Explainable machine learning reveals the challenges of predicting new-onset motor complications in Parkinson’s disease

Predicting new-onset motor fluctuations and levodopa-induced dyskinesias (LID) is crucial for optimizing Parkinson’s disease management. To establish a transparent prognostic framework, we applied explainable machine learning to real-world, multicentric clinical data from 247 patients to forecast the 3-year onset of these complications. Evaluated strictly on complication-free patients, the models achieved moderate predictive power (LID MCC = 0.28; fluctuations MCC = 0.32).

SHAP-based interpretability confirmed predictions aligned accurately with established clinical knowledge, driven primarily by levodopa duration and Levodopa Equivalent Daily Dose, with risk increasing significantly above a 300–400 mg threshold. Crucially, an ablation study revealed that excluding patients with pre-existing complications from training caused model sensitivity to collapse, demonstrating that the full spectrum of disease severity is essential for robust risk stratification. Ultimately, this rigorous methodological stress-test demonstrates that baseline clinical features alone yield limited absolute sensitivity, highlighting the necessity of integrating dynamic, longitudinal data to achieve clinical-grade individualized prediction.

The authors thank all the patients involved in the NeuroArtP3 project (NET-2018-12366666), as well as colleagues from the different institutions and departments that contributed to this initiative: Andrea Falini, Antonella Castellano, Andrea Rossi, Domenico Tortora, Costanza Parodi, Antonio Verrico, Federica Sabatini, Emilio Portaccio, Matteo Betti, Guido Pasquini, Filippo Gerli, Claudia Niccolai, Isabella Cama, Cristina Campi, Alessio Cirone, Sara Garbarino, Michele Piana, Donatella Ottaviani, Raffaella Di Giacopo, Ruggero Bacchin. This research was funded by the Italian Ministry of Health, grant number NET-2018-12366666 (NeuroArtP3). Data Science for Health, Fondazione Bruno Kessler, Trento, Italy Walter Endrizzi, Flavio Ragni, Monica Moroni, Stefano Bovo, Lorenzo Gios, Giuseppe Jurman & Venet Osmani Department of Cellular, Computational and Integrative Biology, University of Trento, Trento, Italy Neurology Unit, Azienda sanitaria universitaria integrata del Trentino (ASUIT), Ospedale Santa Chiara, Trento, Italy Nicole Campese, Chiara Longo, Bruno Giometto, Maria Chiara Malaguti & Ruggero Bacchin Azienda sanitaria universitaria integrata del Trentino (ASUIT), Ospedale Santa Maria del Carmine, Rovereto, Italy Chiara Longo, Maria Chiara Malaguti, Donatella Ottaviani & Raffaella Di Giacopo IRCCS Ospedale Policlinico San Martino, Genoa, Italy Antonio Uccelli, Alessio Cirone & Sara Garbarino Department of Neurosciences, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health, University of Genoa, Genoa, Italy Centro Interdipartimentale di Scienze Mediche (CISMed), Facoltà di Medicina e Chirurgia, Università di Trento, Trento, Italy Department of Biomedical Sciences, Humanitas University, Milan, Italy Digital Environment Research Institute, Queen Mary University of London, London, UK Neuroradiology Unit and CERMAC, IRCCS Ospedale San Raffaele, Milan, Italy Università Vita-Salute San Raffaele, Milan, Italy Neuroradiology Unit, IRCCS Istituto Giannina Gaslini, Genoa, Italy Andrea Rossi, Domenico Tortora & Costanza Parodi Department of Health Sciences (DISSAL), University of Genoa, Genoa, Italy Neuro-Oncology Unit, IRCCS Istituto Giannina Gaslini, Genoa, Italy UOSD Cell Factory, IRCCS Istituto Giannina Gaslini, Genoa, Italy Department of Neuroscience, Psychology, Drug Research and Child Health (NEUROFARBA), University of Florence, Florence, Italy IRCCS Don Carlo Gnocchi Foundation, Florence, Italy Guido Pasquini, Filippo Gerli & Claudia Niccolai Life Science Computational Laboratory (LISCOMP), IRCCS Ospedale Policlinico San Martino, Genoa, Italy Isabella Cama, Cristina Campi, Alessio Cirone, Sara Garbarino & Michele Piana Department of Mathematics (DIMA), University of Genoa, Genoa, Italy Isabella Cama, Cristina Campi & Michele Piana Donatella Ottaviani & Raffaella Di Giacopo The authors declare no competing interests.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it.

The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Endrizzi, W., Campese, N., Ragni, F. et al. Explainable machine learning reveals the challenges of predicting new-onset motor complications in Parkinson’s disease. npj Parkinsons Dis. (2026). https://doi.org/10.1038/s41531-026-01550-1 DOI: https://doi.org/10.1038/s41531-026-01550-1