The Parieto-Frontal Integration Theory (P-FIT) predicts that human intelligence is closely linked to structural and functional properties of several brain regions mainly located in the parietal and frontal cortices. It also proposes that solving abstract reasoning tasks involves multiple processing stages and thus requires the harmonic interplay of these brain regions. However, empirical studies directly investigating the relationship between intellectual performance and the strength of individual functional connections related to the P-FIT network are scarce. Here we demonstrate, in two independent samples comprising a total of 1489 healthy individuals, that fMRI resting-state connectivity, especially between P-FIT regions, is associated with interindividual differences in matrix reasoning performance. Interestingly, respective associations were only present in the overall samples and the female subsamples but not in the male subsamples, indicating a sex-specific effect. We found five statistically significant connections which replicated across both samples. These were constituted by BAs 8, 10, 22, 39, 46, and 47 in the left as well as BAs 44 and 45 in the right hemisphere. Given that many of these brain regions are predominantly involved in language processing, we hypothesized that our results reflect the importance of inner speech for solving matrix reasoning tasks. Complementary to previous research investigating the association between intelligence and functional brain connectivity by means of comprehensive network metrics, our study is the first to identify specific connections from the P-FIT network whose functional connectivity strength at rest can be considered an indicator of intellectual capability.

额叶额叶整合理论（P-FIT）预测，人类智力与主要位于顶叶和额叶皮层的几个大脑区域的结构和功能特性紧密相关。它还提出，解决抽象推理任务涉及多个处理阶段，因此需要这些大脑区域的谐波相互作用。但是，缺乏直接研究智力表现和与P-FIT网络相关的各个功能连接强度之间关系的实证研究。在这里，我们证明，在包括1489名健康个体的两个独立样本中，fMRI静息状态的连通性（尤其是在P-FIT区域之间）与矩阵推理性能的个体差异有关。有趣的是，各自的关联仅出现在总体样本和女性子样本中，而没有出现在男性子样本中，表明存在性别特异性效应。我们发现在两个样本中都有五个具有统计学意义的联系。这些由左侧的BA 8、10、22、39、46和47以及右侧半球的BA 44和45组成。鉴于许多这些大脑区域主要参与语言处理，我们假设我们的结果反映了内部语音对于解决矩阵推理任务的重要性。与以前的研究相辅相成，后者通过综合的网络指标来研究智力与功能性大脑连通性之间的关联，