Arch Dis Child 98:561-563 doi:10.1136/archdischild-2013-304307
  • Archimedes
  • Question 2

Is there effective behavioural treatment for children with chronic fatigue syndrome/myalgic encephalomyelitis?

  1. Esther Crawley2
  1. 1 Medical Student, School of Medicine, Swansea University, Swansea, UK
  2. 2 School of Social and Community Medicine, University of Bristol, Bristol, UK
  1. Correspondence to Simon N Smith, College of Medicine, Swansea University, Singleton Park, Swansea SA2 8PP, UK; 630443{at}
  • Received 22 April 2013
  • Accepted 25 April 2013


A 15-year-old girl comes to your outpatient clinic with a 6-month history of fatigue, headaches, nausea and muscle pain. Clinical examination and investigations are normal and you make a diagnosis of chronic fatigue syndrome/myalgic encephalomyelitis (CFS/ME). She is going to sit her General Certificate of Secondary Education (GCSE) examinations in 6 months, but is currently unable to attend school full time. She and her mother ask you which behavioural treatment (they don't want medication) is the most likely to get her back to school.

Structured clinical question

Do teenagers with CFS/ME (P) benefit from behavioural treatments (I) compared with usual care (C) in improving school attendance and fatigue (O).


A comprehensive list of synonyms was compiled for the subjects (adolescents) and the condition (CFS). These were then combined and applied to the following databases so that all possible combinations of synonyms in article titles would be returned: Cochrane Library, PubMed, Ovid, Medline and EMBASE. This review updates the previous systematic review1 used to inform NICE guidance. We reviewed randomised controlled trials (RCTs) comparing any behavioural intervention with normal care published since 2005. The results are discussed in the context of published case–control and cohort studies.

Abstracts were reviewed for 77 papers, and 20 papers were retrieved. Seven papers were excluded because they were case studies, two were cohort studies, four evaluated non-behavioural treatments and four did not use normal usual care as a control group leaving three RCTs.

Table 2



Although the National Institute for Health and Care Excellence (NICE) recommends cognitive behavioural therapy (CBT), graded exercise therapy (GET) and activity management, trial evidence only exists for CBT. Since the previous systematic review, there have been two further trials bringing the total to three (table 2). All three investigate different modalities of CBT (face-to-face, internet delivered and coupled with biofeedback) and a different number of sessions.2–4

In a study of 71 patients, Stulemijer et al 2 randomised 36 to their CBT arm (10 individual sessions of CBT over 5 months) and 35 to the waiting list for therapy. The CBT protocol varied depending on whether patients were considered to be ‘active’ or ‘passive’ in relation to their normal level of physical activity. Both protocols looked to establish a baseline level of activity before it was gradually increased and had focus on negative beliefs and behaviours. Participants in the CBT arm reported a decrease in the fatigue score of 14.5 (95% CI 7.4 to 21.6) and were attending an additional 18.2% (95% CI 0.8 to 35.5) of school classes compared with controls.

Nijhof et al 3 investigated an internet-delivered CBT compared with usual care. The internet programme consisted of 21 CBT modules. Participants could contact a therapist by email. Young people in the CBT arm attended 32.6% more school (95% CI 21.5 to 43.6) and had a reduction in fatigue scores with a mean difference of −18.3 (95% CI −22.9 to −13.7) compared with controls at 6 months. Despite removing the face-to-face element, this is a high resource intervention. The mean number of e-consults (emails) sent by patients was 66.6 (SD 16.3) and by parents was 22.8 (SD 10.3). The mean number of e-consults sent by the therapists was 28.7 (SD 10.3) per patient and 19.5 (SD 10.5) per parent.

The final RCT from Al-Haggar et al 4 (n=159) compared CBT plus biofeedback with usual care where children were offered symptomatic treatment only. Participants in the CBT arm received 40–60 sessions over 18 months. CBT treatment focused on changing habits, adjusting thinking patterns and gradually increasing activity. The biofeedback element included the use of biofeedback machines to allow participants identify circumstances that trigger symptoms (eg, heart rate) and develop relaxation and coping strategies. The trial reported that those in the intervention group were attending an extra 23.14 (95% CI 20.6 to 26.8) hours of school a month with fatigue scores lower by 12.23 (95% CI 7.4 to 14.8) compared with controls at the end of the trial. However, there are limitations. It was not clear if the biofeedback machine was used before, during or after CBT; 63 patients were lost to follow-up and we do not know how much the biofeedback contributed to effectiveness. Full details of the control group treatment were not included.

The RCTs described are consistent with two RCTs that did not have usual care as a control and two cohort studies. One RCT compared CBT with psychoeducation5 (n=63) and the other compared CBT with pacing6 (N=13). In both trials, school attendance and fatigue scores improved in the CBT arms. Both cohort studies7 ,8 described an improvement in fatigue and school attendance at 6 months; however, without a control arm, it is not clear whether this was due to the treatment given.

Graded exercise therapy

A Cochrane review in 20049 concluded that there was encouraging evidence that some adult patients with CFS may benefit from exercise therapy and found no evidence that exercise therapy may worsen outcomes on average. The NICE guidelines10 recommend that GET is offered to children with mild to moderate CFS. Further research11 in adult patients has established that GET can moderately improve outcomes when added to standard care.

There are currently no published RCTs investigating GET compared with usual care in adolescents.

One small pilot study12 compared GET with progressive resistance training and reported a modest reduction in fatigue severity in both arms (GET group −0.42 (SD ± 0.31), resistance training arm −0.12 (SD ± 0.60)). This study is limited by its sample size (n=22) and it was not powered sufficiently to demonstrate an effect.

Two case series13 ,14 suggest an improvement in fatigue after inpatient stays where patients were offered psychological support as part of a multidisciplinary package of care which included exercise. However, as there is no control group, it is not clear how these treatments compare with usual care, whether they are feasible on a large scale or how they compare with outpatient CBT.


In summary, the strongest evidence for treating children with CFS is for CBT which is effective in reducing fatigue and increasing school attendance. The available evidence for GET while showing some modest fatigue reduction comes from underpowered studies that are too small to draw conclusions from.

The varied approach of some of the CBT studies raises possibilities of using more cost-effective methods to deliver CBT. Research is also needed on the number of sessions required (studies varied between 10 and 60 sessions).

All interventions reviewed were provided by CFS specialists and therefore these results are not generalisable to non-specialists. NICE guidance in relation to adult CFS should be recognised as valid for paediatric cases in its recommendation that healthcare professionals should seek specialist expertise when planning and providing care for patients with severe CFS.

Clinical bottom lines

  • CBT is the best evidenced and most effective behavioural intervention in reducing fatigue and increasing school attendance in children.

  • GET has shown some fatigue reduction, but studies are small and underpowered.

  • CBT delivered over the internet has been shown to be more effective than usual care.


  • Contributors SS and EC had idea for the article. SS completed the literature search and wrote the first draft. SS and EC critically reviewed, redrafted the article and approved the final version. EC is guarantor.

  • Competing interests None.

  • Provenance and peer review Not commissioned; internally peer reviewed.