Accelerometer data reduction rules might affect sitting time. However, the impact of different data reduction decisions on sitting time remains uncertain. This study evaluated the effect of different non-wear time rules on sitting time and changes in sitting time from age 9 to 12 years.

Longitudinal ActiGraph accelerometry data were collected when participants were 9.3 (±0.4) and 12.5 (±0.3) years. Sitting time was defined using an accelerometer cutpoint of 25counts/15s. Non-wear time was defined using manual data reduction (the reference method) as well as rules using 10min, 20min and 60min consecutive zeros. Differences between rules were analyzed using repeated measures Anova with Bonferoni post-hoc analyses.

At age 9, sitting time between the 10min and 60min rules ranged from 364min to 426min per day depending (P<0.05). At age 12 years, the difference between rules increased, with average sitting times ranging from 424min to 518min per day (P<0.05). Changes in daily sitting time over the three years ranged from 60min per day when using the 10min rule to 93min per day when using the 60min rule (P<0.05). When sitting time was adjusted for wear time, differences in daily sitting time between rules decreased from 62min to 27min (age 9), 95min to 32min (age 12) and 33min to 10min (for changes in sitting time between 9–12 years).

Using different data reduction rules results in significant differences in both sitting time and changes in sitting time in children and adolescents. Compared to the reference method, using the 20min zero-string rule, while correcting for wear time, provided the most accurate estimates of sitting time and changes in sitting time in 9 to 12 year olds.