Technical reliability of the CSA activity monitor: The EarlyBird Study

To evaluate the technical performance of the CSA accelerometer-based activity monitor.

Twenty-three CSA monitors were subjected to intra- and inter-instrument variability tests by controlled trials using a motorized turntable. The CSA monitor measures change in acceleration, and precision was tested by producing sinusoidal variations in speed around two fixed baseline speeds (fast and medium). The angle of the monitor to the line of force along the radius of the turntable was varied using tilted blocks. Three sets of tests were carried out. 1. Intra-instrument variability: seven monitors were tested three times in each of the four quadrants. 2. All 23 monitors were used for inter-instrument tests. 3. The effects of tilt at 15 degrees, 30 degrees, and 45 degrees were carried out on six monitors.

Intra-instrument coefficients of variation (CV) never exceeded 2% for fast or medium speed and achieved "between run" intra-class correlation coefficients (ICC) of 0.92 and 0.84 respectively. There were no significant differences between the monitors in terms of repeatability (fast: = 0.97, medium: = 0.77). Although there were significant differences between monitors in terms of mean score, inter-instrument variability did not exceed 5% at either speed. Inter-batch ICCs ranged from 0.87 to 0.98 for fast and from 0.71 to 0.99 for medium. The angle test results corresponded closely to those predicted theoretically, with a loss in mean score of only 6% when the monitor was tilted from 0 degrees to 15 degrees.

The CSA monitor provides a precise tool for measuring changes in acceleration in laboratory settings. Technically, the device performs well, and is likely to prove a useful tool in the assessment of physical activity in children and adults.