This page was translated automatically from Turkish. Read the original

Bu sayfa, 26. Dünya Psikiyatri Kongresi'nde sunulan çalışmanın yapay zekâ desteğiyle hazırlanmış Türkçe özetidir. Çalışmanın sahibi sunan yazarlardır; atıf için orijinal kaynağa başvurun.

EV0457

Feasibility of wearable device–based mental health monitoring for just-in-time interventions: a secondary analysis of a military cohort

THE FEASIBILITY OF WEARABLE-BASED MENTAL HEALTH MONITORING FOR JUST-IN-TIME INTERVENTION: A SECONDARY ANALYSIS OF A MILITARY COHORT

Wearable device–based mental health monitoring in veterans was found to be feasible, with high wear time and good participant retention; significant negative associations were identified between deep sleep duration and body energy expenditure and the severity of post-traumatic stress disorder symptoms.

The aim was to assess the feasibility of using wearable device data to monitor veterans' mental health and to address the problem of treatment disengagement. In a secondary analysis of data from an observational feasibility study, UK veterans scoring 12 or above on the GHQ-12 were passively monitored using a Garmin Vivosmart-5 for 3 months and completed monthly mental health questionnaires; analyses included baseline characteristics, participation and retention rates, questionnaire and mental health score summaries, device wear time, exploratory correlations, simple regression, and univariable mixed-effects models. A total of 63 participants were recruited over 5 months at an average rate of 2.41 participants per week; 53 completed at least two of the four questionnaires, 47 completed all four, 51 provided demographic data (76% male, mean age 47.2 years, standard deviation 9.12), and 49 provided wearable device data (mean wear time 23.3 hours, standard deviation 2.68). Post-traumatic stress disorder symptom scores (PCL-5) decreased from 31.6 at baseline to 26.1 at month 3, while general mental health scores (GHQ-12) decreased from 17.6 to 15.1; both reductions were statistically significant and exceeded the threshold for clinical significance. In the mixed-effects regression model, the negative associations of deep sleep duration and body energy expenditure with PCL-5 scores remained significant after false discovery rate (FDR) correction, whereas minimum respiratory rate and other variables showed only nominal significance. In conclusion, good retention, strong questionnaire adherence, and high average wear time support the feasibility of monitoring; these preliminary associations between wearable device data and mental health data may provide a foundation for future machine learning–based classification approaches.

Tüm WCP 2026 özetleri