Salivary cortisol output before and after cognitive behavioural therapy for chronic fatigue syndrome, 2008, Roberts, Wessely, Chalder, Cleare

[dreampop]

Abstract
Background:There is evidence that patients with chronic fatigue syndrome (CFS) have mild hypocortisolism. One theory about the aetiology of this hypocortisolism is that it occurs late in the course of CFS via factors such as inactivity, sleep disturbance, chronic stress and deconditioning. We aimed to determine whether therapy aimed at reversing these factors–cognitive behavioural therapy for CFS–could increase cortisol output in CFS.

Methods:We measured diurnal salivary cortisol output between 0800 and 2000h before and after 15 sessions (or 6 months) of CBTin 41 patients with CDC-defined CFS attending a specialist, tertiary outpatient clinic.

Results:There was a significant clinical response to CBT, and a significant rise in salivary cortisol output after CBT.

Limitations:We were unable to control for the passage of time using a non-treated CFS group.

Conclusions:Hypocortisolism in CFS is potentially reversible by CBT. Given previous suggestions that lowered cortisol may be amaintaining factor in CFS, CBT offers a potential way to address this.

https://www.kcl.ac.uk/ioppn/depts/pm/research/cfs/publications/assets/2009/Robertssalivary.pdf

 

[dreampop]

This is a small, self-controlled study with a long, variable treatment (CBT ofc) duration up to 6 months. 47% were "responders" to CBT. (sidenote: in this case, I believe as the outcomes are subjective by the therapist rather than the patient). The authors include Wessley and Chalder.

The response to CBT was not consistent with the change in cortisol levels, undermining the hypothesis involving the pathophysiology of the HPA axis. Both non-responders and responders showed an increase in total daily cortisol output. In fact, non-responders showed greater increases in cortisol (+12.8 vs +8)

There was no differential effect of whether patients were judged to have responded to CBT or not apparent in salivary cortisol measures (Table 3). We also calculated the change in the endocrine variables after CBT (i.e. pre-treatment values subtracted from post-treatment values)and performed an independentt-test on these values between responders and non-responders. There were no significant differences detectable (data not shown).We also looked to see if there were any correlations between clinical and cortisol measures after CBT. There were no correlations with total or mean cortisol output.

CBT responders were nearly 6 points higher in total output, almost as much as the increase from CBT, suggesting a slightly better off treatment group to begin with.

However, after CBT, a flatter cortisol day curve (i.e. areduced diurnal change) was associated with worse functioning on the SF-36 social function subscale (r=−0.40,P=0.015) and higher BDI depression scores(r=0.41,P=0.013).

4. Discussion
As outlined, hypocortisolism is one of the most oftenreported biological changes in CFS (Cleare, 2003). We have shown for the first time that cortisol levels can be increased by CBT, with a 16% increase in total cortisol output from 0800–2000h apparent after 6 months of therapy. Our findings are robust due to the careful selection of medication-free subjects, and because patients acted as their own controls before and after CBT, thus excluding other inter-individual factors that might complicate comparisons of HPA axis assessments. Other advantages of the present study include the careful assessment of psychiatric and other comorbidities, and theuse of salivary cortisol, a non-stressful method ofassessing biologically active hormone.

Ok, so +16% total output refers to *responders and non-responders* to CBT.

CBT responders experienced an +11.2% costisol outpot and non-responders experienced a +19.6% change !

Not only is this conclusion very deceitful, the following is a pure contradiction when combined with Wessley's own results. If cortisol increased more in patients that didn't respond to therapy either the therapy or the cortisol finding is being misinterpreted.

We are not directly able to deduce the mechanism bywhich CBT increases cortisol levels from this study.However, there is accumulating evidence that HPA axischanges are not a primary feature of CFS, but instead couldbe a secondary effect of illness. Thus, chronic fatigue sixmonths after Epstein–Barr virus (Candy et al., 2003)orsurgery (Rubin et al., 2005) is not associated with HPA axischanges. However, patients who have had CFS for severalyears, do show HPA axis changes (Cleare, 2003). Since sleep pattern, habitual physical activity levels, physical deconditioning and perceived stress can all exert marked effects on the HPA axis (Cleare, 2003), we hypothesise that these factors, together with potential effects of comorbid psychiatric illness and medication, are largely responsible for the varying degrees of hypocortisolism in longstanding CFS (Cleare, 2004). Since CBT for CFS concentrates on addressing many of these factors, it is plausible that it is via the reversal of these cognitive–behavioural factors that both clinical and endocrine change is produced

Except that the larger change coincided with the unresponsiveness to CBT. So, how can that be. And how can the HPA axis, lifestile and deconditioning *not* be correlated with responsiveness to CBT if BPS model holds.

