Various physical activity monitors have demonstrated respectable step-count accuracy when worn by pregnant women during treadmill walking, particularly at faster speeds. However, it is unclear if this accuracy extends to over ground walking, a far more common free-living modality. PURPOSE: This study examined the step-count accuracy of five commonly-used physical activity monitors among pregnant women during over ground walking at self-paced speeds. METHODS: Twenty-nine pregnant women (19 second trimester, 10 third trimester) completed six over ground walking trials while wearing three consumer-grade activity monitors (FB, OM, NL) and two research-grade activity monitors (AG, SW). All walking trials consisted of 100 steps, as measured by a hand-tally counter which served as our criterion measure. Participants were instructed to walk the first three trials at a “normal pace” and the final three trials at a “brisk pace”. Steps recorded during each trial were recorded or downloaded for each device in order to calculate mean absolute percentage error (MAPE) ([(criterion measure – monitor measure) / criterion measure] x 100). One-way ANOVAs were performed to determine whether MAPE differed between monitors and also whether pregnancy trimester affected monitor accuracy. Pairwise comparisons with Bonferroni adjustments were performed to explore specific differences between monitors. RESULTS: Significant differences were found between monitor MAPE for normal-pace walking trials (F(4, 140)=8.66, pF(4, 140)=30.07, pThe research-grade SW monitor is highly accurate for over ground walking during pregnancy. The OM and NL, while not research-grade, are accurate tools for over ground walking and represent a cost-effective alternative for the consumer or for future pregnancy walking intervention trials.

Supported by WSU College of Education Faculty Funding Award Program

This document is currently not available here.