Objective: To identity the effect of exercise on brain metabolism and energetics in PD using PET/MR imaging.

Background: Aerobic exercise slows PD progression as reflected clinically [1]. Mitochondrial dysfunction may play a role in the onset and progression of PD, and aerobic exercise may positively modulate mitochondrial dynamics [2]. PET/MR imaging can be used to obtain simultaneously cerebral metabolic rates of oxygen (CMRO₂) and glucose (CMR_glu); estimates of mitochondrial function can be inferred through the oxygen-to-glucose ratio.

Method: An ongoing study is performing dual-calibrated BOLD MRI and FDG PET imaging to obtain CMRO₂ and CMR_glu in PD habitual exercisers (PD-HE, n=4) and PD non-exercisers (PD-NE) before (n=6) and after (n=4) six-month stationary cycling. Constrained principal component analysis (CPCA) [3] is used to identify disease- and exercise-related spatial patterns of regional CMRO₂ and CMR_glu. CPCA first regresses Z, the spatially-normalized imaging measures for each subject, onto clinical variables, X, related to disease (disease duration (DD), UPDRS-III) and aerobic fitness (VO₂ max), i.e. Z ≈ XC, where C are the regression coefficients. Standard PCA is then run on XC to obtain spatial patterns rank-ordered by their explanation of the variance-of-interest contained in XC.

Results: Subject scores of the first CMR_glu CPCA pattern are significantly higher in the PD-NE pre-intervention compared to PD-HE (p<0.01, [figure 1]) and PD-NE post intervention (p<0.05). No significant difference was found between PD-HE and PD-NE after intervention. This pattern has significant overlap with the PD disease-related and PD cognition-related [4] patterns ([figure 2], PDRP: ρ=0.57; PDCP: ρ=0.51), known to correlate with disease severity. Subject scores correlate positively with DD (R²=0.36, p<0.05), negatively with VO₂ max (R²=0.70, p<0.001), and have no correlation with UPDRS-III, indicating most of the CPCA pattern variance is governed by exercise-induced changes. A similar initial CMRO₂ analysis did not yield group differences.

Conclusion: A pattern of glucose metabolism positively related to DD and negatively related to aerobic fitness was found. Group differences of pattern expression between a small cohort of PD non-exercisers before and after aerobic exercise suggests causality between exercise and disease progression. More subjects will be added as they complete the intervention. Joint CMR_glu and CMRO₂ pattern analysis will follow.

References: [1] Tsukita K, Sakamaki-Tsukita H, Takahashi R. Long-term Effect of Regular Physical Activity and Exercise Habits in Patients With Early Parkinson Disease. Neurology. 2022 Feb 22;98(8):e859-e871. doi: 10.1212/WNL.0000000000013218. Epub 2022 Jan 12. PMID: 35022304; PMCID: PMC8883509.
[2] Chuang CS, Chang JC, Cheng FC, Liu KH, Su HL, Liu CS. Modulation of mitochondrial dynamics by treadmill training to improve gait and mitochondrial deficiency in a rat model of Parkinson’s disease. Life Sci. 2017 Dec 15;191:236-244. doi: 10.1016/j.lfs.2017.10.003. Epub 2017 Oct 3. PMID: 28986095.
[3] Takane Y & Shibayama, T. Principal component analysis with external information on both subjects and variables. Psychometrika. Mar 1991; 56(1):97–120. https://doi.org/10.1007/BF02294589.
[4] Eidelberg D. Metabolic brain networks in neurodegenerative disorders: a functional imaging approach. Trends Neurosci. 2009 Oct;32(10):548-57. doi: 10.1016/j.tins.2009.06.003. Epub 2009 Sep 16. PMID: 19765835; PMCID: PMC2782537.

« Back to 2022 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/effect-of-exercise-on-brain-metabolism-and-energetics-in-pd/