In recent years, wearable technology has gained traction in monitoring health metrics, particularly in the realm of cardiovascular care. This validation study, published in JMIR Cardio, focuses on the accuracy of a consumer-grade wrist-worn activity tracker, specifically the Fitbit Inspire 3, in measuring sedentary behavior among older outpatients with cardiovascular conditions.
Study Overview
The study involved 51 outpatients diagnosed with heart failure or ischemic heart disease, with an average age of 78.4 years. Participants wore both the consumer-grade wrist-worn tracker and a research-grade waist-worn accelerometer simultaneously for a week. The primary aim was to assess the agreement between the sedentary time recorded by both devices during a standardized 12-hour daytime window.
Key Findings
The results indicated that the mean sedentary time recorded was 497.9 minutes for the research-grade device and 559.9 minutes for the Fitbit tracker. The correlation between the two devices was strong, with a Pearson correlation coefficient of 0.905. However, the Fitbit tracker overestimated sedentary time by approximately 62 minutes. The study also found that while the Fitbit showed a constant bias, there was no proportional bias detected.
To improve the accuracy of the consumer-grade device, a constant bias correction was applied, which significantly enhanced the concordance of the measurements. After correcting for the bias, the concordance correlation coefficient improved to approximately 0.90, indicating a high level of agreement between the devices.
Implications for Use
The findings suggest that while consumer-grade wrist-worn trackers can provide useful insights into sedentary behavior, they may require calibration for accurate individual-level assessments. The wide limits of agreement highlight the need for caution when interpreting data from these devices in clinical settings. The study emphasizes the potential for calibrated consumer wearables to facilitate scalable monitoring in ambulatory cardiac care, although further validation is necessary before applying the correction factor in different populations.
What this study does not show
This study does not demonstrate that the consumer-grade wrist-worn tracker is suitable for precise individual-level assessments of sedentary behavior without calibration.
Evidence scorecard
- Analytic validity: Established
- Clinical validity: Established
- Clinical utility: Partial
The evidence class for this study is primary_research with human status unclassified.