Behavioral Power-law Exponents in the Usage of Electric Appliances Correlate Mood States in the Elderly

The proportion of the population over 65 has been increasing rapidly in industrialized countries. Recently, innovative devices for unconstrained monitoring of health variables at home have been developed. Previous studies focused mainly on the physical aspects, yet mood disturbances are also important in the elderly. To investigate the relationship between subjective symptoms including mood states and behavioral patterns, unconstrained monitoring of the usage of ubiquitous home appliances such as a refrigerator and a television was performed on 100 elderly people for 31 consecutive days. Using inter-event intervals for the use of the appliances, the probability density function (PDF) and cumulative distribution function (CDF) were calculated. The power-law exponents α and β were then obtained as the slopes of the linear regression in log-log plots of the PDF and CDF, respectively. Subjects with higher depression, confusion, fatigue and total mood disturbance scores had significantly smaller α and β. The more fatigued group tended to have a smaller β for the use of the refrigerator. These results suggested that the usage patterns of these ubiquitous appliances could be used to detect changes in mood states in the elderly.