Wearable technology has become increasingly popular in monitoring various health metrics, including sedentary behavior. This validation study, published in JMIR Cardio, investigates the accuracy of a consumer-grade wrist-worn activity tracker, specifically the Fitbit Inspire 3, in measuring sedentary time among Japanese outpatients with cardiovascular disease.
Study Overview
The study involved 51 outpatients, with a mean age of 78.4 years, who were diagnosed with either heart failure or ischemic heart disease. Participants wore both the Fitbit tracker and a research-grade waist-worn accelerometer (ActiveStyle Pro HJA-750C) simultaneously for seven consecutive days. The primary objective was to evaluate the agreement between the sedentary time recorded by the Fitbit and the research-grade device.
Data collection was conducted during a standardized 12-hour daytime window, from 9 AM to 9 PM, with sedentary time assessed in one-minute epochs. The researchers employed various statistical methods, including Pearson correlation, Bland-Altman analysis, and Lin concordance correlation coefficient (CCC), to analyze the data.
Results
The findings revealed that the mean sedentary time recorded by the waist-worn accelerometer was 497.9 minutes, while the Fitbit recorded a higher mean of 559.9 minutes. The correlation between the two devices was strong (r=0.905), indicating a high level of agreement. However, the Fitbit exhibited a significant fixed bias, overestimating sedentary time by approximately 62 minutes.
After applying a constant bias correction, the concordance improved significantly, with a cross-validated CCC of approximately 0.90. This suggests that while the Fitbit can provide a useful estimate of sedentary behavior, it requires calibration to enhance accuracy.
Implications for Use
The study supports the calibrated use of consumer-grade wearables for monitoring sedentary behavior in older adults with cardiac issues. However, the wide limits of agreement indicate that individual-level precision may be constrained. Therefore, the population-specific correction factor identified in this study needs further validation before being applied to different settings or populations.
What this study does not show
This study does not establish the Fitbit’s accuracy for individual-level assessments of sedentary behavior, as the wide limits of agreement suggest variability in measurements.
Evidence scorecard
- Analytic validity: established
- Clinical validity: established
- Clinical utility: partial
The evidence is based on a validation study involving human participants.