Clinical Answers: Are nonpharmacological interventions for migraine effective in children and adolescents?


QUESTIONAre nonpharmacological interventions for migraine effective in children and adolescents?
ANSWERThere is evidence that psychological treatments are effective in reducing pain intensity in children and adolescents with headache, and that therapies such as relaxation and cognitive behavioural therapy (CBT) may have lasting effect for improving mood and reducing pain for chronic headache; however, it is not possible to distinguish effectiveness for migraine versus other types of chronic headache.
IMPLICATIONS FOR PRACTICEPsychological therapies (relaxation, biofeedback and CBT) are likely to reduce pain in children and adolescents with chronic headache, and warrant inclusion as part of management, with the understanding that there is not yet sufficient evidence to select responders amongst the different severities or types of headache (e.g. migraine).
  • Headache is one of the most common type of recurrent pain in children.

  • Young people report pain to be distressing, interfering and potentially debilitating, with effects of untreated pain possibly extending into adulthood.

  • Issue 1, 2013 of the Cochrane Database of Systematic Reviews was searched using the term 'migraine' restricted to the title, abstract or keywords.

  • Of the 33 potential reviews and 10 protocols, 1 review was included in this analysis.

  • The review includes a total of 37 studies and 1938 children for headache and other types of chronic pain; we report results for the headache group (20 studies and 820 children) here.

  • One relevant review on non-invasive physical treatments for chronic/recurrent headache was excluded from this analysis: it contained one trial with mixed (adult and paediatric) data, but did not report numerical data for the paediatric population.

  • Table 1 presents characteristics of the included review.

  • The majority of included studies that reported age ranges contained children aged 7 and older.

  • Table 2 presents a summary of the results for studies examining headache (including migraine).

  • Children who received psychological treatment for headache had significantly lower pain post-treatment [risk ratio (RR) 2.90; 95% confidence interval (CI): 2.25, 3.73; I2: 25%], and this was maintained at follow-up of 3 months or more (RR 3.34; 95% CI: 2.01, 5.53).

  • Table 3 presents a summary of studies with the headache group for each type of psychological intervention.

  • The majority of studies (90%) of CBT, biofeedback and combined (CBT and biofeedback) therapy reported effectiveness in reducing pain due to headache.

  • Across the 18 studies examining the outcome of pain, quality of the trials was rated as moderate because of poor descriptions of randomization, lack of reports on follow-up data or lack of clarity on whether studies used intention-to-treat analyses. Across the 6 studies examining the outcome of pain at follow-up, quality of trials was rated as low because of poor descriptions of randomization, small sample sizes or unexplained heterogeneity or inconsistency in results.

  • Diagnostic criteria for headache and migraine varied significantly across studies, with a paucity of trials reporting baseline or upper limits of headache severity among participants.

  • The majority of trials used headache diaries to determine outcome measurements (headache frequency, intensity, duration, etc.); however, the method of calculating each measure varied across trials (e.g. frequency recorded as number of headaches per day, headaches per hour or as presence of headache at meal times).

  • Three trials contained a mix of children with headache and nonheadache diagnoses (e.g. abdominal pain and musculoskeletal pain) where results should be interpreted with caution because of co-morbidity of symptoms in this group of patients.

  • References

Table Table 1.. Characteristics of included reviews
Review titleNumber of studiesPopulationInterventionComparisonOutcomes for which data were reported
AuthorsPooled sample (range)Definition of migraine   
Assessed as up-to-date     

Psychological therapies for the management of chronic and recurrent pain in children and adolescents

Eccleston C, Palermo TM, Williams ACDC, Lewandowski A, Morley S, Fisher E, Law E

November 2012

20 (headache studies with extracted data) 820 (26–75)

Children and adolescents (<18 years) reporting chronic (nonmalignant) pain, included a psychological intervention as active treatment.

This analysis only reported data from trials for children with headache (including migraine) and three trials with mixed (headache and nonheadache) conditions.

The review examined paediatric chronic or recurrent pain, and did not specify a definition for headache or migraine.

Psychological (behavioural/relaxation-based or cognitive behavioural)Active treatment, treatment as usual or waiting-list control

Pain immediately post-treatment;

Clinically significant change in pain;

Disability at follow-up; and

Mood at follow-up

Table Table 2.. Comparison of psychological therapy versus control: pain, disability and mood during post-treatment and follow-up for children with headache
OutcomeNumber of subjects (studies)Measure of effect (95% CI)I2 (%)
  • *

    Significantly favours psychological treatment (indicated in bold).

  • Small sample size provides insufficient evidence to draw conclusions.

  • CI, confidence interval; RR, risk ratio; SMD, standardized mean difference.

Pain (post-treatment)748 (18)RR: 2.90 [2.25, 3.73]*29
Pain (follow-up)196 (6)RR: 3.34 [2.01, 5.53] *0
Disability (post-treatment)108 (3)SMD: −0.30 [−0.85, 0.24]49
Disability (follow-up)24 (1)SMD: −0.45 [−1.27, 0.36]
Mood (post-treatment)204 (5)SMD: −0.16 [−0.45, 0.13]0
Mood (follow-up)59 (2)SMD: −0.60 [−1.13, −0.07]0
Table Table 3.. Treatment outcomes for studies with headache participants
InterventionStudies reporting effectiveness of treatment group (participants)*Studies reporting no difference between groups (participants)*
  • *

    Number of participants are reported from meta-analyses within the review, and may not reflect the entire study population.

  • Study contains a mix of headache and nonheadache participants.


Connelly 2006 (37)

Fichtel 2001 (36)

Hicks 2006 (37)

Kroener-Herwig 2002 (75)

Larsson 1987a (41)

Larsson 1987b (34)

Larsson 1990 (48)

Larsson 1996 (26)

McGrath 1992 (72)

Osterhaus 1997 (39)

Palermo 2009 (48)

Sartory 1998 (43)

Trautmann 2010 (55)

Wicksell 2009 (29)

Barry 1997 (29)

Griffiths 1996 (42)

Biofeedback + CBT

Bussone 1998 (35)

Scharff 2002 (36)


Labbé 1984 (28)

Labbé 1995 (30)