Pre-reading abilities are predictive of later reading ability and can be assessed before reading begins. However, the neural correlates of pre-reading abilities in young children are not fully understood. To address this, we examined 246 datasets collected in an accelerated longitudinal design from 81 children aged 2–6 years (age = 4.6 ± 0.98 years, 47 males). Children completed pre-reading assessments (NEPSY-II Phonological Processing and Speeded Naming) and underwent a diffusion magnetic resonance imaging (MRI) scan to assess white matter connectivity. We defined a core neural network of reading and language regions based on prior literature, and structural connections within this network were assessed using graph theory analysis. Linear mixed models accounting for repeated measures were used to test associations between children’s pre-reading performance and graph theory measures for the whole bilateral reading network and each hemisphere separately. Phonological Processing scores were positively associated with global efficiency, local efficiency, and clustering coefficient in the bilateral and right hemisphere networks, as well as local efficiency and clustering coefficient in the left hemisphere network. Our findings provide further evidence that structural neural correlates of Phonological Processing emerge in early childhood, before and during early reading instruction.

预阅读能力可以预测以后的阅读能力，可以在阅读开始之前进行评估。然而，幼儿预读能力的神经相关性尚不完全清楚。为了解决这个问题，我们检查了通过加速纵向设计从 81 名 2-6 岁儿童（年龄 = 4.6 ± 0.98 岁，47 名男性）收集的 246 个数据集。儿童完成预阅读评估（NEPSY-II 语音处理和快速命名），并接受扩散磁共振成像 (MRI) 扫描以评估白质连接性。我们根据现有文献定义了阅读和语言区域的核心神经网络，并使用图论分析评估了该网络内的结构连接。使用考虑重复测量的线性混合模型来分别测试儿童的预阅读表现与整个双边阅读网络和每个半球的图论测量之间的关联。语音处理得分与双侧和右半球网络的全局效率、局部效率和聚类系数以及左半球网络的局部效率和聚类系数呈正相关。我们的研究结果提供了进一步的证据，表明语音处理的结构神经关联出现在幼儿期、早期阅读教学之前和期间。