Intra-individual variation in children's physical activity patterns: Implications for measurement.

Area of Science:

• Pediatric Exercise Science
• Behavioral Epidemiology
• Biostatistics

Background:

• Children exhibit considerable day-to-day variability in physical activity (PA) and sedentary behaviors.
• This intra-individual variability poses challenges for precise measurement of PA and sedentary behavior variables.
• Understanding and mitigating random error is crucial for accurate assessment in pediatric populations.

Purpose of the Study:

• To explore strategies for reducing random error in PA and sedentary behavior assessment due to intra-individual variability.
• To demonstrate the impact of varying sampling durations on the precision of PA and sedentary behavior estimates in adolescents.
• To guide researchers in determining optimal sample sizes and durations for assessing children's PA patterns.

Main Methods:

• Utilized self-reported data from 298 adolescents aged 13-14 years.
• Compared PA and sedentary behavior estimates derived from 1 to 7 days of recall data.
• Calculated intra-individual coefficients of variation for key variables including PA level, moderate-to-vigorous physical activity (MVPA), screen time, and sleep.

Main Results:

• Large intra-individual coefficients of variation were observed: PA level (14.5%), MVPA (83.4%), screen time (60.8%), and sleep (12.2%).
• Measurement error decreased as the number of sampled days increased.
• The precision of estimates varied based on the specific PA or sedentary behavior variable being assessed.

Conclusions:

• While increased sampling days enhance precision, shorter sampling periods may be adequate for certain research questions.
• Researchers must conduct power analyses considering inter- and intra-individual variability to determine appropriate sampling durations.
• Optimizing sampling strategies is essential for reliable assessment of children's physical activity and sedentary behavior patterns.