Impact of fine particulate matter on latent tuberculosis infection and active tuberculosis in older adults: a population-based multicentre cohort study

Evidence showed that air pollution was associated with an increased risk of tuberculosis (TB). This study aimed to study the impact of long-term exposure to ambient particulate matter with an aerodynamic diameter less than 2.5 μm (PM2.5) on the acquisition of LTBI and on the risk of subsequent active disease development among rural older adults from a multicentre cohort, which have not yet been investigated to date. A total of 4790 older adults were included in a population-based, multicentre, prospective cohort study (LATENTTB-NSTM) from 2013 to 2018. The level of long-term exposure to PM2.5 for each participant was assessed by aggregating satellite-based estimates. Logistic regression and time-varying Cox proportional hazards models with province-level random intercepts were employed to assess associations of long-term exposures to PM2.5 with the risk of LTBI and subsequent development of active TB, respectively. Out of 4790 participants, 3284 were LTBI-free at baseline, among whom 2806 completed the one-year follow-up and 127 developed newly identified LTBI. No significant associations were identified between PM2.5 and the risk of LTBI. And among 1506 participants with LTBI at baseline, 30 active TB cases were recorded during the 5-year follow-up. Particularly, an increment of 5 μg/m3 in 2-year moving averaged PM2.5 was associated with a 50.6% increased risk of active TB (HR = 1.506, 95% CI: 1.161-1.955). Long-term air pollution might be a neglected risk factor for active TB development from LTBI, especially for those living in developing or less-developed areas where the air quality is poor.

Development of a community severance index for urban areas in the United States: A case study in New York City

BACKGROUND AND AIMS

Traffic-related exposures, such as air pollution and noise, have a detrimental impact on human health, especially in urban areas. However, there remains a critical research and knowledge gap in understanding the impact of community severance, a measure of the physical separation imposed by road infrastructure and motorized road traffic, limiting access to goods, services, or social connections, breaking down the social fabric and potentially also adversely impacting health. We aimed to robustly quantify a community severance metric in urban settings exemplified by its characterization in New York City (NYC).

METHODS

We used geospatial location data and dimensionality reduction techniques to capture NYC community severance variation. We employed principal component pursuit, a pattern recognition algorithm, combined with factor analysis as a novel method to estimate the Community Severance Index. We used public data for the year 2019 at census block group (CBG) level on road infrastructure, road traffic activity, and pedestrian infrastructure. As a demonstrative application of the Community Severance Index, we investigated the association between community severance and traffic collisions, as a proxy for road safety, in 2019 in NYC at CBG level.

RESULTS

Our data revealed one multidimensional factor related to community severance explaining 74% of the data variation. In adjusted analyses, traffic collisions in general, and specifically those involving pedestrians or cyclists, were nonlinearly associated with an increasing level of Community Severance Index in NYC.

CONCLUSION

We developed a high spatial-resolution Community Severance Index for NYC using data available nationwide, making it feasible for replication in other cities across the United States. Our findings suggest that increases in the Community Severance Index across CBG may be linked to increases in traffic collisions in NYC. The Community Severance Index, which provides a novel traffic-related exposure, may be used to inform equitable urban policies that mitigate health risks and enhance well-being.

Measuring long-term exposure to wildfire PM2.5 in California: Time-varying inequities in environmental burden

Wildfires have become more frequent and intense due to climate change and outdoor wildfire fine particulate matter (PM2.5) concentrations differ from relatively smoothly varying total PM2.5. Thus, we introduced a conceptual model for computing long-term wildfire PM2.5 and assessed disproportionate exposures among marginalized communities. We used monitoring data and statistical techniques to characterize annual wildfire PM2.5 exposure based on intermittent and extreme daily wildfire PM2.5 concentrations in California census tracts (2006 to 2020). Metrics included: 1) weeks with wildfire PM2.5 < 5 μg/m3; 2) days with non-zero wildfire PM2.5; 3) mean wildfire PM2.5 during peak exposure week; 4) smoke waves (≥2 consecutive days with <15 μg/m3 wildfire PM2.5); and 5) mean annual wildfire PM2.5 concentration. We classified tracts by their racial/ethnic composition and CalEnviroScreen (CES) score, an environmental and social vulnerability composite measure. We examined associations of CES and racial/ethnic composition with the wildfire PM2.5 metrics using mixed-effects models. Averaged 2006 to 2020, we detected little difference in exposure by CES score or racial/ethnic composition, except for non-Hispanic American Indian and Alaska Native populations, where a 1-SD increase was associated with higher exposure for 4/5 metrics. CES or racial/ethnic × year interaction term models revealed exposure disparities in some years. Compared to their California-wide representation, the exposed populations of non-Hispanic American Indian and Alaska Native (1.68×, 95% CI: 1.01 to 2.81), white (1.13×, 95% CI: 0.99 to 1.32), and multiracial (1.06×, 95% CI: 0.97 to 1.23) people were over-represented from 2006 to 2020. In conclusion, during our study period in California, we detected disproportionate long-term wildfire PM2.5 exposure for several racial/ethnic groups.

