The association between urinary concentrations of phosphorous-containing flame retardant metabolites and semen parameters among men from a fertility clinic.

BACKGROUND: The use of PFRs has steadily increased as brominated compounds have been or are being phased out. Human exposure is widespread and animal studies have shown adverse impacts on male reproduction, but human data are lacking. OBJECTIVE: To study the associations between urinary concentrations of phosphorous-containing flame retardant (PFR) metabolites and semen parameters. METHODS: A subset of 220 men from an existing longitudinal cohort of couples were recruited from Massachusetts General Hospital fertility clinic between 2005 and 2015. Semen parameters included sperm count, concentration, motility, and morphology; some men had samples measured from multiple clinic visits (up to five visits; n = 269 semen samples). Metabolites [bis(1,3-dichloro-2-propyl) phosphate (BDCIPP), diphenyl phosphate (DPHP), isopropylphenyl phenyl phosphate (ip-PPP), tert-butylphenyl phenyl phosphate (tb-PPP) and bis(1-chloro-2-propyl) phosphate (BCIPP)] were measured in urine samples (between one and five urine samples per participant; n = 355 urine samples). Semen parameters were evaluated continuously and dichotomized for models. Metabolites were assessed for associations with semen parameters as continuous and categorized into quartiles using multivariable generalized mixed models, adjusted for specific gravity, age, BMI, smoking, and abstinence period. RESULTS: Metabolites BDCIPP, DPHP, and ip-PPP were detected in a high proportion of urine samples (85%, 86%, and 65% respectively). Concentrations varied by season of collection, particularly for BDCIPP where samples collected in the summer were approximately 2-fold higher than concentrations of other seasons (p < 0.0001). The odds of having a sperm count less than 39 mil/ejaculate decreased by 20% for increasing BDCIPP concentrations (p = 0.04). When regressing semen parameters on PFR metabolite quartiles, some negative associations were observed for individual quartiles among sample volume and morphology, but overall associations were weak and inconsistent. CONCLUSION: Detection rates were high for BDCIPP, DPHP, and ip-PPP. We did not observe consistent associations between PFR metabolites and semen parameters. Due to the high prevalence of exposure, further investigation of other potential health effects should be conducted.