Identifying physicochemical predictors that connect protein aggregation to clinical phenotypes remains challenging. Here, we examine whether simulated aggregation descriptors of the intrinsically disordered protein α-synuclein (αS) relate to the clinical age of onset in Parkinson's disease. Using a multiscale simulation framework, we combine coarse-grained (CG) models to quantify the kinetics of early dimerization with atomistic simulations to estimate dimer interaction energies, as well as CG thermodynamic protofilament binding free energies complemented by atomistic calculations. These analyses were performed for wild-type αS and five familial mutants, considering both homo- and heterodimeric species. The time scale of WT-mutant heterodimer formation emerges as a statistically informative predictor of age of onset, whereas homodimer formation kinetics and heterodimer protofilament binding affinities fail miserably. In contrast, homodimer protofilament binding free energies display only moderate correlation, suggesting that late-stage fibril stability alone does not determine disease initiation. Overall, the results are consistent with a two-stage aggregation process in which early oligomeric dynamics may contribute to clinically relevant aggregation behavior.