Neuromelanin imaging is an emerging biomarker for PD as it captures degeneration of the midbrain, a process which is associated with the motor symptoms of the disease. Currently, it is unknown whether this degeneration also contributes to cognitive dysfunction in PD beyond dysfunction associated with fronto-subcortical systems, as quantitative examination of substantia nigra (SN) degeneration could not be studied until recently.

In the current study, we examine whether neuromelanin signal is associated with broader cognitive dysfunction in PD patients with varying degrees of cognitive impairment: PD with normal cognition (PD-NC), PD with mild cognitive impairment (PD-MCI), and healthy controls (HC).

11 PD-NC, 16 PD-MCI and 14 age and sex-matched healthy controls (HC) participated in the study. PD participants were diagnosed with MCI based on the Movement Disorders Society Task force, Level II assessment (comprehensive assessment). In addition, all participants underwent an MRI scan that included a T1-weighted sequence and a neuromelanin-sensitive (NM-MRI) sequence. Contrast-to-noise-ratio of the substantia nigra pars compacta (SNc) was calculated and a distribution-corrected z-score was used to identify the number of extrema voxels for each individual, suggestive of the number of voxels that have exhibited significant degeneration (extrema_count). An analysis of covariance (ANCOVA) was used to evaluate group differences between HC, PD-NC, and PD-MCI in the extrema_count accounting for age, sex, and education. A multiple regression for each cognitive variable with extrema_count as the dependent variable adjusting for age, sex, and education were conducted.

A significant main effect of group (F(2, 33) = 33.548 ; p < 0.001) indicated that PD-NC (21.55 ± 12.57) and PD-MCI (43.64 ± 32.84) patients exhibited significantly greater extrema_counts relative to HC (3.36 ± 3.61; both p < 0.001). Regression results indicated that higher extrema_counts were associated with worse cognitive performance across cognitive domains, including working memory (Digit Span Backward; R2 = .357, F(1,20) = 5.295, p = .032), (Hopkins Verbal Learning Test - Revised, Trials 1 to 3; R2 = .432, F(1,20) = 5.819, p = .026).

PD patients (PD-NC and PD-MCI) exhibited decreased neuromelanin in the SNc relative to healthy controls, confirming the ability of the NM-MRI sequence to differentiate PD from HC. There was no significant difference in SNc neuromelanin levels between PD-NC patients and PD-MCI patients, however, this is likely due to the small sample size. In addition, significant SNc degeneration was associated with worse cognitive performances in tasks associated with working memory and executive functioning. These results warrant further examination of the role of SN in PD patients with differing levels of cognitive impairment.