Persistent organochlorine pollutants in human serum of 50-65 years old women in the Flanders Environmental and Health Study (FLEHS). Part 1: concentrations and regional differences

In 1999, a campaign of the Flemish Ministry of Health, Belgium was set up to assess pollutant concentrations and related health effect biomarkers in humans living in two regions of Flanders. The study was called the 'Flemish Environment and Health Study' (FLEHS). One of the goals was to measure present concentrations of persistent organochlorine pollutants in a Flemish population and to compare values obtained from pooled and individual serum samples. Concentrations of selected organochlorine pesticides, polychlorinated biphenyls (PCB) and polychlorinated dibenzo-p-dioxins (PCDD) and furans (PCDF) were measured by gas chromatography-mass spectrometry. TEQ values were also assessed by Chemical-Activated LUciferase gene eXpression (CALUX) bioassay. The study population consisted of 200 women between 50 and 65 years living in two areas of Flanders, Belgium. Because of the large volumes serum needed for all measurements, the concentrations of organochlorines were measured in 47 pooled serum samples originating from these women. The concentrations of the indicator PCBs (359.8 ng/g fat) and organochlorine pesticides (hexachlorobenzene, p,p'-dichlorodiphenyldichloroethylene, p,p'-dichlorodiphenyl-trichloroethane, lindane and pentachlorophenol), were comparable to those found in other European countries. The concentrations of PCDD/PCDFs showed another picture. With a median value of 48 pg WHO-TEQ/g fat, the women had 2-fold higher levels than a comparable age group from Germany examined in 1996. The mean total WHO-TEQ including PCDD/F, non-ortho and mono-ortho PCBs was 72.7 pg WHO-TEQ/g fat, whereas the CALUX-TEQ mean value was only 35.0 pg TEQ/g fat. In order to assess the pooling procedure, indicator PCBs and CALUX-TEQs were measured in all 200 individuals that were integrated in the pools. The measured values were comparable to the pool results: 390.0 ng/g fat and 41.6 pg TEQ/g fat respectively. It was concluded that pooling of serum samples offers the possibility to measure exposure in the whole study population on a more cost-effective way. However, because of statistical power loss and no possibility of confounder adjustment, pooling is not the most effective way to study regional differences. (C) 2002 Elsevier Science Ltd. All rights reserved.