Abstract
This umbrella review analysed the relationships between family variables and child/adolescent body weight, diet and physical activity. In line with theories of health behaviour change, it was assumed that behaviour-specific family variables (i.e. beliefs, perceptions and practices referring to food intake or physical activity) would have stronger support than more general family variables (i.e. socio-economic status or general parental practices). Data obtained from 18 systematic reviews (examining 375 quantitative studies) were analysed. Reviews of experimental trials generally supported the effectiveness of reward/positive reinforcement parental strategies, parental involvement in treatment or prevention programmes, and cognitive–behavioural treatment in reducing child/adolescent body mass and/or obesity. Results across reviews of correlational studies indicated that healthy nutrition of children/adolescents was related to only one parental practice (parental monitoring), but was associated with several behaviour-specific family variables (e.g. a lack of restrictive control over food choices, high intake of healthy foods and low intake of unhealthy foods by parents and siblings, low pressure to consume foods). With regard to adolescent physical activity, stronger support was also found for behaviour-specific variables (e.g. physical activity of siblings), and for certain socio-economic variables (e.g. parental education). Child and adolescent obesity prevention programmes should account for behaviour-specific family variables.