Gut bacteria may impact cognitive status in Parkinson’s, study finds
Motor symptom severity doesn't appear to be substantially affected
Written by |
- A new study reveals that gut bacteria variations are linked to cognitive issues in Parkinson's disease, rather than motor symptoms.
- Machine learning models accurately distinguished Parkinson's patients from controls using gut and blood markers, predicting cognitive scores.
- Further large-scale investigations are necessary to validate these findings.
Variations in gut bacteria do not appear to substantially affect motor symptom severity in Parkinson’s disease, but they may be linked to cognitive issues in Parkinson’s, according to a new study.
“Gut-derived features, including both [gut bacteria themselves and the molecules that are made by gut bacteria], are more closely linked to cognitive function than with motor impairment in [Parkinson’s],” researchers wrote. They called for further investigation into these associations.
The study, “Multi-omics of host–gut metabolic crosstalk in Parkinson’s disease,” was published in npj Parkinson’s Disease.
Models distinguished between Parkinson’s patients, controls
The human gut is home to billions of bacteria, which influence human health in ways that are not fully understood. A growing body of research has indicated that the activity of gut bacteria is disrupted in people with Parkinson’s — but the details remain incompletely understood, and it’s unclear how differences in gut bacteria may be associated with specific manifestations of the disease. Parkinson’s is defined by motor symptoms such as slowness and tremor, but nonmotor symptoms like cognitive problems are also common.
In this study, scientists in Malaysia collected both stool and blood samples from 67 people with Parkinson’s and 64 people without the disease, who served as controls. The researchers analyzed blood biomarkers alongside the composition of gut bacteria, as well as their metabolites (biologically active molecules produced by gut bacteria).
Using all this data, the researchers constructed machine learning models. Machine learning is a type of artificial intelligence that involves feeding a computer a large data set, along with detailed mathematical models it uses to identify patterns in the data. For their studies, the researchers fed all the gut and blood data into computers, and the machine learning systems would then determine which factor(s) to use when interpreting the data.
The researchers found that machine learning models based on both gut and blood markers could distinguish people with Parkinson’s from controls with very high accuracy — the models correctly identified more than 97% of Parkinson’s patients and more than 90% of controls. Notably, these models incorporated data from both the gut and the blood — consistent with findings from other studies showing that Parkinson’s disrupts many biological systems throughout the body.
Researchers call for additional studies
The researchers then used the same approach to see if they could predict the severity of Parkinson’s symptoms, as measured using standardized tests. For motor symptoms, however, the machine learning models showed little-to-no accuracy, suggesting that the gut and blood markers evaluated in this study have little impact on motor symptoms.
For cognitive scores, on the other hand, the machine learning systems accurately predicted outcomes. The accuracy was modest, but the result was statistically meaningful. Notably, almost all the factors retained by the machine learning system were gut data, including counts of gut bacteria and certain specific bacterial metabolites.
These findings suggest that gut bacteria may influence cognitive status in Parkinson’s, at least to a greater extent than motor symptoms. The researchers stressed, however, that it’s not possible to draw firm cause-and-effect conclusions from these data. The team also noted that this analysis was limited to a fairly small number of patients, highlighting the need for additional studies to validate and expand upon these results.
Leave a comment
Fill in the required fields to post. Your email address will not be published.