Metabolomic Effects of Folic Acid Supplementation in Adults: Evidence from the FACT Trial

BACKGROUND

Folic acid (FA) is the oxidized form of folate found in supplements and FA-fortified foods. Most FA is reduced by dihydrofolate reductase to 5-methyltetrahydrofolate (5mTHF); the latter is the form of folate naturally found in foods. Ingestion of FA increases the plasma levels of both 5mTHF and unmetabolized FA (UMFA). Limited information is available on the downstream metabolic effects of FA supplementation, including potential effects associated with UMFA.

OBJECTIVE

We aimed to assess the metabolic effects of FA-supplementation, and the associations of plasma 5mTHF and UMFA with the metabolome in FA-naïve Bangladeshi adults.

METHODS

Sixty participants were selected from the Folic Acid and Creatine Trial; half received 800 μg FA/day for 12 weeks and half placebo. Plasma metabolome profiles were measured by high-resolution mass spectrometry, including 170 identified metabolites and 26,541 metabolic features. Penalized regression methods were used to assess the associations of targeted metabolites with FA-supplementation, plasma 5mTHF, and plasma UMFA. Pathway analyses were conducted using Mummichog.

RESULTS

In penalized models of identified metabolites, FA-supplementation was associated with higher choline. Changes in 5mTHF concentrations were positively associated with metabolites involved in amino acid metabolism (5-hydroxyindoleacetic acid, acetylmethionine, creatinine, guanidinoacetate, hydroxyproline/n-acetylalanine) and 2 fatty acids (docosahexaenoic acid and linoleic acid). Changes in 5mTHF concentrations were negatively associated with acetylglutamate, acetyllysine, carnitine, propionyl carnitine, cinnamic acid, homogentisate, arachidonic acid, and nicotine. UMFA concentrations were associated with lower levels of arachidonic acid. Together, metabolites selected across all models were related to lipids, aromatic amino acid metabolism, and the urea cycle. Analyses of nontargeted metabolic features identified additional pathways associated with FA supplementation.

CONCLUSION

In addition to the recapitulation of several expected metabolic changes associated with 5mTHF, we observed additional metabolites/pathways associated with FA-supplementation and UMFA. Further studies are needed to confirm these associations and assess their potential implications for human health.

TRIAL REGISTRATION NUMBER

This trial was registered at https://clinicaltrials.gov as NCT01050556.

An environmental justice analysis of air pollution emissions in the United States from 1970 to 2010

Over the last decades, air pollution emissions have decreased substantially; however, inequities in air pollution persist. We evaluate county-level racial/ethnic and socioeconomic disparities in emissions changes from six air pollution source sectors (industry [SO2], energy [SO2, NOx], agriculture [NH3], commercial [NOx], residential [particulate organic carbon], and on-road transportation [NOx]) in the contiguous United States during the 40 years following the Clean Air Act (CAA) enactment (1970-2010). We calculate relative emission changes and examine the differential changes given county demographics using hierarchical nested models. The results show racial/ethnic disparities, particularly in the industry and energy generation source sectors. We also find that median family income is a driver of variation in relative emissions changes in all sectors-counties with median family income >$75 K vs. less generally experience larger relative declines in industry, energy, transportation, residential, and commercial-related emissions. Emissions from most air pollution source sectors have, on a national level, decreased following the United States CAA. In this work, we show that the relative reductions in emissions varied across racial/ethnic and socioeconomic groups.

