Document 8ZZ1JJJJm1GarEY6K0Z2LORa

MICHIGAN DIOXIN EXPOSURE STUDY FACTORS THAT PREDICT SERUM DIOXIN CONCENTRATIONS IN MICHIGAN, USA Garabrant D1, Hong B1, Chen Q2, Franzblau A1, Lepkowski J3, Adriaens P4, Demond A4, Hedgeman E1, Knutson K1, Zwica L1, Chang C-W1, Lee S-Y2, Olson K3, Towey T4, Trin H4, Wenger Y1, Luksemburg W5, Maier M5, Gillespie BW2 1Department of Environmental Health Sciences, University of Michigan School of Public Health, 109 S Observatory, Ann Arbor, MI 48109; 2Department of Biostatistics, University of Michigan School of Public Health, 109 S Observatory, Ann Arbor, MI 48109; 3Survey Research Center, Institute for Social Research, University of Michigan, 426 Thompson Street, Ann Arbor, Michigan 48104; 4Department of Civil and Environmental Engineering, University of Michigan College of Engineering, 1351 Beal, Ann Arbor, MI 48109; 5Vista Laboratories, El Dorado Hills, CA 95762 Abstract Linear regression models were performed to identify factors that explain variation in serum TEQ and 2,3,7,8TCDD concentrations measured from the 946 participants in the University of Michigan Dioxin Exposure Study (UMDES). The regression analyses accounted for sampling weights, stratification, and clustering to insure the inferences from the regression models were applicable to the population from which participants were selected. We found that demographic factors were by far the most important contributors to the population variation in both serum TEQ and TCDD. Residing in Midland/Saginaw and living with contaminated soil and household dust contributed little to the variation in serum TEQ, but explained small percentages of the serum TCDD. Food consumption was a small contribution to the variation in serum TEQ, but was slightly larger for TCDD. Introduction The University of Michigan Dioxin Exposure Study (UMDES) was undertaken in response to concerns among the population of Midland and Saginaw Counties (Michigan, USA) that the discharge of dioxin-like compounds from the Dow Chemical Company facilities in Midland, Michigan (USA) has resulted in contamination of soils in the Tittabawassee River flood plain and areas of the City of Midland, leading to an increase in residents' body burdens of PCDDs, PCDFs and PCBs1. To understand the factors that predict residents' body burdens, 946 people, sampled from five geographically-defined populations by using a two-stage area probability household sample design, participated in an interview and gave blood samples for analysis of the WHO 29 dioxin-like compounds3. Soil and household dust samples were analyzed for the same set of congeners. The objective of this presentation is to discuss the factors that explain variation in the blood serum dioxin concentrations in this population. Materials and Methods The entire protocol for the University of Michigan Dioxin Exposure Study can be found on our study website.2 Briefly, adults age 18 and over who had lived in their current residence for five or more years were eligible to participate. Eligible subjects were randomly selected from the populations of five counties in Michigan, USA and invited to complete an interview, donate an 80 milliliter whole blood sample, have their household dust collected, and have their soil sampled. Three counties (Midland, Saginaw, and part of Bay Counties, MI) were chosen because of their proximity to the Dow Chemical Company and two counties (Jackson and Calhoun Counties, MI) were chosen as a reference population. Serum, household dust, and soil were analyzed for the 29 congeners recognized by the World Health Organization as having dioxin-like activity, including TCDD. Samples that fell below the limit of detection were estimated using LOD/2. All serum results are lipid adjusted and survey weighted to reflect the entire referent population region. Multiple imputation procedures were used (five imputations) to impute missing values in explanatory variables. Organohalogen Compounds Vol 69 (2007) O-047 206 MICHIGAN DIOXIN EXPOSURE STUDY Figure1. Linear regression modeling strategy A multi-stage backwards selection was applied to identify factors that predict the serum dioxin levels (Figure 1). In each step of backwards selection: (1) The point estimates from the 5 imputed datasets were averaged to get a combined estimate. (2) The combined p-values were calculated across the 5 imputed dataset in a manner that accounted for variation between datasets and variation within datasets. (3) The backwards selection then eliminated from the pool of covariates the least significant variable (the variable having the largest combined p-value). This procedure was repeated until all variables with p-values greater than the significance