994 Training & Testing Santos-Lozano A et al. Intermonitor Variability of GT3X Accelerometer. Int J Sports Med 2012; 33: 994–999 accepted after revision March 12, 2012 Bibliography DOI http://dx.doi.org/ 10.1055/s-0032-1312580 Published online: July 12, 2012 Int J Sports Med 2012; 33: 994–999 © Georg Thieme Verlag KG Stuttgart · New York ISSN 0172-4622 Correspondence Dr. Nuria Garatachea Department of Physioteraphy and Nursing Faculty of Health and Sport Science Universidad de Zaragoza 22001 Huesca Spain Tel.: +34/60/6777 435 Fax: +34/97/6761 720 nuria.garatachea@unizar.es Key words ● ▶ physical activity ● ▶ activity assessment ● ▶ accelerometry ● ▶ motion sensor Intermonitor Variability of GT3X Accelerometer GT3X device during standardized PA. Hence, the main purpose of this study was to investigate the reliability and inter-monitor variability of the GT3X, along each orthogonal axis of motion, over a range of standardized PA. Methods ▼ Instrumentation Actigraph GT3X activity monitor The GT3X (Actigraph, Pensacola, FL, USA) is a lightweight, and compact accelerometer. The unit must be worn at the waist using a belt clip or elastic belt. This tri-axial accelerometer collects motion data on 3 orthogonal axes, known as ver- tical ( Y), horizontal right-left ( X), and horizontal front-back axis ( Z). The Actigraph GT3X also includes the vector summed value X Y Z 2 2 2   , known as ‘vector magnitude’ (VM). This device accurately and consistently records time varying accelerations ranging in magnitude from ~0.05 to 2.5 Gs. Each sample is summed over a user speci- fied interval of time called ‘epoch’. A epoch dura- tion of 60 s is generally used in the field and for all vectors that were needed in our study. The output of the Actigraph is in ‘counts’, with one count equaling 16.6 milliGs per second at 0.75 Hz. Introduction ▼ Activity monitors (i. e., accelerometers) are com- monly used for measuring free-living physical activity (PA) in humans [7]. Despite their limita- tions, e. g., inability to assess the metabolic cost associated with standing, load carrying, static work, vertical lift, changes in gradient and most upper body movements [2, 14], accelerometers are valid tools for PA assessment [9, 10, 20]. Acti- graph accelerometers have been widely used over the last 10 years to assess PA in free-living populations [6]. A tri-axial model was recently developed, the GT3X activity monitor (Actigraph, Pensacola; FL, USA), which collects motion data on 3 axes, i. e., vertical ( Y), horizontal right-left ( X), and horizontal front-back ( Z). A design that is commonly used to assess inter- monitor variability is a mechanical device that provides a standardized amount of acceleration [18]. Yet it is important to evaluate the behaviour of the measurements recorded by accelerometers while worn by humans because studies using mechanical devices cannot provide a true evalu- ation of how accelerometers perform during PA [22]. No published data are available on the reliability and inter-monitor variability of the Actigraph Authors A. Santos-Lozano 1 , G. Torres-Luque 2 , P. J. Marín 3 , J. R. Ruiz 4, 5 , A. Lucia 6 , N. Garatachea 1 Abstract ▼ The main purpose of this study was to assess the inter-monitor reliability of the tri-axial GT3X Actigraph accelerometer over a range of physical activities (PA). This device collects motion data on each of the vertical ( Y), horizontal right-left ( X), and horizontal front-back ( Z) axes and also calculates the vector summed value X Y Z 2 2 2 + + known as ‘vector magnitude’ (VM). 8 GT3X accele- rometers were worn at the same time by the same participant. Accelerometers were placed back-to- front, all facing forward and in sets of 4 securely taped together, attached to a belt and allocating each block above either left or right hip at waist level. Inter-monitor reliability was assessed dur- ing 6 conditions: rest, walking (4 and 6 km·h − 1 ), running (8 and 10 km·h − 1 ) and repeated sit-to- stand (40 times·min − 1 ). The intra-class corre- lation coefficients were high for X, Y and Z axes (i. e., all ≥ 0.925) and for VM ( ≥ 0.946). In conclu- sion, we found good inter-instrument reliability of the GT3X accelerometer across all planes, yet our results also suggest that the X and Z axes do not provide further benefits over the ‘traditional’ Y-axis to assess the movement in typical PA. Affiliations Affiliation addresses are listed at the end of the article