Commentary: on the importance of using prospective diary data in the assessment of recurrent headaches, stressors, and health behaviors in children and adolescents.

It has long been known among clinicians that certain foods (e.g., chocolate, cheese, and nuts), stroboscopic and flickering lights (TV and cinema), physical exertion and mental stress may elicit migraine attacks in children and adolescents (Barlow, 1984). In the first study of potential triggers, Vahlquist (1955) reported that most school children (87%) aged 10–12 years with migraine experienced mental stress as the prime elicitor. In a subsequent and pioneering survey of about 9,000 school children aged 7–15 years in Uppsala (99.3% of the parents responded), Bille (1962) found school work, especially related to examinations, to be the most common trigger of pronounced migraine. Since these two early studies, relationships between various types of stressors (daily hassles and negative life events) and recurrent headaches of migraine and unspecified types among children and adolescents in clinic and community samples have been examined in numerous studies. However, less is known about triggers in individuals with tension-type headaches, although these are the most common headaches in children and adolescents. In spite of the existing and rather extensive information from previous research, our understanding of triggers in children with headaches is hampered by the use of certain study designs (i.e., cross-sectional case-control designs of headache sufferers vs. headache-free controls with between-group comparisons or the use of within-group correlations in single groups) and retrospective reports using questionnaires and interviews with children/parents and adolescents (Alfvén, Östberg, & Hjern, 2008; Karwautz et al., 1999; Kröner-Herwig, Morris, & Heinrich, 2008; Passchier & Orlebeke, 1985). Focusing specifically on potential school stressors, Passchier and Orlebeke (1985) found that about 40–50% of the 10to 17-year olds reported that they knew what caused their headaches. The most common self-reported triggers consisted of various types of mental and school-related stressors being related to headache frequency and intensity. Similarly, in a recent Swedish study in which interviews were used to elicit information (Alfvén, Östberg, & Hjern, 2008), the authors reported school stressors such as harassment by peers, schoolwork pressure (too many tests), and disturbance in the classroom to be associated with both frequent headaches and abdominal pain. However, in a large-scale questionnaire-based survey of school children aged 9–14 years, Kröner-Herwig and her colleagues (2008) found nonsignificant relationships between school stressors and frequency of headaches. Instead, parental headaches, family stressors and having fewer friends were the strongest predictors. While Karwautz and collaborators (1999), in an extensive review of community and clinic studies, found few differences between migraine and tension-type headaches in children and adolescents in regard to psychosocial factors, parental divorce was more common and number of friends fewer among those with tension-type headaches as compared to migraine and headache-free controls. In most comparisons, the factors pertained to more longstanding risk factors such as relationships to parents, housing conditions, socio-economic status, and the presence of emotional–behavioral problems in the child. Sleep problems (Bruni et al., 1997; Kowal &