Motor phenotype as a risk factor for dementia
Although a number of risk factors have been reported for incident dementia in PD, three have been consistently found in almost all studies: advanced age, motor phenotype, and baseline cognitive performance [30, 41-47]. Advanced age is the single most significant risk factor, in both cross-sectional and prospective studies, although older age and more severe motor symptoms are synergistic in predicting dementia . When a cohort of patients was divided into four groups by dichotomizing around their median age and UPDRS motor scores, the group with older age and severe disease had a 12-fold increased risk for incident dementia compared with the younger patients with mild disease . As younger patients with greater disease severity and older patients with less severe motor symptoms did not show a significantly increased risk, a combined effect of age and disease severity was assumed. In addition to the overall severity of extrapyramidal features influencing the likelihood of developing PD-D, a non-TD motor phenotype also predicts a greater risk for incident dementia. In one prospective study of 40 PD patients, 25% of 16 patients with the PIGD phenotype developed dementia over 2 years, compared with none of 18 patients with the TD phenotype or six with indeterminate phenotype .
As mentioned, the predominant motor phenotype may not necessarily remain stable throughout the disease course. Thus a patient may evolve from a TD to a PIGD phenotype, presumably reflecting the underlying pathological progression. This, in turn, may influence the risk of developing dementia. In a Norwegian community-based sample of 171 PD patients without dementia followed prospectively over 8 years, logistic regression was used to analyse the relationship between subtype of parkinsonism and dementia . The transition from the TD to PIGD subtype was associated with a more than three-fold increase in the rate of decline measured by the Mini-Mental State Examination (MMSE). Compared with patients with a persistent TD or indeterminate subtype, the odds ratio for dementia was 56.7 [95% confidence interval (CI): 4.0-808] for patients changing from a TD or indeterminate subtype to the PIGD subtype, and 80.0 (95% CI: 4.6-1400) for patients with a persistent PIGD subtype. Furthermore, patients with a TD subtype at baseline did not develop dementia until they developed a PIGD motor subtype, and dementia did not occur among patients with persistent tremor dominance. Although the CIs were wide for these estimates, reflecting relatively small numbers, this work emphasizes the importance of current motor phenotype and associated risk of dementia.
In a community-based study performed in Cambridgeshire, UK, incident cases of PD were recruited, thereby removing much of the bias associated with selective mortality in prevalence cohorts [30, 47]. Bivariate comparisons of baseline demographic, clinical, and neuropsychological variables versus rate of cognitive decline showed that, in addition to older age, a non-TD motor phenotype, a higher UPDRS motor score, and below average performance on tests of semantic fluency, pentagon copying, spatial recognition memory, and the Tower of London test were associated with a more rapid rate of cognitive decline. Multivariate analysis revealed that a non-TD motor phenotype together with poor semantic fluency and inaccurate pentagon copying were the most significant predictors of cognitive decline, independent of age. Patients with a non-TD phenotype were 4.1 times more likely to develop dementia than TD patients.
Axial symptoms and PD-D may have an overlapping pathogenesis, with distinct loci of dysfunction different from those underlying TD PD. Specifically, the postural instability of PD tends to be refractory to dopaminergic therapy, and may relate to the loss of subcortical neurons within the cholinergic system which also plays an important role in the cognitive and neuropsychiatric symptoms of PD-D . In addition, recent work  that examined cerebrospinal fluid (CSF) in de novo PD has suggested that neurotransmitter dysfunction may not be the only overlapping pathophysiological process between cognition and motor phenotype. Alves et al.  looked at the CSF of 99 untreated patients with PD and 46 age-matched controls. Those with PD and a PIGD phenotype (n = 39) had significantly lower amyloid-beta markers than those who had TD disease (n = 60); there were no significant differences in CSF marker levels between the TD PD group and controls. Furthermore, multivariate linear regression models with adjustments for age, white matter hyperintensities, and cognition demonstrated significant associations of CSF amyloid-beta with PIGD score and lower limb bradykinesia, highlighting that abnormal deposition of amyloid-beta may contribute to motoric disease beyond its effect on cognition.