level were eliminated from the model. For the initial backwards selection step, variables were kept in the model if the pvalue was < 0.1. For all subsequent steps, variables were kept in the model if the p-value was < 0.05. Results (Table 2) are presented for TEQ (based on WHO 2005 TEFs) and for 2,3,7,8-TCDD. All regression analyses were performed using SAS version 9.1. Results The overall model explained 73% of the variance in the serum TEQ and 67% of the variance in TCDD. Demographic factors (age, age2, BMI, BMI loss in the past 12 months, gender, months a woman breast fed, pack-years of smoking, and the interaction term for gender*age) were the most important predictors of both TEQ and TCDD (explaining 40% and 30% of the variance in the serum TEQ and TCDD, respectively) (see Table 1). The region variable, representing the population from which each participant was selected (floodplain, near flood plain, out of flood plain, or Midland plume), was forced into the model as a 0,1 variable to examine whether there was a difference in the mean serum TEQ in comparison to the Jackson/Calhoun (J/C) population, after adjustment for all other factors. The results showed no significant differences. However, living in any area of Midland and Saginaw counties in 1960-1979 was associated with increased serum TEQ (the parameter estimate of 0.0029 indicates that the log10 serum TEQ increased by 0.0029 pg/g for each year of residence). Similar results were seen for TCDD. Table 1: Adjusted R2 Contribution to Adjusted R2 (%) Overall model Demographic factors Residence factors Soil and household dust Property use factors Food consumption, fishing, and hunting TEQ-2005 72.56 40.32 0.22 0.15 2.21 1.83 2,3,7,8-TCDD 67.17 29.55 2.85 0.53 1.25 2.64 A principal goal of this study was to determine whether soil or household dust contamination was associated with increased serum TEQ. The top 1" house perimeter soil showed no significant relationship to the serum TEQ or TCDD. The garden soil TCDD was significantly related to serum TCDD. The highest TEQ level Organohalogen Compounds Vol 69 (2007) O-047 207 MICHIGAN DIOXIN EXPOSURE STUDY found in any soil sample on each property (referred to as the maximum soil concentration) was significantly associated with the increased serum TEQ but the magnitude of this association was small (parameter estimate of 0.0000099). Water activities on the Tittabawassee River were positively associated with serum TEQ and TCDD. Consumption of fish from the contaminated area (Tittabawassee River, Saginaw River, Saginaw Bay) showed little relationship to the serum TEQ but showed a clear association with serum TCDD. Fishing in 1980-2005 around the Saginaw River and Bay was positively associated with serum TEQ and TCDD. Table 2: Results of important predictors from the linear regression model with the outcome variable of log10 Serum dioxin concentration. Blue-shaded values indicate significant positive estimates and yellow shaded values indicate significant negative estimates (p < 0.05). TEQ-2005 2,3,7,8-TCDD Important Predictors Estimate P-value Estimate P-value Demographics Age at interview Age2 BMI (Unit: kg/m2) 0.0203 -0.0001 0.0073 0.000 0.000 0.003 0.0135 0.0000 0.0051 0.001 0.316 0.015 BMI loss in the past 12 months 0.0106 0.000 0.0182 0.003 Gender (1 for female; 0 for male) 0.1033 0.229 -0.3760 0.000 Num. of months the first child was breast-fed -0.0051 0.001 -0.0104 0.002 Num. of months for all children except first one were breast-fed -0.0018 0.009 Pack-years of smoking -0.0021 0.000 -0.0038 0.000 At least High school graduate (Y vs. N) -0.1142 0.001 The participant is white in addition to being Hispanic (Y vs. N) -0.1113 0.010 Interaction term: Gender x Age 0.0033 0.000 0.0097 0.000 Interaction term: BMI x Gender -0.0083 0.004 Residence M/S Floodplain vs. Jackson/Calhoun -0.0337 0.223 -0.0235 0.634 M/S Near Floodplain vs. Jackson/Calhoun -0.0051 0.851 0.0851 0.087 M/S Out Floodplain vs. Jackson/Calhoun -0.0251 0.244 -0.0059 0.884 M/S Plume vs. Jackson/Calhoun -0.0374 0.183 -0.0012 0.988 Num. of yrs lived in Midland/Saginaw in 1940-59 0.0059 0.012 Num. of yrs lived in Midland/Saginaw in 1960-79 0.0029 0.036 0.0084 0.005 Soil and household dust Soil dioxin concentrations for house perimeter 0-1" -1.7E-04 0.190 -9.8E-04 0.653 Soil dioxin concentrations for soil contact 0-6" 5.3E-04 0.163 7.1E-03 0.001 Maximum soil concentration Household dust dioxin loading pg/m2 9.9E-06 -2.3E-05 0.009 0.041 Property Use Num. of yrs in 40-59 lived in a farm or property where crops, livestock or poultry were raised 0.0064 0.001 Num. of yrs in 40-59 lived in a property ever damaged by a fire -0.0788 0.041 -0.1321 0.024 Num. of yrs in 60-79 lived