Exercise-Induced Metaplasticity: Breaking the Addiction Cycle in Methamphetamine Use Disorder.

Sense of agency (SoA) is the subjective experience of having control over one’s own actions and influencing the external world through intentional effort []. The SoA is critical for distinguishing one’s actions from those of others []. It also functions as an internal reward, promoting movement selection and execution [], and can influence decision-making related to behavioral change and goal achievement [,]. Multiple task-level and social context factors influence SoA, including the fluency of action selection and the extent of choice or coercion [-]. As many behavioral interventions manipulate these determinants, such as autonomy and choice support, feedback, and social collaboration, the SoA can function as a proximate, reactive indicator of mechanisms related to perceived control and engagement. Therefore, we consider the SoA as an indicator for evaluating the acceptability and effectiveness of interventions, as detailed below. SoA is shaped by autonomy vs coercion, environmental regularities, and multisensory congruence and feedback. Discrepancies in these signals have been shown to reduce perceived control [-]. Neurocomputational models suggest that SoA is caused by hierarchical inference processes such as active inference and predictive coding and that disruptions in these processes are associated with neuropsychiatric and functional disorders []. Taken together, these accounts suggest 2 related but distinct layers: an explicit, reportable contextual appraisal shaped by autonomy or coercion, and an implicit multisensory sensorimotor process based on timing and sensory match. In this study, we frame SoA as a 2-layered construct in which explicit appraisals (integrating background information about the environment, internal knowledge about the world, and background beliefs) and implicit sensorimotor coupling shape behavior via motivational and affective routes [-]. Here, we used the Sense of Agency Scale (SoAS) to index general explicit sense of agency (a general sense of control, regardless of the situation; hereafter referred to as GenExp-SoA) [,] and outcome (temporal) binding in the Libet clock task as an indirect marker of local, implicit sensorimotor coupling (a momentary, task-dependent, nonconscious link between action and outcome; LocalImp-SoA hereafter) [,]. Hereafter, we refer to these as the 2 forms of the SoA. As these 2 measures target different layers of agency, they may show dissociable dynamics in free-living settings. Therefore, assessing both provides complementary information. With the increasing use of digital devices such as smartphones in health promotion and therapeutic interventions [,], many digital therapeutic software programs and health care apps have been developed [-]. Additionally, the SoA can act as an inner psychological strength that protects individuals from the harmful effects of stress and adversity by helping to maintain motivation, purposeful action, and a sense of control [,]. Affective processing may also mediate the relationship between SoA and action regulation, influencing health behaviors []. In this context, SoA is a relevant construct that can be measured and potentially targeted through digital interventions. We assessed both layers to better capture how SoA changes in real-life digital contexts. Given this operationalization, modeling SoA dynamics can inform the timing and manner of support delivery. In particular, when the explicit agency is low or declining, interventions can shift toward autonomous support, and when it is high or increasing, task difficulty can increase gradually. When the implicit agency is unstable or declining, the task and notification load can be temporarily reduced, and the reliance on passive sensing can increase until stability returns. Frequent active assessments are burdensome in real-life settings; thus, passively sensed behavioral signals (eg, activity, phone use, heart rate variability [HRV], and location variability) can serve as low-burden proxies for tracking SoA dynamics between periodic reanchoring and direct assessments, and preliminary evidence indicates that they encode relevant aspects of SoA dynamics []. When validated rigorously and reanchored periodically to direct SoA assessments, such models could support just-in-time adaptive interventions that are contingent on agency [].