of CBT includes an arm with healthy controls also receiving CBT and that any gains in measures such as 'general well being' or 'optimism' be discounted from any claimed therapeutic gains in the patient group?


People are coached to think differently, so they will report differently. This should be expected. The reporting is mediated by their thoughts, not a direct reflection of their actions. It would not surprise me if the primary outcome of this type of CBT was to teach people how to more favourably complete patient surveys.

On the other hand I do think traditional CBT has some benefits. If this whole school were not doggedly pursuing the idea that wrong illness behaviour underlies CFS, and focussed on how to improve patient lives, then CBT might be of some small help. The way it is used is not CBT as understood around the world, just as their CFS is not CFS and so on.

Bye, Alex
I agree Alex. But if I were asked to choose between coping strategies I think I would choose mindfulness over CBT for the sole reason that mindfulness encourages you to try to accept and live as best as you can with your current situation rather suggesting that your interpretation of the situation is in some way flawed.

I would honestly like to think that someone who has lived through decades of chronic illness has a better handle on 'reality' than someone who has just completed a diploma in CBT. That said I'm not saying that there's no benefit it maintaining a generally positive attitude, but reality is reality and sooner or later reality will manifest itself despite any good intentions.
This might surprise you Marco, but traditional CBT is very much like an attempt to make the meditative process of mindfulness into a cognitive one. They both do more or less the same thing, although one has a spiritual basis and the other a linguistic one. I do not think they are incompatible. However, I wouldn't touch the biopsychosocial version with a hundred foot pole, not even if I were allowed to burn the pole afterward.

Blog entry information

Last update

More entries in User Blogs