Birgit Kröner-Herwig1, Lisette Morris, Marion Heinrich. 1. Georg-Elias-Müller-Institute of Psychology, Department of Clinical Psychology and Psychotherapy, University of Göttingen, Göttingen, Germany.
Abstract
OBJECTIVE: The study aims at identifying biopsychosocial risk factors for headache in children and adolescents aged 9 to 14. METHODS: An epidemiological survey was conducted in a randomly drawn population sample of families with children in the above age group. Questionnaires were mailed to parents and children (n=6400), on whose data this report is based. The objective of the study was to establish a profile of risk factors regarding the occurrence of headache. Headache, as the criterion variable, was ranked according to its frequency in the last 6 months (no, rare, monthly, weekly). Independent variables came from 6 domains: health, socioeconomic, family, school, leisure/peers, and psychological factors. Data analysis was conducted via multinomial regression analyses in a 4-step strategy: (1) analysis of age and sex as control variables; (2) analysis of single variables from each of the 6 domains (controlled for age and sex); (3) domain analyses; and (4) comprehensive analysis including all significant variables from the domain analyses. RESULTS: Age and sex explained a small but significant proportion of the variance in headache frequency (3.5%). All health variables, several socioeconomic, and most family- and school-related as well as the psychological variables demonstrated a significant association with the criterion in the single variable models. However, only a few of the variables related to leisure/peer activities reached significance. The domain model comprising health variables explained 27% of the variance, achieving the best model fit, followed by the psychological model with 13%. CONCLUSIONS: The comprehensive model was able to explain one third of the total variance in headache occurrence. Contrary to our hypothesis, the addition of psychosocial variables to health-related predictors did not markedly improve model fit.
OBJECTIVE: The study aims at identifying biopsychosocial risk factors for headache in children and adolescents aged 9 to 14. METHODS: An epidemiological survey was conducted in a randomly drawn population sample of families with children in the above age group. Questionnaires were mailed to parents and children (n=6400), on whose data this report is based. The objective of the study was to establish a profile of risk factors regarding the occurrence of headache. Headache, as the criterion variable, was ranked according to its frequency in the last 6 months (no, rare, monthly, weekly). Independent variables came from 6 domains: health, socioeconomic, family, school, leisure/peers, and psychological factors. Data analysis was conducted via multinomial regression analyses in a 4-step strategy: (1) analysis of age and sex as control variables; (2) analysis of single variables from each of the 6 domains (controlled for age and sex); (3) domain analyses; and (4) comprehensive analysis including all significant variables from the domain analyses. RESULTS: Age and sex explained a small but significant proportion of the variance in headache frequency (3.5%). All health variables, several socioeconomic, and most family- and school-related as well as the psychological variables demonstrated a significant association with the criterion in the single variable models. However, only a few of the variables related to leisure/peer activities reached significance. The domain model comprising health variables explained 27% of the variance, achieving the best model fit, followed by the psychological model with 13%. CONCLUSIONS: The comprehensive model was able to explain one third of the total variance in headache occurrence. Contrary to our hypothesis, the addition of psychosocial variables to health-related predictors did not markedly improve model fit.