Purpose: To validate the 7-d Physical Activity Recall (PAR) telephone interview version and its activity intensity categories.
Methods: Seventy-four adults (47 women, 27 men), ranging in age (18-67) and activity levels, were interviewed by phone and in-person using the same PAR protocol. Each participant wore a TriTrac-R3D accelerometer for 10 d. Validity was assessed by comparing the phone and in-person PAR interviews with the TriTrac-R3D data.
Results: Sixty-nine adults (44 women, 25 men) were used for all statistical analyses. Intraclass correlations between the two PAR interviews for total minutes per week of activity were r = 0.96, and r = 0.94 for moderate, r = 0.97 for hard, and r = 0.97 for very hard intensity activities. Pearson product moment correlations between the phone PAR and TriTrac-R3D for total minutes per week of physical activity were r = 0.43, and r = 0.31 for moderate, r = 0.39 for hard, and r = 0.78 for very hard intensity activities. Pearson correlations between the in-person PAR and TriTrac-R3D for total minutes per week of physical activity were r = 0.41, and r = 0.33 for moderate, r = 0.43 for hard, and r = 0.74 for very hard intensity activities. Participants overestimated the amount of physical activity in both interviews as compared with the TriTrac-R3D.
Conclusion: The phone and in-person versions of the PAR are equivalent measures for self-reported physical activity. Regardless of age, body mass index, or physical activity level both interview methods had similar estimates for total minutes per week of moderate, hard, and very hard activity. Correlations between each interview method and the TriTrac-R3D were lower for moderate and hard activities as compared with very hard activities.