Global Screening of Environmental Contaminants and Reproductive Function in Greenlandic and European Men Using Partial Least Squares Regression

摘要

Background: Numerous environmental contaminants have been suggested to adversely affect reproductive health. However, the complex correlation structure of contaminants and multiple testing issues limit interpretation of existing evidence. Aims: We evaluated exposure-response relationships (ERRs) for a large set of biomarkers of exposure and biomarkers of male reproductive health using a multivariate partial least squares (PLS) regression approach to account for correlation among exposures and multiple testing. Methods: Male partners of pregnant women were recruited from 2002 to 2004 during routine antenatal care visits from 3 locations: Greenland; Warsaw, Poland; and Kharkiv, Ukraine (n=602). Fifteen contaminants were measured in blood, including metabolites of diethylhexyl and diisononyl phthalates, perfluorinated compounds, 3 metals and 3 organochlorines. Over 20 biomarkers of reproductive function were assessed, including serum levels of reproductive hormones, markers of semen quality, sperm chromatin integrity, epididymal and accessory sex gland function, and X:Y chromosome ratio. We tested ERRs using PLS regression and applied a permutation procedure to derive a suitable significance threshold. Results: Of the 330 ERRs examined in 22 PLS regression models, we detected 6 significant ERRs. The most robust associations were a negative association between phthalates and testosterone, and a positive association between cadmium and testosterone. Conclusions: The extended-PLS regression appropriately accounted for correlated exposures and multiple testing. Results provide further epidemiologic support for a negative effect of phthalates on male reproductive function.