Self-selected exercise intensity during household/garden activities and walking in 55 to 65-year-old females.

This study determined whether some of the more vigorous household and garden tasks (sweeping, window cleaning, vacuuming and lawn mowing) were performed at a moderate intensity (3-6 METs or metabolic equivalents) by a representative sample of 50, 55 to 65-year-old women (X +/- SD; 59.3 +/- 3.1 years, 161.5 +/- 5.2 cm, 69.4 +/- 12.4 kg, 38.4 +/- 7.3% BF). Data collection was conducted in a standardised laboratory environment and in the subjects' homes. Energy expenditure during self-perceived moderate paced walking around a quadrangle was also used as a marker of exercise intensity. Energy expenditure measured via indirect calorimetry was also predicted from: HR, CSA accelerometer counts, Quetelet's index and the Borg rating of perceived exertion. Ninety-six percent of the subjects walked at an intensity of >or= 3.0 METs. Except for vacuuming in the laboratory (X = 2.9 METs; P = 0.19), the intensity of each of the other activities was significantly (P </or= 0.002) greater than 3.0 METs. Subjects swept (3.7 vs. 3.3 METs) and vacuumed (3.6 vs. 2.9 METs) at greater intensities in the home than in the laboratory, whereas the converse applied to window cleaning (3.3 vs. 3.6 METs) and lawn mowing (4.9 vs. 5.5 METs). Eighty-six percent (172 out of 200) of the VO2 measurements were >or= 3.0 METs when the four household/garden activities were performed in the subjects' homes. These activities therefore have the potential to contribute to the 30 min day(-1) of moderate intensity physical activity required to confer health benefits but there was much inter-individual variability in the intensity at which these tasks were performed. Random intercept regression analyses yielded prediction equations with 95% confidence intervals of +/- 0.80 and +/- 0.84 METs for the laboratory and home based equations, respectively. Considering the means for the five activities ranged from 2.9 to 5.5 METs, these 95% confidence intervals lack predictive precision at the individual level. Nevertheless, the laboratory and home-based equations predicted with correct classification rates of 89 and 90%, respectively, whether energy expenditure was < 3.0 or >or= 3.0 METs.