Nutritional Intake and Gut Microbiome Composition Predict Parkinson’s Disease Original paper

What was studied?

This study investigated whether integrating gut microbiome composition and nutritional intake data could predict Parkinson’s disease (PD) status, aiming to establish a robust, clinically relevant biomarker signature. The researchers performed a cross-sectional analysis of GM diversity and composition using 16S rRNA gene amplicon sequencing on stool samples, pairing these results with detailed dietary assessments and clinical parameters. Two predictive models (Random Forest and support-vector machine) were constructed, incorporating both microbial taxonomic profiles and macronutrient intake, to determine their combined utility in discriminating PD from controls.

Who was studied?

The study enrolled 103 patients with clinically diagnosed idiopathic Parkinson’s disease and 81 household controls from Sydney, Australia. PD participants were recruited from specialist neurology clinics and met strict diagnostic criteria, while HCs were cohabitating relatives or spouses with similar dietary habits and without clinical signs of PD. Exclusion criteria included secondary Parkinsonism, tube feeding, inability to complete questionnaires, significant cognitive impairment, and recent antibiotic or probiotic use. Both groups underwent extensive clinical, dietary, and laboratory assessments, ensuring well-characterized phenotyping.

Most important findings

The study identified significant differences in gut microbiome composition between PD patients and HCs, despite no significant difference in alpha diversity (species richness and evenness). Beta diversity (overall compositional dissimilarity) was significantly higher in PD, indicating a disease-related shift in GM structure. At the family level, Lactobacillaceae was notably overrepresented in PD (2.7-fold increase), with similar enrichment observed at the order (Lactobacillales) and genus levels (e.g., Bifidobacterium, Butyricimonas, [Ruminococcus] gnavus group). Conversely, genera such as Butyricicoccus, Fusicatenibacter, Lachnospiraceae ND3007 group, and Agathobacter were underrepresented in PD. Several of these taxa are known short-chain fatty acid (SCFAs) producers, suggesting a potential link between their depletion and PD pathogenesis.

Correlation analyses revealed that multiple clinical features—including constipation severity, physical activity, and certain pharmacological therapies (levodopa, COMT inhibitors)—were significantly associated with GM beta diversity. Notably, device-assisted therapies (like levodopa-carbidopa intestinal gel) correlated with increased Enterococcus and Klebsiella, both relevant to levodopa metabolism in the gut. Incorporating dietary macronutrient data, particularly carbohydrate intake as a percentage of total energy, improved the predictive accuracy of the Random Forest model for PD (AUC increased from 0.71 to 0.74 at the genus level). This two-stage model, integrating both clinical and microbial variables, outperformed models based solely on microbiome or diet data.

Key implications

This study underscores the importance of considering both microbiome composition and nutritional intake in developing predictive biomarkers for Parkinson’s disease. The identification of a PD-specific gut microbial signature—characterized by increased Lactobacillaceae and Bifidobacterium and decreased SCFA-producing genera—offers valuable candidates for a microbiome signature database and potential clinical biomarker development. The findings also highlight the strong influence of clinical features (particularly constipation, physical activity, and pharmacotherapy) on the gut ecosystem in PD. Importantly, dietary factors, especially carbohydrate intake, modulate these microbiome signatures and enhance disease prediction. These results support integrative, multivariate biomarker models and suggest that dietary and microbial interventions could be clinically significant for PD risk assessment, diagnosis, and management. Further large-scale, longitudinal studies are warranted to validate these findings and elucidate causality.

Citation

Lubomski M, Xu X, Holmes AJ, Muller S, Yang JYH, Davis RL, Sue CM. Nutritional Intake and Gut Microbiome Composition Predict Parkinson’s Disease. Front Aging Neurosci. 2022;14:881872. doi:10.3389/fnagi.2022.881872