UC Berkeley

Although approximately 123 million people may be exposed to high levels of insecticides through the use of indoor residual spraying (IRS) for malaria control, very little data exists on exposure levels and risk to residents. In addition, certain populations may be more susceptible to the unintended health effects of insecticide exposure from IRS including the developing fetus. The aims of this dissertation were as follows: 1) build indoor transport and fate models to estimate insecticide exposure and risk to individuals living in homes sprayed for malaria control, 2) measure the serum levels of dichlorodiphenyl-trichloroethane (DDT) and its breakdown product dichlorodiphenyl-dichloroethylene (DDE) in a community where IRS occurs and test for intervenable characteristics to reduce exposure, and 3) evaluate the relationship between prenatal DDT and DDE exposure and children's cognitive development.

Of the 12 insecticides recommended for IRS by the World Health Organization (WHO), only DDT and deltamethrin have been measured longitudinally after IRS to assess exposure to residents of sprayed homes. In Chapter 2, we developed two dynamic indoor fugacity models, representing two of the common building types of homes in rural Africa, to track insecticide transport and fate in indoor compartments. For three age groups (birth to < 1 year old, 6 to 11 years old, and > 21 years old), we calculated insecticide uptake via inhalation, dust ingestion, and dermal absorption of dust and compared dose estimates with chronic and acute health-based benchmark doses set by the Food and Agriculture Organization and WHO. We accounted for model variability and uncertainty with Latin hypercube sampling. We found simulated indoor air concentrations, dust concentrations, and loading levels generally agreed with longitudinal measurements previously reported in homes sprayed with DDT and deltamethrin. While indoor air concentrations typically peaked on the day of IRS, dust concentrations often peaked days after IRS. At least one simulation of the average daily dose over the year after IRS for DDT, fenitrothion, pirimiphos-methyl, bendiocarb, and propoxur exceeded the Acceptable Daily Intake (ADI). The use of fenitrothion and DDT for IRS posed the greatest risk to residents of sprayed homes. For example, 67% and 33% of the simulated average dose estimates for fenitrothion and DDT exceeded the ADI in children less than one year old, respectively. None of the daily dose estimates exceeded the Acute Reference Doses. Results from this chapter indicate the feasibility and utility of using fugacity models to estimate exposure and risk to insecticides from IRS. In addition, model results indicate long-term home contamination from IRS and residents of IRS homes may be exposed to insecticides at levels that exceed chronic health-based benchmark doses. Given that dust concentrations often peaked days after IRS, residents should be informed that contact with the floor should be avoided during this period, especially for children with high hand-to-mouth behavior.

The use of DDT as an IRS insecticide has contributed to uniquely high DDT and DDE body burden in residents of sprayed homes. In Chapter 3, we described the Venda Health Examination of Mothers, Babies, and their Environment (VHEMBE) cohort and presented data on DDT and DDE serum concentrations measured in VHEMBE mothers. In addition, we applied targeted maximum likelihood estimation (TMLE) procedures to understand the change in marginal p,p'-DDT and p,p'-DDE body burden given seven hypothetical exposure interventions. A total of 751 mothers completed a baseline questionnaire and provided a serum sample. The majority of mothers enrolled in the VHEMBE cohort study were between 18 and 24 years of age (50.2%), completed at least grade 12 education (68.3%), lived below the South African poverty line of $40 per household member per month (58.3%), and were multiparous (56.8%). p,p'-DDT and p,p'-DDE serum concentrations were above the limit of quantification (LOQ) in 90.7 and 97.2% of the blood samples, respectively, while o,p'-DDT and o,p'-DDE serum concentrations were above the LOQ in 43.3 and 17.2% of the blood samples, respectively. Median (inter-quartile range) p,p'-DDT and p,p'-DDE serum concentrations were 56.8 (19.6-261.1) and 76.5 (27.9-271.5) ng/g-lipid, respectively. Of the seven interventions tested, three significantly reduced DDT and DDE exposures. If all VHEMBE mothers never lived in a DDT sprayed home, they would have 69.4% (95% CI: -76.3, -60.4) and 67.1% (95% CI: -73.6, -58.9) lower marginal p,p'-DDT and p,p'-DDE serum concentrations, respectively, than if all mothers ever lived in a DDT sprayed home. If all mothers lived in a household that wet mopped their floors at least seven times per week, they would have 25.2% (95% CI: -40.4, -6.1) and 21.9% (95% CI: -36.1, -4.6) lower marginal p,p'-DDT and p,p'-DDE serum concentrations, respectively, than if all mothers lived in a household that wet mops their floors less than seven times per week. In addition, if all mothers lived in a household with piped water, they would have 22.1% (95% CI: -38.0, -2.0) lower marginal p,p'-DDT serum concentrations than if all mothers lived in a household without piped water. Our findings suggest that DDT/E exposure is decreasing in IRS areas and several intervenable factors may exist to reduce DDT/E exposure in IRS communities.

Animal studies have shown DDT and DDE to be neurodevelopmental toxicants, but epidemiological studies have reported inconsistent findings between prenatal DDT and DDE exposure and child neurodevelopment. In Chapter 4, we investigated the association between prenatal DDT and DDE exposure and child neurodevelopment in the Center for the Health Assessment of Mothers and Children of Salinas (CHAMACOS) cohort study. We combined the original prospective CHAMACOS cohort with a retrospective cohort of mother/child pairs and estimated prenatal DDT and DDE exposure with measured or predicted maternal concentrations during pregnancy. Using generalized estimating equation and linear regression models, we evaluated the relationship of prenatal DDT and DDE exposure with Full Scale Intelligence Quotient (IQ) and 4 subtests (Working Memory, Perceptual Reasoning, Verbal Comprehension, and Processing Speed) assessed in children 7 and 10.5-years after birth using the Wechsler Intelligence Scale for Children (WISC). In the longitudinal analyses (n = 619), we found an inverse association between prenatal DDT and DDE exposure and Processing Speed scores (p-value < 0.05). In the cross sectional analyses when the children were 7 and 10.5 years old, prenatal DDT exposure was inversely associated with Processing Speed at age 7 years (n = 316), but prenatal DDT and DDE exposure were not associated with Full Scale IQ or any of the WISC subscales at age 10.5 years (n = 595). We found evidence for effect modification by sex as prenatal DDE exposure was inversely associated with Processing Speed in the longitudinal analysis and both Full Scale IQ and Processing Speed at age 7 years in females, but not males. We conclude that prenatal DDT and DDE exposure may be associated with delayed Processing Speed in school-aged children, but with no other intelligence metric, and that the child's sex may modify this relationship.