Ram B. Jain

Observed differentials in the levels of selected environmental contaminants among Mexican and other Hispanic American children, adolescents, adults, and senior citizens

Starting with the 2007–2008 cycle, the National Health and Nutrition Examination Survey (NHANES) also oversampled Hispanics other than Mexicans (OHISP) making it possible to treat OHISP as a separate demographic group along with Mexican Americans (MAs), non-Hispanic whites (NHWs), and non-Hispanic blacks (NHBs). Yet, more often than not, OHISP have been merged with MA to form an all-Hispanic demographic group (HISP) thus limiting comparisons between NHW, NHB, and HISP. Consequently, for the first time, this study was undertaken to evaluate differences in the observed levels of selected environmental contaminants between MA and OHISP from five groups of environmental contaminants, namely, polycyclic aromatic hydrocarbons (PAHs), iodine uptake inhibitors (IUIs), environmental phenols (EPHs), priority pesticides (PPs), and perfluoroalkyl acids (PFAAs). Data for 2007–2010 from NHANES were used to conduct this study. OHISP children born in USA had higher levels of selected PAH metabolites than USA-born MA, and Mexican-born MA adolescents had higher levels of selected PAH metabolites than USA-born MA adolescents. USA-born adolescent MA had higher levels of selected parabens than USA-born adolescent OHISP, and OHISP adults born in another Spanish-speaking country had higher levels of selected parabens than USA-born OHISP adults. USA-born MA adults and seniors had higher levels of selected dichlorophenols than Mexico-born MA adults and seniors, respectively. Females had higher levels of selected PAH metabolites, EPHs, and PPs than males among children, adolescents, adults, and seniors, but the reverse was true for the levels of selected IUIs and PFAAs among adolescents and seniors. Smokers had higher levels of almost all PAH metabolites than non-smokers for adolescents, adults, and seniors. The same was true for urinary thiocynate for adolescents, adults, and seniors. OHISP is a multiracial multiethnic demographic group substantially different from MA with possibly different smoking behavior and with possibly differential levels of exposure to certain environmental contaminants and as such should be treated as a demographic group by itself.

Ratio-based vs. model-based methods to correct for urinary creatinine concentrations

Creatinine-corrected urinary analyte concentration is usually computed as the ratio of the observed level of analyte concentration divided by the observed level of the urinary creatinine concentration (UCR). This ratio-based method is flawed since it implicitly assumes that hydration is the only factor that affects urinary creatinine concentrations. On the contrary, it has been shown in the literature, that age, gender, race/ethnicity, and other factors also affect UCR. Consequently, an optimal method to correct for UCR should correct for hydration as well as other factors like age, gender, and race/ethnicity that affect UCR. Model-based creatinine correction in which observed UCRs are used as an independent variable in regression models has been proposed. This study was conducted to evaluate the performance of ratio-based and model-based creatinine correction methods when the effects of gender, age, and race/ethnicity are evaluated one factor at a time for selected urinary analytes and metabolites. It was observed that ratio-based method leads to statistically significant pairwise differences, for example, between males and females or between non-Hispanic whites (NHW) and non-Hispanic blacks (NHB), more often than the model-based method. However, depending upon the analyte of interest, the reverse is also possible. The estimated ratios of geometric means (GM), for example, male to female or NHW to NHB, were also compared for the two methods. When estimated UCRs were higher for the group (for example, males) in the numerator of this ratio, these ratios were higher for the model-based method, for example, male to female ratio of GMs. When estimated UCR were lower for the group (for example, NHW) in the numerator of this ratio, these ratios were higher for the ratio-based method, for example, NHW to NHB ratio of GMs. Model-based method is the method of choice if all factors that affect UCR are to be accounted for.