Monitoring Activity With a Wrist-Worn Actigraph: Effects of Amplifier Passband and threshold Variations

This project investigated some properties of the AMA-32 wrist-worn activity monitor. In the first phase of this project, a software program for the analysis of activity data was developed. The program .permits rapid statistical and graphical summaries of activity data. The second phase of the project investigated effects of varying the sensitivity of the actigraph. Three different frequency passbands were investigated. The ability to differentiate sleep from wake was optimized with a 2 tO 3 Hz passband, particularly with a high trigger threshold. Increasing sensitivity by lowering the trigger threshold or by broadening the passband reduced sleepwake discriminability. The optimal sleepwake discrimination parameters for the AMA-32 approximate those of earlier generations of wrist-worn actigraphs. The work in this report suggests research to investigate the utility of other actigraph sensitivity settings for differentiating movement patterns during sleep or for assessments of daytime workload. Activity, Actigraph, Sleep, Circadian rhythms, Ambulatory monitoring