Abstract Background and aims The International Physical Activity Questionnaire - Short Form (IPAQ-SF) and ActiGraph accelerometers are widely used and internationally recommended for assessing physical activity (PA) and sedentary time (ST) after stroke. However, their agreement on chronic stroke survivors remains unclear, which is crucial to avoid misinforming clinical and research monitoring. This study quantifies the agreement between IPAQ-SF and ActiGraph measurements of PA and ST and compares agreement across different epoch lengths. Methods Fifty community-dwelling chronic stroke survivors with independent walking wore an ActiGraph GT3X+ accelerometer for seven consecutive days and completed the IPAQ-SF on the following day. Agreement between IPAQ-SF and accelerometer-derived weekly moderate-to-vigorous PA (MVPA), moderate PA (MPA), and ST was assessed using Bland-Altman analysis. Agreement was compared across four epoch lengths (1-, 10-, 15-, 60-second). Results For MVPA, the geometric mean ratio was 1.02 (95%CI: 0.49 to 2.1; limits of agreement (LoA): 0.007 to 154.6), with a proportional bias (β=0.832, p=0.001). For MPA, the geometric mean ratio was 0.91 (95%CI: 0.45 to 1.87; LoA: 0.0066 to 127.6), with a proportional bias (β=0.793, p=0.003). For ST, the geometric mean ratio was 0.85 (95%CI: 0.78 to 0.92; LoA: 0.47 to 1.53), with a proportional bias (β=0.773, p0.001). Agreement was higher with 60-second epochs. Conclusions IPAQ-SF overestimated MVPA and MPA, and underestimated ST compared to the accelerometer, with wide limits of agreement across outcomes. Proportional bias suggests that discrepancies increase at higher magnitudes. IPAQ-SF and accelerometer data are not interchangeable for assessment of PA and ST in chronic stroke survivors. Conflict of interest Teresa Dias: nothing to disclose; Carla Pareira: nothing to disclose; Fernando Ribeiro: nothing to disclose.
Dias et al. (2026) studied this question.