The question of whether the aforementioned pathological changes and their distribution actually follow a distinct course of progression is central to understanding the pathophysiology of LB-related dementia. Braak et al.  meticulously studied a series of brains from patients with clinical PD, incidental LBs and LNs, and controls. Working under the assumption that there was a coherent progression of disease in these cases, they organized the samples in a sequential manner, and subsequently developed a staging system for LRP in this patient population. The staging system that emerged proposed a caudo-rostral progression of pathological change, which began in the dorsal motor nucleus of the glossopharyngeal and vagus nerves, as well as the olfactory nucleus, and ascended to mesocortical and then neocortical structures. The first two stages are related to incidental Lewy body disease, the third and fourth stages to motor symptoms, and the later stages (five and six) to concurrent cognitive dysfunction. Later studies by the same authors were able to correlate cognitive dysfunction with this staging system of LRP in patients with PD .
Multiple counter-examples to this staging system have since been proposed, whereby sparing of caudal regions occurs with more rostral involvement, which implies that other pathological changes, apart from Lewy body pathology, may be contributing . Further evidence of this notion is that the existence of LRP—to the extent that it meets the pathological criteria for diagnosis of DLB—may occur in the absence of a clinical dementing syndrome . More recent revisions of this staging system propose that the early sites of LRP are the olfactory bulb and enteric plexus of the stomach, suggesting that a neurotrophic pathogen enters through the nasal or gastric mucosa and proceeds to the central nervous system via anterograde progression to temporal structures or through retrograde transport via peripheral ganglia, respectively .
A competing theory, which not only challenges Braak’s proposed theory of disease progression but is also a unitary theory of pathophysiology for Lewy body diseases, identifies three distinct clinico-pathological subgroups. The first consists of a younger group of patients with a predominantly early motor phenotype, which is consistent with caudo-rostral progression of pathological change. The second subgroup consists of patients who suffered a more aggressive course of disease, with an early onset of dementia and concomitant early neocortical involvement. The third subgroup identified represents patients with later disease onset, and a subsequently shorter course, with a greater burden of pathological change that also tended to be more heterogeneous. This theory, and the data that support it, implies that a caudo-rostral progression of pathological change is only applicable to a subgroup of patients, and does not explain progression of disease in other subgroups, particularly those who follow a dementia-dominant clinical course .
In a critical evaluation Jellinger  questioned the predictive value of the current staging system, noting that there is little correlation between Braak LB stages and the neuronal cell loss as well as the clinical severity of dementia. Others have suggested a more informative semiquantitative and quantitative assessment of Lewy body pathology over the existing topographical staging for a better correlation of the cognitive impairment and the underlying pathology . A unified staging system was proposed during 2009 by the BrainNet Europe Consortium , which combined the brainstem, limbic, and neocortical stages as well as the Braak staging. As a result of a study involving 22 observers analysing 31 autopsy cases, this system incorporates a modified Braak staging  and a modified McKeith typing . The staging process involves an assessment of 10 different areas of the brain—medulla, pons, midbrain, basal forebrain, striatum, hippocampus, cingulate gyri, temporal cortex, frontal cortex, and parietal cortex—for the presence or absence of a-synuclein immunoreactivity, LBs, and LNs. The pattern of distribution to assess the typical/atypical nature and presence or absence of amygdala-predominant type to assess the severity of LRP is also taken into consideration .