Multiple Prenatal Exposures and Acute Care Clinical Encounters for Asthma among Children Born to Mothers Living near a Superfund Site.

Prenatal exposures are associated with childhood asthma, and risk may increase with simultaneous exposures. Pregnant women living in lower-income communities tend to have elevated exposures to a range of potential asthma risk factors, which may interact in complex ways. We examined the association between prenatal exposures and the risk of childhood asthma acute care clinical encounters (hospitalization, emergency department visit, observational stay) using conditional logistic regression with a multivariable smooth to model the interaction between continuous variables, adjusted for maternal characteristics, and stratified by sex. All births near the New Bedford Harbor (NBH) Superfund site (2000-2006) were followed through 2011 using the Massachusetts Pregnancy to Early Life Longitudinal data system to identify children ages 5-11 with asthma acute care clinical encounters (265 cases among 7,787 with follow-up). Hazard ratios (HRs) were higher for children living closer to the NBH with higher cord blood Pb levels than children living further away from the NBH with lower Pb levels (P<0.001). HRs were highest for girls (HR=4.17, 95% CI: 3.60, 4.82) compared to boys (HR=1.72, 95% CI: 1.46, 2.02). Our results suggest that prenatal Pb exposure in combination with residential proximity to the NBH is associated with childhood asthma acute care clinical encounters.

Unmatched spatially stratified controls: A simulation study examining efficiency and precision using spatially-diverse controls and generalized additive models.

Unmatched spatially stratified random sampling (SSRS) of non-cases selects geographically balanced controls by dividing the study area into spatial strata and randomly selecting controls from all non-cases within each stratum. The performance of SSRS control selection was evaluated in a case study spatial analysis of preterm birth in Massachusetts. In a simulation study, we fit generalized additive models using controls selected by SSRS or simple random sample (SRS) designs. We compared mean squared error (MSE), bias, relative efficiency (RE), and statistically significant map results to the model results with all non-cases. SSRS designs had lower average MSE (0.0042-0.0044) and higher RE (77-80%) compared to SRS designs (MSE: 0.0072-0.0073; RE across designs: 71%). SSRS map results were more consistent across simulations, reliably identifying statistically significant areas. SSRS designs improved efficiency by selecting controls that are geographically distributed, particularly from low population density areas, and may be more appropriate for spatial analyses.

Association between traffic-related air pollution exposure and fertility-assisted births.

Previous studies have suggested that traffic-related air pollution is associated with adverse fertility outcomes, such as reduced fecundability and subfertility. The purpose of this research is to investigate if PM2.5 exposure prior to conception or traffic-related exposures (traffic density and distance to nearest major roadway) at birth address is associated with fertility-assisted births. We obtained all live and still births from the Massachusetts state birth registry with an estimated conception date between January 2002 through December 2008. All births requiring fertility drugs or assisted reproductive technology were identified as cases. We randomly selected 2000 infants conceived each year to serve as a common control group. PM2.5 exposure was assessed using 4 km spatial satellite remote sensing, meteorological and land use spatiotemporal models at geocoded birth addresses for the year prior to conception. The mean PM2.5 level was 9.81 µg m-3 (standard deviation = 1.70 µg m-3), with a maximum of 14.27 µg m-3. We calculated crude and adjusted fertility treatment odds ratios (ORs) and 95% confidence intervals (CI) per interquartile range of 1.72 µg m-3 increase in PM2.5 exposure. Our final analyses included 10 748 fertility-assisted births and 12 225 controls. After adjusting for parental age, marital status, race, maternal education, insurance status, parity, and year of birth, average PM2.5 exposure during the year prior to conception was weakly associated with fertility treatment (OR: 1.01; 95% CI: 0.97, 1.05). Fertility-assisted births were inversely associated with traffic density (highest quartile compared to lowest quartile, OR: 0.92; 95% CI: 0.83, 1.02) and positively associated with distance from major roadway (OR per 100 m: 1.01; 95% CI: 1.00, 1.02) in adjusted analyses. We did not find strong evidence to support an adverse relationship between traffic-related air pollution exposure and fertility-assisted births.

