Joshua L. Warren

Spatial patterning of supermarkets and fast food outlets with respect to neighborhood characteristics.

A large body of literature has reported differences in exposure to environments supporting either healthy (e.g. supermarkets) or unhealthy (e.g. fast food outlets) dietary choices by neighborhood characteristics. We explored the associations of both supermarkets and fast food outlets availability with neighborhood characteristics, and clustering of these two outlet types in a largely rural state. Compared to block groups without a supermarket, those with a supermarket had a significantly higher income, higher housing value, larger population with high school education and above, lower minority population and lower population living below poverty even after controlling for urbanicity and population density of census block groups. Surprisingly, a similar relationship was found for block groups with and without fast food outlets. This was due to spatial co-occurrence and clustering of fast food outlets around supermarket locations. Hence, future studies exploring the associations of food environment with diet or diet-related health outcome should concurrently examine all aspects of food environment (healthy and unhealthy).

Spatial‐Temporal Modeling of the Association between Air Pollution Exposure and Preterm Birth: Identifying Critical Windows of Exposure

Exposure to high levels of air pollution during the pregnancy is associated with increased probability of preterm birth (PTB), a major cause of infant morbidity and mortality. New statistical methodology is required to specifically determine when a particular pollutant impacts the PTB outcome, to determine the role of different pollutants, and to characterize the spatial variability in these results. We develop a new Bayesian spatial model for PTB which identifies susceptible windows throughout the pregnancy jointly for multiple pollutants (PM(2.5) , ozone) while allowing these windows to vary continuously across space and time. We geo-code vital record birth data from Texas (2002-2004) and link them with standard pollution monitoring data and a newly introduced EPA product of calibrated air pollution model output. We apply the fully spatial model to a region of 13 counties in eastern Texas consisting of highly urban as well as rural areas. Our results indicate significant signal in the first two trimesters of pregnancy with different pollutants leading to different critical windows. Introducing the spatial aspect uncovers critical windows previously unidentified when space is ignored. A proper inference procedure is introduced to correctly analyze these windows.

Association between arsenic, cadmium, manganese, and lead levels in private wells and birth defects prevalence in North Carolina: a semi-ecologic study

BackgroundToxic metals including arsenic, cadmium, manganese, and lead are known human developmental toxicants that are able to cross the placental barrier from mother to fetus. In this population-based study, we assess the association between metal concentrations in private well water and birth defect prevalence in North Carolina.MethodsA semi-ecologic study was conducted including 20,151 infants born between 2003 and 2008 with selected birth defects (cases) identified by the North Carolina Birth Defects Monitoring Program, and 668,381 non-malformed infants (controls). Maternal residences at delivery and over 10,000 well locations measured for metals by the North Carolina Division of Public Health were geocoded. The average level of each metal was calculated among wells sampled within North Carolina census tracts. Individual exposure was assigned as the average metal level of the census tract that contained the geocoded maternal residence. Prevalence ratios (PR) with 95% confidence intervals (CI) were calculated to estimate the association between the prevalence of birth defects in the highest category (≥90th percentile) of average census tract metal levels and compared to the lowest category (≤50th percentile).ResultsStatewide, private well metal levels exceeded the EPA Maximum Contaminant Level (MCL) or secondary MCL for arsenic, cadmium, manganese, and lead in 2.4, 0.1, 20.5, and 3.1 percent of wells tested. Elevated manganese levels were statistically significantly associated with a higher prevalence of conotruncal heart defects (PR: 1.6 95% CI: 1.1-2.5).ConclusionsThese findings suggest an ecologic association between higher manganese concentrations in drinking water and the prevalence of conotruncal heart defects.

Assessment of critical exposure and outcome windows in time-to-event analysis with application to air pollution and preterm birth study

In reproductive epidemiology, there is a growing interest to examine associations between air pollution exposure during pregnancy and the risk of preterm birth (PTB). One important research objective is to identify critical periods of exposure and estimate the associated effects at different stages of pregnancy. However, population studies have reported inconsistent findings. This may be due to limitations from the standard analytic approach of treating PTB as a binary outcome without considering time-varying exposures together over the course of pregnancy. To address this research gap, we present a Bayesian hierarchical model for conducting a comprehensive examination of gestational air pollution exposure by estimating the joint effects of weekly exposures during different vulnerable periods. Our model also treats PTB as a time-to-event outcome to address the challenge of different exposure lengths among ongoing pregnancies. The proposed model is applied to a dataset of geocoded birth records in the Atlanta metropolitan area between 1999-2005 to examine the risk of PTB associated with gestational exposure to ambient fine particulate matter [Formula: see text]m in aerodynamic diameter (PM[Formula: see text]). We find positive associations between PM[Formula: see text] exposure during early and mid-pregnancy, and evidence that associations are stronger for PTBs occurring around week 30.

