Smart Wearable Technologies for Fall Detection, Risk Assessment, and Prevention in Older Adults: A Review

Abstract

Falls remain the leading cause of injury-related morbidity and mortality among people aged 65 and older, and the wearable sensor industry has responded with a proliferating set of accelerometer- and gyroscope-based products marketed as fall prevention solutions. This review narrows that broad claim to a specific, testable object: on-body inertial measurement unit (IMU) systems—accelerometers, gyroscopes, and combined IMUs worn on the trunk, waist, wrist, or limb—that perform either (a) real-time post-event fall detection, (b) sub-second pre-impact fall detection intended to trigger a protective response, or (c) prospective, gait-based fall-risk classification of community-dwelling older adults. Drawing on systematic reviews, meta-analyses, and individual validation studies published through early 2026, this review (1) organizes recent advances along a clinical nursing closed-loop pathway—risk identification, real-time monitoring and early-warning response, injury protection, and nursing collaboration; (2) traces the sensor-to-algorithm-to-system pipeline from raw inertial signal to alarm; (3) tabulates sensor type, algorithm class, sample composition, accuracy, sensitivity/specificity, and false-alarm rate across a representative set of studies; and (4) discusses four adoption-limiting factors: the digital divide, device adherence, privacy, and alarm fatigue. A central finding is that laboratory-simulated-fall accuracy figures (often >95%) do not transfer to real-world, unscripted settings, where sensitivity for actual falls in older cohorts has been reported as low as 55%—80% and false-alarm rates vary by two orders of magnitude across studies and algorithms.