Comparing physical activity and sedentary time in children and adolescents with type 1 diabetes using different accelerometer data processing protocols.

Background

Regular physical activity (PA) is crucial for managing type 1 diabetes (T1D), yet many children fail to achieve the recommended 60 minutes of daily moderate-to-vigorous physical activity (MVPA). Device-based (accelerometry) measurement of PA and sedentary time (SED) may be influenced by analytical decisions (e.g., cut-points), but less is known about these methodological effects in children with T1D.

Objectives

To (1) compare the effects of different accelerometer data processing protocols on PA and SED estimates and (2) examine how these variations influence the proportion of children with T1D meeting PA guidelines.

Methods

Participants aged 7-17 years with T1D (n=43; 49% female; 13.7±2.3 years) wore an ActiGraph accelerometer on their right hip for 7 days. ActiLife software processed data according to eight published protocols to classify SED, light PA (LPA), and MVPA. Linear mixed-effects models compared activity time estimates across protocols, and Cochran's Q test assessed differences in the proportion meeting PA guidelines.

Results

SED, LPA, and MVPA estimates differed significantly across protocols (p < 0.001), with mean daily SED ranging from 438-700 minutes, LPA from 82-329 minutes, and MVPA from 16-193 minutes. Protocol choice also influenced the proportion of participants meeting the MVPA guidelines (Q(7) =172.0, n=31, p<0.001), with some methods consistently classifying more participants as meeting recommendations than others.

Conclusion

This study addresses gaps in accelerometer-based research by identifying protocol-specific variability in activity measurement in children with T1D. Without criterion reference measures, it remains unclear which protocols provide the most accurate estimates. Future research should incorporate validation methods to strengthen methodological decision-making.