Identification and Prioritization of Relationships between Environmental Stressors and Adverse Human Health Impacts

Abstract

Background: There are > 80,000 chemicals in commerce with few data available describing their impacts on human health. Biomonitoring surveys, such as the NHANES (National Health and Nutrition Examination Survey), offer one route to identifying possible relationships between environmental chemicals and health impacts, but sparse data and the complexity of traditional models make it difficult to leverage effectively.

Objective: We describe a workflow to efficiently and comprehensively evaluate and prioritize chemical-health impact relationships from the NHANES biomonitoring survey studies.

Methods: Using a frequent itemset mining (FIM) approach, we identified relationships between chemicals and health biomarkers and diseases.

Results: The FIM method identified 7,848 relationships between 219 chemicals and 93 health outcomes/biomarkers. Two case studies used to evaluate the FIM rankings demonstrate that the FIM approach is able to identify published relationships. Because the relationships are derived from the vast majority of the chemicals monitored by NHANES, the resulting list of associations is appropriate for evaluating results from targeted data mining or identifying novel candidate relationships for more detailed investigation.

Conclusions: Because of the computational efficiency of the FIM method, all chemicals and health effects can be considered in a single analysis. The resulting list provides a comprehensive summary of the chemical/health co-occurrences from NHANES that are higher than expected by chance. This information enables ranking and prioritization on chemicals or health effects of interest for evaluation of published results and design of future studies.

Citation: Bell SM, Edwards SW. 2015. Identification and prioritization of relationships between environmental stressors and adverse human health impacts. Environ Health Perspect 123:1193-1199; http://dx.doi.org/10.1289/ehp.1409138.

Figure 1

General workflow using FIM. A brief overview of the workflow from Bell and Edwards (2014) is described. Values used in the Discretization step are a key determinant of the generated rules and should be reviewed.

Figure 2

Heat map of associations across the different slices. Colors are based on odds ratio of the associations, associations not found or below the threshold are in gray (NA). Abbreviations: MexAmer, Mexican American; NHBlack, non-Hispanic black; NHWhite, non-Hispanic white. Color labels on the left highlight the different chemical groups for the association. From bottom to top: Light yellow = drinking water volatile organic compounds; light green = urinary perchlorate, nitrate, and thiocyanate; gray = phytoestrogens; light cyan = phthalates; midnight blue = urinary metals; cyan = current use pesticides; salmon = urinary arsenic; tan = smoking; green yellow = environmental pesticides; purple = PFC (perfluorinated compound); magenta = PCB; pink = PAH; black = organophosphate pesticides; red = organochlorine pesticides; green = environmental phenols; yellow = DFP (diisopropyl fluorophosphate); brown = carbamates; blue = blood volatile organic compounds; turquoise = blood metals.

Figure 3

Odds ratios for rules generated from all data (first column from Figure 2). Row color labels indicate the chemical groupings as in Figure 2, and column labels indicated groupings for the health variables. Gray indicates no rule present for the data set (NA). Column colors (top) from left to right: turquoise = allergies; blue = anemia; brown = arthritis; yellow = asthma; green = cancer; red = cardiovascular health; black = complete blood count; pink = diabetes; magenta = iron; purple = kidney; green yellow = liver; tan = multiple associations; salmon = parathyroid; cyan = respiratory; midnight blue = thyroid; light cyan = vitamins and minerals; gray = weight.

Figure 4

Sample workflow using PFOA. The general workflow for data processing (Figure 1) was employed leveraging additional data strata to obtain a set of rules. Screenshots are generated using Supplemental Material, File S4 (“Spreadsheet”). Ex., example. Rules containing PFOA were extracted (panel 1) then those meeting criteria for strong associations were filtered out (panel 2). Use of additional filters (panel 3) helps to identify and prioritize the relationships for further follow-up.