Quantification of free-living activity patterns using accelerometry in adults with mental illness

Area of Science:

• Digital health
• Psychiatry
• Movement science

Background:

• Physical activity is frequently altered in psychiatric disorders.
• Wearable technology, like smartphone accelerometers, offers new avenues for unobtrusive activity monitoring.
• Validating analytical methods is crucial for interpreting complex movement data.

Purpose of the Study:

• To demonstrate a statistical approach for characterizing accelerometer-derived physical activity patterns.
• To compare simple, complex, and composite models in explaining activity and inactivity.
• To explore differences in activity patterns based on psychiatric diagnoses.

Main Methods:

• Utilized statistical methods to analyze accelerometer data from 99 community-based adults with mental illnesses.
• Screened psychiatric diagnoses using the Mini International Neuropsychiatric Interview.
• Collected one-week accelerometer data during daily life.

Main Results:

• Activity during wakefulness comprised both brief random and complex heavy-tailed movements.
• Movement during sleep lacked the heavy-tailed component, while inactivity followed a heavy-tailed process.
• Significant differences in activity patterns were observed between individuals with bipolar disorder and primary psychotic disorders.

Conclusions:

• Complex statistical models can effectively quantify detailed human movement patterns from accelerometry.
• These models capture nuances in activity during wakefulness and sleep, and inactivity.
• Movement pattern analysis may offer insights into the interaction between psychiatric diagnosis and health.