Fine particulate matter composition in American Indian vs. Non-American Indian communities

BACKGROUND

Fine particulate matter (PM2.5) exposure is a known risk factor for numerous adverse health outcomes, with varying estimates of component-specific effects. Populations with compromised health conditions such as diabetes can be more sensitive to the health impacts of air pollution exposure. Recent trends in PM2.5 in primarily American Indian- (AI-) populated areas examined in previous work declined more gradually compared to the declines observed in the rest of the US. To further investigate components contributing to these findings, we compared trends in concentrations of six PM2.5 components in AI- vs. non-AI-populated counties over time (2000-2017) in the contiguous US.

METHODS

We implemented component-specific linear mixed models to estimate differences in annual county-level concentrations of sulfate, nitrate, ammonium, organic matter, black carbon, and mineral dust from well-validated surface PM2.5 models in AI- vs. non-AI-populated counties, using a multi-criteria approach to classify counties as AI- or non-AI-populated. Models adjusted for population density and median household income. We included interaction terms with calendar year to estimate whether concentration differences in AI- vs. non-AI-populated counties varied over time.

RESULTS

Our final analysis included 3108 counties, with 199 (6.4%) classified as AI-populated. On average across the study period, adjusted concentrations of all six PM2.5 components in AI-populated counties were significantly lower than in non-AI-populated counties. However, component-specific levels in AI- vs. non-AI-populated counties varied over time: sulfate and ammonium levels were significantly lower in AI- vs. non-AI-populated counties before 2011 but higher after 2011 and nitrate levels were consistently lower in AI-populated counties.

CONCLUSIONS

This study indicates time trend differences of specific components by AI-populated county type. Notably, decreases in sulfate and ammonium may contribute to steeper declines in total PM2.5 in non-AI vs. AI-populated counties. These findings provide potential directives for additional monitoring and regulations of key emissions sources impacting tribal lands.

The association between temperature and alcohol- and substance-related disorder hospital visits in New York State

BACKGROUND

Limited evidence exists on how temperature increases are associated with hospital visits from alcohol- and substance-related disorders, despite plausible behavioral and physiological pathways.

METHODS

In the present study, we implemented a case-crossover design, which controls for seasonal patterns, long-term trends, and non- or slowly-varying confounders, with distributed lag non-linear temperature terms (0-6 days) to estimate associations between daily ZIP Code-level temperature and alcohol- and substance-related disorder hospital visit rates in New York State during 1995-2014. We also examined four substance-related disorder sub-causes (cannabis, cocaine, opioid, sedatives).

RESULTS

Here we show that, for alcohol-related disorders, a daily increase in temperature from the daily minimum (-30.1 °C (-22.2 °F)) to the 75th percentile (18.8 °C (65.8 °F)) across 0-6 lag days is associated with a cumulative 24.6% (95%CI,14.6%-34.6%) increase in hospital visit rates, largely driven by increases on the day of and day before hospital visit, with an association larger outside New York City. For substance-related disorders, we find evidence of a positive association at temperatures from the daily minimum (-30.1 °C (-22.2 °F)) to the 50th percentile (10.4 °C (50.7 °F)) (37.7% (95%CI,27.2%-48.2%), but not at higher temperatures. Findings are consistent across age group, sex, and social vulnerability.

CONCLUSIONS

Our work highlights how hospital visits from alcohol- and substance-related disorders are currently impacted by elevated temperatures and could be further affected by rising temperatures resulting from climate change. Enhanced social infrastructure and health system interventions could mitigate these impacts.

Exploring Relevant Time Windows in the Association Between PM2.5 Exposure and Amyotrophic Lateral Sclerosis: A Case-Control Study in Denmark

