Exercise dose architecture as a precision-informed developmental framework in autism spectrum disorder: a narrative review of mechanisms, sex- and development-related heterogeneity, and clinical translation.

Autism spectrum disorder (ASD) is a heterogeneous neurodevelopmental condition characterized not only by differences in social communication and restricted or repetitive behaviors, but also by variable profiles of motor function, sleep, cognition, sensory processing, and adaptive participation. Exercise and structured physical activity have emerged as promising non-pharmacological interventions in ASD, with growing evidence supporting benefits across multiple outcome domains, including executive function, stereotyped behaviors, sleep, motor competence, social communication, and quality of life. However, this field is constrained by fragmented concepts, inconsistent dose documentation, lack of mechanistic integration, and inadequate clinical translation. Exercise is often treated as a generic intervention rather than as a well-defined and physiologically meaningful exposure. This narrative review proposes a precision-informed framework in which exercise in ASD is conceptualized as a structured developmental exposure whose potential therapeutic value may depend on dose architecture rather than modality labels alone. Beyond traditional descriptors such as frequency, intensity, time, and type, exercise dose in ASD should be understood as a multidimensional construct that also includes motor complexity, cognitive demand, social embeddedness, sensory load, predictability, progression rate, and adherence burden. We synthesize ASD-specific clinical evidence together with indirect evidence from broader exercise neuroscience and animal models, while explicitly distinguishing established clinical findings from candidate or hypothesis-generating mechanisms. We also examine how exercise responses may differ according to developmental stage, pubertal changes, gender-associated traits, and the lack of female representation in existing evidence. Overall, we investigate how efficacy findings can inform individualized exercise plans, mechanism-linked trials, and scalable interventions in real-world settings. Taken together, this review offers a conceptual and translational framework for future ASD exercise research, emphasizing standardized dose reporting, evidence-level distinctions, developmental stratification, safety monitoring, and clinically meaningful outcomes.