How fatigue researchers reason away nil results and negative findings regarding CBT in CFS and QFS.

Studies that show treatment using cognitive behavioral therapy with activity-building program does not work, are now presented as if they were a success and an underpinning for the claimed favorable effect. A recent example is the Qure study and the corresponding follow-up study, showing that the treatment is not working. An earlier example is an investigation into group CGT. Both publications show how zero results and even negative findings are reasoned away.

Research 1. QFS  

At the end of 2017, the publication describing the Qure study[i] was presented, a study into the effects of CBT (Cognitive Behavioral Therapy) with a CFS approach and the effect of antibiotics on patients who are supposedly suffering with QFS.

QFS (Q Fever Fatigue Syndrome) occurs when symptoms persist and no active form of the bacterium Coxiella burnetii is found during lab tests. The original study found a very questionable outcome with regard to the effect of CBT. As I described earlier [ii], according to this study, there is indeed a statistically significant improvement in the CIS F score in the CBT group, but what is the relevance of this statistically significant improvement? The score of 31.6 still does not match the score <27 that is used as normal in other studies. (Knoop et al, 2007). What is the value of such a result if the level of activity is not increased when that is precisely the objective? It is striking that the only score (SIP 8) that indicates at least something about physical functioning does not lead to a substantial improvement. The score of 786.8 in the CBT group is so poor that in other studies patients are considered to be severely limited and it will lead to inclusion in research. The score is also completely inadequate compared to the normal value chosen in this study (450). As a normal value, a score of 203 is used in other studies [iii] (Knoop et al. 2007)

At the end of 2018, the follow-up study [iv] was published which clearly showed that CBT did not work. On the contrary, the CBT scores are significantly worse in the follow-up measurement, while the Placebo score on CIS-F even improves (although modestly). The outcome scores on SIP8 (physical functioning), which are completely inadequate for all three groups in the endpoint measurement, deteriorate at the follow-up for the CBT group but improve for the Doxycycline and Placebo group. The result at follow-up is that the Placebo group has better scores on CIS-F and on SIP8 (although still insufficient) than both treatment groups.

Quotes from the follow-up study:

  1. “Due to its initial positive effect and side effects of long-term doxycycline use, CBT is still advised as therapy of choice for QFS patients. At present, it is still the only well-investigated treatment modality for QFS patients with a positive effect.”
  2. “Throughout this study, all patients were functionally impaired. It is conceivable that persisting functional impairment leads to a constant confrontation with limitations caused by an impaired health status which could eventually lead to an increase in fatigue severity.” 
  3. “In addition, it can be noted that patients received a median of 9 CBT sessions in the original trial, which, although effective for fatigue severity, might be insufficient to address perceived functional impairment. As noted in the original trial, there was a trend towards a beneficial effect of CBT on functional impairment.”
  4. “An alternative explanation could be the recurrent negative media attention in the Netherlands for Q fever since the outbreak of 2007, frequently reminding patients of their complaints and possible unfavourable long-term outcome of QFS.”
  5. “Moreover, a large lawsuit, in which patients collectively sued the Dutch government for negligence during the Q fever outbreak, was still pending during the follow-up period. This encouraged patients to supply attorneys with proof of diagnosis and constantly reminded them of their often dire financial situation and perceived lack of social support.”
  6. “It is likely that not all relevant perpetuating factors for QFS have yet been identified and were therefore not addressed during CBT. It could be postulated that such inappropriately addressed perpetuating factors contributed to the relapse in fatigue severity at 1-year follow-up. The fact that there was a positive effect of CBT directly following CBT however speaks against this hypothesis. One could however also assume that the unknown and unaddressed perpetuatingfactors in QFS are responsible for the relapse.”

Notes to these statements:

point 1. There is no positive effect of CBT at all. The authors chose their words considerately. Although there seems to be some improvement on the CIS F score immediately after completion of the treatment, the effect disappears quickly and the CBT group scores during follow-up are even worse than the control group (although not statistically significant).

point 2. The reasoning that the patients experienced a limitation in physical functioning during the entire study shows that the treatment does not work.

point 3. The results with regard to physical functioning showed a tendency towards improvement. This too is cleverly described. One forgets to report that the outcomes were very poor. The authors acknowledge this in statement 2, because that is precisely what is being presented here as an explanation for the measured fatigue at follow-up. Moreover, there was a so-called “tendency towards improvement” in all groups and not just in the CBT group.

point 4. Returning negative news coverage in the media would cause the long-term effects to be worse. Well, and that was not the case during the endpoint assessment? Could it be that the known bias effects [v] that the patient has learned what are desirable answers and want to please his therapist, has faded out over time?

point 5. An ongoing lawsuit affects the scores. With this argument, the authors show that such subjective measures as the CIS F are very sensitive to the circumstances and that they therefore do not give a reliable picture of reality (see also my remark above at point v4). An objective outcome measure does not show this shortcoming or much less. This is an additional argument for the possibility that the “favorable score” on the CIS F is based on bias upon completion of the treatment. This is reinforced by the fact that the control group showed an improvement at follow-up. The logic in the authors’ arguments is therefore lacking; they contradict themselves.

point 6. The authors suggest that there may be a still unknown perpetuating factor. This is highly speculative, especially since they have dismissed another possible factor in the same statement. Moreover, in Keijmel’s research they have found that the underlying model of perpetuating factors is not applicable. The possible explanations why the treatment does not work are colorful and speculative. The consideration that cognitive behavioral therapy is neither an appropriate treatment nor will be, has apparently no place in their repertoire.

