自閉症類群障礙症兒童語言發展預測因素：縱貫研究

研究背景與目的：語言能力是自閉症類群障礙症患者的重要指標，影響社會能力與適應功能的發展。共享式注意力和模仿能力在早期社會溝通扮演重要的角色，能促進語言能力的發展。過去少有縱貫研究針對36個月以下的自閉症類群障礙症幼兒，探討共享式注意力、模仿能力和語言能力的同時與長期關係。因此，本研究以縱貫研究的方式，探討36個月以下自閉症類群障礙症幼兒的共享式注意力與模仿能力對語言發展的預測。方法：參與者為74位的自閉症類群障礙症幼兒，於生理年齡17至35個月（平均生理年齡24個月）進行第一次評估，並於18個月後進行第二次評估。修訂台灣版學步期自閉症篩檢工具（Taiwan Version of the Screening Tool for Autism in Two-Year-Olds, T-STAT），測量完整主動性共享式注意力（full-initiating joint attention, F-IJA）、部分主動性共享式注意力（partial-initiating joint attention, P-IJA）、反應性共享式注意力（responding joint attention, RJA）、物體模仿和動作模仿能力。穆林早期學習量表（Mullen Scales of Early Learning, MSEL）測量語言理解、語言表達和整體語言能力。參與者兩次評估皆測量共享式注意力、模仿和語言能力。結果：第一次評估的RJA、物體模仿和動作模仿，皆與語言理解、語言表達和整體語言能力存在同時相關；F-IJA和P-IJA與語言理解存在同時相關，但F-IJA與語言表達則無顯著相關。控制生理年齡後，第一次評估的動作模仿能預測所有語言能力指標；RJA能預測語言理解和整體語言能力。第二次評估的F-IJA、P-IJA、RJA、物體模仿和動作模仿，皆與語言理解、語言表達和整體語言能力存在同時相關。控制生理年齡後，第二次評估的物體模仿能預測所有語言能力指標；動作模仿能預測語言表達和整體語言能力；RJA能預測語言表達。第一次評估的RJA、物體模仿和動作模仿，皆與第二次評估的語言理解、語言表達和整體語言能力存在長期相關；F-IJA和P-IJA與語言能力則無顯著相關。控制生理年齡後，第一次評估的動作模仿能預測所有第二次評估的語言能力指標，第一次評估的RJA能預測第二次評估的語言理解和整體語言能力。結論：第一次評估的動作模仿為全面性的預測因子，具有同時和長期預測力，能預測第一次和第二次評估的語言能力。動作模仿涉及高度的社會互動能力，具有象徵性表徵與理解意圖的能力，對語言學習相當重要。此外，第一次評估的RJA亦能預測第一次和第二次評估的語言理解、整體語言能力，代表幼兒理解大人溝通意圖，以有效掌握大人的注意力的焦點且語言配對。本研究可以提供自閉症類群障礙症幼兒介入的參考。

關鍵字

自閉症類群障礙症；共享式注意力；預測；語言能力；模仿能力；縱貫

並列摘要

Background and objectives: Previous studies have shown that early language ability is related to long-term outcomes and adaptive function in children with autism spectrum disorder (ASD). It is an important issue to explore predictors of language development in young children with ASD. Joint attention and imitation were supported that they could enhance language development in young children with ASD. However, there were few longitudinal studies to examine that both joint attention and imitation contribute to language development in young children with ASD under 36-months-old. Therefore, the purpose of this longitudinal study was to examine that joint attention and imitation predict language development in young children with ASD less than 36 months. Method: The participants were 74 young children with ASD aged 17-35 months (M = 24.2, SD = 4.4) at initial assessment who were followed up 18 months after their initial assessment. Taiwan version of the Screening Tool for Autism in Two-Year-Olds, (T-STAT; Chiang et al., 2013) was modified for measuring joint attention and imitation, including initiating joint attention (IJA), responding joint attention (RJA), object imitation and manual imitation. The IJA was divided into full-IJA (F-IJA) and partial-IJA (P-IJA). In addition, Mullen Scales of Early Learning (MSEL; Mullen, 1995) was used for assessing language abilities, including receptive language, expressive language and overall language. Joint attention, imitation and language abilities were assessed at two time points. Result: Language abilities at Time 1, RJA, object imitation and manual imitation were concurrently correlated with all of receptive language, expressive language and overall language. F-IJA and P-IJA concurrently correlated with receptive language, but F-IJA no significant correlated with expressive language. After contorlling chronological age, the significant concurrent predictors of all language abilities were manual imitation. In addition, the significant concurrent predictors of receptive language and overall language were RJA. Language abilities at Time 2, F-IJA, P-IJA, RJA, object imitation and manual imitation were concurrently correlated with all of receptive language, expressive language and overall language. After controlling chronological age, the significant concurrent predictors of all language abilities were object imitation. In addition, the significant concurrent predictors of expressive and overall language were manual imitation, whereas the significant concurrent predictor of expressive language was RJA. Language abilities at Time 2, all of receptive language, expressive language and overall language were longitudinally correlated with RJA, object imitation and manual imitation, but there were no significant correlated with F-IJA and P-IJA. After controlling chronological age, the significant longitudinal predictors of all language abilities were manual imitation. In addition, the significant longitudinal predictors of receptive language and overall language were RJA. Conclusion: Results of this study showed that manual imitation was the concurrent and longitudinal predictors for all language abilities at Time 1 and Time 2 in young children with ASD. Manual imitation involved complex social interaction, with representation and intentional understanding ability, which is an important factor of early language development. In addition, the results also showed that RJA was the concurrent and longitudinal predictors for receptive language and overall language at Time 1 and at Time 2 in young children with ASD. RJA reflected understanding communicative intentions, so young children could correctly follow the adult's attention and acquire language. The current findings provide implications for early intervention in young children with ASD.