From disinfectant to neurodegeneration: Integrating machine learning and mendelian randomization reveals triclosan as a novel environmental risk factor for Alzheimer’s disease

Comprehensive Summary

Cheng et al. investigated the potential association between triclosan (TCS), a common antimicrobial agent, and Alzheimer’s disease (AD). Machine learning models (LASSO, SVM-RFE, RF) were used to screen and validate key TCS-AD genes. Three genes—APP, SLC6A3, and DRD2—emerged as central to this relationship, with DRD2 identified as the most robust core gene. Functional analyses suggested that TCS downregulates these genes, which are critical for neuroplasticity, cognition, and dopaminergic signaling. A logistic regression model incorporating these genes was developed to predict AD risk. However, the study was limited to computational predictions using public datasets and did not account for complex human exposure scenarios. The authors recommend future integration of epigenomic and proteomic data to strengthen biological plausibility.

Outcomes and Implications

This study aims to build understanding of TCS-induced neurotoxicity. The findings suggest that TCS increases AD risk by diminishing APP’s neuroprotective role in human brains. Clinically, this highlights the necessity to improve risk management of daily chemical exposures that contribute to AD development, offering a basis for formulation of public health policies that reduce long-term TCS exposure.