Abstract
The increasing burden of dementia highlights the need for sensitive and ecologically relevant tools for early detection and intervention. Dual-task (DT) paradigms have emerged as promising approaches for assessing cognitive-motor alterations. This study aimed to map the 22-year evolution of DT research in cognitive aging using bibliometric analysis. We retrieved 1347 articles from the Web of Science Core Collection (January 2003-February 15, 2025) and analyzed them using CiteSpace. Co-citation, keyword burst, and collaboration network analysis were performed to identify research evolution, hotspots, knowledge structures, and international collaborations. Research showed a progressive shift from gait-focused behavioral assessment (2003-2012) to neuroimaging-based investigations (2013-2018), and subsequently toward artificial intelligence-driven, ecologically valid approaches integrating virtual reality and multimodal sensing (2019-2025). The dual-task cost emerged as a potential candidate digital biomarker associated with neural alterations and cognitive decline, while its complementary value to traditional cognitive assessment was highlighted. Dual-task training demonstrated potential benefits for improving functional outcomes and reducing fall risk; however, heterogeneity in treatment effects indicates the need for personalized intervention strategies. DT paradigms are evolving toward potential applications in precision gerontology. The proposed assessment-intervention continuum provides a conceptual framework for future research. Standardized protocols and prospective validation of DT-derived digital phenotypes are required before clinical implementation.