Threshold effect.

Background
There are limited epidemiological investigations of blood metal levels related to hyperlipidemia, and results indicating the association between blood lead (Pb), cadmium (Cd) and selenium (Se), and lipid biomarkers have been conflicting.
Methods
We included populations for which NHANES collected complete data. Multivariate logistic regression and subgroup analyses were conducted to ascertain the relationship between blood Pb, Cd, and Se levels and hyperlipidemia. Nonlinear relationships were characterized by smoothed curve fitting and threshold effect analysis.
Results
5429 participants in all, with a mean age of 53.70 ± 16.63 years, were included; 47.1% of the subjects were male, and 3683 (67.8%) of them had hyperlipidemia. After modifying for variables with confounders in a multivariate logistic regression model, we discovered a positive correlation between blood Pb and Se levels and hyperlipidemia (Pb: OR:2.12, 95% CI:1.56–2.88; Se: OR:1.84, 95% CI:1.38–2.45). Gender, age, smoking status, alcohol use status, hypertension, diabetes, and body mass index were not significantly linked with this positive correlation, according to subgroup analysis and interaction test (P for interaction>0.05). Positive correlations between blood Pb, Cd, and Se levels and the risk of hyperlipidemia have been found using smooth curve fitting.
Conclusions
This study demonstrates that higher blood levels of Pb, Cd, and selenium are linked to an increased risk of hyperlipidemia.