Abstract
BACKGROUND: The increasing adoption of accelerometers for the assessment of sedentary behaviour and physical activity among dialysis patients demands robust validation of these monitoring devices. This study aims to determine the comparability of wrist- versus waist-worn ActiGraph GT3X accelerometers, using placement-specific cut-points for peritoneal dialysis patients, to refine research and clinical practices. METHODS: This was a cross-sectional study. Thirty-one participants wore two ActiGraph GT3X accelerometers, positioned on the right waist and nondominant wrist, and monitored over a seven-day period in a naturalistic setting. Data were processed with ActiLife v6.13.3 and analysed using intraclass correlation coefficients (ICC), limits of agreement, and pairwise 90% equivalence test within a ± 10% threshold. RESULTS: The sedentary time measurements from both wrist- and waist-worn GT3X accelerometers were deemed equivalent, with high ICC values (0.98, 95% confidence intervals (CI) 0.97-0.99) and a ratio of 1.0 within the 90% CI of 0.9 to 1.0. Although agreement between accelerometers was good for classification of light-intensity activity (ICC = 0.76), the waist-worn device's estimates exceeded the equivalence criteria compared to the wrist-worn device (ratio 1.4; 90% CI 1.2-1.6). Conversely, the waist-worn device reported a significantly lower duration of moderate-to-vigorous physical activity (MVPA) than the wrist-worn device (Ln transformed ratio 0.3; 90% CI 0.1-0.4). CONCLUSIONS: The use of placement-specific cut-points did not ensure equivalence in physical activity parameter estimates between wrist- and waist-worn ActiGraph GT3X devices. The findings underscore the necessity for consistent accelerometer placement for reliable monitoring of physical activity in peritoneal dialysis patients. CLINICAL TRIAL NUMBER: Not applicable.