Role of mammography accessibility, deprivation and spatial effect in breast cancer screening participation in France: an observational ecological study.

BACKGROUND: The detection of cancer in its early latent stages can improve patients' chances of recovery and thereby reduce the overall burden of the disease. Our objectives were to investigate factors (geographic accessibility and deprivation level) affecting mammography screening participation variation and to determine how much geographic variation in participation rates can be explained by spillover effects between adjacent areas, while controlling for covariates. METHODS: Mammography screening participation rates between 2015 and 2016 were calculated by census blocks (CB), for women aged 50-74 years, residing in Lyon metropolitan area. Global spatial autocorrelation tests were applied to identify the geographic variation of participation. Spatial regression models were used to incorporate spatial structure to estimate associations between mammography participation rate and the combined effect (geographic accessibility and deprivation level) adjusting for modes of travel and social cohesion. RESULTS: The mammography participation rate was found to have a statistically significant and positive spatial correlation. The participation rate of one CB was significantly and positively associated with the participation rates of neighbouring CB. The participation was 53.2% in residential and rural areas and 46.6% in urban areas, p < 0.001. Using Spatial Lag models, whereas the population living in most deprived CBs have statistically significantly lower mammography participation rates than lower deprived ones, significant interaction demonstrates that the relation differs according to the degree of urbanization. CONCLUSIONS: This study makes an important methodological contribution in measuring geographical access and understanding better the combined effect of deprivation and the degree of urbanization on mammography participation and other contextual factors that affect the decision of using mammography screening services -which is a critical component of healthcare planning and equity.

Spatial-Temporal Trends in Ovarian Cancer Outcomes in California.

BACKGROUND: Research suggests that geographic location may affect ovarian cancer (OC) outcomes. Insurance status often remains an important predictor of outcomes. The Affordable Care Act was enacted in 2010 to expand access to affordable health insurance. Our objective was to examine spatiotemporal trends in OC treatment nonadherence and disease-specific mortality in California (USA) among women diagnosed with OC. METHODS: Newly diagnosed epithelial OC cases between 1996 and 2017 were identified from the California Cancer Registry. Spatiotemporal trends in adherence to treatment guidelines were examined using generalized additive models and OC-specific mortality using Cox proportional hazards additive models. Prediction grids covering California were used to display the odds ratios (ORs) and hazard ratios of location using the median value for the study area as the referent value. Seven overlapping 5-year periods and 2 larger ones (pre- and post-2013) were assessed. Analyses were stratified according to stage (early vs advanced) and used P = .05 to determine statistical significance. RESULTS: Statistically significant spatial patterns in treatment nonadherence were observed for every time period examined (P < .001). Odds of treatment nonadherence associated with geographic location were highest among women with early-stage OC in southern Los Angeles County during 2014-2017 (OR max = 3.89, confidence interval = 1.04 to 7.61). For women with advanced-stage OC, residing in northern California was generally associated with lower odds ratios, whereas southern California was associated with higher odds ratios, with higher odds in the latter time period (OR range = 0.53-1.84 in 1996-2012 vs 0.49-2.37 in 2013-2017). Geographic location was not a statistically significant predictor of mortality. CONCLUSIONS: Residential location was statistically significantly associated with treatment received in California, with spatial patterns varying over time but not OC-specific mortality. Changes in insurance status over time were accompanied by shifts in population demographics and increased travel distances to receive care.

Concentration of Cannabis and Tobacco Retailers in Los Angeles County, California: A Spatial Analysis of Potential Effects on Youth and Ethnic Minorities.

