Skip to main content
. 2023 Dec 13;25:e46778. doi: 10.2196/46778

Table 5.

Anxiety disorder: retrieved studies and their main outcomes.

Study Country (region) Data Psychiatric disorder Symptoms investigated Assessment technology Sample, n Data collection time Statistics Synthesis of main results
Jacobson et al [33], 2020 Lebanon, New Hampshire (United States) 2020 Social anxiety disorder Severity of anxiety and depressive symptoms and positive and negative affects; movement and social contact Passive (smartphone): accelerometer, incoming and outgoing calls, SMS text messages
Active: SIAS (Social Interaction Anxiety Scale), DASS-21 (Depression, Anxiety, Stress Scale), self-report PANASa 		59

  • Period of collection: 2 weeks

  • Frequency of collection: passive (continuously) and active (twice)

 Correlation between predicted and observed symptoms severity: r=0.702 The results suggest that these passive detection data can be used to accurately predict the severity of participants’ social anxiety symptoms, specifically demonstrating a strong correlation between the predicted and observed severity of social anxiety symptoms.
Jacobson et al [34], 2021 United States 2021 Generalized Anxiety Disorder and Panic Disorder Wake-sleep rhythms (sleep duration, wake duration, number of wake periods, and number of sleep periods), latency at sleep onset, sleep repetition time before waking up, sleep quality, and time to get up after waking up Passive (wearable) through smartphone app: actigraphy 265

  • Period of collection: phase 1, phase 2: 9-14-years later phase 1; phase 3: 17-18-years later phase 1

  • Frequency of collection: passive (continuously for 1 week in phase 2) and active (twice in phase 1 and in phase 3)

 Prediction of symptoms deterioration: (AUCb=0.696; 95% CI 0.598-0.793; 84.6% sensitivity; 52.7% specificity; balanced accuracy=68.7%) The results show that through the use of wearable motion-sensing tools, such as the ActiGraph, it is indeed possible to significantly predict which individuals will experience symptom deterioration over a 17-18 years period.
Jacobson et al [35], 2022 Lebanon, New Hampshire (United States) 2022 Generalized Anxiety Disorder or Social Anxiety Disorder Physiological activation (heart rate and heart rate variability), light exposure, social contact, and location Passive (smartphone): GPS, Google Places, National Weather Service, finger pressure on the rear camera.
Active: Self-Report of PANAS-X (positive and negative affect tab, fear, and sadness subscales) and Self-Report of MEAQ (Multidimensional Experiential Avoidance Questionnaire).		 32

  • Period of collection: 1 week

  • Frequency of collection: passive (continuously for 1 week) and active (hourly for 1 week)

 Future changes in anxiety symptoms model: R2=0.748
Changes hour-by-hour within-person: R2=0.385		 Customized deep learning models using smartphone sensor data can accurately predict future changes in anxiety disorder symptoms and even changes in the same participant from hour to hour.
Meyerhoff et al [36], 2021 Chicago, United States 2021 Mood disorder, social anxiety disorder, and Generalized Anxiety Disorder Movement, social interactions, location Passive (smartphone): GPS, app use, calls, messages
Active: PHQ-8c, GAD-7 (Generalized Anxiety Disorder 7-item scale), SPIN (Social Phobia Inventory)		 282

  • Period of collection: 16 weeks

  • Frequency of collection: passive (every 5 min in 1 day every 3 weeks) and active (every 3 weeks)

 Multimorbidity groups: changes in depression are predicted by changes in GPS features Time: r=−0.23; P=.02, locations: r=−0.36; P<.001, exercise duration: r=0.39; P=.03, and use of active apps (r=−0.31; P<.001)
Depression and anxiety groups: changes in depression are predicted by changes in GPS features for locations (r=−0.20; P=.03) and transitions (r=−0.21; P=.03)		 Changes in sensor-derived behavioral characteristics are associated with subsequent changes in depression, but not vice versa, suggesting a unidirectional relationship in which changes in detected behaviors are associated with subsequent changes in symptoms.

aPANAS: Positive and Negative Affect Schedule.

bAUC: area under the curve.

cPHQ-8: Patient Health Questionnaire-8.