Studies suggest a link between particulate matter less than or equal to 2.5 μm in diameter (PM2.5) and amyotrophic lateral sclerosis (ALS), but to our knowledge critical exposure windows have not been examined. We performed a case-control study in the Danish population spanning the years 1989-2013. Cases were selected from the Danish National Patient Registry based on International Classification of Diseases codes. Five controls were randomly selected from the Danish Civil Registry and matched to a case on vital status, age, and sex. PM2.5 concentration at residential addresses was assigned using monthly predictions from a dispersion model. We used conditional logistic regression to estimate odds ratios (ORs) and 95% confidence intervals (CIs), adjusting for confounding. We evaluated exposure to averaged PM2.5 concentrations 12-24 months, 2-6 years, and 2-11 years pre-ALS diagnosis; annual lagged exposures up to 11 years prediagnosis; and cumulative associations for exposure in lags 1-5 years and 1-10 years prediagnosis, allowing for varying association estimates by year. We identified 3,983 cases and 19,915 controls. Cumulative exposure to PM2.5 in the period 2-6 years prediagnosis was associated with ALS (OR = 1.06, 95% CI: 0.99, 1.13). Exposures in the second, third, and fourth years prediagnosis were individually associated with higher odds of ALS (e.g., for lag 1, OR = 1.04, 95% CI: 1.00, 1.08). Exposure to PM2.5 within 6 years before diagnosis may represent a critical exposure window for ALS.

Short-term excess mortality following tropical cyclones in the United States

Knowledge of excess deaths after tropical cyclones is critical to understanding their impacts, directly relevant to policies on preparedness and mitigation. We applied an ensemble of 16 Bayesian models to 40.7 million U.S. deaths and a comprehensive record of 179 tropical cyclones over 32 years (1988-2019) to estimate short-term all-cause excess deaths. The deadliest tropical cyclone was Hurricane Katrina in 2005, with 1491 [95% credible interval (CrI): 563, 3206] excess deaths (>99% posterior probability of excess deaths), including 719 [95% CrI: 685, 752] in Orleans Parish, LA (>99% probability). Where posterior probabilities of excess deaths were >95%, there were 3112 [95% CrI: 2451, 3699] total post-hurricane force excess deaths and 15,590 [95% CrI: 12,084, 18,835] post-gale to violent storm force deaths; 83.1% of post-hurricane force and 70.0% of post-gale to violent storm force excess deaths occurred more recently (2004-2019); and 6.2% were in least socially vulnerable counties.

Can traffic-related air pollution trigger myocardial infarction within a few hours of exposure? Identifying hourly hazard periods

INTRODUCTION

Traffic-related air pollution can trigger myocardial infarction (MI). However, the hourly hazard period of exposure to nitrogen dioxide (NO2), a common traffic tracer, for incident MI has not been fully evaluated. Thus, the current hourly US national air quality standard (100 ppb) is based on limited hourly-level effect estimates, which may not adequately protect cardiovascular health.

OBJECTIVES

We characterized the hourly hazard period of NO2 exposure for MI in New York state (NYS), USA, from 2000 to 2015.

METHODS

For nine cities in NYS, we obtained data on MI hospitalizations from the NYS Department of Health Statewide Planning and Research Cooperative System and hourly NO2 concentrations from the US Environmental Protection Agency's Air Quality System database. We used city-wide exposures and a case-crossover study design with distributed lag non-linear terms to assess the relationship between hourly NO2 concentrations over 24 h and MI, adjusting for hourly temperature and relative humidity.

RESULTS

The mean NO2 concentration was 23.2 ppb (standard deviation: 12.6 ppb). In the six hours preceding MI, we found linearly increased risk with increasing NO2 concentrations. At lag hour 0, a 10 ppb increase in NO2 was associated with 0.2 % increased risk of MI (Rate Ratio [RR]: 1.002; 95 % Confidence Interval [CI]: 1.000, 1.004). We estimated a cumulative RR of 1.015 (95 % CI: 1.008, 1.021) for all 24 lag hours per 10 ppb increase in NO2. Lag hours 2-3 had consistently elevated risk ratios in sensitivity analyses.

CONCLUSIONS

We found robust associations between hourly NO2 exposure and MI risk at concentrations far lower than current hourly NO2 national standards. Risk of MI was most elevated in the six hours after exposure, consistent with prior studies and experimental work evaluating physiologic responses after acute traffic exposure. Our findings suggest that current hourly standards may be insufficient to protect cardiovascular health.

