Graham Glen

The national exposure research laboratory's consolidated human activity database.

EPAs National Exposure Research Laboratory (NERL) has combined data from 12 U.S. studies related to human activities into one comprehensive data system that can be accessed via the Internet. The data system is called the Consolidated Human Activity Database (CHAD) and is available at E-mail: http://www.epa.gov/nerl/. CHAD contains 22,968 person days of activity and is designed to assist exposure assessors and modelers in constructing population “cohorts” of people with specified characteristics that are suitable for subsequent analysis or modeling. This paper describes the studies comprising CHAD and the various intellectual foundations that underlay the gathering of human activity pattern data. Next, it provides a brief overview of the Internet version of CHAD, and discusses how the program was formulated. Emphasis is placed on how activity-specific energy expenditure estimates were developed. Finally, the paper recommends steps that should be taken to improve the collection of activity data that would improve energy expenditure estimates and related information needed for physiologically based exposure–dose modeling efforts.

A Probabilistic Arsenic Exposure Assessment for Children who Contact CCA-Treated Playsets and Decks, Part 1: Model Methodology, Variability Results, and Model Evaluation

Concerns have been raised regarding the safety of young children who may contact arsenic residues while playing on and around chromated copper arsenate (CCA)-treated wood playsets and decks. Although CCA registrants voluntarily canceled the production of treated wood for residential use in 2003, the potential for exposure from existing structures and surrounding soil still poses concerns. The EPAs Office of Research and Development developed and applied the probabilistic Stochastic Human Exposure and Dose Simulation model for wood preservatives (SHEDS-Wood) to estimate childrens absorbed dose of arsenic from CCA. Skin contact with, and nondietary ingestion of, arsenic in soil and wood residues were considered for the population of children in the United States who frequently contact CCA-treated wood playsets and decks. Model analyses were conducted to assess the range in population estimates and the impact of potential mitigation strategies such as the use of sealants and hand washing after play events. The results show predicted central values for lifetime annual average daily dose values for arsenic ranging from 10(-6) to 10(-5) mg/kg/day, with predicted 95th percentiles on the order of 10(-5) mg/kg/day. There were several orders of magnitude between lower and upper percentiles. Residue ingestion via hand-to-mouth contact was determined to be the most significant exposure route for most scenarios. Results of several alternative scenarios were similar to baseline results, except for the scenario with greatly reduced residue concentrations through hypothetical wood sealant applications; in this scenario, exposures were lower, and the soil ingestion route dominated. SHEDS-Wood estimates are typically consistent with, or within the range of, other CCA exposure models.

A Probabilistic Arsenic Exposure Assessment for Children Who Contact Chromated Copper Arsenate (CCA)‐Treated Playsets and Decks, Part 2: Sensitivity and Uncertainty Analyses

A probabilistic model (SHEDS-Wood) was developed to examine childrens exposure and dose to chromated copper arsenate (CCA)-treated wood, as described in Part 1 of this two-part article. This Part 2 article discusses sensitivity and uncertainty analyses conducted to assess the key model inputs and areas of needed research for childrens exposure to CCA-treated playsets and decks. The following types of analyses were conducted: (1) sensitivity analyses using a percentile scaling approach and multiple stepwise regression; and (2) uncertainty analyses using the bootstrap and two-stage Monte Carlo techniques. The five most important variables, based on both sensitivity and uncertainty analyses, were: wood surface residue-to-skin transfer efficiency; wood surface residue levels; fraction of hand surface area mouthed per mouthing event; average fraction of nonresidential outdoor time a child plays on/around CCA-treated public playsets; and frequency of hand washing. In general, there was a factor of 8 for the 5th and 95th percentiles and a factor of 4 for the 50th percentile in the uncertainty of predicted population dose estimates due to parameter uncertainty. Data were available for most of the key model inputs identified with sensitivity and uncertainty analyses; however, there were few or no data for some key inputs. To evaluate and improve the accuracy of model results, future measurement studies should obtain longitudinal time-activity diary information on children, spatial and temporal measurements of residue and soil concentrations on or near CCA-treated playsets and decks, and key exposure factors. Future studies should also address other sources of uncertainty in addition to parameter uncertainty, such as scenario and model uncertainty.

