For the substantial population of veterans who disengage from conventional mental health services, continuous physiological monitoring through consumer-grade wearables could offer a passive, low-friction alternative to clinical assessment — a possibility with real implications for a group at disproportionate risk of depression and anxiety. Understanding whether smartwatch data can reliably track mental health trajectories in this population would reframe how we think about monitoring outside the clinic.
The V-MIND protocol deploys Garmin vívosmart-5 wristwatches across a UK veteran cohort over 84 days, capturing continuous streams of physical activity, heart rate variability, sleep architecture, and stress indicators. Validated psychiatric questionnaires are administered at four time points — baseline, day 28, day 56, and day 84 — creating parallel subjective and objective data tracks. Machine learning models will then be trained on this merged dataset to detect shifts in mental health and functioning outcomes, with the explicit goal of assessing feasibility before any large-scale inferential study is designed.
This is a protocol paper, meaning the science is prospective and no results yet exist — an important distinction for readers expecting clinical conclusions. The study's observational, convenience-sampled design limits causal inference, and small feasibility cohorts often lack power to validate machine learning classifiers meaningfully. That said, digital phenotyping — inferring psychiatric states from passively collected device data — is a rapidly maturing field. Prior work in general populations has shown moderate success linking HRV and step-count variability to depressive episodes, but veteran-specific validation is sparse. If V-MIND demonstrates adequate recruitment retention and signal quality over three months, it would establish groundwork for a much-needed powered trial. For now, the study is best understood as infrastructure-building: incremental but directionally important for extending mental health monitoring to those least likely to seek it.