OBJECTIVE: Cannabis and tobacco retailers are believed to cluster in areas with more racial/ethnic minorities, which could account for the disproportionate use of blunts in Black and Hispanic communities. The current study examined the spatial relationship between cannabis and licensed tobacco retailers in Los Angeles County, California, and assessed whether various neighborhood and business factors influenced the spatial patterning. METHOD: Generalized additive models were used to test the association between the location of cannabis retailers (N = 429) and their accessibility potential (AP) to tobacco retailers (N = 8,033). The covariates included cannabis licensure status, median household income, population density, percentages of racial/ ethnic minorities and young adults (18-34), unemployment status, families living in poverty, minimum completion of high school/General Educational Development (GED) credential, and industrial businesses by census tract. RESULTS: The location of cannabis retailers was significantly associated with AP in all adjusted models (p < .005). The percentage of racial/ethnic minorities, age (18-34 years), and nonlicensure of cannabis retailers, which were positively correlated with AP (p < .05), confounded the association between AP and cannabis retailer location. CONCLUSIONS: The concentration of unlicensed cannabis retailers and tobacco retailers in young and racially/ethnically diverse neighborhoods may increase access to and use of cigarillos for blunt smoking. Jurisdictions within Los Angeles County should consider passing ordinances requiring minimum distances between cannabis and tobacco retailers.

Effect of socioeconomic factors during the early COVID-19 pandemic: a spatial analysis.

BACKGROUND: Spatial variability of COVID-19 cases may suggest geographic disparities of social determinants of health. Spatial analyses of population-level data may provide insight on factors that may contribute to COVID-19 transmission, hospitalization, and death. METHODS: Generalized additive models were used to map COVID-19 risk from March 2020 to February 2021 in Orange County (OC), California. We geocoded and analyzed 221,843 cases to OC census tracts within a Poisson framework while smoothing over census tract centroids. Location was randomly permuted 1000 times to test for randomness. We also separated the analyses temporally to observe if risk changed over time. COVID-19 cases, hospitalizations, and deaths were mapped across OC while adjusting for population-level demographic data in crude and adjusted models. RESULTS: Risk for COVID-19 cases, hospitalizations, and deaths were statistically significant in northern OC. Adjustment for demographic data substantially decreased spatial risk, but areas remained statistically significant. Inclusion of location within our models considerably decreased the magnitude of risk compared to univariate models. However, percent minority (adjusted RR: 1.06, 95%CI: 1.06, 1.07), average household size (aRR: 1.06, 95%CI: 1.05, 1.07), and percent service industry (aRR: 1.05, 95%CI: 1.04, 1.06) remained significantly associated with COVID-19 risk in adjusted spatial models. In addition, areas of risk did not change between surges and risk ratios were similar for hospitalizations and deaths. CONCLUSION: Significant risk factors and areas of increased risk were identified in OC in our adjusted models and suggests that social and environmental factors contribute to the spread of COVID-19 within communities. Areas in north OC remained significant despite adjustment, but risk substantially decreased. Additional investigation of risk factors may provide insight on how to protect vulnerable populations in future infectious disease outbreaks.

Arsenic in private well water and birth outcomes in the United States.

BACKGROUND: Prenatal exposure to drinking water with arsenic concentrations >50 µg/L is associated with adverse birth outcomes, with inconclusive evidence for concentrations ≤50 µg/L. In a collaborative effort by public health experts, hydrologists, and geologists, we used published machine learning model estimates to characterize arsenic concentrations in private wells-federally unregulated for drinking water contaminants-and evaluated associations with birth outcomes throughout the conterminous U.S. METHODS: Using several machine learning models, including boosted regression trees (BRT) and random forest classification (RFC), developed from measured groundwater arsenic concentrations of ∼20,000 private wells, we characterized the probability that arsenic concentrations occurred within specific ranges in groundwater. Probabilistic model estimates and private well usage data were linked by county to all live birth certificates from 2016 (n = 3.6 million). We evaluated associations with gestational age and term birth weight using mixed-effects models, adjusted for potential confounders and incorporated random intercepts for spatial clustering. RESULTS: We generally observed inverse associations with term birth weight. For instance, when using BRT estimates, a 10-percentage point increase in the probability that private well arsenic concentrations exceeded 5 µg/L was associated with a -1.83 g (95% CI: -3.30, -0.38) lower term birth weight after adjusting for covariates. Similarly, a 10-percentage point increase in the probability that private well arsenic concentrations exceeded 10 µg/L was associated with a -2.79 g (95% CI: -4.99, -0.58) lower term birth weight. Associations with gestational age were null. CONCLUSION: In this largest epidemiologic study of arsenic and birth outcomes to date, we did not observe associations of modeled arsenic estimates in private wells with gestational age and found modest inverse associations with term birth weight. Study limitations may have obscured true associations, including measurement error stemming from a lack of individual-level information on primary water sources, water arsenic concentrations, and water consumption patterns.

