Early detection of the health lesions induced by chronic arsenic exposure (HLICAE) are crucial to prevent permanent arsenic-induced damage. If HLICAE can be identified in time, appropriate preventive and therapeutic measures may be provided without various avoidable lesions. The present study aims to assess the probability of HLICAE early recognition with metabolomics. Applying a case-control study, 94 participants with HLICAE (cases) and other 94 subjects without HLICAE (controls) were matched with gender and age (±1 year), coming from a previous chronic arsenic exposure cohort. Serum metabolomic profiles were assessed by ultra-high-performance liquid chromatography-quadrupole time-of-flight mass spectrometry (UPLC-QTOF-MS) and analyzed with univariate and multivariate statistics. A total of 210 and 364 features were detected in positive and negative ion modes (ESI+/ESI-), respectively. The altered metabolic pathways included lipid and amino acid metabolisms. 28 metabolomics-based biomarkers were significantly associated with HLICAE and provided areas under the curve (AUC, 95% confidence interval) of 0.898 (0.836, 0.960) and 0.908 (0.855, 0.960) in the discovery phase, 78.6% and 86.4% of positive predictive values in the validation phase, in distinguishing HLICAE from controls in ESI+/ESI-, respectively. This study provides novel insights on mechanisms of health effects probably induced by chronic arsenic exposure, and these biomarkers may be applied in HLICAE early detection.
