Immediate and longer-term effects of a cooking intervention on adolescents' mental well-being, fruit and vegetable intake, and cooking involvement, self-efficacy and attitudes

Abstract

Background: Practical cooking interventions give adolescents the ability to prepare healthy food. However, few adolescent interventions have had control groups or follow-ups beyond immediately post-intervention. Although many interventions have measured behavioural determinants such as cooking self-efficacy, cooking behaviour has seldom been assessed. Adolescent cooking is also positively associated with mental well-being in observational studies, but this has not been investigated experimentally. Aim: To investigate the immediate and longer-term effects of a cooking intervention on adolescents’ mental well-being, fruit and vegetable consumption, and cooking behaviours, self-efficacy and attitudes. Secondary aims were to explore whether the intervention’s effectiveness differed by participant characteristics, and whether cooking behaviour influenced other outcomes, as well as to contextually describe adolescents’ home cooking involvement. Design: The COOK Study was a randomised controlled trial with a 12-month follow-up with adolescents in Dunedin, New Zealand. The intervention was a five-day intensive cooking program followed by a support phase, where participants received six weekly meal kits to prepare for their families. Questionnaires were administered at baseline, after the support phase (post-intervention) and at the 12-month follow-up. Group interviews were also held after post-intervention and follow-up measures were completed. For continuous variables, mixed regression models analysed between-group differences in changes from baseline to post-intervention and follow-up. Logistic regression compared odds ratios for meeting fruit and vegetable intake guidelines. For exploratory analyses, the sample was divided into two strata by gender, deprivation index, weight status, and cooking involvement at 12 months, and effect sizes were compared. Group interview data were summarised for each discussion topic. Results: Of the 118 participants who completed baseline measures (intervention n=91, control n=27), 96% were retained at follow-up. Participants’ mean age at baseline was 13.6 ± 0.8 years and 64% were female. Mental well-being increased more for the intervention compared to the control group at post-intervention only. The odds of meeting fruit and vegetable intake guidelines did not differ between groups, however produce variety scores increased more for the intervention group at post-intervention. Overall cooking frequency increased by 0.4 evenings/week more for the intervention group at post-intervention, although changes in cooking main meals from basic ingredients were not significantly different between groups. Cooking self-efficacy and positive cooking attitude increased more for intervention participants at post-intervention, remaining significant at 12 months. Exploratory analyses suggested that the intervention was more effective for participants who were overweight or had deprivation indices below the median value, and that cooking involvement may moderate how well other outcomes were maintained. Feedback from group interviews showed that adolescent cooking behaviours were strongly influenced by parent and family factors, which could promote or inhibit continued cooking after the intervention. Conclusion: Improvements from cooking interventions in outcomes adolescents have less control over (dietary, cooking frequency and mental well-being outcomes) may be less sustainable than non-behavioural outcomes (self-efficacy and attitude). Home environments and practices should be considered when trying to increase adolescent cooking involvement, especially as cooking behaviours may influence other outcomes. Adolescents who are overweight or less socioeconomically advantaged may particularly benefit from cooking interventions.