Validity of the SenseWear® Armband to Predict Energy Expenditure in Pregnant Women

SMITH, K. M., L. M. LANNINGHAM-FOSTER, G. J. WELK, and C. G. CAMPBELL. Validity of the SenseWear (R) Armband to Predict Energy Expenditure in Pregnant Women. Med. Sci. Sports Exerc., Vol. 44, No. 10, pp. 2001-2008, 2012. Few valid, objective methods exist to quantify physical activity and predict energy expenditure (EE) during pregnancy. Purpose: The purpose of this study was to evaluate the validity of the SenseWear (R) Mini armband monitor (SWA) (BodyMedia, Pittsburgh, PA) to estimate EE in pregnant women. Methods: Thirty healthy pregnant women (22-24 wk of gestation) completed a series of activities of daily living (typing, laundry, sweeping, and treadmill walking: 2.0, 2.5, 3.0, and 3.0 mph, 3% incline) while EE was estimated by the SWA and measured by indirect calorimetry (IC). The SWA data were processed using both the v2.2 algorithm and the newer v5.2 algorithm. The estimated EE values were compared with the measured EE values using a three-way (method x algorithm x activity) mixed model ANOVA. Least square means T SE were estimated in the model. Significance was set at P < 0.05. Results: The analyses revealed a significant method (IC vs. SWA) x algorithm (v5.2 vs. v2.2) interaction with significantly smaller error (IC-SWA) for the newer v5.2 algorithm (-0.57 +/- 0.06 kcal.min(-1)) than the older v2.2 algorithm (-0.78 +/- 0.06 kcal.min(-1)). The SWA significantly overestimated EE for all activities, except inclined walking. The average mean absolute percentage error was considerably lower for the new algorithm (22%) than that for the older algorithm (35%). The average individual correlation coefficients revealed good overall agreement between the SWA and the IC (v5.2, mean r = 0.93; v2.2, mean r = 0.87). Conclusion: Overall, the SWA correlated well with IC; however, EE was significantly overestimated during most activities. Future studies should develop pregnancy-specific algorithms and assess validity of the SWA at all stages of pregnancy to further improve prediction of EE in this population.