Does school environment affect 11-year-olds' fruit and vegetable intake in Denmark?

It is often found that adolescents eat too little fruit and vegetables. We examined the importance of school for 11-year-olds' daily intake measured by food frequency- and 24-h recall questionnaires in Danish data from the European 2003 Pro Children Survey. Multilevel logistic regression analyses included matched student-parent-school questionnaire data (N=1410) from a random sample of 59 schools and were conducted for fruit and vegetables separately: 1) without explanatory variables, to decompose the between-school and within-school variance; 2) with individual level covariates (socioeconomic position, parental intake, etc.) to examine if the between-school variance was attributable to different student compositions of schools; and 3) with individual- and school-level covariates (school availability of fruit/vegetables and unhealthy food) to examine the effect of context. Additional analyses stratified by gender and home availability of fruit/vegetables examined if school food availability influenced subgroups differently. Between-school variations were quantified by intra class correlations and median odds ratios. We found that 40% of the students ate > or = 200 g fruit/day and 25% ate > or = 130 g vegetables/day. Most of the total variance in students' intake occurred at the individual level (93-98%). There were larger between-school variations in vegetable intake than in fruit intake. Fruit and vegetable consumption clustered within schools to a larger degree for boys than girls. The between-school variance did not differ by home availability. Boys and students from high availability homes consumed more fruit and/or vegetables if enrolled in schools with access to fruit/vegetables and unhealthy food or contrarily with no food available versus schools with only fruit/vegetables available. The small school-level effects on 11-year-olds' fruit and vegetable intake imply that family level interventions may be more important and that the success of school interventions will rely on the degree of parental involvement.