Estimating Sobol sensitivity indices using correlations

Sensitivity analysis is a crucial tool in the development and evaluation of complex mathematical models. Sobols method is a variance-based global sensitivity analysis technique that has been applied to computational models to assess the relative importance of input parameters on the output. This paper introduces new notation that describes the Sobol indices in terms of the Pearson correlation of outputs from pairs of runs, and introduces correction terms to remove some of the spurious correlation. A variety of estimation techniques are compared for accuracy and precision using the G function as a test case.

Modeled estimates of soil and dust ingestion rates for children.

Daily soil/dust ingestion rates typically used in exposure and risk assessments are based on tracer element studies, which have a number of limitations and do not separate contributions from soil and dust. This article presents an alternate approach of modeling soil and dust ingestion via hand and object mouthing of children, using EPAs SHEDS model. Results for children 3 to <6 years old show that mean and 95th percentile total ingestion of soil and dust values are 68 and 224 mg/day, respectively; mean from soil ingestion, hand-to-mouth dust ingestion, and object-to-mouth dust ingestion are 41 mg/day, 20 mg/day, and 7 mg/day, respectively. In general, hand-to-mouth soil ingestion was the most important pathway, followed by hand-to-mouth dust ingestion, then object-to-mouth dust ingestion. The variability results are most sensitive to inputs on surface loadings, soil-skin adherence, hand mouthing frequency, and hand washing frequency. The predicted total soil and dust ingestion fits a lognormal distribution with geometric mean = 35.7 and geometric standard deviation = 3.3. There are two uncertainty distributions, one below the 20th percentile and the other above. Modeled uncertainties ranged within a factor of 3-30. Mean modeled estimates for soil and dust ingestion are consistent with past information but lower than the central values recommended in the 2008 EPA Child-Specific Exposure Factors Handbook. This new modeling approach, which predicts soil and dust ingestion by pathway, source type, population group, geographic location, and other factors, offers a better characterization of exposures relevant to health risk assessments as compared to using a single value.

Quantifying children's aggregate (dietary and residential) exposure and dose to permethrin: application and evaluation of EPA's probabilistic SHEDS-Multimedia model

Reliable, evaluated human exposure and dose models are important for understanding the health risks from chemicals. A case study focusing on permethrin was conducted because of this insecticides widespread use and potential health effects. SHEDS-Multimedia was applied to estimate US population permethrin exposures for 3- to 5-year-old children from residential, dietary, and combined exposure routes, using available dietary consumption data, food residue data, residential concentrations, and exposure factors. Sensitivity and uncertainty analyses were conducted to identify key factors, pathways, and research needs. Model evaluation was conducted using duplicate diet data and biomonitoring data from multiple field studies, and comparison to other models. Key exposure variables were consumption of spinach, lettuce, and cabbage; surface-to-skin transfer efficiency; hand mouthing frequency; fraction of hand mouthed; saliva removal efficiency; fraction of house treated; and usage frequency. For children in households using residential permethrin, the non-dietary exposure route was most important, and when all households were included, dietary exposure dominated. SHEDS-Multimedia model estimates compared well to real-world measurements data; this exposure assessment tool can enhance human health risk assessments and inform childrens health research. The case study provides insights into childrens aggregate exposures to permethrin and lays the foundation for a future cumulative pyrethroid pesticides risk assessment.

A new method of longitudinal diary assembly for human exposure modeling.

Human exposure time-series modeling requires longitudinal time–activity diaries to evaluate the sequence of concentrations encountered, and hence, pollutant exposure for the simulated individuals. However, most of the available data on human activities are from cross-sectional surveys that typically sample 1 day per person. A procedure is needed for combining cross-sectional activity data into multiple-day (longitudinal) sequences that can capture day-to-day variability in human exposures. Properly accounting for intra- and interindividual variability in these sequences can have a significant effect on exposure estimates and on the resulting health risk assessments. This paper describes a new method of developing such longitudinal sequences, based on ranking 1-day activity diaries with respect to a user-chosen key variable. Two statistics, “D” and “A”, are targeted. The D statistic reflects the relative importance of within- and between-person variance with respect to the key variable. The A statistic quantifies the day-to-day (lag-one) autocorrelation. The user selects appropriate target values for both D and A. The new method then stochastically assembles longitudinal diaries that collectively meet these targets. On the basis of numerous simulations, the D and A targets are closely attained for exposure analysis periods >30 days in duration, and reasonably well for shorter simulation periods. Longitudinal diary data from a field study suggest that D and A are stable over time, and perhaps over cohorts as well. The new method can be used with any cohort definitions and diary pool assignments, making it easily adaptable to most exposure models. Implementation of the new method in its basic form is described, and various extensions beyond the basic form are discussed.