Predictors of Test Positivity, Mortality, and Seropositivity during the Early Coronavirus Disease Epidemic, Orange County, California, USA.

We conducted a detailed analysis of coronavirus disease in a large population center in southern California, USA (Orange County, population 3.2 million), to determine heterogeneity in risks for infection, test positivity, and death. We used a combination of datasets, including a population-representative seroprevalence survey, to assess the actual burden of disease and testing intensity, test positivity, and mortality. In the first month of the local epidemic (March 2020), case incidence clustered in high-income areas. This pattern quickly shifted, and cases next clustered in much higher rates in the north-central area of the county, which has a lower socioeconomic status. Beginning in April 2020, a concentration of reported cases, test positivity, testing intensity, and seropositivity in a north-central area persisted. At the individual level, several factors (e.g., age, race or ethnicity, and ZIP codes with low educational attainment) strongly affected risk for seropositivity and death.

A spatial analysis of birth defects in Texas, 1999-2011.

BACKGROUND: The etiologies of major birth defects are still unclear and few spatial analyses have been conducted in the United States. Spatial analyses of individual-level data can help elucidate environmental and social risk factors. METHODS: We used generalized additive models to analyze 52,955 cases of neural tube defects, congenital heart defects (CHDs), gastroschisis, and orofacial cleft defects, and sampled from 642,399 controls born between 1999 and 2011 in Texas. The effect of geographic location was measured using a bivariable smooth term of geocoded birth address within a logistic regression framework. We calculated and mapped odds ratios (ORs) and 95% confidence intervals (CIs) for birth defects subtypes across Texas, and adjusted for maternal characteristics, environmental indicators, and community-level covariates. We also performed time-stratified spatiotemporal analyses for more prevalent birth defects. RESULTS: Location was significantly associated with crude odds of all birth defects except hypoplastic left heart syndrome. After adjusting for maternal characteristics, environmental indicators, and community-level factors, ORs in many geographic areas were no longer statistically significant for most defects, especially CHDs. However, areas of significant and insignificant elevated risk remained for defects in all groups in North and South Texas, with ORs for ventricular septal defects increasing over time. Low risk of birth defects was often present in the northern part of East Texas. CONCLUSION: Significant spatial patterns of birth defects were identified and varied depending on adjustment of different categories of covariates. Further investigation of areas with increased risks may aid in our understanding of birth defects.

Ambient air pollution and ovarian cancer survival in California.

