Rethinking the role of animal models in adaptive deep brain stimulation: From feasibility to mechanistic and disease-modifying insights

BACKGROUND: Adaptive deep brain stimulation (aDBS) has emerged as an advanced neuromodulation approach capable of dynamically adjusting stimulation parameters based on real-time biomarkers, thereby overcoming the limitations of conventional continuous DBS (cDBS). While early animal studies were instrumental in establishing the feasibility and symptomatic efficacy of aDBS, its availability in humans poses the question of whether animal studies are still needed in this field. OBJECTIVE: This review aims to critically reassess the role of animal models in the development and future evolution of aDBS, with particular emphasis on their contribution to mechanistic insights, potential disease-modifying effects, biomarker identification, and the integration of machine learning for next-generation adaptive neuromodulation systems. FINDINGS: Animal studies have provided the foundational evidence for closed-loop neuromodulation, demonstrating the feasibility of real-time sensing-stimulation systems and the initial superiority of aDBS over cDBS in modulating motor symptoms. Preclinical models have enabled the identification and causal validation of electrophysiological and neurochemical biomarkers while also serving as indispensable platforms for hardware innovation, causal circuit interrogation using optogenetics, and validation of machine learning-based adaptive control algorithms. Moreover, animal models offer a unique platform for investigating disease progression, neural plasticity, and stage-dependent effects of stimulation, which remain difficult to assess in human studies. In parallel, machine learning approaches have increasingly leveraged preclinical and clinical data to decode physiological states, optimize stimulation parameters, and dynamically adapt therapy, thus supporting the development of personalized aDBS systems. CONCLUSION: The role of animal models in aDBS research is undergoing a conceptual transition, from demonstrating feasibility and symptomatic efficacy to addressing mechanistic and disease-modifying questions. By enabling causal investigation of neural circuits, hardware development, biomarker validation, and integration with machine learning frameworks, preclinical research remains essential for advancing aDBS toward more precise, adaptive, and potentially disease-modifying neuromodulation strategies.