Determination of phenolic organohalogens in human serum from a Belgian population and assessment of parameters affecting the human contamination.

Many in vitro or in vivo studies highlighted the potential deleterious effects of phenolic organohalogenated compounds (POHs) on the health, particularly on the thyroid system homeostasis, however few large scale human epidemiological studies have been carried out, especially in Europe. Further studies monitoring the human contamination by POHs, the sources of exposure and the influence of these compounds on thyroid health are still needed. Therefore we determined the concentrations of 16 POHs (pentachlorophenol (PCP), tetrabromobisphenol A (TBBPA), 4 bromophenols (BPs), 3 hydroxy-polybromodiphenylethers (OH-PBDEs) and 7 hydroxy-polychlorobiphenyls (OH-PCBs)) in serum from 274 people aged from 18 to 76years old living in Liege (Belgium) and the surrounding area. A questionnaire about their alimentary habits, life style and home environment was also administered to the volunteers. The predominant compound measured in the population was PCP (median concentration of 593.0pgmL-1). 4-OH-CB 107, 4-OH-CB 146 and 4-OH-CB 187 were detected in all samples and contributed for 75% of the sum of OH-PCBs (ΣOH-PCBs). The median measured in our population for ΣOH-PCBs was 143.7pgmL-1. TBBPA and 2,4,6-tribromophenol were detected in 31% and 63.8% of the samples respectively while the detection frequency observed for the other BPs and the OH-PBDEs was close to zero. We computed multivariate regression models in order to assess the influence of demographic and lifestyle parameters on the PCP and ΣOH-PCBs contamination levels. Significant correlation was found between the PCP concentration and sex, smoker status, sea fish consumption and level of education, although the model seemed to be a poor (R2=0.14) predictor of the PCP concentration. The model computed for ΣOH-PCBs was more explanatory (R2=0.61) and involved age, BMI and sea fish consumption. Finally, we assessed the parameters affecting the ΣOH-PCBs/ΣPCBs ratio. The model proposed involved age, BMI, smoker status and parent PCB level, and explained 41% of the variability of the ΣOH-PCBs/ΣPCBs ratio.