OBJECTIVE: To examine whether exposure to ambient ozone, particulate matter with diameter less than 2.5 µm (PM2.5), nitrogen dioxide (NO2), and distance to major roadways (DTR) impact ovarian cancer-specific survival, while considering differences by stage, race/ethnicity, and socioeconomic status. METHODS: Women diagnosed with epithelial ovarian cancer from 1996 to 2014 were identified through the California Cancer Registry and followed through 2016. Women's geocoded addresses were linked to pollutant exposure data and averaged over the follow-up period. Pollutants were considered independently and in multi-pollutant models. Cox proportional hazards models assessed hazards of disease-specific death due to environmental exposures, controlling for important covariates, with additional models stratified by stage at diagnosis, race/ethnicity and socioeconomic status. RESULTS: PM2.5 and NO2, but not ozone or DTR, were significantly associated with survival in univariate models. In a multi-pollutant model for PM2.5, ozone, and DTR, an interquartile range increase in PM2.5 (Hazard Ratio [HR], 1.45; 95% Confidence Interval [CI], 1.41-1.49) was associated with worse prognosis. Similarly, in the multi-pollutant model with NO2, ozone, and DTR, women with higher NO2 exposures (HR for 20.0-30.0 ppb, 1.30; 95% CI, 1.25-1.36 and HR for >30.0 ppb, 2.48; 95% CI, 2.32-2.66) had greater mortality compared to the lowest exposed (<20.0 ppb). Stratified results show the effects of the pollutants differed by race/ethnicity and were magnified among women diagnosed in early stages. CONCLUSIONS: Our analyses suggest that greater exposure to NO2 and PM2.5 may adversely impact ovarian cancer-specific survival, independent of sociodemographic and treatment factors. These findings warrant further study.

Critical review on PFOA, kidney cancer, and testicular cancer.

The carcinogenicity of perfluorooctanoic acid (PFOA) has been reviewed previously by several different regulatory agencies and researchers, with contradictory conclusions-especially regarding epidemiological findings on kidney cancer and testicular cancer. In addition, previous dose-response assessments have focused primarily on evidence from animal studies. This critical review summarizes peer reviewed epidemiological studies on PFOA and cancers of the kidneys and testes, using modified Hill's criteria to assess the evidence for causation. We converted exposures to a common scale based on serum PFOA concentrations and applied meta-analysis to estimate the average increase in cancer risk reported by the studies with sufficient information to estimate serum PFOA. Using random effects meta-analysis, we found that the average relative increase in cancer risk per 10 ng/mL increase in serum PFOA for these studies is 16% (95% CI: 3%, 30%) for kidney cancer and 3% (95% CI: 2%, 4%) for testicular cancer. These associations are most likely causal, but results are limited by the small number of studies for testicular cancer, the overlapping study populations for several studies, and the lack of measured or modeled serum PFOA concentrations for several studies.Implications: Our review meta-analysis indicates an average increase in cancer risk per 10 ng/mL increase in serum PFOA for kidney and testicular cancers. These associations are most likely causal, but results are limited by the small number of studies for testicular cancer, the overlapping study populations for several studies, and the lack of measured or modeled serum PFOA concentrations for several studies. The weight of evidence could be even stronger with the addition of future studies conducted in large cohorts.

Estimated seroprevalence of SARS-CoV-2 antibodies among adults in Orange County, California.

Clinic-based estimates of SARS-CoV-2 may considerably underestimate the total number of infections. Access to testing in the US has been heterogeneous and symptoms vary widely in infected persons. Public health surveillance efforts and metrics are therefore hampered by underreporting. We set out to provide a minimally biased estimate of SARS-CoV-2 seroprevalence among adults for a large and diverse county (Orange County, CA, population 3.2 million). We implemented a surveillance study that minimizes response bias by recruiting adults to answer a survey without knowledge of later being offered SARS-CoV-2 test. Several methodologies were used to retrieve a population-representative sample. Participants (n = 2979) visited one of 11 drive-thru test sites from July 10th to August 16th, 2020 (or received an in-home visit) to provide a finger pin-prick sample. We applied a robust SARS-CoV-2 Antigen Microarray technology, which has superior measurement validity relative to FDA-approved tests. Participants include a broad age, gender, racial/ethnic, and income representation. Adjusted seroprevalence of SARS-CoV-2 infection was 11.5% (95% CI: 10.5-12.4%). Formal bias analyses produced similar results. Prevalence was elevated among Hispanics (vs. other non-Hispanic: prevalence ratio [PR] = 1.47, 95% CI 1.22-1.78) and household income < $50,000 (vs. > $100,000: PR = 1.42, 95% CI: 1.14 to 1.79). Results from a diverse population using a highly specific and sensitive microarray indicate a SARS-CoV-2 seroprevalence of ~ 12 percent. This population-based seroprevalence is seven-fold greater than that using official County statistics. In this region, SARS-CoV-2 also disproportionately affects Hispanic and low-income adults.

