Yeah cause in his blog he refers to CBT being the therapy that caused these effects with no mention of GET, but that seems weird. In the study they talk about GET. Seems more likely that that caused the improvement indeed.
CBT for this group explicitly incoporated an activity programme:
CBT comprised 6 possible modules based on the known fatigue-perpetuating factors3 and previous research.8 These modules were directed at insufficient coping with the disease; dysfunctional cognitions regarding fatigue, activity, pain, or other symptoms; fatigue catastrophizing; dysregulation of sleep or activity; poor social support; and negative social interactions (appendix e-2). To account for interindividual differences in these factors, the intervention was adapted to the specific needs of each participant, including an individually tailored structured
activity program. Each session was 50 minutes in duration and was conducted at the nearest participating center by a cognitive-behavioral therapist. The total number of sessions for each participant was based on the number of modules to be addressed, which were identified by the therapist by performing an interview and specific tests. Acceptable compliance with the CBT program was defined as completion of a minimum of 3 sessions.
Also:
A limitation of this study is the potentially low generalizability as only 74 of the 377 invited patients took part.
I don't know why it says 74 patients, when 94 patients were assessed, and then 37 excluded, No severe fatigue (25), wheelchair bound (8), illiteracy (2), other reasons (2). This left 57, not 74 patients for randomisation.
Also, finally this was a crossover study and 20/24 (4 discontinued) patients were randomised for CBT/AET, so no long term followups comparisons to a control group are possible.
Also, on adherence by the participants:
Eleven AET participants (39%), six CBT participants (24%) did not achieve the level of acceptable adherence. (Acceptable adherence with the AET program was defined as completion of a minimum of 40 training sessions. Acceptable adherence with the CBT program was defined as completion of a minimum of 3 sessions.)
All of this limits the generalisability of the findings as I mentioned above.