Adaptation et validation en population générale et clinique chez l’enfant d’une échelle de signes neurologiques mineurs (SNM) : l’échelle de Krebs et al. (2000).

Après une étude préliminaire qui a examiné 42 enfants, des modifications ont été apportées pour permettre une adaptation à l’enfant et faciliter sa pratique en particulier chez les enfants atteints de troubles du spectre autistique (TSA). Une étude de validation a été menée en examinant les propriétés psychométriques de l’échelle chez 86 enfants âgés de 6 à 12ans dont 26 enfants atteints de TSA et 60 enfants au développement normal.

L’adaptation de l’échelle a consisté en une modification de l’ordre des items, en une utilisation de supports concrets pour l’évaluation de la latéralité, et en l’élimination de l’item praxie constructive. La cohérence interne était bonne avec un alpha de Cronbach de 0,87. La fiabilité interjuge donnait un coefficient de corrélation intraclasse de 0,91. L’analyse en composantes principales rapportait cinq facteurs représentant 62,96 % de la variance totale. L’évaluation des SNM a montré des différences significatives entre patients et témoins pour le score total de SNM (p=0,000) et dans différents sous-scores.

La validité de l’échelle est confirmée pour l’évaluation des SNM chez les enfants, en particulier pour les enfants atteints de TSA.

Neurological soft signs (NSS) include anomalies in motor integration, coordination, sensory integration and lateralization and could be endophenotypic markers in autism spectrum disorders (ASD). Their characterization provides a more precise phenotype of ASD and more homogeneous subtypes to facilitate clinical and genetic research. Few scales for NSS have been adapted and validated in children including children with ASD. Our objective was to perform an adaptation to the child of a scale assessing neurological soft signs and a validation study in both general and clinical populations.

We have selected the NSS scale of Krebs et al. (2000) already validated in adults. It encompasses 5 dimensions: motor coordination, motor integration, sensory integration, involuntary movement, laterality. After a preliminary study that examined 42 children, several changes have been made to the original version to adapt it to the child and to increase its feasibility, particularly in children with ASD. Then we conducted a validation study by assessing the psychometric properties of this scale in a population of 86 children including 26 children with ASD (DSM 5 Criteria) and 60 typically developing children. Children's ages ranged between 6 and 12 years, and patients and controls were matched for gender, age and intelligence. Patients were assessed using the Autism diagnostic Interview-revised and the Childhood Autism Rating Scale to confirm diagnosis. Typically developing children were assessed using the semi-structured Mini International Neuropsychiatric Interview for Children and Adolescents to eliminate any psychiatric disorder. All children with neurological pathologies (history of cerebral palsy, congenital anomaly of the central nervous system, epilepsy, tuberous sclerosis, neurofibromatosis, antecedent of severe head trauma) and obvious physical deformities or sensory deficits that could interfere with neurological assessment were excluded from the study. Both patients and controls were assessed using the Raven Progressive Matrices to exclude intellectual disability, and the adapted Krebs’ scale for the assessment of NSS.

Adaptation of the scale consisted of a modification in the order of items, in the use of concrete supports for the assessment of laterality and in the elimination of item constructive praxis. The internal consistency was good with a Cronbach alpha of 0.87. Inter-rater reliability was good, kappa coefficient was greater than 0.75 for 16 items, 3 items had a kappa value between 0.74 and 0.60, only 1 item had a kappa coefficient between 0.4 and 0.59. Good inter-rater reliability was also checked for the total score with a value of intra-class correlation coefficient (ICC) of 0.91. Principal component analysis found five factors accounting for 62.96 % of the total variance. About the comparison between patients and controls, significant differences were found for NSS total score (P=0.000) and all subscores.

The adaptation for children of the Krebs et al.’ NSS scale proved to be valid, especially in children with ASD.