Prenatal environmental exposures and associations with teen births.

BACKGROUND: Children's prenatal exposure to multiple environmental chemicals may contribute to subsequent deficits in impulse control, predisposing them to risk-taking. OBJECTIVE: Our goal was to investigate associations between prenatal exposure mixtures and risk of teen birth, a manifestation of high-risk sexual activity, among 5865 girls (1st generation) born in southeast Massachusetts from 1992-1998. METHODS: Exposures included prenatal modeled polychlorinated biphenyls (PCBs), ρ,ρ'-dichlorodiphenyl dichloroethylene (DDE), hexachlorobenzene (HCB), lead (Pb), and mercury (Hg). We fit adjusted generalized additive models with multivariable smooths of exposure mixtures, 1st generation infant's birth year, and maternal age at 1st generation birth. Predicted odds ratios (ORs) for teen birth were mapped as a function of joint exposures. We also conducted sensitivity analyses among 1st generation girls with measured exposure biomarkers (n = 371). RESULTS: The highest teen birth risk was associated with a mixture of high prenatal HCB, Hg, Pb, and PCB, but low DDE exposure, with similar associations in sensitivity analyses. The highest OR predicted for girls born in 1995 to mothers of median age (26 years) was at the 95th percentile of the HCB and PCB exposure distributions (OR = 3.09; 95% confidence interval: 0.29, 32.4). Additionally, girls born earlier in the study period or to teen mothers were at increased risk of teen birth. SIGNIFICANCE: Prenatal environmental chemical exposures and sociodemographic characteristics may interact to substantially increase risk of teen births.

Birth defects and unconventional natural gas developments in Texas, 1999-2011.

Unconventional natural gas developments (UNGD) may release air and water pollutants into the environment, potentially increasing the risk of birth defects. We conducted a case-control study evaluating 52,955 cases with birth defects and 642,399 controls born between 1999 and 2011 to investigate the relationship between UNGD exposure and the risk of gastroschisis, congenital heart defects (CHD), neural tube defects (NTDs), and orofacial clefts in Texas. We calculated UNGD densities (number of UNGDs per area) within 1, 3, and 7.5 km of maternal address at birth and categorized exposure by density tertiles. For CHD subtypes with large case numbers, we also performed time-stratified analyses to examine temporal trends. We calculated adjusted odds ratios (aOR) and 95% confidence intervals (CI) for the association with UNGD exposure, accounting for maternal characteristics and neighborhood factors. We also included a bivariable smooth of geocoded maternal location in an additive model to account for unmeasured spatially varying risk factors. Positive associations were observed between the highest tertile of UNGD density within 1 km of maternal address and risk of anencephaly (aOR: 2.44, 95% CI: 1.55, 3.86), spina bifida (aOR: 2.09, 95% CI: 1.47, 2.99), gastroschisis among older mothers (aOR: 3.19, 95% CI: 1.77, 5.73), aortic valve stenosis (aOR: 1.90, 95% CI: 1.33, 2.71), hypoplastic left heart syndrome (aOR: 2.00, 95% CI: 1.39, 2.86), and pulmonary valve atresia or stenosis (aOR: 1.36, 95% CI: 1.10, 1.66). For CHD subtypes, results did not differ substantially by distance from maternal address or when residual confounding was considered, except for atrial septal defects. We did not observe associations with orofacial clefts. Our results suggest that UNGDs were associated with some CHDs and possibly NTDs. In addition, we identified temporal trends and observed presence of spatial residual confounding for some CHDs.

Assessing the relation of chemical and non-chemical stressors with risk-taking related behavior and adaptive individual attributes among adolescents living near the New Bedford Harbor Superfund site.

