Modelling techniques for analysis of human activity patterns

We investigate the complexity and recurrence properties of human activity measured from forearm movement in individuals carrying out their normal daily routines. We show that complexity in activity decreases with age and speculate that this coincides with a reduction in healthy behaviour and could also indicate that one of the underlying physiological systems that help control and maintain a person's activity levels could also be declining. We also uncover a threshold that exists in the 60-70 age-group and hypothesise that this be a possible peak in human activity levels.