Activity measurment for elderly population using smart phone: Criterion validity in daily life

The accelerometers implanted into smart phones can be used as a self monitoring device for daily physical activity in elderly population. The study goal is to validate the smart phone-based activity monitoring application in adults aged below and above 65 years old. Ten young adults (mean of 33, SD 13.7) and ten elderly (mean 76, SD 5.5) were asked to monitor their daily physical activity with a smart phone and an ActiGraph GT3X for 7 consecutive days. Each minute of activity for both devices was classified as one of four different levels: sedentary, light, moderate and high. Minutes count from each activity level between both devices was compared using spearman correlations for each group of adults separately. Analysis of Pearson's correlation and paired t-tests between both devices was applied for separately for two groups. Spearman correlations in adults aged below 65 years varied between .62 and .89 and correlations in adults aged above 65 years varied between .73 and .98. Pearson's correlations between the two devices for minutes spent in every activity level per day and per participant were found high ranged .79–.98. Paired t-tests showed an underestimation for the count of minutes spent in sedentary activity level in elderly and adults participants detected from the smart phone with 6.45% and 6.78% respectively.