The school health promotion study is a classroom survey focusing on adolescent health, health behaviour, and behaviour in school; it has been carried out annually in Finland every April since 1995. The health promotion studyhasbeen approved by the ethical committee of Tampere University Hospital.
In 1997, students in the 8th and 9th grades of secondary school (ages 14-16 years) in two regions in Finland (Vaasa and Pirkanmaa) participated in the study. Out of a total of 20
213 pupils in these schools, 2570 (13%) were absent on the day of the survey. Altogether, 17
643 pupils (87%) returned the questionnaire (8695 girls, 8948 boys). A total of 1179 students (6.7%) gave incomplete responses on the Beck inventory and their questionnaires were excluded from the analysis. An additional 54 respondents did not answer the questions about bullying and were also excluded from the analysis. We were thus able to analyse the responses of 16
(81% of the target population, 93% of those present at school).
Involvement in bullying either as a bully or as the person being bullied was evaluated using two questions derived from a World Health Organisation study on youth health.12
The subject was introduced as follows:
The next questions are about bullying. We say a pupil is being bullied when another pupil, or a group of pupils, says or does nasty and unpleasant things to him or her. It is also bullying when a pupil is teased repeatedly in a way he or she doesn’t like. But it is not bullying when two pupils of about the same strength quarrel or fight.
The students were asked how frequently they had been bullied during the current school term (from the beginning of January until the end of May) and how frequently they had bullied others. They were asked to indicate whether these actions had occurred many times a week, about once a week, less frequently, or not at all. Those who bullied others at least once a week were classed as frequently being bullies; those who were bullied at least once a week were classed as frequently being bullied. Respondents were classed as not participating in bullying or being bullied, as being bullied or being a bully less than weekly, as frequently acting as a bully and not being bullied, as frequently being bullied and not a bully, or as frequently being both bullied and a bully.
Depression was measured using a modified, 13 item version of the Beck depression inventory13,14
which had been validated in Finnish.15
The Beck inventory has been shown to be a valid measure for detecting depression among adolescents.16
It has good psychometric properties in this population.17
Students who scored from 0 to 7 were classed as having no depression or mild depression, and those who had scores of 8 to 39 were classed as having moderate to severe depression.13
One of the items on the inventory asks about thoughts of self harm. We classed the students as having severe suicidal ideation if they chose either “I have definite plans about committing suicide” or “I would kill myself if I had the chance.”
The sociodemographic variables evaluated were age, sex, years since moving to current area, educational level attained by parents,whether one or both parents had been unemployed duringthe past 12 months, and family structure (whether the adolescent was living with both parents, with one parent, with a step-parent, or apart from the parents). School performance was measured using the child’s grade point average; perceived lack of social support from parents, friends, and teachers were used as independent variables. Grade point average and perceived social support have been shown to be determinants of depression in the sample studied (unpublished data). Age and grade point average were analysed as continuous variables, others were dichotomised.
Incomplete responses to the Beck inventory occurred more often among boys (7.8% (698/8948) v 5.5% (481/8695), P<0.0001). The item most frequently unanswered was the item about suicidal ideation which was omitted by 4.9% (438/8948) of boys and 3.7% (322/8695) of girls.
Associations between bullying and depression (dichotomous variables) and severe suicidal ideation were tested using χ2. Multivariate associations were tested with logistic regression using depression and severe suicidal ideation as dependent variables. The models were fitted using a stepwise forward method starting from an empty model. When using depression as a dependent variable, sex, age, and involvement in bullying (being bullied or being a bully) were tested first. Other variables were added later. When severe suicidal ideation was used as a dependent variable, logistic regression was carried out in three steps. Firstly, sex, age, and involvement in bullying were analysed. Secondly, depressive symptoms (defined as the total score of items on the Beck inventory excluding those on suicidal ideation) were added to the analysis. Finally, the other independent variables were included.