Tracking Subjective Sleep Quality and Mood With Mobile Sensing: Multiverse Study

Abstract

Background Sleep influences moods and mood disorders. Existing methods for tracking the quality of people’s sleep are laborious and obtrusive. If a method were available that would allow effortless and unobtrusive tracking of sleep quality, it would mark a significant step toward obtaining sleep data for research and clinical applications. Objective Our goal was to evaluate the potential of mobile sensing data to obtain information about a person’s sleep quality. For this purpose, we investigated to what extent various automatically gathered mobile sensing features are capable of predicting (1) subjective sleep quality (SSQ), (2) negative affect (NA), and (3) depression; these variables are associated with objective sleep quality. Through a multiverse analysis, we examined how the predictive quality varied as a function of the selected sensor, the extracted feature, various preprocessing options, and the statistical prediction model. Methods We used data from a 2-week trial where we collected mobile sensing and experience sampling data from an initial sample of 60 participants. After data cleaning and removing participants with poor compliance, we retained 50 participants. Mobile sensing data involved the accelerometer, charging status, light sensor, physical activity, screen activity, and Wi-Fi status. Instructions were given to participants to keep their smartphone charged and connected to Wi-Fi at night. We constructed 1 model for every combination of multiverse parameters to evaluate their effects on each of the outcome variables. We evaluated the statistical models by applying them to training, validation, and test sets to prevent overfitting. Results Most models (on either of the outcome variables) were not informative on the validation set (ie, predicted R2≤0). However, our best models achieved R2 values of 0.658, 0.779, and 0.074 for SSQ, NA, and depression, respectively on the training set and R2 values of 0.348, 0.103, and 0.025, respectively on the test set. Conclusions The approach demonstrated in this paper has shown that different choices (eg, preprocessing choices, various statistical models, different features) lead to vastly different results that are bad and relatively good as well. Nevertheless, there were some promising results, particularly for SSQ, which warrant further research on this topic.

Opportunities for Smartphone Sensing in E-Health Research: A Narrative Review

“…Six studies collected and examined smartphone data as a proxy indicator of a person’s sleep [ 19 , 51 , 56 , 57 , 58 , 59 ]. Given that a smartphone cannot directly record a person’s sleep, studies used a combination of passively sensed data and user reports to detect sleeping behaviors.…”

“…Given that a smartphone cannot directly record a person’s sleep, studies used a combination of passively sensed data and user reports to detect sleeping behaviors. Identifying smartphone-usage behaviors during bedtime, such as putting the phone on charge [ 58 ] and lack of screen activity [ 19 ] and sound [ 57 ] assisted in estimating the user’s sleeping behavior.…”

“…The majority of articles (59%) used smartphone sensing to infer mental health conditions. While six articles examined overall mental health [ 13 , 24 , 25 , 26 , 27 , 68 ], other studies examined specific factors such as mood [ 15 , 16 , 17 , 18 , 19 , 20 , 21 , 69 , 70 , 71 , 72 , 73 , 74 , 75 ] and stress [ 76 , 77 , 78 ]. Additionally, studies also examined specific mental health conditions such as depression [ 7 , 8 , 67 , 79 , 80 , 81 , 82 , 83 , 84 , 85 , 86 , 87 , 88 , 89 ], schizophrenia [ 90 , 91 , 92 , 93 ], and bipolar disorder [ 94 , 95 ].…”

“…For example, users may not carry their devices at all or keep their devices in handbags, which may affect the accelerometer and gyroscope readings, leading to inaccuracies in human-activity recognition systems [ 2 , 11 ]. Another example (using the light sensor) might be in situations where the device is kept facing downwards, which would give inaccurate readings of users’ surrounding light conditions [ 19 , 84 ]. Such factors must be accounted for while setting up passive sensing.…”

“…Several studies have utilized smartphones to remotely collect data from users and gain health insights from these data. Studies have explored how smartphone-based sensing can be used to infer mood [ 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 ], physical activity [ 3 , 11 , 22 , 23 ], and mental health [ 13 , 24 , 25 , 26 , 27 ], and also to track the effects of chronic health conditions [ 28 , 29 , 30 , 31 ]. Moreover, software frameworks have also been developed to help researchers use smartphone sensing to conduct their own studies, such as AWARE [ 32 ], Beiwe [ 33 ], and EARS [ 34 ] to name a few.…”