A final limitation was that we did not use an untreated group of CFS sufferers: since there is clear evidence of theefficacy of CBT we did not feel it would be ethical to delaythis for 6 months to act as a control for this experiment.

...

Thus, we cannot exclude the increase in cortisol output being due to factors independent of CBT such as illness progression or repeated testing.

Final conclusion:

Although previous research suggests that some CFS patients may benefit from low dose hydrocortisone replacement, we suggest that our results, together with the superior evidence base for CBT in CFS, point to CBT being the treatment of choice for the observed mild hypocortiso-lism in CFS at this time

 

[Saz94]

Doesn't cortisol reflect stress? So CBT made them more stressed

 

[dreampop]

Anyone on twitter (I am not) might want to show this paper to Dr. Keith Geraghty as it seems very relevant to the question he asked. Ironically, it's a BPS paper that seems to undermine the position he was querying.

 

[Snow Leopard]

Doesn't cortisol reflect stress? So CBT made them more stressed

Cortisol is a feed-forward metabolic hormone, it's purpose is to provide sufficient blood glucose upon wakening and smooth the blood glucose level over the day given anticipated activity demands (insulin is a feed-back metabolic hormone...).

Despite all the pop-psychology on the contrary, it's more correct to say that cortisol levels reflect sleep-wake and activity patterns than "stress".

 

[obeat]

Anyone on twitter (I am not) might want to show this paper to Dr. Keith Geraghty as it seems very relevant to the question he asked. Ironically, it's a BPS paper that seems to undermine the position he was querying.

Done

 

[Midnattsol]

Done

Could you also share the protocol for the CEBA project with the "sustained arousal model"?

It is the "Forskningsprotokoll" found when one clicks on "Relevant sentral informasjon" at this link:

https://www.ahus.no/fag-og-forsknin...esyke-hos-ungdom#relevant-sentral-informasjon

 

[Hutan]

The Kings College London link in the first post doesn't seem to work anymore. Here's another link to an In Press version
https://www.simonwessely.com/Downloads/Publications/CFS/192.pdf

Journal of Affective Disorders
Amanda Roberts, Andrew Papadopoulos, Simon Wessely, Trudie Chalder, Anthony Cleare

 

[Hutan]

Chronic fatigue syndrome (CFS) is probably a multifactorial condition in which psychological and social factors act alongside biological changes (Wessely et al., 1998). Much biological research has focussed on the hypothalamo-pituitary-adrenal (HPA) axis, with evidence for reduced cortisol output, detectable by sequential salivary samples (Cleare, 2003). Low cortisol may contribute to symptoms since cortisol replacement can ameliorate fatigue and other features of CFS (Cleare et al., 1999, McKenzie et al., 1998).

However, most patients studied have been ill for many years; it is not clear whether HPA-axis disturbances are also present before onset or early in the course of CFS, nor to what extent observed HPA-axis disturbances are secondary to inactivity, sleep disturbance, deconditioning or stress (Cleare, 2004). Indeed, it is possible that there is a vicious cycle present in which low levels of cortisol may exacerbate fatigue and other symptoms, leading to a worsening of some of the factors that had initially contributed to lowered cortisol levels.

 

[Hutan]

Cognitive behavioural therapy (CBT) is an effective treatment for CFS (Whiting et al., 2001). The primary aim of therapy is to identify, modify and change factors that may be maintaining symptoms. Therapy is individually tailored; components include changing unhelpful patterns of rest and activity, improving sleep patterns, increasing exercise capacity, identifying unhelpful cognitions about the illness or the coping strategies used, using problem solving techniques to reduce stress, and treating anxiety and depression if present (Wessely et al., 1998).

In this study we hypothesised that hypocortisolism in CFS is at least partly secondary to inactivity, sleep disturbance, deconditioning and perceived stress, and, therefore, that CBT, which aims to reverse these factors, would lead to increased salivary cortisol output.

 

[Hutan]

Suitability for CBT was judged by the assessing clinician using the framework laid out by Sharpe et al. (1997); essentially the process was to exclude factors that could interfere with CBT (such as untreated severe anxiety or depression, personality disorder or unwillingness to change) and seek patient agreement.

41 patients (26 female) entered the study (mean age 38.4 years (SD 11.3). The mean length of illness at assessment was 57 (SD 48) months.

Patients: filled out the following instruments before and after CBT.