Combining spectral domain optical coherence tomography structural parameters for the diagnosis of glaucoma with early visual field loss.

PURPOSE To create a multivariable predictive model for glaucoma with early visual field loss using a combination of spectral-domain optical coherence tomography (SD-OCT) parameters, and to compare the results with single variable models. METHODS Two hundred fifty-three subjects (149 healthy controls and 104 with early glaucoma) underwent optic disc and macular scanning using SD-OCT in one randomly selected eye per subject. Sixteen parameters (rim area, cup-to-disc area ratio, vertical cup-to-disc diameter ratio, average and quadrant RNFL thicknesses, average, minimum, and sectoral ganglion cell inner-plexiform layer [GCIPL] thicknesses) were collected and submitted to an exploratory factor analysis (EFA) followed by logistic regression with the backward elimination variable selection technique. Area under the curve (AUC) of the receiver operating characteristic (ROC), sensitivity, specificity, Akaikes information criterion (AIC), predicted probability, prediction interval length (PIL), and classification rates were used to determine the performances of the univariable and multivariable models. RESULTS The multivariable model had an AUC of 0.995 with 98.6% sensitivity, 96.0% specificity, and an AIC value of 43.29. Single variable models yielded AUCs of 0.943 to 0.987, sensitivities of 82.6% to 95.7%, specificities of 88.0% to 94.0%, and AICs of 113.16 to 59.64 (smaller is preferred). The EFA logistic regression model correctly classified 91.67% of cases with a median PIL of 0.050 in the validation set. Univariable models correctly classified 80.62% to 90.48% of cases with median PILs 1.9 to 3.0 times larger. CONCLUSIONS The multivariable model was successful in predicting glaucoma with early visual field loss and outperformed univariable models in terms of AUC, AIC, PILs, and classification rates.

The burden of typhoid fever in low- and middle-income countries: A meta-regression approach

Background Upcoming vaccination efforts against typhoid fever require an assessment of the baseline burden of disease in countries at risk. There are no typhoid incidence data from most low- and middle-income countries (LMICs), so model-based estimates offer insights for decision-makers in the absence of readily available data. Methods We developed a mixed-effects model fit to data from 32 population-based studies of typhoid incidence in 22 locations in 14 countries. We tested the contribution of economic and environmental indices for predicting typhoid incidence using a stochastic search variable selection algorithm. We performed out-of-sample validation to assess the predictive performance of the model. Results We estimated that 17.8 million cases of typhoid fever occur each year in LMICs (95% credible interval: 6.9–48.4 million). Central Africa was predicted to experience the highest incidence of typhoid, followed by select countries in Central, South, and Southeast Asia. Incidence typically peaked in the 2–4 year old age group. Models incorporating widely available economic and environmental indicators were found to describe incidence better than null models. Conclusions Recent estimates of typhoid burden may under-estimate the number of cases and magnitude of uncertainty in typhoid incidence. Our analysis permits prediction of overall as well as age-specific incidence of typhoid fever in LMICs, and incorporates uncertainty around the model structure and estimates of the predictors. Future studies are needed to further validate and refine model predictions and better understand year-to-year variation in cases.

Residual and Dynamic Range of Retinal Nerve Fiber Layer Thickness in Glaucoma: Comparison of Three OCT Platforms.

PURPOSE To estimate visual field (VF) sensitivity at which retinal nerve fiber layer (RNFL) thinning reaches the measurement floor and at which RNFL stops thinning (change points), the dynamic range of RNFL thickness, and the number of steps from normal to RNFL floor among three optical coherence tomography (OCT) devices. METHODS Glaucomatous patients (n = 58) and healthy subjects (n = 55-60) prospectively underwent VF testing and RNFL thickness measurement with Cirrus, Spectralis, and RTVue. Change points and corresponding RNFL thicknesses were estimated with simple linear regression (SLR) and Bayesian change point (BCP) analyses. The dynamic range and number of steps to RNFL floor were determined. RESULTS The average VF change points and corresponding residual thickness at the time RNFL stopped thinning were -22.2 dB and 57.0 μm (Cirrus), -25.3 dB and 49.2 μm (Spectralis), and -24.6 dB and 64.7 μm (RTVue). The RNFL dynamic ranges derived from SLR values were wider on Spectralis (52.6 μm) than on Cirrus (35.4 μm) and RTVue (35.5 μm); the corresponding number of steps to reach the RNFL floor were 9.0 on Cirrus, 10.6 on Spectralis, and 8.3 on RTVue. CONCLUSIONS The relative VF sensitivity at which average RNFL thickness reaches the measurement floor, the residual layer thickness, and RNFL dynamic measurement range differ among the three devices. However, the number of steps from normal to the RNFL thickness floor is comparable.