Cohort Network: A Knowledge Graph toward Data Dissemination and Knowledge-Driven Discovery for Cohort Studies

Contemporary environmental health sciences draw on large-scale longitudinal studies to understand the impact of environmental exposures and behavior factors on the risk of disease and identify potential underlying mechanisms. In such studies, cohorts of individuals are assembled and followed up over time. Each cohort generates hundreds of publications, which are typically neither coherently organized nor summarized, hence limiting knowledge-driven dissemination. Hence, we propose a Cohort Network, a multilayer knowledge graph approach to extract exposures, outcomes, and their connections. We applied the Cohort Network on 121 peer-reviewed papers published over the past 10 years from the Veterans Affairs (VA) Normative Aging Study (NAS). The Cohort Network visualized connections between exposures and outcomes across different publications and identified key exposures and outcomes, such as air pollution, DNA methylation, and lung function. We demonstrated the utility of the Cohort Network for new hypothesis generation, e.g., identification of potential mediators of exposure-outcome associations. The Cohort Network can be used by investigators to summarize the cohort's research and facilitate knowledge-driven discovery and dissemination.

Unintended impacts of the Open Streets program on noise complaints in New York City

BACKGROUND

During the COVID-19 pandemic, several cities allocated more public spaces for physical activity and recreation instead of road transport through Open Streets. This policy locally reduces traffic and provides experimental testbeds for healthier cities. However, it may also generate unintended impacts. For instance, Open Streets may impact the levels of exposure to environmental noise but there are no studies assessing these unintended impacts.

OBJECTIVES

Using noise complaints from New York City (NYC) as a proxy of annoyance caused by environmental noise, we estimated associations at the census tract level between same-day proportion of Open Streets in a census tract and noise complaints in NYC.

METHODS

Using data from summer 2019 (pre-implementation) and summer 2021 (post-implementation), we fit regressions to estimate the association between census tract-level proportion of Open Streets and daily noise complaints, with random effects to account for within-tract correlation and natural splines to allow non-linearity in the estimated association. We accounted for temporal trends and other potential confounders, such as population density and poverty rate.

RESULTS

In adjusted analyses, daily street/sidewalk noise complaints were nonlinearly associated with an increasing proportion of Open Streets. Specifically, compared to the mean proportion of Open Streets in a census tract (0.11%), 5% of Open Streets had a 1.09 (95% CI: 0.98, 1.20) and 10% had a 1.21 (95% CI: 1.04, 1.42) times higher rate of street/sidewalk noise complaints. Our results were robust to the choice of data source for identifying Open Streets.

CONCLUSION

Our findings suggest that Open Streets in NYC may be linked to an increase in street/sidewalk noise complaints. These results highlight the necessity to reinforce urban policies with a careful analysis for potential unintended impacts to optimize and maximize the benefits of these policies.

Clean fleets, different streets: evaluating the effect of New York City's clean bus program on changes to estimated ambient air pollution

BACKGROUND

Motor vehicles, including public transit buses, are a major source of air pollution in New York City (NYC) and worldwide. To address this problem, governments and transit agencies have implemented policies to introduce cleaner vehicles into transit fleets. Beginning in 2000, the Metropolitan Transit Agency began deploying compressed natural gas, hybrid electric, and low-sulfur diesel buses to reduce urban air pollution.

OBJECTIVE

We hypothesized that bus fleet changes incorporating cleaner vehicles would have detectable effects on air pollution concentrations between 2009 and 2014, as measured by the New York City Community Air Survey (NYCCAS).

METHODS

Depot- and route-specific information allowed identification of areas with larger or smaller changes in the proportion of distance traveled by clean buses. Data were assembled for 9670 300 m × 300 m grid cell areas with annual concentration estimates for nitrogen oxide (NO), nitrogen dioxide (NO2), and black carbon (BC) from NYCCAS. Spatial error models adjusted for truck route presence and total traffic volume.

RESULTS

While concentrations of all three pollutants declined between 2009 and 2014 even in the 39.7% of cells without bus service, the decline in concentrations of NO and NO2 was greater in areas with more bus service and with higher proportional shifts toward clean buses. Conversely, the decline in BC concentration was slower in areas with more bus service and higher proportional clean bus shifts.

SIGNIFICANCE

These results provide evidence that the NYC clean bus program impacted concentrations of air pollution, particularly in reductions of NO2. Further work can investigate the potential impact of these changes on health outcomes in NYC residents.

IMPACT STATEMENT

Urban air pollution from diesel-burning buses is an important health exposure. The New York Metropolitan Transit Agency has worked to deploy cleaner buses into their fleet, but the impact of this policy has not been evaluated. Successful reductions in air pollution are critical for public health.