Fatigue: Chalder Fatigue Scale (Chalder et al., 1993); Psychiatric symptoms: General Health Questionnaire- 12 (Goldberg and Blackwell, 1970); Beck

Depression Inventory (Beck et al., 1961). Functional capacity: Medical Outcomes Survey Short Form-36 (Ware and Sherbourne, 1992); Work and Social Adjustment Scale. (Mundt et al.,2002)

Sleepittsburgh Sleep Quality Index (Buysse et al., 1989).

Response to therapy was defined using a therapist- rating of “very much improved” or “much improved” on the clinical global impression improvement (CGI) scale (Guy, 1976) blind to cortisol results.

 

[Hutan]

Samples were taken at 0800h, 1200h, 1600h and 2000h, kept refrigerated overnight and returned by post in the morning. On arrival at the laboratory, they were frozen at −20 °C until assay.

Results of the CBT : Clinical Global Impression (assessed by therapist)
Very much or much improved 18 (47%)
Minimally improved 13 (34%)
No change 4 (11%)
Minimally, much or very much worse 3 (8%)
Missing data 3 (8%)

Note the lumping the negative outcomes together. We don't know how many people were very much worse. Also note the people who gave up on the study.

 

[Hutan]

I'm trying to work out if the levels of cortisol recorded were abnormal.



AI tells me
The normal range for salivary cortisol in adults in the morning is 5–46 nmol/L, and it's usually measured between 8–9 AM.
A normal late-night salivary cortisol level, e.g. 11 pm is 2.6–3.2 nmol/L.

The levels before and after CBT therefore look completely normal. There does not seem much of a rationale for applying CBT to increase cortisol levels if the cortisol levels are normal to start with.

 

[Hutan]

The discussion is full of references to that fact that CFS involves hypocortisolism.

As outlined, hypocortisolism is one of the most often reported biological changes in CFS (Cleare, 2003).

However, there is accumulating evidence that HPA axis changes are not a primary feature of CFS, but instead could be a secondary effect of illness. Thus, chronic fatigue six months after Epstein–Barr virus (Candy et al., 2003) or surgery (Rubin et al., 2005) is not associated with HPA axis changes. However, patients who have had CFS for several years, do show HPA axis changes (Cleare, 2003). Since sleep pattern, habitual physical activity levels, physical deconditioning and perceived stress can all exert marked effects on the HPA axis (Cleare, 2003), we hypothesise that these factors, together with potential effects of comorbid psychiatric illness and medication, are largely responsible for the varying degrees of hypocortisolism in longstanding CFS (Cleare, 2004).

This study did not assess the response of the HPA axis to challenge. However, as yet no specific HPA axis abnormality has been identified (Cleare, 2003) and the explanation for lowered cortisol levels is most likely multifactorial (Cleare, 2004). Nevertheless, we cannot say what aspects of the HPA axis were changed by CBT to lead to increased cortisol output. Although we interpret the results as indicating a normalisation of whatever has led to hypocortisolism, it is also possible that CBT induces some form of compensation for an underlying deficit.

Although previous research suggests that some CFS patients may benefit from low dose hydrocortisone replacement, we suggest that our results, together with the superior evidence base for CBT in CFS, point to CBT being the treatment of choice for the observed mild hypocortiso- lism in CFS at this time.

I'm pretty astounded that nowhere do they actually acknowledge that their sample of CFS patients had normal cortisol levels for sedentary adults. Nowhere do they note what normal cortisol ranges look like and compare the results they found. The closest they come is suggesting that there is only mild hypocortisolism and that they think it is a downstream result.

 

[Hutan]

Then there is the finding that the CBT increase was higher in the people who were assessed by therapists unaware of the cortisol levels as having not responded to the CBT!

The non-responders' total day-time cortisol increased from 65.2 to 78.0, i.e. about 13
The responders' total day-time cortisol increased from 71.2 to 79.2 i.e. about 8

It seems pretty obvious that increasing cortisol levels was not the answer to curing CFS.

 

[forestglip]

As outlined, hypocortisolism is one of the most often reported biological changes in CFS (Cleare, 2003).

For the cited review, I didn't look too closely, but Table 1 has 17 studies. Out of these, 9 found lower cortisol in CFS, 7 found no difference, and 1 found higher cortisol.

 

[Hutan]

Cleare is one of the authors of this 2008 paper, the senior author. This paper quotes him liberally as having found that cortisol is low in CFS (see, for example the quotes in post #14 above). That provides one reason for the 2008 paper to not report that actually their sample had normal cortisol values - it would have contradicted Cleare's earlier work.

If we don't have a thread for that Cleare study, I'll make a thread for it.

