National Health and Nutrition Examination Survey

The Family Smoking Prevention and Tobacco Control Act of 2009 gave the Food and Drug Administration regulatory authority over tobacco products. In Section 907(a), the Act states that the Secretary of Health and Human Services shall establish tobacco product standards. These product standards shall also include provisions for the testing of the tobacco products. In order to determine the effectiveness of these product standards, it is critical that surveillance systems be in place before the standards are enacted to determine whether they impact exposure of people who continue to use tobacco product. NHANES is an ideal source of human samples for assessing smokers in multiple locations around the country. NHANES will provide the most robust dataset possible for identifying changes in exposure of smokers that should result from enactment of tobacco product standards.

We anticipate that the results of this survey of tobacco exposure biomarkers will be used by FDA, other government agencies, and research institutions. The NHANES data will be used to determine the effectiveness of initial product standards and direct whether further changes to product standards need to be made.

Exposure to volatile organic compounds (VOCs) is ubiquitous. Chronic exposure to high levels of some VOCs can lead to cancer and neurocognitive dysfunction. Nearly 200 VOCs have been associated with adverse health effects in occupational studies or laboratory studies, but have not been monitored in general population groups. Metabolites of some VOCs can be detected in urine following exposure events. Therefore, information on exposure to these compounds, as measured by their metabolite concentrations in urine, is essential to understand the levels of hazardous VOCs to which the general population is exposed.

Urine collected from participants in NHANES 2005-2006 aged 12 years and older from a one-half subsample were eligible. Tests were conducted on residual urine samples.

This method is a quantitative procedure for the measurement of VOC metabolites in human urine using ultra performance liquid chromatography coupled with electro spray tandem mass spectrometry (UPLC-ESI-MS/MS) (Alwis et al., 2012). Chromatographic separation is achieved using an Acquity UPLC® HSS T3 (Part no. 186003540, 1.8 mm x 2.1 mm x 150 mm, Waters, Inc.) column with 15 mM ammonium acetate and acetonitrile as the mobile phases. The eluent from the column is ionized using an electrospray interface to generate and transmit negative ions into the mass spectrometer. Comparison of relative response factors (ratio of native analyte to stable isotope labeled internal standard) with known standard concentrations yields individual analyte concentrations.

The analytical measurements were conducted following strict quality control/quality assurance CLIA guidelines. Along with the study samples, each analytical run included high- and low-concentration quality control materials (QCs) and reagent blanks to assure the accuracy and reliability of the data. The concentrations of the high-concentration QCs and the low-concentration QCs, averaged to obtain one measurement of high-concentration QC and low-concentration QC for each run, were evaluated using standard statistical probability rules (Caudill et al. 2008).

The data were reviewed. Incomplete data or improbable values were sent to the performing laboratory for confirmation.

Special Analytic Note: Bias Adjustments Applied to VOC Metabolites in Urine Assay in NHANES 2005–2006 cycle (Surplus)

Adjustments have been applied to the previously published data.

During routine review, several systematic biases were discovered in measurements of certain urinary VOC metabolites in the 2005 – 2006 and 2011 – 2012 NHANES cycles. The systematic nature of these biases permitted retrospective adjustment through a statistical approach based on linear regression modeling where adjusted measurements are predicted from unadjusted measurements using suitable data. The “statistical recalibrations” implemented here have the general form of a univariate linear regression equation (accounting for dilution factors routinely used during measurement):

where exp[·] and ln(·) are the natural exponential and natural logarithm functions, respectively. Natural log-transformation of both the un-adjusted and adjusted measurements moderated adverse influence on model fit from extreme measurements. The estimated regression slope and intercept are m and b, respectively, which are tabulated below for each analyte. The dilution factor was accounted for only once for each analyte that underwent more than one statistical recalibration (e.g., URXMB3). Statistical recalibration for the mass fraction of salt (URXHP2, URXGAM) only required a regression slope (i.e., ).

The explanatory power of the linear regressions was excellent (), and residuals exhibited distributions with good approximation to normality, as well as stable variance over the observed range of measurements. These favorable diagnostics demonstrate the efficacy of statistical recalibration for the retrospective amelioration of systematic bias.

Correction for Change in Internal Standard

Isotopically labeled internal standards were not commercially available for URXMB3 and URXHP2 during analysis of NHANES 2005–2006. Instead, the lab used surrogate internal standards.  Isotopically labeled internal standards later became available for purchase and were used in the NHANES 2011–2012 cycle. A systematic bias between these two cycles was noted and confirmed to be due to the internal standard change.

Correction for Variation in Preparation of Neat Native Metabolites Standards

During the 2005–2006 and 2011–2012 NHANES cycles, calibrators were prepared from a multi-analyte stock solution formulated from neat materials in-house. Beginning with the NHANES 2015–2016 cycle, calibrators and the multi-analyte stock solution were prepared externally. A systematic bias between the cycles that used in-house calibration materials and the cycle using externally prepared calibration materials was observed for analytes listed in the table below. Further validation of the externally prepared calibrators suggested that the in-house calibrators were inaccurate due to special requirements for handling and preparing accurate solutions from neat materials, some of which are highly hygroscopic.

Correction for Mass Fraction of Salt in Neat Compound Standard

During the 2005–2006 and 2011–2012 NHANES cycles, calibrators were prepared from a multi-analyte stock solutions formulated from neat materials in-house. Beginning with the NHANES 2015–2016 cycle, calibrators and the multi-analyte stock solutions were prepared externally. A systematic bias between the cycles that used in-house calibration materials and the cycle using externally prepared calibration materials was observed for analytes listed in the table below. Further validation confirmed that the large mass fraction of salt in the neat materials was not accounted for during formulation of in-house calibration materials. This discrepancy between the calculated concentration of the calibrators and the actual concentration led to a systematic bias in results for URXHP2 and URXGAM.

Correction for Stereoisomerism in Urine Samples

URXMB3 occurs in human urine predominantly in the cis isomer, but quantitation in the NHANES 2005–2006 and NHANES 2011–2012 cycles used standards containing the trans isomer. This discrepancy led to a systematic bias in the URXMB3 results. Neat material that is purely cis isomer cannot be readily obtained, but a predictive regression equation can account for these quantitative differences.

Detection Limits

 
The detection limits were constant for the analytes in the data set. Two variables are provided for each of these analytes. The variable named ending in “LC” (ex., URDAAMLC) indicates whether the result was below the limit of detection: the value “0” means that the result was at or above the limit of detection, “1” indicates that the result was below the limit of detection. For analytes with analytic results below the lower limit of detection (ex., URDAAMLC=1), an imputed fill value was placed in the analyte results field. This value is the lower limit of detection divided by the square root of 2 (LLOD/sqrt [2]). The other variable prefixed URX (ex., URXAAM) provides the analytic result for the analyte.

The lower limit of detection (LLOD, in µg/L) for urine VOC metabolites: