Exposure assessment of French women and their newborn to brominated flame retardants: Determination of tri- to deca- polybromodiphenylethers (PBDE) in maternal adipose tissue, serum, breast milk and cord serum

Abstract

In the frame of a French monitoring program, tri- to deca- polybromodiphenylethers (PBDE) have been measured in maternal and cord serum, adipose tissue, and breast milk samples, collected from 93 volunteer women during caesarean deliveries. The seven major tri- to heptaBDE (BDE-28, 47, 99, 100, 153, 154, and 183) were detected in adipose tissue and breast milk with cumulated median values of 2.59 and 2.51 ng g⁻¹ l w. Nine highly brominated octa- to decaBDE (BDE-196, 197, 201, 202, 203, 206, 207, 208 and 209) was performed in the same samples, with cumulated median values of 2.73 and 3.39 ng g⁻¹ l w in adipose tissue and breast milk, respectively. At this opposite, median levels of octa- to decaBDE in maternal and cord serum appeared significantly higher than the levels of tri- to heptaBDE in the same matrices, i.e. 8.85 and 12.34 versus 0.98 and 0.69 ng g⁻¹ l w, respectively.

Introduction

Polybromodiphenylethers (PBDE) are synthetic compounds used as brominated flame retardants (BFR) which are integrated in many industrial products. Plastic, electric and electronic industries are among the main users of BFR (IPCS, 1994). PBDE are incorporated as inert additives at the end of the industrial process, in opposition to other BFR (such as tetrabromobisphenol-A) which are used as reactive substances, chemically interacting with the processed material during its fabrication (BSEF, 2003). Three industrial PBDE mixtures have been extensively marketed (namely penta-, octa-, and decaMix) which mainly consist to pentaBDE, hepta- and octaBDE, and decaBDE, respectively (Darnerud et al., 2001). Recently, the production and use of the pentaMix and the octaMix have been banned in Europe (2002/95/EC) and in some other countries. Consequently, the decaMix (containing around 97% of decaBDE) is more than ever the main PBDE mixture in use nowadays.

PBDE are environmental contaminants (de Wit, 2002). They share close physico-chemical properties with other organohalogenated persistent organic pollutants (POPs) such as dioxins or polychlorobiphenyls (PCBs). Their bioaccumulation and biomagnification potential have already been documented (Gustafsson et al., 1999, Burreau et al., 1999). From a toxicological point of view, highly brominated PBDE were shown to produce hepatotoxic effects in rodents (Norris et al., 1973, Norris et al., 1975, NTP, 1986, Hardy, 2002). Some PBDE are also considered to be endocrine disruptors (Fowles et al., 1994), with particular concerns related to thyroid homeostasis (Zhou et al., 2002, Stoker et al., 2004) and possible neurobehavioral effects (Branchi et al., 2002, Timme-Laragy et al., 2006). Potential adverse consequences of human exposure to such biologically active chemicals are of specific high concern when the exposure occurs at critical stages of foetal life or at early stages of life, i.e. during the perinatal period (Branchi et al., 2003, Darnerud, 2003, Colborn, 2004). A first study showing a possible relation between PBDE contamination level in breast milk and congenital cryptorchidism in newborn boys has been recently published (Main et al., 2007).

The presence of PBDE in the environment has been extensively reviewed (de Wit, 2002). Their occurrence in various animal species, such as fish (Akutsu et al., 2001, Christensen et al., 2002, Anderson and MacRae, 2006) and marine mammals (Law et al., 2005, Dietz et al., 2007) was also demonstrated. At the opposite, relevant information related to PBDE residual concentrations in human tissues is more limited, and even scarcer are the studies where the presence of highly brominated PBDE congeners (octaBDE, nonaBDE and decaBDE) has been examined. Available studies, mainly focusing on the major tri- to heptaBDE congeners, concern United Kingdom (Thomas et al., 2006), Northern European countries (Main et al., 2007, Meironyté et al., 1999, Sjödin et al., 1999, Thomsen et al., 2001, Thomsen et al., 2002, Thomsen et al., 2007, van Bavel et al., 2002, Covaci et al., 2002, Smeds and Saukko, 2003), Asia (Ohta et al., 2002, Akutsu et al., 2003, Takasuga et al., 2004, Koizumi et al., 2005, Li et al., 2005, Bi et al., 2006), or the United States (Päpke et al., 2001, Mazdai et al., 2003, Schecter et al., 2003, Schecter et al., 2007). Some data have also been published for Southern European countries like Spain (Meneses et al., 1999, Fernandez et al., 2007, Schuhmacher et al., 2007), and more recently for Eastern Europe (Tsydenova et al., 2007, Kazda et al., 2004). But such data were never provided for France. One of the reasons explaining this lack of expanded data is the high technical difficulty associated to the measurement of these substances (de Boer and Wells, 2006, Papke et al., 2004). Yet, the monitoring of highly brominated PBDE is of major importance, considering that decaMix now remains the main additive BFR mixture used worldwide with a total demand estimated to 50,000 metric tons in 2003 (BSEF, 2003), and that DecaMix components may, to some extent, be released from the manufactured products into the environment (Birnbaum and Staskal, 2004, Kim et al., 2006). Moreover, the existence of various degradation processes (photolytic, microbial and/or metabolic) affecting highly brominated compounds in the environment was clearly demonstrated, such debromination reactions finally leading to the formation of lower brominated PBDE congeners (Chen et al., 2007, Rayne et al., 2006, Hagberg et al., 2006, Sanchez-Prado et al., 2005, Rupp and Metzger, 2005). Several authors have hypothesized that these degradation processes implicate, de facto, a very poor probability to find high molecular weight PBDE in human tissues. In reality, very little is known about the real occurrence and tissue distribution of these compounds in human biological tissues.

In this context, an analytical strategy was developed for the simultaneous measurement of tri- to decaBDE from various human biological matrices (serum, adipose tissue and breast milk). This procedure, which was fully validated according to current European standards, was detailed elsewhere (Cariou et al., 2005). The methodology was applied within the frame of a French monitoring survey, Human biological samples being provided by the Centre Hospitalier Universitaire (CHU) of Toulouse. This study concerned 93 volunteer women (mother/newborn pairs). Biological samples were collected during caesarean deliveries, and included maternal serum, adipose tissue, breast milk, and cord serum.