JMIR Publications

ABSTRACT

Background:

Sleep is an important determinant of our health and behavior during the wake phase. To facilitate the monitoring of sleep over prolonged time and across a large number of persons, novel research methods for field assessments are required. The ubiquity of smartphones offers new avenues to detect rest-activity patterns in everyday-life situations in a non-invasive and inexpensive manner and at a large scale. Recent studies provided evidence for the potential of smartphone interactions monitoring as a novel tracking method to approximate rest-activity patterns based on the timing of periods of smartphone activity and inactivity throughout the 24-h day. These findings require further replication and more detailed insights in interindividual variations in the associations and deviations with commonly used metrics to monitor rest-activity patterns in everyday life.

Objective:

The present study aimed to replicate and expand on earlier findings regarding the associations and deviations between smartphone keyboard-derived and self-reported estimates of the timing of the onset of the rest and active periods and the duration of the rest period. Moreover, we aimed to quantify interindividual variations in the associations and time differences between the two assessment modalities, and investigate to what extent general sleep quality, chronotype and trait self-control moderate these associations and deviations.

Methods:

Students were recruited to participate in a 7-day experience sampling study with parallel smartphone keyboard interactions monitoring. Multilevel modelling was employed to analyze the data.

Results:

In total, 157 students participated in the study, with an overall response rate for the diaries of 88.9%. Results revealed moderate to strong relations between the keyboard-derived and self-reported estimates, with stronger associations for the timing-related estimates (beta ranging from .61 to .78) than the duration-related estimates (beta=.51 and beta=.52). The relational strength between the time-related estimates was lower, but did not significantly differ for the duration-related estimates, among students experiencing more disturbances in general sleep quality. Time differences between the keyboard-derived and self-reported estimates were on average small (<.5 h), while large discrepancies between the two assessment modalities were also registered for quite some nights. The time differences between the two assessment modalities were larger for both the timing-related and rest duration-related estimates among students who reported more disturbances in their general sleep quality. Chronotype and trait self-control did not significantly moderate the associations and deviations between the two assessment modalities.

Conclusions:

We were able to replicate earlier findings regarding the potential of smartphone keyboard interaction monitoring as a method to approximate rest-activity patterns. Complementing earlier research findings, the current study showed that the behavioral proxies obtained from smartphone interactions might be less powerful among students who experienced disturbances in their general sleep quality. The generalization and underlying process of these findings remain to be further investigated.