The National Health and Nutrition Examination Survey (NHANES), conducted continuously since 1999 by the CDC, assesses the health and nutritional status of the civilian noninstitutionalized U.S. population ≥ 2 months of age (CDC 2003 ). The survey includes household interviews, medical histories, standardized physical examinations, and collection of biologic specimens, some of which can be used to assess exposure to environmental chemicals. NHANES 2003–2004 included examinations of 9,282 people (CDC 2006a ). We measured BP-3 by analyzing a random one-third subset of urine samples (n = 2,517) collected from NHANES participants ≥ 6 years of age. Because this subset was randomly selected from the entire set, it maintained the representativeness of the survey. Participants provided informed written consent; parents provided informed written consent for their children.
Urine specimens were shipped on dry ice to the CDC’s National Center for Environmental Health and stored frozen at or below –20°C until analyzed. We measured total (free plus conjugated species) concentrations of BP-3 in urine by online solid-phase extraction coupled to high-performance liquid chromatography–tandem mass spectrometry described in detail elsewhere (Ye et al. 2005a (link)). Briefly, conjugated BP-3 in 100 μL of urine was hydrolyzed using β-glucuronidase/ sulfatase (Helix pomatia; Sigma Chemical Co., St. Louis, MO). After hydrolysis, samples were acidified with 0.1 M formic acid; BP-3 was preconcentrated by online solid-phase extraction, separated by reversed-phase high-performance liquid chromatography, and detected by atmospheric pressure chemical ionization–tandem mass spectrometry. Because a stable isotope-labeled BP-3 was not available, we used 13C12-bisphenol A as internal standard (Ye et al. 2005a (link)). The limit of detection (LOD), calculated as 3S0, where S0 is the standard deviation as the concentration approaches zero (Taylor 1987 ), was 0.34 μg/L, and the precision ranged from 17.6% (at 18.5 μg/L) to 16.2% (at 46 μg/L). Low-concentration (~ 20 μg/L) and high-concentration (~ 45 μg/L) quality control materials, prepared from pooled human urine, were analyzed with standard, reagent blank, and NHANES samples (Ye et al. 2005a (link)).
We analyzed the data using Statistical Analysis System (version 9.1.3; SAS Institute, Inc., Cary, NC) and SUDAAN (version 9.0.1; Research Triangle Institute, Research Triangle Park, NC). SUDAAN calculates variance estimates after incorporating the sample population weights, nonresponse rates, and sample design effects. We calculated the percentage of detection and the geometric mean and distribution percentiles for both the volume-based (in micrograms per liter urine) and creatinine-corrected (in micrograms per gram creatinine) concentrations. For concentrations below the LOD, as recommended for the analysis of NHANES data (CDC 2006b ), we used a value equal to the LOD divided by the square root of 2 (Hornung and Reed 1990 ).
A composite racial/ethnic variable based on self-reported data defined three major racial/ethnic groups: non-Hispanic black, non-Hispanic white, and Mexican American. We included participants not defined by these racial/ethnic groups only in the total population estimate. Age, reported in years at the last birthday, was stratified in groups (6–11, 12–19, 20–59, and ≥ 60 years of age) for calculation of the geometric mean and the various percentiles.
We used analysis of covariance to examine the influence of several variables, selected on the basis of statistical, demographic, and biologic considerations, on the concentrations of BP-3. For the multiple regression models, we used the variables described below and all possible two-way interactions to calculate the adjusted least square geometric mean (LSGM) concentrations. LSGM concentrations provide geometric mean estimates (in micrograms per liter) after adjustment for all covariates in the model. Because the distributions of BP-3 and creatinine concentrations were skewed, these variables were log transformed. We analyzed two separate models: one for adults (≥ 20 years of age) and one for children and teenagers (≤ 19 years of age). We considered age (continuous), age squared, sex, race/ethnicity, and log-transformed crea-tinine concentration for both models. When the model included both age and age squared, we centered age by subtracting 50 from each participant’s age, thus avoiding multi-collinearity and obtaining the least weighted correlation between these two variables (Bradley and Srivastava 1979 ). Additionally, to further evaluate the relation between the log-transformed BP-3 concentration and age, we used age group (20–29, 30–39, 40–49, and ≥ 50 years of age) as a categorical variable in the model and generated a bar chart of LSGM concentrations by age group.
To reach the final reduced model, we used backward elimination with a threshold of p < 0.05 for retaining the variable in the model, using Satterwaite-adjusted F statistics. We evaluated for potential confounding by adding each of the excluded variables back into the final model one by one and examining changes in the β coefficients of the statistically significant main effects or interactions. If addition of one of these excluded variables caused a change in a β coefficient by ≥ 10%, we re-added the variable to the model.
