A Better Inter-Instrument Reliability of Uniaxial vs Triaxial Accelerometry in Free Living Conditions

Introduction

The most popular accelerometers in physical activity research are the Actigraph uniaxial accelerometer (AGR, Model GT1M, Actigraph, Shalimar, CA) and the RT3 triaxial accelerometer (Stayhealthy Inc, Monrovia, CA). When a large population is evaluated at the same time, the use of several accelerometers is necessary. To be able to compare data between subjects or repeated measures, accelerometers have to be highly reliable. To date, only one study assessed the inter-instrument reliability of Actigraph accelerometer in free living conditions (FLC), and none on the RT3 accelerometer [1]. Also, to the best of our knowledge, no study compared inter-instrument CV of uniaxial and triaxial accelerometers in FLC. Therefore, the purpose of the present study was to evaluate the inter-instrument reliability of RT3 and Actigraph accelerometers at different levels of PA in FLC.

Methods

Fifteen healthy adults, aged 22 to 33 year-old, participated in this study. On the test day, the participant arrived at the laboratory. The research team explained instruction on an appropriate use of the two accelerometers. Participants then attached eight Actigraph accelerometers to the level of the back with an elastic belt and adjustable buckle, and five RT3 accelerometers to the back with, also, an elastic belt. The same accelerometers were used for all the participants. Accelerometers were worn during 24 hours in free living conditions. All time spent in motorized vehicles was removed in the statistical analysis. The next day, the participant came back to the laboratory to return accelerometers. Inter-instrument coefficient of variation (CV) for each intensity was assessed using the formula: CV = standard deviation of the measure × 100/mean of the measure for each intensity. The U-Mann & Whitney test was used to compare the difference between the both accelerometers. A p value of .05 was chosen for significance.

Results

Participants wore accelerometers for an average of 768 ± 164 min. For both accelerometers, the reliability increased when PA intensity increased. Inter-instrument CV of Actigraph (3 to 10.5%) was lower than those of RT3 (12.6 to 35.5%) at each PA level (p<0.05). Therefore, the Actigraph shows a better inter-instrument reliability compared to the RT3 accelerometer at each PA level.

Discussion and Conclusion

In free-living conditions, a good inter-instrument reliability is needed when assessing PA in large population. This study shows some limits in assessing sedentary and light PA levels, particularly for the RT3 accelerometer. Even if triaxial accelerometers are very precise to assess PA in FLC, the assessment of sedentary behavior may be inaccurate because of the poor reliability.

References

[1] McClain JJ, Sisson SB, Tudor-Locke C. Actigraph accelerometer interinstrument reliability during free-living in adults. Med Sci Sports Exerc, 2007; 39: 1509-14.