Functional and structural connectome-based predictive modelling of balance in elderly adults

Balance control is fundamental to the quality of life among older adults, yet its neural underpinnings remain only partially understood. Despite advances in neuroimaging techniques, the neural correlates of balance are often examined at a regional level and typically restricted to either the functional or structural connectome alone. In this study, we employed connectome-based predictive modelling (CPM) for a large-scale discovery of brain connections predictive of individual balance abilities using both structural and functional connectomes in a cohort of 54 older adults. The test-retest reliability and specificity of the constructed models was evaluated using repeated-measurement data and strength performance data. Our results show that both structural and functional connectomes can successfully predict balance performance on an unstable device measured using mean sway area. A comprehensive system, encompassing motor-subcortical connections, medial-frontal and fronto-parietal networks emerged from both connectome types as consistent predictors of balance. Notably, connections with visual networks uniquely contributed to prediction in the structural but not in the functional connectomes. Structural connectomes also showed better prediction performance and test-retest reliability compared to functional connectomes. The specificity of constructed models was validated using strength performance data. In summary, our study shows that structural and functional connectomes are strong predictors of motor control abilities in challenging conditions in the elderly highlighting their interdependency and complementary roles in balance control.