Association of cholesterol and glycemic state biomarkers with phenotypic variation and Parkinson's disease progression: the Oxford Discovery cohort

Nicholas Kwan Ho Chiu, Michael A Lawton, Tanja Zerenner, Alireza Morovat, Jessica Welch, Jamil Razzaque, Michele Hu, Yoav Ben-Shlomo*

*Corresponding author for this work

Research output: Contribution to journalArticle (Academic Journal)peer-review

Abstract

Background
Parkinson's disease (PD) has marked phenotypic variability. Increased lipids have been suggested as being neuroprotective whilst hyperglycemia may increase α-synuclein aggregation.

Objective
We have tested whether high total cholesterol and high-density lipoprotein cholesterol (HDL-C) and low levels of fructosamine are associated with better PD phenotypes and predict less rapid progression

Methods
Non-fasting serum HDL-C, total cholesterol, and fructosamine were measured at baseline in 866 patients with early PD (median duration, 0.96; IQR, 0.43–1.98 years) from the Oxford Discovery cohort. These biomarkers were compared against our data-derived PD subtypes using multinomial logistic regression. We used multilevel models to predict longitudinal motor and non-motor outcomes (e.g., cognition, mood).

Results
HDL-C and total cholesterol differed across baseline PD phenotype clusters, with reduced levels associated with the most severe motor and non-motor phenotypes (psychological well-being, cognitive impairment, REM sleep behavior disorder, and daytime sleepiness). Higher HDL-C and total cholesterol, although the latter was attenuated after adjustment for statin use, were associated with better baseline activities of daily living (e.g., UPDRS-II score with 1 SD increase in HDL-C −0.74, 95%CI −1.22 to −0.26, p = 0.002) and non-motor features. Neither predicted the rate of motor or non-motor progression. Fructosamine levels were not associated with phenotypic variability or rate of disease progression.

Conclusions
Hypercholesterolemia was associated with a better motor/non-motor disease subtype and daily living impairment at presentation, but did not predict longitudinal change. Future research needs to determine if these associations are causally related or secondary to disease onset by examining prodromal subjects.
Original languageEnglish
Pages (from-to)522-530
Number of pages9
JournalJournal of Parkinson's Disease
Volume15
Issue number3
Early online date13 Apr 2025
DOIs
Publication statusPublished - 1 May 2025

Fingerprint

Dive into the research topics of 'Association of cholesterol and glycemic state biomarkers with phenotypic variation and Parkinson's disease progression: the Oxford Discovery cohort'. Together they form a unique fingerprint.

Cite this