Which items of the movement assessment battery for children are most sensitive for identifying children with probable developmental coordination disorder? Results from a large-scale study

Abstract

Introduction: Despite the widespread use of the Movement Assessment Battery for Children, 2nd edition (MABC-2), little is known about the sensitivity or specificity of the individual items to detect probable Developmental Coordination Disorder (p-DCD). This study examined which specific MABC-2 items were most sensitive to identify children with p-DCD and which items would predict p-DCD. Methods: Based on a large dataset including European and African children aged 3-16 years (n = 4916, typically developing (TD, 49.6% boys); n = 822 p-DCD (53.1 % boys), Hedges' g was calculated to establish the standardized mean difference (SMD) between p-DCD/TD. SMDs were considered substantial when absolute values at or above 1.4. Sensitivity and specificity of the raw MABC-2 item scores predicting p-DCD/TD per age band (AB) were established with logistic regression analysis. Results: AB1: Children with p-DCD performed substantially poorer on threading beads (SMD:-1.61) and jumping on mats (SMD: 1.61). By combining all items and the country of origin, the sensitivity was 61.7% and specificity 98.6%. AB2: Walking heel-to-toe forwards (SMD: 1.65) was substantially poorer in p-DCD. By combining all items and the country of origin, the sensitivity was 79.0% and specificity 97.6%. AB3: Catching a ball with the preferred (SMD: 1.8) or non- preferred (SMD: 1.61) hand, and for walking heel-to-toe backwards (SMD: 1.78) were substantially poorer in p-DCD. All items combined resulted in a sensitivity of 94.4% and specificity of 99.6%. Conclusion: Not all MABC-2 items are equally sensitive to distinguish between performances of pDCD and TD. Despite the good specificity, the sensitivity was only moderate in AB1-2, the age at which children learn culturally influenced motor skills.