Detecting bouts of physical activity in a field setting

The purposes of this study were to assess the TRITRAC and CSA for: (a) interaccelerometer agreement; (b) agreement in detecting patterns of moderate-intensity physical activity; and (c) agreement in detecting walking patterns recorded in a diary. Thirty-one women wore both the TRITRAC and CSA accelerometers for three consecutive days. Interaccelerometer agreement (measured with generalizability coefficients) ranged from .88 to .99. In total, 71.3% of the accelerometers' patterns agreed in length, with CSA patterns being on average significantly longer. Interaccelerometer agreement in detecting patterns of brisk walking, as recorded in a diary, was comparable (69.4%). Interaccelerometer discrepancies may be related in part to the threshold employed by each instrument for classifying moderate intensity patterns.