Project Summary/Abstract Early detection of Alzheimer's Disease and Related Dementias (ADRD) is essential for implementing interventions that can slow cognitive decline, benefiting individuals, caregivers, and society. Gait speed emerges as a powerful predictor of ADRD, especially in the early stages of Mild Cognitive Impairment (MCI), often preceding cognitive symptoms. However, the underlying mechanisms between gait speed and cognitive decline remain unclear, limiting the specificity of gait as an ADRD predictor. This research project aims to investigate gait deficits during tasks requiring cognitive input into walking control (Aim 1) and to unveil the neural processes behind these cognitive contributions (Aim 2). The study focuses on two crucial gait markers: locomotor learning and attentional need for gait control. Locomotor learning involves encoding and retrieving walking motor memories. Attentional need for gait control indicates the requirement for explicit attentional resources, primarily from the prefrontal cortex (PFC), during walking, often assessed through dual-tasking. Both markers are vital for the mobility of older adults in the community, yet their neural mechanisms remain unclear. We will specifically determine the contribution of basal ganglia dysfunction on these gait markers. Growing evidence suggests that BG pathology, characterized by factors like iron deposition, white matter hyperintensities (WMH), and compromised BG intra-connectivity, negatively affects mobility. Furthermore, we investigate the specific contribution of BG's role in compensatory mechanisms for attentional control during mobility, particularly through its connectivity with executive control networks (ECN), including the PFC. Our central hypothesis is that individuals at high risk for ADRD (i.e., those with MCI) have diminished locomotor learning and higher attentional need for gait control (Aim 1) due to BG pathology and weak BG-ECN inter-connectivity (Aim 2). Preliminary results support this hypothesis, showing lower locomotor learning and higher attentional need for gait control in MCI compared to age- and sex-matched controls. Interestingly, these gait markers prove to be more sensitive to MCI than gait speed itself. Thus, we anticipate that BG pathology and BG-ECN inter-connectivity, which are related to slow walking, will be strongly associated with the performance of locomotor learning and attentional need for gait control. The expected findings from this research hold the potential to yield significant insights into both behavioral and mechanistic deficits in cognitive contributions to gait in individuals with MCI. Such insights can greatly enhance the specificity and validity of novel gait measures as preclinical indicators for risk of ADRD.