Insights into PBDE Uptake, Body Burden, and Elimination Gained from Australian Age–Concentration Trends Observed Shortly after Peak Exposure

Background Population pharmacokinetic models combined with multiple sets of age–concentration biomonitoring data facilitate back-calculation of chemical uptake rates from biomonitoring data. Objectives We back-calculated uptake rates of PBDEs for the Australian population from multiple biomonitoring surveys (top-down) and compared them with uptake rates calculated from dietary intake estimates of PBDEs and PBDE concentrations in dust (bottom-up). Methods Using three sets of PBDE elimination half-lives, we applied a population pharmacokinetic model to the PBDE biomonitoring data measured between 2002–2003 and 2010–2011 to derive the top-down uptake rates of four key PBDE congeners and six age groups. For the bottom-up approach, we used PBDE concentrations measured around 2005. Results Top-down uptake rates of Σ4BDE (the sum of BDEs 47, 99, 100, and 153) varied from 7.9 to 19 ng/kg/day for toddlers and from 1.2 to 3.0 ng/kg/day for adults; in most cases, they were—for all age groups—higher than the bottom-up uptake rates. The discrepancy was largest for toddlers with factors up to 7–15 depending on the congener. Despite different elimination half-lives of the four congeners, the age–concentration trends showed no increase in concentration with age and were similar for all congeners. Conclusions In the bottom-up approach, PBDE uptake is underestimated; currently known pathways are not sufficient to explain measured PBDE concentrations, especially in young children. Although PBDE exposure of toddlers has declined in the past years, pre- and postnatal exposure to PBDEs has remained almost constant because the mothers’ PBDE body burden has not yet decreased substantially. Citation Gyalpo T, Toms LM, Mueller JF, Harden FA, Scheringer M, Hungerbühler K. 2015. Insights into PBDE uptake, body burden, and elimination gained from Australian age–concentration trends observed shortly after peak exposure. Environ Health Perspect 123:978–984; http://dx.doi.org/10.1289/ehp.1408960

Least-squares optimization 2. Modeled and measured cross-sectional age-concentration profile for the male population Figure S2. Modeled age-concentration profiles (blue: scenario A; green: scenario B; red: scenario C) fitted to the biomonitoring data (dots) from the male population.
3. Input data for the PBDE bottom-up approach Table S2. Congener-specific parameters. Table S3. Age-dependent parameters.

Parameterization of the time-variant population pharmacokinetic (PK) model
Equation 1 defines the time course of the chemical concentration in a representative individual born at time t birth : where t age (years) is the age of the individual; C(t age ) (ng g lip -1 ) is the lipid-normalized concentration of chemical in the body; U ref (t) (ng kg -1 d -1 ) is the reference daily uptake of the chemical for an adult and depends on the year of sampling, t; M bw (t age ) (kg) and M lip (t age ) (kg lip ) are the body weight and the body lipid weight as a function of age, respectively; P(t age ) (dimensionless) is a proportionality factor adapting U ref (t) to younger ages; F (kg lip -1 g lip ) is a unit conversion factor; k elim (d -1 ) is the first-order rate constant describing intrinsic elimination.
Importantly, U ref represents the absorbed amount of chemicals (= uptake) from all sources and pathways (excluding breast milk) that contribute to the PBDE concentration in the body.
The model was programmed in Matlab R2013a and solved with a 3-day resolution.

Transfer of chemical via breast milk
The daily human milk consumption rate, r bm (t age ) (g d -1 ) and the lipid fraction of the human milk, f lip,bm (t age ) (dimensionless) are described dependent on the age of the infant (t age ) (years) and his/her body weight (M bw ) (kg) according to Verner et al. (2013) (Equations 2 and 3):

Proportionality factor
We derived the proportionality factor, P(t age ) in Equation 1, by dividing the uptake rates of younger age groups by the uptake rate of adults (Table S1). The empirical proportionality factors show steps, because they represent whole age groups, i.e. 1-6, 6-12, 12-20, and >20 years ( Figure S1, black and blue diamonds). Since no exposure estimate is given for infants < 1 years in Lorber (2008), we assumed it to be 50% of that of the group of 1-6 years. We used a Weibull function to interpolate the proportionality factors for uptakes of the different age groups (black and blue lines). We used data from Lorber (2008) as base case (panel A in Figure S1). As an alternative, we used the median uptake rates for the US population from Trudel et al. (2011) (panel B in Figure S1).

Least-squares optimization
For each optimization, we used 29 empirical data points (see main text). By minimizing the sum of squared residuals weighted (SSRW), we maximized R 2 (Equation 4): where n is the number of empirical data points, here n = 29, y i is the empirical data point i, f i is the equivalent modeled value, and is the empirical sample mean. Figure S2. Modeled age-concentration profiles (blue: scenario A; green: scenario B; red: scenario C) fitted to the biomonitoring data (dots) from the male population.

Input data for the PBDE bottom-up approach
The lipid fraction of breast milk was set to 3.3% (Toms et al. 2012). The transfer fraction from dust to skin was set to 13% (Trudel et al. 2011); the dermal absorption fraction was set to 3% (Roper et al. 2006).  Table S4. Congener-specific uptake rates (ng kg -1 d -1 ).