Validity of the Actigraph-GT9X accelerometer for measuring steps and energy expenditures in heart failure patients

Figures

• 

  Fig 1. Correlations between ActiGraph-derived total step counts (per 10-sec) and manual total step counts (per 10-sec). (a) Ankle location; (b) Waist location. The diagonal line indicates a linear regression line.

• 

  Fig 2. Correlations between ActiGraph-derived mean energy expenditure (METs) using six different equations and indirect calorimetry-derived mean energy expenditure (METs). (a) Freedson; (b) Freedson+Williams; (c) Refined Crouter using 10-sec epoch data; (d) Refined Crouter using 60-sec epoch data; (e) Sasaki; (f) Santos-Lozano VT; (g) Santos-Lozano VM. The Refined Crouter equation used 10-sec and 60-sec epoch data, while other equations used 60-sec epoch data. The diagonal line indicates a linear regression line.

• 

  Fig 3. Bland-Altman plots representing the agreement between ActiGraph-derived step counts and manual step counts. (a) ankle location; (b) waist location. The horizontal line in the middle represents mean bias, followed by lines representing 95% of the limits of agreement (LOA) and the maximum limits of LOA. The diagonal line represents the linear regression line.

• 

  Fig 4. Bland-Altman plots representing the agreement between ActiGraph-derived and indirect calorimetry-derived energy expenditure using six different equations. (a) Freedson; (b) Freedson+Williams; (c) Refined Crouter using 10-sec epoch data; (d) Refined Crouter using 60-sec epoch data; (e) Sasaki; (f) Santos-Lozano VT; (g) Santos-Lozano VM. All equations except for the Refined Crouter used 60-sec epoch data. The horizontal line in the middle represents mean bias followed by lines representing 95% of the limits of agreement (LOA) and maximum limits of LOA. The diagonal line represents the linear regression line.