in a property ever damaged by a fire 0.0306 0.000 0.0289 0.001 Num. of yrs in 40-59 lived in a property where trash or yard waste was burned 0.0073 0.007 Num. of yrs in 60-79 used weed killers on the property -0.0054 0.001 Personally did work in the flower or other garden (Y vs. N) -0.0293 0.036 Flood waters from the Tittabawassee R. entered into the home (Y vs. N) 0.1131 0.018 Water Activities Did water activities near the Tittabawassee R. in 60-79 (1 per month vs. never) 0.2516 0.012 0.2992 0.045 Did water activities near the Tittabawassee R. after 80 (1 per month vs. never) 0.2404 0.024 Did water activities near the Tittabawassee R. after 80 (<1 per month but ever did vs. never) 0.1169 0.001 Did water activities near the Saginaw R. or Bay after 80 (1 per month vs. never) -0.3186 0.000 Food consumption, Fishing and Hunting Activities Num. of yrs ate fish after 80 0.0056 0.000 Num. of yrs ate fish caught from the Tittabawassee R., Saginaw R. or Bay after 80 0.0083 0.001 Ate walleye or perch that were caught from the Saginaw R. or Bay during the last 5 years (1 per month vs. never) -0.2583 0.001 Ate walleye or perch that were caught from the Saginaw R. or Bay during the last 5 years (<1 per month but ever ate vs. never) -0.1583 0.019 Organohalogen Compounds Vol 69 (2007) O-047 208 MICHIGAN DIOXIN EXPOSURE STUDY Important Predictors Ate walleye or perch that were caught from the Kalamazoo R., somewhere else, store-bought or bought in a restaurant during the last 5 yrs (1 per month vs. never) Ate walleye or perch that were caught from the Kalamazoo R., somewhere else, store-bought or bought in a restaurant during the last 5 yrs (<1 per month but ever ate vs. never) Ate any fish other than walleye and perch that were caught from the Saginaw R. or Bay during the last 5 years (1 per month vs. never) Ate the skin of the Wild Turkey, Pheasant, Grouse, Quail, or Woodcock during the last 5 yrs (Y vs. N) Did fishing activities in the Saginaw R. or Bay after 80 (1 per month vs. never) Did hunting activities near the Saginaw R. or Bay in 60-79 (Y vs. N) Did hunting activities in the surrounding areas of the Saginaw R. or Bay after 80 (1 per month vs. never) Did hunting activities in the surrounding areas of the Saginaw R. or Bay after 80 ( < 1 per month but ever did vs. never) Ate eggs, milk or other dairy products from cows that were home-raised in the Tittabawassee R. during the last 5 yrs (1 per month vs. never) TEQ-2005 Estimate P-value 0.0622 0.001 0.0528 0.023 -0.2873 0.000 0.0608 0.0959 0.1200 -0.2244 0.032 0.001 0.012 0.001 -0.0894 0.043 2,3,7,8-TCDD Estimate P-value 0.1845 0.002 -0.2604 -0.1549 0.2109 0.109 0.042 0.016 Discussion This study was large and was capable of finding small associations that are statistically significant. Inferences regarding these associations should include consideration not only of the statistical significance of the parameter estimate, but also the magnitude of the effect, and the amount of variance in serum TEQ explained by the factor. A number of the significant findings above are both small in magnitude and explain little variation in serum TEQ. We found that demographic factors were by far the most important contributors to the population variation in both serum TEQ and TCDD. Residing in Midland/Saginaw and living with contaminated soil and household dust contributed very little to the variation in serum TEQ, but explained small percentages of the serum TCDD. Food consumption was a small contribution to the variation in serum TEQ, but was slightly larger for TCDD. Acknowledgements Financial support for this study comes from the Dow Chemical Company through an unrestricted grant to the University of Michigan. The authors acknowledge Ms. Sharyn Vantine for her continued assistance and Drs. Linda Birnbaum, Ron Hites, Paolo Boffetta and Marie Haring Sweeney for their guidance as members of our Scientific Advisory Board. References 1. Franzblau A, Garabrant D, Adriaens P, Gillespie BW, Demond A, Olson K, Ward B, Hedgeman E, Knutson K, Zwica L, Towey T, Chen Q, Ladronka K, Sinibaldi J, Chang S-C, Lee S-Y, Gwinn D, Sima C, Swan S, Lepkowski J. Organohalogen Comp 2006; 68:205 2. Garabrant DH, Franzblau A, Gillespie B, Lin X, Lepkowski J, Adriaens P, Demond A. The University of Michigan Dioxin Exposure Study. Study Protocol. www.umdioxin.org. 3. Van den Berg M, Birnbaum L, Bosveld ATC, Brunstrom B, Cook P, Feeley M, Giesy JP, Hanberg A, Hasegawa R, Kennedy SW, Kubiak T, Larsen JC, Van Leeuwen FXR, Liem AKD, Nolt C, Peterson RE, Poellinger L, Safe S, Schrenk D, Tillitt D, Tysklind M, Younes M, Waern F, Zacharewski T. Environmental Health Perspectives 1998; 106:775. Organohalogen Compounds Vol 69 (2007) O-047 209