Use of Wearable Technology for Measuring and Characterizing Sedentary Behavior in People With Mild Cognitive Impairment and Dementia: Systematic Review.

Abstract

Background: Sedentary behavior (SB) is a critical, modifiable risk factor for adverse health outcomes. Evidence suggests that SB is higher among individuals with cognitive impairment relative to their cognitively healthy peers. However, the nature and extent of SB across cognitive impairments remains unclear, largely due to the reliance on self-report data and the lack of synthesized evidence from more accurate methodology, such as wearable devices. Wearable device-based methodologies offer a reliable means of capturing SB in real-world settings, circumventing the recall bias inherent to self-report methods. Continuous remote monitoring of SB, via wearable devices, may provide nuanced insights important for understanding SB's contribution to cognitive impairment and health consequences.

Objective: This review aims to synthesize evidence on the volume, patterns, and variability of SB across cognitive impairment and critically appraise the wearable device-based methodology used to capture SB in this population.

Methods: Following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, we searched 5 databases (Embase, MEDLINE, PsycInfo, Scopus, and Web of Science) up to January 2025 for peer-reviewed English-language studies using wearable devices to measure SB in community-dwelling or aged residential care residents aged 50 years or older with cognitive impairment (PROSPERO: CRD42024616523). Study quality was assessed using an adapted version of the National Institute of Health Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies. Data were extracted on SB outcomes (eg, volume, pattern, and variability) and methodological characteristics (eg, device type, placement, SB classification/processing, and its corresponding validation).

Results: From 2824 screened records, 17 studies (2016-2025) were included. Most studies (n=11, 65%) were of "good" quality (scoring ≥5 on bias assessment). Synthesis revealed inconsistent evidence for differences in SB volume across cognitive impairment. However, individuals with dementia consistently exhibited a unique SB pattern, engaging in significantly fewer but longer sedentary bouts than other forms of cognitive impairment and cognitively intact controls. All (n=17, 100%) studies used volume metrics to describe SB, followed by pattern metrics (n=7, 41%); only 1 study reported on SB variability. Methodological appraisal found significant heterogeneity: 13 different device models across 6 body placements were used. Most studies quantified SB using count-based thresholds (counts per minute), which were largely unvalidated in cognitively impaired or older adult populations.

Conclusions: This review found that participants with dementia consistently exhibited a unique pattern of SB compared to other forms of cognitive impairment and healthy controls, while evidence for differences in SB volume was inconsistent. This may indicate that differences in SB volume are not inherent to dementia pathology but may be mediated by other factors, such as neuropsychiatric symptoms or environmental influences. Furthermore, methodological heterogeneity and unvalidated thresholds were observed throughout the review, highlighting a need for standardized protocols to enhance the validity and clinical applicability of future research.

Trial registration: PROSPERO CRD42024616523; https://www.crd.york.ac.uk/PROSPERO/view/CRD42024616523.