Parkinson’s disease (PD) is a progressive disorder of the nervous system that affects movement by causing shaking, stiffness, and difficulty with walking, balance, and coordination. PD is the second most common chronic progressive neurodegenerative disorder in the elderly after Alzheimer’s disease, affecting 1%–2% of individuals ages 65 years and older worldwide. There are currently no blood or laboratory tests to diagnose non-genetic cases of PD and it is mostly diagnosed based on a person’s medical history and a neurological examination. Recently, machine learning (ML) has been used to analyse the medical and laboratory history of the patient and predict the probability of PD. If ML technology develops an ability to predict PD with reasonable accuracy before symptoms develop, this will be a significant leap forward for the early diagnosis and management of PD.

According to a study conducted by J Diana Zhang and colleagues, published in ACS Central Science Journal in May 2023, the ML tool can predict PD up to 15 years before a clinical diagnosis by analysing chemicals in the blood. Researchers also claimed that PD can be predicted with up to 96% accuracy. The study was a collaboration between Boston University and Sydney University, in which they analysed blood samples of 39 healthy individuals from the registry of Spanish European Prospective Investigation into Cancer and Nutrition. They used a tool called CRANK-MS (classification and ranking analysis using neural network-generated knowledge from mass spectrometry) to analyse detailed information about metabolites present in the blood. By analysing fluctuation in PD-specific biomarkers such as triterpenoids, diacylglycerols, and a poly-fluoroalkyl that predate clinical PD diagnosis, early disease prediction was possible.

PD is a significant health burden, and GlobalData epidemiologists forecast that there are currently more than 2.4 million diagnosed prevalent cases of PD in the seven major pharmaceutical markets (7MM: US, France, Germany, Italy, Spain, UK, and Japan). In the 7MM combined, the diagnosed prevalent cases of PD are expected to increase to 2.9 million cases in 2029 at an annual growth rate of 2.30%. The increase in the number of cases can be attributed to the projected increase in the population, as the rates for PD have remained unchanged over the previous decade, based on historical data analysed by GlobalData.

PD is difficult to diagnose early, especially in the preclinical stages. The finding of this study is significant as early diagnosis using ML tools will help in better treatment and improve the quality of life for patients. Further research on large populations is needed to validate the findings of this ML research.