A hypothetical model for biomarkers/clinical stages in neurodegenerative diseases. The model involves a two-step causal pathway in the progression of neurodegenerative diseases: 1) a causal relationship of pathogenic biomarkers (PB) to a neurodegenerative biomarker (NB); and 2) a causal relationship of the NB to the symptomatic/clinical stages. As evidenced by the current study, the PB can interpret the kinetics of the NB as shown in the left lower graph (model 1, risk-based stochastic kinetic model). The heterogeneity in time for converting from a mild to a more severe clinical stage can be explained by the kinetics of the NB (model 2, neurodegeneration-symptomatic model). Optimal biomarkers for PB or NB may differ across different kinds of neurodegenerative diseases. For example, in glaucoma, the PB may be IOP and the NB may be DTI or structural MRI scans, as evidenced by the current study, whereas in Alzheimer’s disease , PBs may be low Aβ1-42 and high p-tau in cerebrospinal fluid (CSF), APOE genotype, PIB-PET and FDG-PET scans, while NBs may be structural MRI scans DTI findings; in Parkinson’s disease , PBs may be α-synuclein in CSF, sympathetic denervation in the heart (MIBG-SPECT), and LRRK2 gene mutation, while NBs may be F-DOPA PET and DaTSCAN (β-CIT) scans. From a therapeutic point of view, the earlier phase of the model may be the optimum time window for neuroprotection (blue arrow in left graph) and the later phase may be optimal for neuro-regeneration and neuroplastic therapies (blue arrow in right graph).