BACKGROUND: To increase the knowledge about physical activity (PA) patterns and correlates among children under the age of 4, there is a need for study's using objective measurements. The aim of this study was therefore to investigate if objectively measured PA among 3-year-old children differed between day of week and time of day and whether it correlated to child weight status and sex as well as parental weight status and education.

METHODS: Totally 61 children (51% girls) aged 3, participating in Early Stockholm Obesity Prevention Project were included. PA was measured with a tri-axial accelerometer (ActiGraph GT3X+) worn on the non-dominant wrist for one week. The main outcome was average PA expressed as counts per minute from the vector magnitude. PA and demographics/family-related factors were collected at baseline and at age 3. To analyze the results simple linear regression, ANOVA and paired t-tests were performed.

RESULTS: The mean number of valid days was 6.7 per child. The children were more active on weekdays than weekends (p < 0.01) and the hourly pattern differed over the day with children being most active midmorning and midafternoon (p = 0.0001). Children to parents with low education were more active (p = 0.01) than those with highly educated parents. No differences in PA by child weight status, sex nor parental weight status were found.

CONCLUSIONS: PA in 3-year-old children was lower during weekends than weekdays and varied over the day. Boys and girls had similar PA patterns, these patterns were independent of child or parental weight status. Children to parents with low education were more active than their counterparts. The fact that PA differed between weekdays and weekends indicates that PA might be affectable in 3-year-old children.