BACKGROUND: Early life exposure to neurotoxicants and non-chemical psychosocial stressors can impede development of prefrontal cortical functions that promote behavioral regulation and thereby may predispose to adolescent risk-taking related behaviors (e.g., substance use or high-risk sexual activity). This is particularly concerning for communities exposed to multiple stressors. METHODS: This study examined the relation of exposure to mixtures of chemical stressors, non-chemical psychosocial stressors, and other risk factors with neuropsychological correlates of risk-taking. Specifically, we assessed psychometric measures of both adverse behavioral regulation and adaptive attributes among adolescents (age ∼ 15 years) in the New Bedford Cohort (NBC), a sociodemographically diverse cohort of 788 children born 1993-1998 to mothers residing near the New Bedford Harbor Superfund site. The NBC includes biomarkers of prenatal exposure to organochlorines and metals; sociodemographic, parental and home characteristics; and periodic neurodevelopmental assessments. We modelled exposure mixtures using multi-dimensional smooths within generalized additive models. RESULTS: Children of younger mothers with lower IQ who were exposed prenatally to higher polychlorinated biphenyls and lead had poorer anger control. This pattern was not apparent for children of older mothers with higher IQs. Direction of associations between increased hyperactivity and prenatal levels of organochlorine mixtures differed by maternal age and depression symptoms. Higher cord blood Pb levels, in conjunction with poorer HOME scores, were associated with poorer self-esteem when mothers had fewer depression symptoms. CONCLUSIONS: Analyses suggest that prenatal chemical exposures and non-chemical factors interact to contribute to neuropsychological correlates of risk-taking behaviors in adolescence. By simultaneously considering multiple factors associated with adverse behavioral regulation, we identified potential high-risk combinations that reflect both chemical and psychosocial stressors amenable to intervention.

Contribution of socioeconomic and environmental factors to geographic disparities in breast cancer risk in the Nurses' Health Study II.

BACKGROUND: Evidence of geographic disparities in breast cancer incidence within the U.S. and spatial analyses can provide insight into the potential contribution of environmental exposures or other geographically-varying factors to these disparities. METHODS: We applied generalized additive models (GAMs) to smooth geocoded residential coordinates while adjusting for covariates. Our analysis included 3,478 breast cancer cases among 24,519 control women from the Nurses' Health Study II (NHSII). We first examined associations with residential location during adolescence (high school address) or early adulthood (address in 1991). We then assessed the contribution from known individual-level risk factors, measures of socioeconomic status (SES), and occupational and environmental factors that vary spatially and have been linked to breast cancer. Secondary analyses by estrogen receptor (ER) and menopausal status were also conducted. RESULTS: We identified geographic patterns of breast cancer risk associated with location during adolescence, with increased risk apparent in Michigan, the Northwest, and the New York City area, that shifted to southern New England when addresses during early adulthood were analyzed. Similar results were observed after adjustment for individual- and area-level factors, although spatial associations were no longer statistically significant. CONCLUSION: Breast cancer risk is not spatially uniform across the U.S. and incidence patterns varied depending on the timing during life of the residence considered. Geographic disparities persisted even after accounting for established and suspected breast cancer risk factors, suggesting that unmeasured environmental or lifestyle risk factors may explain geographic variation in risk in different parts of the country.

Ovarian cancer in California: Guideline adherence, survival, and the impact of geographic location, 1996-2014.

