Social cognitions measured in 4 to 6 year olds are predictive of objectively measured physical activity.

Objective: This study tested whether social cognitions from the reasoned action approach (RAA) can be reliably measured in a sample of primary school children (aged 4-6 years) and whether these RAA measures are predictive of physical activity.

Design: Longitudinal observational study with repeated measures over six months.

Measures: RAA variables were measured using a novel choice paradigm between physically active and sedentary pastimes. Relative reinforcing value, covariates (sex, age, deprivation and BMI) and physical activity by accelerometry (primary outcome) were also measured.

Results: RAA cognitions could be measured with acceptable retest reliability and discriminant validity and social norms regarding parents and teachers (but not friends) were correlated with physical activity and change in physical activity when relative reinforcing value and other covariates were controlled for.

Conclusions: RAA cognitions can be reliably measured in primary school children aged 4-6 and RRA measures appear useful in understanding children's activity choices and may potentially inform future interventions.