 

[bobbler]

https://www.kcl.ac.uk/ioppn/depts/pm/research/cfs/publications/assets/2009/Robertssalivary.pdf

The worst thing about all this is that there is a basic standard fir testing of hypocortisolism - the synacthen test

and that has nothing to do with either timings or behaviour. It’s ’how Much is the max your body can actually produce in response to acth’.

if people don’t have that it’s a sign that’s nothing to do with training

just like those with Addisons or diabetes just don’t have what is needed being produced. They’ll have low morning cortisol too even if they live in a retreat in utopia doing nothing but mindfulness. It’s not a useful measure but a misleading one used as a stooge fir sophism unless taken with other proper measures over long time periods on a 24hr basis with huge samples they have someone who isn’t biased taking the history and diary of.

m but no one’s ever done proper designs on these ever so somehow they’ve entitled themselves as that being permission to make up truisms as being true when they aren’t even for healthy people. So this is nonsense. A snapshot cortisol level only tells you how much stress ie exertion that body is under as a proportion of their max cortisol and an idea of if it’s enough that you’d literally be unable to get up out of bed easily is the reason for the 9am. If it’s not that suggesting it’s because you behave wrong is outrageous but somehow accepted as an assumption that brutalises some people yet is allowed to slip thru because it’s only some people.

I can understand due to their agenda why this bunch pretend it doesn’t exist and gloss over that basic existing but why isn’t it being the standard being mentioned in all our conversations?

this is all just psychosomatic narrative with selective ‘props’ being chosen just like the old days alchemist might have a vandergraff generator on their stage to give the air/impression of being sciencey.

this is just a manifesto for ‘one must do routines we approve of’ justified by lies that makes anyone ‘healthy’ or ‘happy’ (also not proven) taken to an utter disability bigotry level by trying to use sophism to suggest people being ill just need to behave right is ‘proven to work’ . I would have called it a hypothesis rather than a lie if anyone phrased it as a question and then tested it properly in an open-minded way. But on this area it feels it is never called out.

the simple point that it has nothing to do with causing me/cfs is that those lucky enough their routine or way their body behaves fits with the approved timetable when it feels ok aren’t well and recovered. And they should care that those forced thru this get much worst incredibly quickly. So this is very dangerous stuff veiled as if it’s harmless and a starting point tick box. The most dangerous kind.

As I’ve seen many times elsewhere on this forum there are likely to be upstream and downstream things some of which are the body breaking more in certain individuals others are important adaptations to stop other bits breaking more/keep it ‘standing’ despite the broken bits.

all this awful manifesto stuff is making me feel so dreadfully sad about the world and how this vicious brutal hold of people who write this sort of thing will never be ignored and go away and will continue to leave me with never having any basic rights others don’t even realise are privileges they enjoy because they are so fundamental

I’m hoping this has been put up to be critiqued but finding these all being dumped in a big sudden batch very confronting and worrying as if it gives the impression it’s ’well researched’ given the incitement this crud has caused.

 

[bobbler]

Then there is the finding that the CBT increase was higher in the people who were assessed by therapists unaware of the cortisol levels as having not responded to the CBT!

The non-responders' total day-time cortisol increased from 65.2 to 78.0, i.e. about 13
The responders' total day-time cortisol increased from 71.2 to 79.2 i.e. about 8

It seems pretty obvious that increasing cortisol levels was not the answer to curing CFS.

The shocking thing is they are pretending as if the measure they are using is some biological measure of ‘fatigue’ and fatigue alone - like it’s not even in storytelling bs land at least in a curve and needing to be calibrated (because cortisol has other functions so if you need x amount to stay breathing whilst you sleep cos of injuries or other illness the daytime reading being less than that but higher than other peoples means what?)

then of course there’s the me/cfs is about nit deteriorating long term not getting people to push thru and feel upbeat for a few hours at the expense of their future heakth then call it a cure because you won’t measure the actual heakth impact. Which is the long term but we all can’t undo thanks to them and this tosh if thinking beliefs.

 

[bobbler]

For the cited review, I didn't look too closely, but Table 1 has 17 studies. Out of these, 9 found lower cortisol in CFS, 7 found no difference, and 1 found higher cortisol.

All of which even in their bs sophism works could have been interpreted a different way: it’s not scientific but if you follow their warped logic they could/should if they are honest be just as much saying that it shows pwcfs therefore manage stress better than norms and are still ill despite that so it’s nothing to do with theur thinking or coping

but they never do report such which I think is indicative of a dishonest fishing type mindset towards ‘research’ - even if it is still bs it’s just as ‘valid’ as what they chose so stinks if cherrypicking and doing research only to put a one-liner alongside a claim they already want to make?