Integrated causal-predictive machine learning models for tropical cyclone epidemiology

Strategic preparedness reduces the adverse health impacts of hurricanes and tropical storms, referred to collectively as tropical cyclones (TCs), but its protective impact could be enhanced by a more comprehensive and rigorous characterization of TC epidemiology. To generate the insights and tools necessary for high-precision TC preparedness, we introduce a machine learning approach that standardizes estimation of historic TC health impacts, discovers common patterns and sources of heterogeneity in those health impacts, and enables identification of communities at highest health risk for future TCs. The model integrates (i) a causal inference component to quantify the immediate health impacts of recent historic TCs at high spatial resolution and (ii) a predictive component that captures how TC meteorological features and socioeconomic/demographic characteristics of impacted communities are associated with health impacts. We apply it to a rich data platform containing detailed historic TC exposure information and records of all-cause mortality and cardiovascular- and respiratory-related hospitalization among Medicare recipients. We report a high degree of heterogeneity in the acute health impacts of historic TCs, both within and across TCs, and, on average, substantial TC-attributable increases in respiratory hospitalizations. TC-sustained windspeeds are found to be the primary driver of mortality and respiratory risks.

Bias Amplification and Variance Inflation in Distributed Lag Models Using Low-Spatial-Resolution Data

Distributed lag models (DLMs) are often used to estimate lagged associations and identify critical exposure windows. In a simulation study of prenatal nitrogen dioxide (NO2) exposure and birth weight, we demonstrate that bias amplification and variance inflation can manifest under certain combinations of DLM estimation approaches and time-trend adjustment methods when using low-spatial-resolution exposures with extended lags. Our simulations showed that when using high-spatial-resolution exposure data, any time-trend adjustment method produced low bias and nominal coverage for the distributed lag estimator. When using either low- or no-spatial-resolution exposures, bias due to time trends was amplified for all adjustment methods. Variance inflation was higher in low- or no-spatial-resolution DLMs when using a long-term spline to adjust for seasonality and long-term trends due to concurvity between a distributed lag function and secular function of time. NO2-birth weight analyses in a Massachusetts-based cohort showed that associations were negative for exposures experienced in gestational weeks 15-30 when using high-spatial-resolution DLMs; however, associations were null and positive for DLMs with low- and no-spatial-resolution exposures, respectively, which is likely due to bias amplification. DLM analyses should jointly consider the spatial resolution of exposure data and the parameterizations of the time trend adjustment and lag constraints.

Metabolomic changes associated with chronic arsenic exposure in a Bangladeshi population

Chronic exposure to arsenic (As) remains a global public health concern and our understanding of the biological mechanisms underlying the adverse effects of As exposure remains incomplete. Here, we used a high-resolution metabolomics approach to examine how As affects metabolic pathways in humans. We selected 60 non-smoking adults from the Folic Acid and Creatine Trial (FACT). Inorganic (AsIII, AsV) and organic (monomethylarsonous acid [MMAs], dimethylarsinous Acid [DMAs]) As species were measured in blood and urine collected at baseline and at 12 weeks. Plasma metabolome profiles were measured using untargeted high-resolution mass spectrometry. Associations of blood and urinary As with 170 confirmed metabolites and >26,000 untargeted spectral features were modeled using a metabolome-wide association study (MWAS) approach. Models were adjusted for age, sex, visit, and BMI and corrected for false discovery rate (FDR). In the MWAS screening of confirmed metabolites, 17 were associated with ≥1 blood As species (FDR<0.05), including fatty acids, neurotransmitter metabolites, and amino acids. These results were consistent across blood As species and between blood and urine As. Untargeted MWAS identified 423 spectral features associated with ≥1 blood As species. Unlike the confirmed metabolites, untargeted model results were not consistent across As species, with AsV and DMAs showing distinct association patterns. Mummichog pathway analysis revealed 12 enriched metabolic pathways that overlapped with the 17 identified metabolites, including one carbon metabolism, tricarboxylic acid cycle, fatty acid metabolism, and purine metabolism. Exposure to As may affect numerous essential pathways that underlie the well-characterized associations of As with multiple chronic diseases.

Widespread Clean Cooking Fuel Scale-Up and under-5 Lower Respiratory Infection Mortality: An Ecological Analysis in Ecuador, 1990-2019

BACKGROUND

Nationwide household transitions to the use of clean-burning cooking fuels are a promising pathway to reducing under-5 lower respiratory infection (LRI) mortality, the leading cause of child mortality globally, but such transitions are rare and evidence supporting an association between increased clean fuel use and improved health is limited.