It is also questionable whether the patients were rightly diagnosed a QFS patient. In the Qure study, it is firmly stated that the QFS patients the active form of Coxiella burnetii cannot be found and therefore they do not suffer with chronic Q fever. There are three reasons though, why they cannot proclaim that with certainty:

  1. the research methods used are characterized by an above-average risk factor that the finding is not correct. See the article by Wielders et. al. [vi].
  2. The values found are only on border statistically significant in a small study population.
  3. Also the fact that the 1-year follow-up study shows that the CGT has deteriorated (as well as the fact that in the Qure study the SF36PF in the CGT group also still indicates underperforming physical functioning), is an indication that there is an incorrect diagnosis.



[i] Keijmel SP, Delsing CE, Bleijenber G et al (2017) Effectiveness of long-term doxycycline treatment and cognitive behavioral therapy on fatigue severity in patients with Q-fever fatigue syndrome Clinical Infectious Diseases, 64(8):998-1005


[iii] Knoop H, Bleijenberg G, Gielissen MF (2007) Is a full recovery possible after cognitive behavioural therapy for chronic fatigue syndrome? Psychother Psychosom 76: 171-176.

[iv] Raijmakers, PH, Keijmel, SP, Breukers, MC, Bleijenberg, G et al (2018) Long-term effect of cognitive behavioural therapy and doxycycline treatment for patients with Q fever fatigue syndrome: One-year follow-up of the Qure study Journal of Psychosomatic Research 116:62-67


[vi] Wielders CCH, Wijnbergen PCA, Renders NHM (2013) High Coxiella burnetii DNA Load in Serum during Acute Q Fever Is Associated with Progression to a Serologic Profile Indicative of Chronic Q Fever. Journal of Clinical Microbiology 51/10: p. 3192–3198


Research 2. CGT group program for CVS

In the study by Bazelmans et al., 2005 [i] on the effectiveness of group CBT with a graded activity program, CBT and no intervention had a very modest effect on CIS F scores after 6 months. The CIS F score in the CBT group changed from 51.0 at baseline,  to 45.6 at endpoint and that was a significant difference compared to the control group that received no treatment. The CIS F in the control group was at the start: 50.8, and after 6 months: 48.4. These outcomes for both groups  are absolutely inadequate and do not reach normal values ​​by far (<27).

Strangely enough, there was a positive effect on physical functioning in the control group, where no treatment was given. In the control group, the SIP 8 score at the start was 1.710, and after 6 months: 1.417. The CBT group showed a slight deterioration on the SIP 8 score. At the start it was 1.707, and after 6 months: 1.736.

The authors produced the following statements and recommendations (quote):

  1. “ In group therapy CFS patients may reinforce dysfunctional behaviour and resistance against psychological treatment. Furthermore, in Group therapy it is much harder to individualise CBT treatment to individual needs.”
  2. “In trials on individual CBT for CFS it was found that engagement in a claim for a disability related benefit (…) cognitive behaviour group therapy for chronic fatigue syndrome during CBT predicted less improvement after individual CBT for CFS. Our CBGT started beforecompletion of these studies.”
  3. “In the present study it was found that CFS patients with less severe complaints did profit most of CBGT. In future research on the effectiveness of CBGT for CFS these findings will have to be taken into account.”
  4. “Another explanation for the moderate effect of CBGT might be that the therapists had no prior experience with CB(G)T for CFS.(…) CFS patients are more difficult to treat than other patients with somatic complaints’. For our study this might count even more, since the therapists were inexperienced both in group therapy and in CBT for CFS.”  
  5. “Finally, based on this and former studies the treatment protocol CB(G)T for CFS has been improved. Rest and relaxation are less emphasised, and for passive CFS patients the treatment protocol has been adapted.”  
  6. “Recently, lack of social support has been identified as an important determinant of CFS and a new perpetuating factor. Dealing with a lack of social support may also have to become a more prominent aspect of CBT for CFS.”

point 1. In the text preceding this comment, the authors mention the benefits of group treatment, namely that they can be reinforced by the group effect to reach better results. So there is both an explanation in favor and against the group effects. The question remains why this is presented as an explanation that the therapy did not work.

point 2. The statement that the treatment is less effective in patients who are involved in a legal claim has been used more than once. The data do not show for how many patients in this study such a situation was at issue and whether it led to a deviation in the scores.

point 3. Patients with less severe complaints benefit more from the treatment. Do we have a heterogenious group of patients then? The recommendation to involve only less  what is at the bottom of the differences between the two groups.

point 4. The fact that the therapists had no experience is presented as an explanation. However, it concerns therapists who were specially selected for this research.

point 5. One refers to other research showing that the lack of social support is an important determinant for CFS and a new sustaining factor. Nothing shows that this has been of any influence in this research. There is no reporting about it. The comment is, again, speculative.

Despite all the negative findings, the authors conclude suggesting this research indicates a favorable effect of cognitive behavioral therapy, while there is no scientific underpinning for any such effect.

“In the current study we found a moderate effect of CBGT on fatigue, in an unselected group or CFS patients, and with newly-trained therapists.”

The fact that the CIS F outcomes point to severely affected patients is completely avoided in this text. The possibility that CBT could be an inadequate treatment is not considered.

[i] Bazelmans E, Prins JB, Lulofs R et al. (2005) Cognitive behaviour group therapy for chronic fatigue syndrome: a non-randomised waiting list controlled study. Psychother Psychosom 74: 218-224.

Auteur: Lou Corsius

MSc Health Sciences

Geef een reactie

Vul je gegevens in of klik op een icoon om in te loggen. logo

Je reageert onder je account. Log uit /  Bijwerken )

Google photo

Je reageert onder je Google account. Log uit /  Bijwerken )


Je reageert onder je Twitter account. Log uit /  Bijwerken )

Facebook foto

Je reageert onder je Facebook account. Log uit /  Bijwerken )

Verbinden met %s