Type 2 diabetes (T2D) may contribute to the pathogenesis of PD and IR is one of the putative links between them. However, assessment of brain IR is problematic. Fludeoxyglucose (18F) positron emission tomography (FDG-PET) imaging can provide a surrogate marker for brain tissue glucose metabolism. Abnormal FDG-PET patterns have been described in PD patients, but never correlated to insulinemic state.

Brain FDG-PET scans were acquired using a PET/MR integrated scanner (Siemens Biograph mMR). Mean standardized uptake value (SUV) was measured from fused FDG-PET images using a quantification software (FusionQuant V1.0). Frontal, parietal, lateral temporal, and occipital lobes as well as the sensorimotor cortices, putamen, thalamus, locus coeruleus, and cerebellum were analyzed. A total of 84 circular regions of interest (ROIs) - 42 per hemisphere - were manually drawn. ROIs were 10-mm in diameter and 5-mm in height for each region, except for the locus coeruleus (LC) which was 6-mm in diameter and 3-mm in height. The reader ensured that ROIs did not overlap. Correlations between regional SUV mean, HbA1C, and MDS-UPDRS were calculated using linear regression analysis.

Thirty-seven non-diabetic PD patients (26M, 11F) with a mean age of 65.33±8.29 participated in this study. SUV mean in bilateral sensorimotor and temporal cortices, putamen, LC, and cerebellum in addition to the left parietal region showed a strong positive correlation with off-medication MDS-UPDRS part III scores (p<0.05). HbA1C had a negative correlation in all studied brain regions except left LC and right putamen (p<0.05).