Depression Prediction by Using Ecological Momentary Assessment, Actiwatch Data, and Machine Learning: Observational Study on Older Adults Living Alone

Abstract

BackgroundAlthough geriatric depression is prevalent, diagnosis using self-reporting instruments has limitations when measuring the depressed mood of older adults in a community setting. Ecological momentary assessment (EMA) by using wearable devices could be used to collect data to classify older adults into depression groups.ObjectiveThe objective of this study was to develop a machine learning algorithm to predict the classification of depression groups among older adults living alone. We focused on utilizing diverse data collected through a survey, an Actiwatch, and an EMA report related to depression.MethodsThe prediction model using machine learning was developed in 4 steps: (1) data collection, (2) data processing and representation, (3) data modeling (feature engineering and selection), and (4) training and validation to test the prediction model. Older adults (N=47), living alone in community settings, completed an EMA to report depressed moods 4 times a day for 2 weeks between May 2017 and January 2018. Participants wore an Actiwatch that measured their activity and ambient light exposure every 30 seconds for 2 weeks. At baseline and the end of the 2-week observation, depressive symptoms were assessed using the Korean versions of the Short Geriatric Depression Scale (SGDS-K) and the Hamilton Depression Rating Scale (K-HDRS). Conventional classification based on binary logistic regression was built and compared with 4 machine learning models (the logit, decision tree, boosted trees, and random forest models).ResultsOn the basis of the SGDS-K and K-HDRS, 38% (18/47) of the participants were classified into the probable depression group. They reported significantly lower scores of normal mood and physical activity and higher levels of white and red, green, and blue (RGB) light exposures at different degrees of various 4-hour time frames (all P<.05). Sleep efficiency was chosen for modeling through feature selection. Comparing diverse combinations of the selected variables, daily mean EMA score, daily mean activity level, white and RGB light at 4:00 pm to 8:00 pm exposure, and daily sleep efficiency were selected for modeling. Conventional classification based on binary logistic regression had a good model fit (accuracy: 0.705; precision: 0.770; specificity: 0.859; and area under receiver operating characteristic curve or AUC: 0.754). Among the 4 machine learning models, the logit model had the best fit compared with the others (accuracy: 0.910; precision: 0.929; specificity: 0.940; and AUC: 0.960).ConclusionsThis study provides preliminary evidence for developing a machine learning program to predict the classification of depression groups in older adults living alone. Clinicians should consider using this method to identify underdiagnosed subgroups and monitor daily progression regarding treatment or therapeutic intervention in the community setting. Furthermore, more efforts are needed for researchers and clinicians to diversify data collection methods by using a survey, EMA, and a sensor.

Screening for Depression in Daily Life: Development and External Validation of a Prediction Model Based on Actigraphy and Experience Sampling Method

“…While it is assumed that self-reports of depression-related affect and behaviors as assessed with ESM predict depression better than behavioral sensor data, this assumption has never been tested before. To our knowledge, there is only 1 recent study that attempted to predict depression by using both ESM and actigraphy data; however, it was focused on the elderly, had a smaller sample size (N=47), and had no external validation [ 32 ]. Currently, it is not yet clear how these approaches perform and if they can be used for screening purposes, both separately and in combination.…”

Assessing the relationship between routine and schizophrenia symptoms with passively sensed measures of behavioral stability

“…Cognitive, affective, and behavioral data collected with passive sensors and EMA may signal changes in symptoms and functioning. Such predictive models have been tested in major depressive disorder [19][20][21] , bipolar disorder 22,23 , schizophrenia 24,25 , and older adults with depression 26 . These technologies provide the capacity to examine specific questions about the relationships between structured routine and symptoms.…”

Mobile Internet Technologies, Ecological Momentary Assessment, and Intervention—Poison and Remedy for New Online Problematic Behaviors in ICD-11

“…Due to its high cost-effectiveness, EMI offers the possibility to improve access to evidence-based treatment for various populations, democratizing it ( 1 , 5 ). EMI seems to be promising owning the possibility of identifying contextual (social interaction, location) and intra-individual (craving, mood, physiological responses) precipitating factors of lapses through the employment of machine learning algorithms and data mining ( 44 , 45 ). EMI has the potential to tailor the intervention to the demographic, psychological, and behavioral characteristics of a person and specific symptoms experienced ( 46 ), meta-analytic evidence shows that such adaptive features increase the effectiveness of the interventions ( 47 ).…”