BACKGROUND: Evidence suggests that geographic location may independently contribute to ovarian cancer survival. We aimed to investigate how the association between residential location and ovarian cancer-specific survival in California varies by race/ethnicity and socioeconomic status. METHODS: Additive Cox proportional hazard models were used to predict hazard ratios (HRs) and 95% confidence intervals (CI) for the association between geographic location throughout California and survival among 29,844 women diagnosed with epithelial ovarian cancer between 1996 and 2014. We conducted permutation tests to determine a global P-value for significance of location. Adjusted analyses considered distance traveled for care, distance to closest high-quality-of-care hospital, and receipt of National Comprehensive Cancer Network guideline care. Models were also stratified by stage, race/ethnicity, and socioeconomic status. RESULTS: Location was significant in unadjusted models (P = 0.009 among all stages) but not in adjusted models (P = 0.20). HRs ranged from 0.81 (95% CI: 0.70, 0.93) in Southern Central Valley to 1.41 (95% CI: 1.15, 1.73) in Northern California but were attenuated after adjustment (maximum HR = 1.17, 95% CI: 1.08, 1.27). Better survival was generally observed for patients traveling longer distances for care. Associations between survival and proximity to closest high-quality-of-care hospitals were null except for women of lowest socioeconomic status living furthest away (HR = 1.22, 95% CI: 1.03, 1.43). CONCLUSIONS: Overall, geographic variations observed in ovarian cancer-specific survival were due to important predictors such as receiving guideline-adherent care. Improving access to expert care and ensuring receipt of guideline-adherent treatment should be priorities in optimizing ovarian cancer survival.

A stratified generalized additive model and permutation test for temporal heterogeneity of smoothed bivariate spatial effects.

Generalized additive models (GAMs) with bivariate smoothers are frequently used to map geographic disease risks in epidemiology studies. A challenge in identifying health disparities has been the lack of intuitive and computationally feasible methods to assess whether the pattern of spatial effects varies over time. In this research, we accommodate time-stratified smoothers into the GAM framework to estimate time-specific spatial risk patterns while borrowing information from confounding effects across time. A backfitting algorithm for model estimation is proposed along with a permutation testing framework for assessing temporal heterogeneity of geospatial risk patterns across two or more time points. Simulation studies show that our proposed permuted mean squared difference (PMSD) test performs well with respect to type I error and power in various settings when compared with existing methods. The proposed model and PMSD test are used geospatial risk patterns of patent ductus arteriosus (PDA) in the state of Massachusetts over 2003-2009. We show that there is variation over time in spatial patterns of PDA risk, adjusting for other known risk factors, suggesting the presence of potential time-varying and space-related risk factors other than the adjusted ones.

Dioxin exposure and breast cancer risk in a prospective cohort study.

BACKGROUND: Dioxins are persistent organic pollutants generated from industrial combustion processes such as waste incineration. To date, results from epidemiologic studies of dioxin exposure and breast cancer risk have been mixed. OBJECTIVES: To prospectively examine the association between ambient dioxin exposure using a nationwide spatial database of industrial dioxin-emitting facilities and invasive breast cancer risk in the Nurses' Health Study II (NHSII). METHODS: NHSII includes female registered nurses in the US who have completed self-administered biennial questionnaires since 1989. Incident invasive breast cancer diagnoses were self-reported and confirmed by medical record review. Dioxin exposure was estimated based on residential proximity, duration of residence, and emissions from facilities located within 3, 5, and 10 km around geocoded residential addresses updated throughout follow-up. Cox regression models adjusted for breast cancer risk factors were used to calculate hazard ratios (HRs) and 95% confidence intervals (CIs). RESULTS: From 1989 to 2013, 3840 invasive breast cancer cases occurred among 112,397 participants. There was no association between residential proximity to any dioxin facilities (all facilities combined) and breast cancer risk overall. However, women who resided within 10 km of any municipal solid waste incinerator (MSWI) compared to none had increased breast cancer risk (adjusted HR = 1.15, 95% CI: 1.03, 1.28), with stronger associations noted for women who lived within 5 km (adjusted HR = 1.25, 95% CI: 1.04, 1.52). Positive associations were also observed for longer duration of residence and higher dioxin emissions from MSWIs within 3, 5, and 10 km. There were no clear differences in patterns of association for ER + vs. ER-breast cancer or by menopausal status. DISCUSSION: Results from this study support positive associations between dioxin exposure from MSWIs and invasive breast cancer risk.