Jørn Hetland

Studies on adolescent smoking indicate that the smoking behaviours of their parents, siblings and friends are significant micro-level predictors. Parents' socioeconomic status (SES) is an important macro-level predictor. We examined the longitudinal relationships between these predictors and the initiation and development of adolescents' smoking behaviour in Norway.

The aim of this study is to investigate dimensional structures in subjective health complaints in adolescents and to examine the variation in levels and dimensionality across gender and age groups. Data from two studies were used: (1) a study based on a convenience sample, consisting of 1427 Norwegian students (11- to 15-year-olds) from schools participating in the European Network of Health Promoting Schools; (2) a nationwide survey amongst 7,059 Norwegian students (aged 11, 13, 15, and 16) from the Health Behaviour in School-aged Children (HBSC) study. Subjective health complaints were measured by revised versions of the HBSC symptoms checklist (HBSC-SCL). In study 1 confirmatory factor analysis revealed that a model of two highly correlated factors, which can be labelled somatic and psychological, fitted the data reasonably well (CFI = 0.91). This two-factor model was applied in study 2 and latent means were tested across sub-samples defined by gender and age. The results indicate that girls show higher mean levels compared with boys on both factors. There is also an indication of an increase in these factors with age amongst girls, while amongst boys there is less, if any, difference across age groups. The correlation between the somatic and psychological factors was virtually constant across age groups and gender. The identification of a somatic and a psychological factor indicates the existence of two different dimensions that may have different aetiologies. The gender difference in latent means across age groups may suggest a different developmental pattern amongst girls and boys.

Aims: The aim of this study is to investigate dimensional structures in subjective health complaints in adolescents and to examine the variation in levels and dimensionality across gender and age groups. Methods: Data from two studies were used: (1) a study based on a convenience sample, consisting of 1427 Norwegian students (11- to 15-year-olds) from schools participating in the European Network of Health Promoting Schools; (2) a nationwide survey amongst 7,059 Norwegian students (aged 11, 13, 15, and 16) from the Health Behaviour in School-aged Children (HBSC) study. Subjective health complaints were measured by revised versions of the HBSC symptoms checklist (HBSC-SCL). Results: In study 1 confirmatory factor analysis revealed that a model of two highly correlated factors, which can be labelled somatic and psychological, fitted the data reasonably well (CFI = 0.91). This two-factor model was applied in study 2 and latent means were tested across sub-samples defined by gender and age. The results indicate that girls show higher mean levels compared with boys on both factors. There is also an indication of an increase in these factors with age amongst girls, while amongst boys there is less, if any, difference across age groups. The correlation between the somatic and psychological factors was virtually constant across age groups and gender. Conclusions: The identification of a somatic and a psychological factor indicates the existence of two different dimensions that may have different aetiologies. The gender difference in latent means across age groups may suggest a different developmental pattern amongst girls and boys.