Read the General Documentation on Laboratory Data file for detailed data processing and editing protocols.
The analytical methods are described in the Description of the Laboratory Methodology section.
In the dataset, the whole weight detection limit (ng/g serum) is a variable and dependent on the available sample size, however, the variation in sample size was low. Lipid adjusted detection limits are variable (ng/g lipid) due to differences of the lipid levels measured in individual pools/specimens. In cases where the result was below the limit of detection, the value for that variable is the detection limit divided by the square root of 2.
Rationale and Methods used to Create Pooled Results
The National Centers for Disease Control and Prevention (CDC) provides an ongoing assessment of the US population's exposure to environmental chemicals by using biomonitoring in conjunction with CDC's National Health and Nutrition Examination Survey (NHANES). Characterizing the distributions of concentrations of environmental compounds or their metabolites in the US population is a primary objective of CDC's biomonitoring program. Historically, this characterization has been based on individual measurements of these compounds in body fluid or tissue from representative samples of the population. Pooling samples allows for larger sample volumes which can result in lower limits of detection and reduces the number of measurements and costs. In NHANES 2007-2008, a weighted pooled-sample design was implemented to facilitate pooling samples before making analytical measurements. Table 1 lists the IUPAC (International Union of Pure and Applied Chemistry (IUPAC) names, common abbreviation, and NHANES variable name for the analytes included in this data file.
Pools were prepared from serum collected from a random one-third subset of the NHANES 2007–2008 participants aged 12 years and older. Samples were pooled based on gender, race and Hispanic origin, and age. To implement the pooled-sample design, each participant sample was identified as belonging to one of 32 demographic groups based on race and Hispanic origin (non-Hispanic white: NHW, non-Hispanic black: NHB, Mexican American: MA,: not non-Hispanic black, non-Hispanic white or Mexican-American: OTHER), gender (Male, Female), and age group (12-19, 20-39, 40-59, and 60+ years of age and older The number of pools created for each of the 32 demographic groups varied depending on the total number of individual samples available in a demographic group. The one-third subset of NHANES 2007-2008 represents 2282 individual samples, but because the pooled-sample design requires that all samples be of sufficient volume and that there be the same number of samples in each pool, only 2070 samples were available to create 264 pools with 8 samples per pool (Exceptions are in Table 2).
In order to incorporate sample weighting into the pooled-sample design it was necessary to use a different volume of material from each sample contributing to a pool. The volume chosen for each sample in a pool was based on the ratio of its sampling weight to the sum of the sampling weights of all samples in the pool. To physically accomplish the pooling in the laboratory required that the ratio of the largest to the smallest sampling weight of samples in the same pool be no larger than about 4 or 5. The individual samples were sorted/stratified by sampling weight within each of the 32 demographic groups and pools were formed using samples with sampling weights adjacent to one another in the sorted list. The number of samples in the one-third subset, number of samples available, the number of these samples that were usable, and the number of pools formed in each demographic group are presented in Table 2. Once the pools were created, summed sampling weights were further adjusted to account for the unused samples. These adjusted summed sampling weights are represented by the variable named WTSMSMPA.
Table 1. NHANES Variable Name, Common Abbreviation and IUPAC Name for each analyte reported
|NHANES Variable Name
Table 2. Number of subjects per demographic group in the NHANES 2007–2008 one-third subsample, number of individual serum samples available, number of usable samples, and number of pools formed from usable samples.
The original pooled-sample weight created for this file released in July 2013 did not accurately take into account the new sample design for this NHANES survey cycle and it was not correctly stratified to the U.S. population total. This new file contains the corrected sample weight (WTSMSMPA). The corrected sample weight was created so that analyses using race and Hispanic origin would be comparable to the three groups used in NHANES 2005-2006 (non-Hispanic white, non-Hispanic black and Mexican-American). Therefore, analyses of the data for ”Total Hispanics” is not appropriate. No changes or corrections were made to the lab analyte data in this new release. Any analyses of the data using the old public use data file should be repeated using the corrected sample weight on this new file.
PBDEs and PBB-153 were measured in a one third subsample of persons 12 years and over, and samples were pooled in groups of 8 samples per pool within 32 demographic groups. The analysis of NHANES 2007-2008 pooled-sample data must be conducted with the basic demographic variables provided in this data file. This pooled-sample data file cannot be linked to other NHANES 2007-2008 data. Because each sample person does not have an equal probability of selection, sample weighting is needed to produce correct population estimates of means, percentiles, and other descriptive statistics.
The pooled-sample weights required to produce estimates from these data are included in this data file. The analysis of pooled-samples from survey data is a relatively new field of study, and there is currently no established method to produce variance estimates for these pooled results. These data cannot be analyzed using software designed for complex surveys because design features are not available.
Methodological Issues for Analysis
There are a few pooled-sample methodological publications which address issues related to the following topics (See References.)
• Variance estimation and bias correction when pooling samples from log-normally distributed populations (Caudill, 2010a; Caudill, 2010b; Caudill et al., 2007);
• Models that incorporate measurement error when analyzing pooled-samples from normally or log-normally distributed populations (Caudill, 2010a);
• Incorporation of sample weighting into a pooled-sample design (Caudill, 2012; Caudill, 2010a);
• Estimation of standard errors and confidence limits for point estimates from pooled-samples (Caudill, 2012; Caudill, 2010a).