Validity and repeatability of a short pregnancy leisure time physical activity questionnaire

Abstract

Purpose: To examine the validity and repeatability of a short leisure time physical activity (LTPA) questionnaire during pregnancy.

Methods: Seventy-nine women with uncomplicated pregnancies and weeks' gestation </= 33 participated. After the first questionnaire (LTPAQ1) they entered 7-day LTPA and pedometer counts in a logbook and completed the second LTPA questionnaire (LTPAQ2). Validity was assessed with Spearman's rank correlation coefficients by comparing LTPAQ2 with pedometer counts and logbook. For describing repeatability, change in the mean, geometric mean ratio, typical error, coefficient of variation (CV, %) and Bland-Altman plots were used.

Results: Forty-five (57%) and 47 (59%) women were available for pedometer and logbook comparisons and 49 (62%) for repeatability assessment. LTPAQ2 showed no correlation with pedometer but moderate correlation with the logbook for the frequency of moderate to vigorous-intensity LTPA (r(s) = 0.68, 95% CI 0.48 to 0.81). In repeatability, the typical error for frequency estimates varied from 1.2 to 3.7 sessions and CV for duration from 119 to 369%. The corresponding values for systematic error were -1.0 to 0.3 sessions and 4 to 36%. The 95% limits of agreement for single variables were large.

Conclusions: The questionnaire was valid for assessing moderate to vigorous-intensity LTPA but its individual repeatability proved weak.