Energy Expenditure in Institutionalized Older Adults: Validation of SenseWear Mini.

Purpose: Objective methods to quantify physical activity (PA) and predict energy expenditure (EE) are needed in the old and old-old. The aim of the present study was to evaluate the validity of EE estimates by the SenseWear Mini (SWMini) compared with indirect calorimetry during daily life activities in institutionalized older adults.

Methods: Sixty nursing home residents (mean age = 85.5 ± 5.5 yr) wore the SWMini during rest (sitting quietly) and three activity tasks (walking, sitting/rising/walking, and moving objects). SWMini data were processed using software version 7.0. The criterion EE (kcal·min⁻¹) was estimated by a portable gas analyzer, Oxycon Mobile (OM).

Results: The analyses revealed high correlations (rsitting = 0.68, ractivity tasks = 0.88) between EE estimated by OM and SWMini. EE increased between sitting periods and activity tasks for EE estimated by OM (mean difference = 61.5% ± 8.9%), as well as for EE estimated by SWMini (mean difference = 58.2% ± 7.4%) (P < 0.001). However, SWMini significantly underestimated EE, with an overall absolute percent error of 14.1% ± 7.9%. The largest absolute percent differences were observed during sitting periods compared with activity tasks (P < 0.05). Older age significantly reduced accuracy, explaining 12% of the variance in total percent error (β = 0.42, t = 2.84, P < 0.05).

Conclusions: The high percent error scores indicate that the SWMini is of limited value for quantifying EE in the old and old-old. The accuracy could be improved by developing accurate age- and activity-specific algorithms. On the other hand, the SWMini can be used as a suitable device for researchers interested in specific levels and patterns of PA and sedentary behavior.