We also conducted weighted univariate and multiple logistic regressions to examine the association of BP-3 concentrations above the 95th percentile with sex, age group, and race/ ethnicity for all ages.
Urine specimens were shipped on dry ice to the CDC’s National Center for Environmental Health and stored frozen at or below –20°C until analyzed. We measured total (free plus conjugated species) concentrations of BP-3 in urine by online solid-phase extraction coupled to high-performance liquid chromatography–tandem mass spectrometry described in detail elsewhere (Ye et al. 2005a (link)). Briefly, conjugated BP-3 in 100 μL of urine was hydrolyzed using β-glucuronidase/ sulfatase (Helix pomatia; Sigma Chemical Co., St. Louis, MO). After hydrolysis, samples were acidified with 0.1 M formic acid; BP-3 was preconcentrated by online solid-phase extraction, separated by reversed-phase high-performance liquid chromatography, and detected by atmospheric pressure chemical ionization–tandem mass spectrometry. Because a stable isotope-labeled BP-3 was not available, we used 13C12-bisphenol A as internal standard (Ye et al. 2005a (link)). The limit of detection (LOD), calculated as 3S0, where S0 is the standard deviation as the concentration approaches zero (Taylor 1987 ), was 0.34 μg/L, and the precision ranged from 17.6% (at 18.5 μg/L) to 16.2% (at 46 μg/L). Low-concentration (~ 20 μg/L) and high-concentration (~ 45 μg/L) quality control materials, prepared from pooled human urine, were analyzed with standard, reagent blank, and NHANES samples (Ye et al. 2005a (link)).
We analyzed the data using Statistical Analysis System (version 9.1.3; SAS Institute, Inc., Cary, NC) and SUDAAN (version 9.0.1; Research Triangle Institute, Research Triangle Park, NC). SUDAAN calculates variance estimates after incorporating the sample population weights, nonresponse rates, and sample design effects. We calculated the percentage of detection and the geometric mean and distribution percentiles for both the volume-based (in micrograms per liter urine) and creatinine-corrected (in micrograms per gram creatinine) concentrations. For concentrations below the LOD, as recommended for the analysis of NHANES data (CDC 2006b ), we used a value equal to the LOD divided by the square root of 2 (Hornung and Reed 1990 ).
A composite racial/ethnic variable based on self-reported data defined three major racial/ethnic groups: non-Hispanic black, non-Hispanic white, and Mexican American. We included participants not defined by these racial/ethnic groups only in the total population estimate. Age, reported in years at the last birthday, was stratified in groups (6–11, 12–19, 20–59, and ≥ 60 years of age) for calculation of the geometric mean and the various percentiles.
We used analysis of covariance to examine the influence of several variables, selected on the basis of statistical, demographic, and biologic considerations, on the concentrations of BP-3. For the multiple regression models, we used the variables described below and all possible two-way interactions to calculate the adjusted least square geometric mean (LSGM) concentrations. LSGM concentrations provide geometric mean estimates (in micrograms per liter) after adjustment for all covariates in the model. Because the distributions of BP-3 and creatinine concentrations were skewed, these variables were log transformed. We analyzed two separate models: one for adults (≥ 20 years of age) and one for children and teenagers (≤ 19 years of age). We considered age (continuous), age squared, sex, race/ethnicity, and log-transformed crea-tinine concentration for both models. When the model included both age and age squared, we centered age by subtracting 50 from each participant’s age, thus avoiding multi-collinearity and obtaining the least weighted correlation between these two variables (Bradley and Srivastava 1979 ). Additionally, to further evaluate the relation between the log-transformed BP-3 concentration and age, we used age group (20–29, 30–39, 40–49, and ≥ 50 years of age) as a categorical variable in the model and generated a bar chart of LSGM concentrations by age group.
To reach the final reduced model, we used backward elimination with a threshold of p < 0.05 for retaining the variable in the model, using Satterwaite-adjusted F statistics. We evaluated for potential confounding by adding each of the excluded variables back into the final model one by one and examining changes in the β coefficients of the statistically significant main effects or interactions. If addition of one of these excluded variables caused a change in a β coefficient by ≥ 10%, we re-added the variable to the model.
We also conducted weighted univariate and multiple logistic regressions to examine the association of BP-3 concentrations above the 95th percentile with sex, age group, and race/ ethnicity for all ages.
Free full text: Click here
