Comment on: Optimal objective measurement of physical function and its predictive capacity for mortality among community-dwelling older women.