Previous studies have identified several genetic and environmental risk factors for chronic kidney disease (CKD). However, little is known about the relationship between serum metals and CKD risk.
We investigated associations between serum metals levels and CKD risk among 100 medical examiners and 443 CKD patients in the medical center of the First Hospital Affiliated to China Medical University. Serum metal concentrations were measured using inductively coupled plasma mass spectrometry (ICP-MS). We analyzed factors influencing CKD, including abnormalities in Creatine and Cystatin C, using univariate and multiple analysis such as Lasso and Logistic regression. Metal levels among CKD patients at different stages were also explored. The study utilized machine learning and Bayesian Kernel Machine Regression (BKMR) to assess associations and predict CKD risk based on serum metals. A chained mediation model was applied to investigate how interventions with different heavy metals influence renal function indicators (creatinine and cystatin C) and their impact on diagnosing and treating renal impairment.
Serum potassium (K), sodium (Na), and calcium (Ca) showed positive trends with CKD, while selenium (Se) and molybdenum (Mo) showed negative trends. Metal mixtures had a significant negative effect on CKD when concentrations were all from 30th to 45th percentiles compared to the median, but the opposite was observed for the 55th to 60th percentiles. For example, a change in serum K concentration from the 25th to the 75th percentile was associated with a significant increase in CKD risk of 5.15(1.77,8.53), 13.62(8.91,18.33) and 31.81(14.03,49.58) when other metals were fixed at the 25th, 50th and 75th percentiles, respectively.
Cumulative metal exposures, especially double-exposure to serum K and Se may impact CKD risk. Machine learning methods validated the external relevance of the metal factors. Our study highlights the importance of employing diverse methodologies to evaluate health effects of metal mixtures.