Retinal nerve fibre layer thickness floor and corresponding functional loss in glaucoma

Aim To estimate the floor of retinal nerve fibre layer (RNFL) thickness measurements and the corresponding retinal sensitivity loss in glaucoma. Methods Visual field (VF), Spectralis RNFL (83 patients and 37 healthy subjects) and RTVue RNFL data obtained separately (56 patients and 36 healthy subjects) were reviewed. Global and quadrant residual layer thicknesses and corresponding VF losses were estimated using two Bayesian change point models. Results The respective residual thicknesses from change point model 1 (CPM1) on Spectralis and RTVue (respectively) were 49.9 and 70.6 µm globally, 57.1 and 83.7 µm superiorly, 55.2 and 79.0 µm inferiorly, 43.1 and 60.5 µm nasally, and 40.1 and 59.5 µm temporally. Corresponding VF losses ranged between −25.1 and −21.7 dB (Spectralis) and between −21.8 and −3.4 dB (RTVue). From CPM2, RNFL thinning reached horizontal asymptotes at VF losses between −18.0 and −10.7 dB (Spectralis) and between −12.1 and −2.5 dB (RTVue). There were no significant differences between postchange point residual layer thicknesses from CPM1 and CPM2 on Spectralis (37.0–50.8 µm vs 38.3–56.0 µm) and RTVue (60.6–80.5 µm vs 58.4–88.8 µm). Conclusions Global RNFL thinning reaches the floor at a smaller VF loss level with Spectralis than with RTVue. The nasal and temporal quadrants retain thinner residual layers than superior and inferior quadrant RNFL. Measuring RNFL below their minimums will not yield useful clinical information.

Maternal residential exposure to agricultural pesticides and birth defects in a 2003 to 2005 North Carolina birth cohort

BACKGROUND Birth defects are responsible for a large proportion of disability and infant mortality. Exposure to a variety of pesticides have been linked to increased risk of birth defects. METHODS We conducted a case-control study to estimate the associations between a residence-based metric of agricultural pesticide exposure and birth defects. We linked singleton live birth records for 2003 to 2005 from the North Carolina (NC) State Center for Health Statistics to data from the NC Birth Defects Monitoring Program. Included women had residence at delivery inside NC and infants with gestational ages from 20 to 44 weeks (n = 304,906). Pesticide exposure was assigned using a previously constructed metric, estimating total chemical exposure (pounds of active ingredient) based on crops within 500 meters of maternal residence, specific dates of pregnancy, and chemical application dates based on the planting/harvesting dates of each crop. Logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals for four categories of exposure (<10(th) , 10-50(th) , 50-90(th) , and >90(th) percentiles) compared with unexposed. Models were adjusted for maternal race, age at delivery, education, marital status, and smoking status. RESULTS We observed elevated ORs for congenital heart defects and certain structural defects affecting the gastrointestinal, genitourinary and musculoskeletal systems (e.g., OR [95% confidence interval] [highest exposure vs. unexposed] for tracheal esophageal fistula/esophageal atresia = 1.98 [0.69, 5.66], and OR for atrial septal defects: 1.70 [1.34, 2.14]). CONCLUSION Our results provide some evidence of associations between residential exposure to agricultural pesticides and several birth defects phenotypes. Birth Defects Research (Part A) 106:240-249, 2016.

Air Pollution Metric Analysis While Determining Susceptible Periods of Pregnancy for Low Birth Weight

Multiple metrics to characterize air pollution are available for use in environmental health analyses in addition to the standard Air Quality System (AQS) pollution monitoring data. These metrics have complete spatial-temporal coverage across a domain and are therefore crucial in calculating pollution exposures in geographic areas where AQS monitors are not present. We investigate the impact that two of these metrics, output from a deterministic chemistry model (CMAQ) and from a spatial-temporal downscaler statistical model which combines information from AQS and CMAQ (DS), have on risk assessment. Using each metric, we analyze ambient ozones effect on low birth weight utilizing a Bayesian temporal probit regression model. Weekly windows of susceptibility are identified and analyzed jointly for all births in a subdomain of Texas, 2001–2004, and results from the different pollution metrics are compared. Increased exposures during weeks 20–23 of the pregnancy are identified as being associated with low birth weight by the DS metric. Use of the CMAQ output alone results in increased variability of the final risk assessment estimates, while calibrating the CMAQ through use of the DS metric provides results more closely resembling those of the AQS. The AQS data are still preferred when available.