APOE ε4, physical activity, and the brain: a review of systematic reviews.

Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by persistent deficits in social communication and interaction across multiple contexts, along with restricted, repetitive patterns of behavior, interests, or activities (). A lack of empathy is one of the core social communication impairments observed in individuals with ASD (, ). Empathy is defined as the ability to recognize, understand, and share the emotional states of others, serving as the foundation for sincere and mutually beneficial relationships (). Numerous definitions of empathy encompass a broad range of emotional states. The main types of empathy include cognitive empathy, affective empathy, and somatic empathy (). In this study, we examined both affective and cognitive aspects of empathy. Because ASD is a neurodevelopmental condition associated with advanced cognitive impairments, characterizing functional differences between participants with ASD and control subjects can provide a direct and efficient way to identify the underlying mechanisms of ASD. Functional activity across brain regions that operate simultaneously is considered to constitute the brain network, which can be measured using intrinsic functional connectivity (FC) (). FC analysis involves identifying brain regions whose blood oxygenation level-dependent (BOLD) signals obtained from resting-state functional magnetic resonance imaging (rs-fMRI) co-fluctuate (). These co-fluctuating regions are considered to form a brain network supporting a specific cognitive or mental processing task. Functional MRI (fMRI) studies have reported that the theory of mind (ToM) involves the medial prefrontal cortex (mPFC), temporoparietal junction (TPJ), middle temporal gyrus, temporal pole, and posterior cingulate cortex/precuneus (–). Other studies using rs-fMRI have reported that brain regions within the default mode network (DMN) overlap with the brain regions associated with ToM. Thus, using rs-fMRI-derived FC may provide quantitative insights into the social-cognitive capabilities of individuals with ASD (–). For example, Assaf et al. () reported that FC values in specific brain regions were highly correlated with Social Responsiveness Scale scores in a group of individuals with ASD. Weng et al. () reported that FC strength between the posterior cingulate cortex and the temporal lobe was negatively correlated with social impairment based on the Autism Diagnostic Interview–Revised (ADI-R) system. These studies have suggested the potential value of using rs-fMRI to investigate complex cognitive functions in individuals with ASD. Prior resting-state fMRI studies of ASD have characterized functional connectivity across whole-brain scales (–), including distributed correlational networks (, ) and graph-theoretical measures (, ). However, the majority of studies have focused on individual brain regions or a limited number of preselected brain regions or networks. Increasing evidence indicates that human brain function arises from the concurrent interaction of multiple distributed, connectome-scale neural networks (–). Identifying such networks in ASD necessitates defining a set of functionally meaningful, group-consistent regions of interest (ROIs) that serve as a common reference framework for cross-subject comparisons. This challenge extends beyond spatial normalization to a standard template space, which is routinely addressed in rs-fMRI studies, and instead concerns the functional validity of the chosen parcellation scheme (). Widely used atlas-based approaches, including anatomical templates such as the Anatomical Automatic Labeling (AAL) atlas, which disregards functional boundaries, and functional atlases such as the seven-network parcellation (), which is derived from large cohorts of neurotypical adults, impose a fixed, , parcellation that may not fully capture the functional network topology specific to atypical developmental populations, such as children and adolescents with ASD. Consequently, standard atlases may lack sensitivity to subtle, cohort-specific variations in functional architecture.