• The Millennium Cohort study contains 5 year follow-up of 17,561 singleton children recruited between 2000-2002 in the UK. All participants were interviewed in their own homes at 9 months, 3 years and 5 years follow-up and all measures were self reports. Logistic regression and likelihood ratio tests were used.
• Children are less likely to walk to school as income and parental education increase [Adjusted odds: 0.7 (95%CI: 0.6-0.8) for higher income/education compared to low income/no qualifications]. However, if the parent plays with the child in high income families the child is more likely to walk to school [Adjusted odds: 1.67 (95%CI: 1.3-2.1)]. Children taking part in organized activities are from higher income, higher education families, with a car, in a "good" area with non-working mothers. However, in low socio-economic families where the parent plays with the child the child is more likely to take part in organized activities [Adjusted odds: 2.0 (95% CI: 1.5-2.7)].
• Income is an important determinant of the type of activity available to children. Families that report good health behaviors (non-smoking, low TV viewing) and play with their children show higher levels of physical activity. Thus, parenting practice appears to have a strong impact on their child’s physical activity.

Brophy, S., R. Cooksey, et al. (2011). "Parental factors associated with walking to school and participation in organised activities at age 5: Analysis of the Millennium Cohort Study." BMC Public Health 11(1): 14.