Global prostate cancer risk associated with microplastic exposure: a statistical and machine learning analysis
TL;DR
The findings support the need to integrate exposure pathways, biomonitoring indicators, and biological-response markers when studying microplastic-related cancer risk, but are limited by its ecological design, cross-sectional structure, and small sample size.
Abstract
Introduction Prostate cancer is one of the most commonly diagnosed malignancies among men worldwide, with higher reported incidence in many high-income countries. Environmental factors are receiving increasing attention as potential contributors to cancer development. Microplastics, which are found in air, water, food, and personal care items, are one possible risk factor. Methodology Data from 22 nations were investigated to examine whether an association exists between exposure to microplastics and the rate of prostate cancer. Data on exposure were combined from several sources, such as stool particles, breathed air, drinking water, seafood intake, and personal care products. Statistical and machine learning methods, such as K-means clustering, principal component analysis, and random forest modeling, were applied to find the most important exposure variables linked to cancer risk. Results Stool microplastic concentrations and heavy metal burden showed the strongest model-based associations with prostate cancer incidence. Countries with higher external exposure indicators did not consistently show higher reported prostate cancer incidence. This pattern suggests that external exposure metrics alone may be insufficient to explain country-level variation. Internal retention and tissue-response pathways remain plausible hypotheses, but they require direct validation using individual-level and tissue-based data. Discussion The findings support the need to integrate exposure pathways, biomonitoring indicators, and biological-response markers when studying microplastic-related cancer risk. However, this study was limited by its ecological design, cross-sectional structure, and small sample size of 22 countries. Therefore, the results should be interpreted as exploratory and hypothesis-generating rather than causal. Further longitudinal and individual-level studies are required to validate these associations.