Background Cognitive decline and incident dementia in ageing populations have been linked to brain parenchymal injury and the ALPS index (ALPS-I). Objective This investigation aimed to elucidate whether the baseline ALPS-I could predict incident dementia and cognitive decline in this population. Methods In total, 973 dementia-free participants from the Shunyi Study (mean age, 57 years; 37% male) received MRI between 2013 and 2016 to quantify the ALPS-I. The longitudinal relationships between the ALPS-I and cognitive deterioration in various cognitive areas were evaluated via linear mixed models. Cox proportional hazard models were utilized to explore the link between the index and incident dementia. Mediation assessments were carried out to identify the potential mediating effects of brain parenchymal injury on the link between the ALPS-I and cognition. Results The baseline ALPS-I predicted longitudinal changes in global cognition (Montreal Cognitive Assessment), language (verbal fluency test), visuospatial perception (block design subtest of Wechsler intelligence scale), and executive function (Trail Making Test). A lower score was markedly associated with a higher incident dementia risk. Mediation analysis revealed that fractional anisotropy mediated the associations between the ALPS-I and executive function (mediation effect: 21.9%) and visuospatial perception (mediation effect: 68.8%). The white matter hyperintensity fraction was found to mediate the link between the ALPS-I and global cognition (mediation effect: 55.0%). Conclusions This longitudinal evidence supports a link between the ALPS-I and cognitive degeneration. Furthermore, the link is mediated by subcortical parenchymal injury.
Li et al. (2026) studied this question.