Statistical properties of longitudinal time-activity data for use in human exposure modeling

Understanding the longitudinal properties of the time spent in different locations and activities is important in characterizing human exposure to pollutants. The results of a four-season longitudinal time-activity diary study in eight working adults are presented, with the goal of improving the parameterization of human activity algorithms in EPA’s exposure modeling efforts. Despite the longitudinal, multi-season nature of the study, participant non-compliance with the protocol over time did not play a major role in data collection. The diversity (D)—a ranked intraclass correlation coefficient (ICC)— and lag-one autocorrelation (A) statistics of study participants are presented for time spent in outdoor, motor vehicle, residential, and other-indoor locations. Day-type (workday versus non-workday, and weekday versus weekend), season, temperature, and gender differences in the time spent in selected locations and activities are described, and D & A statistics are presented. The overall D and ICC values ranged from approximately 0.08–0.26, while the mean population rank A values ranged from approximately 0.19–0.36. These statistics indicate that intra-individual variability exceeds explained inter-individual variability, and low day-to-day correlations among locations. Most exposure models do not address these behavioral characteristics, and thus underestimate population exposure distributions and subsequent health risks associated with environmental exposures.

Comparison of four probabilistic models (CARES ® , Calendex™, ConsExpo, and SHEDS) to estimate aggregate residential exposures to pesticides

Two deterministic models (US EPAs Office of Pesticide Programs Residential Standard Operating Procedures (OPP Residential SOPs) and Draft Protocol for Measuring Childrens Non-Occupational Exposure to Pesticides by all Relevant Pathways (Draft Protocol)) and four probabilistic models (CARES®, Calendex™, ConsExpo, and SHEDS) were used to estimate aggregate residential exposures to pesticides. The route-specific exposure estimates for young children (2–5 years) generated by each model were compared to evaluate data inputs, algorithms, and underlying assumptions. Three indoor exposure scenarios were considered: crack and crevice, fogger, and flying insect killer. Dermal exposure estimates from the OPP Residential SOPs and the Draft Protocol were 4.75 and 2.37 mg/kg/day (crack and crevice scenario) and 0.73 and 0.36 mg/kg/day (fogger), respectively. The dermal exposure estimates (99th percentile) for the crack and crevice scenario were 16.52, 12.82, 3.57, and 3.30 mg/kg/day for CARES, Calendex, SHEDS, and ConsExpo, respectively. Dermal exposure estimates for the fogger scenario from CARES and Calendex (1.50 and 1.47 mg/kg/day, respectively) were slightly higher than those from SHEDS and ConsExpo (0.74 and 0.55 mg/kg/day, respectively). The ConsExpo derived non-dietary ingestion estimates (99th percentile) under these two scenarios were higher than those from SHEDS, CARES, and Calendex. All models produced extremely low exposure estimates for the flying insect killer scenario. Using similar data inputs, the model estimates by route for these scenarios were consistent and comparable. Most of the models predicted exposures within a factor of 5 at the 50th and 99th percentiles. The differences identified are explained by activity assumptions, input distributions, and exposure algorithms.

Modeling energy expenditure and oxygen consumption in human exposure models: accounting for fatigue and EPOC

Human exposure and dose models often require a quantification of oxygen consumption for a simulated individual. Oxygen consumption is dependent on the modeled individuals physical activity level as described in an activity diary. Activity level is quantified via standardized values of metabolic equivalents of work (METS) for the activity being performed and converted into activity-specific oxygen consumption estimates. However, oxygen consumption remains elevated after a moderate- or high-intensity activity is completed. This effect, which is termed excess post-exercise oxygen consumption (EPOC), requires upward adjustment of the METS estimates that follow high-energy expenditure events, to model subsequent increased ventilation and intake dose rates. In addition, since an individuals capacity for work decreases during extended activity, methods are also required to adjust downward those METS estimates that exceed physiologically realistic limits over time. A unified method for simultaneously performing these adjustments is developed. The method simulates a cumulative oxygen deficit for each individual and uses it to impose appropriate time-dependent reductions in the METS time series and additions for EPOC. The relationships between the oxygen deficit and METS limits are nonlinear and are derived from published data on work capacity and oxygen consumption. These modifications result in improved modeling of ventilation patterns, and should improve intake dose estimates associated with exposure to airborne environmental contaminants.