Comparing activity scores by Actiwatch tied to leg and arm in preschool children

R. Juyall³, C.P. Khokhar³, S. Sazawal^{1, 2}, U. Dhingra², P. Verma¹, A. Sarkar¹, P. Dhingra¹, M. Osmany¹, V.P. Menon ¹, M. Black⁴, R.M. Angulo-Kinzler⁵ and R.E. Black²

¹Center for Micronutrient Research, Annamalai University, India
²Johns Hopkins Bloomberg School of Public Health, Baltimore, USA
³Departments of Psychology, Gurukul Kangri University, Haridwar, India
⁴University of Maryland School of Medicine, Baltimore, Maryland, USA
⁵Division of Kinesiology, Center for Human Motor Research, University of Michigan, USA

With the advent of small accelerometer-based activity monitors, the ability to monitor physical activity in children has become possible. Data from few studies in adults and older children validating the scores and site of placement of watch at hip or leg have been reported. There are no data on either validation or comparison of various sites among preschool children. In order to evaluate arm and ankle as possible sites two Actiwatch (Mini-Mitter co. Inc), one on arm and another on ankle were tied for 1 hour, in 126 children aged 13-69 months residing in peri urban colony of Delhi. For the same time, activity of child was recorded by observation using The Observer^® (Noldus Information Technology bv, The Netherlands). From The Observer, time spent with 16 gross motor activities was estimated in 1 min intervals to coincide with activity scores of 1 min epochs estimated from Actiwatch.

Overall leg scores were nearly twice as much as arm scores. In separate models evaluating gross motor activity with leg active watch score and arm active watch scores, leg active watch score was associated with 6 out of 18 gross activities i.e crawling, running, climbing, walking alone, standing alone and walk with support. The arm actiwatch score was associated with 12 of 16 gross motor activities i.e. squatting, hanging, crawling, running, climbing, rough-tumble, dancing, kneeling, stand with support, walk alone, stand alone, swinging and walk with support.

Among epochs with 6 gross motor activities associated with leg monitors, mean activity score in leg monitor was 896.29±1070.99 compared to arm monitor 419.82±460.56 while for variables not associated with leg monitors, leg scores were 190.93±381.89 compared to 278.97±364.84 for arm.

Actiwatch tied to leg records more of activities associated with leg movements, while Actiwatch tied to arm records more broader range of movements including leg movements. However for leg movement scores for leg monitor is significantly higher then arm monitor. Using a combined score of two monitors will provide a better estimation of overall activity.

Paper presented at Measuring Behavior 2005 , 5^th International Conference on Methods and Techniques in Behavioral Research, 30 August - 2 September 2005, Wageningen, The Netherlands.

© 2005 Noldus Information Technology bv