Self-reporting compared to motion monitor in the measurement of physical activity during pregnancy

Most pregnancy-studies have relied on retrospective, cross-sectional surveys to measure physical activity level. Questionnaires are cost-effective, but validity of the data may be questionable. Objective. The aim of the present study was to validate a physical activity and pregnancy questionnaire (PAPQ) with a portable activity monitor (ActiReg(R)). Design. Prospective comparison study. Setting. Healthy pregnant women recruited in a capital area. Population. Seventy-seven pregnant women wore the ActiReg(R) sensors during waking hours for seven consecutive days and answered the PAPQ. Main outcome measures. Agreement between the two methods was analyzed by Bland-Altman plots and Spearman correlation coefficients. Results. The results indicated only small differences between the PAPQ and the ActiReg(R) in cross-tabulation of total physical activity level and proportion of participants meeting the current exercise guidelines. The correlation between the methods was good (r = 0.59) for time spent in activities with high intensity (METS > 6), moderate for time spent standing/moving (r = 0.36) and fair for sitting/lying (r = 0.29). The Bland-Altman plot of the activity patterns, showed a mean difference near zero with no apparent trends and wide scatter of individual observations. Conclusions. The PAPQ may be considered an acceptable method for assessing habitual physical activity and exercise among pregnant women at group level. However, as questionnaires and portable activity monitors have their strengths in measuring different aspects of physical activity, there may be advantages in combining these two types of instruments for registrations of physical activity level during pregnancy.