OBJECTIVES

This study aimed to investigate the association between increased primary clean cooking fuel use and under-5 LRI mortality in Ecuador between 1990 and 2019.

METHODS

We documented cooking fuel use and cause-coded child mortalities at the canton (county) level in Ecuador from 1990 to 2019 (in four periods, 1988-1992, 1999-2003, 2008-2012, and 2015-2019). We characterized the association between clean fuel use and the rate of under-5 LRI mortalities at the canton level using quasi-Poisson generalized linear and generalized additive models, accounting for potential confounding variables that characterize wealth, urbanization, and child health care and vaccination rates, as well as canton and period fixed effects. We estimated averted under-5 LRI mortalities accrued over 30 y by predicting a counterfactual count of canton-period under-5 LRI mortalities were clean fuel use to not have increased and comparing with predicted canton-period under-5 LRI mortalities from our model and observed data.

RESULTS

From 1990 to 2019, the proportion of households primarily using a clean cooking fuel increased from 59% to 95%, and under-5 LRI mortality fell from 28 to 7 per 100,000 under-5 population. Canton-level clean fuel use was negatively associated with under-5 LRI mortalities in linear and nonlinear models. The nonlinear association suggested a threshold at approximately 60% clean fuel use, above which there was a negative association. Increases in clean fuel use between 1990 and 2019 were associated with an estimated 7,300 averted under-5 LRI mortalities (95% confidence interval: 2,600, 12,100), accounting for nearly 20% of the declines in under-5 LRI mortality observed in Ecuador over the study period.

DISCUSSION

Our findings suggest that the widespread household transition from using biomass to clean-burning fuels for cooking reduced under-5 LRI mortalities in Ecuador over the last 30 y. https://doi.org/10.1289/EHP11016.

Air pollution and decreased bone mineral density among Women's Health Initiative participants

Background

Osteoporosis heavily affects postmenopausal women and is influenced by environmental exposures. Determining the impact of criteria air pollutants and their mixtures on bone mineral density (BMD) in postmenopausal women is an urgent priority.

Methods

We conducted a prospective observational study using data from the ethnically diverse Women's Health Initiative Study (WHI) (enrollment, September 1994-December 1998; data analysis, January 2020 to August 2022). We used log-normal, ordinary kriging to estimate daily mean concentrations of PM10, NO, NO2, and SO2 at participants' geocoded addresses (1-, 3-, and 5-year averages before BMD assessments). We measured whole-body, total hip, femoral neck, and lumbar spine BMD at enrollment and follow-up (Y1, Y3, Y6) via dual-energy X-ray absorptiometry. We estimated associations using multivariable linear and linear mixed-effects models and mixture effects using Bayesian kernel machine regression (BKMR) models.

Findings

In cross-sectional and longitudinal analyses, mean PM10, NO, NO2, and SO2 averaged over 1, 3, and 5 years before the visit were negatively associated with whole-body, total hip, femoral neck, and lumbar spine BMD. For example, lumbar spine BMD decreased 0.026 (95% CI: 0.016, 0.036) g/cm2/year per a 10% increase in 3-year mean NO2 concentration. BKMR suggested that nitrogen oxides exposure was inversely associated with whole-body and lumbar spine BMD.

Interpretation

In this cohort study, higher levels of air pollutants were associated with bone damage, particularly on lumbar spine, among postmenopausal women. These findings highlight nitrogen oxides exposure as a leading contributor to bone loss in postmenopausal women, expanding previous findings of air pollution-related bone damage.

Funding

US National Institutes of Health.

Linking Air Pollution Exposure to Blood-Based Metabolic Features in a Community-Based Aging Cohort with and without Dementia

BACKGROUND

Long-term exposure to air pollution has been associated with changes in levels of metabolites measured in the peripheral blood. However, most research has been conducted in ethnically homogenous, young or middle-aged populations.

OBJECTIVE

To study the relationship between the plasma metabolome and long-term exposure to three air pollutants: particulate matter (PM) less than 2.5μm in aerodynamic diameter (PM2.5), PM less than 10μm in aerodynamic diameter (PM10), and nitrogen dioxide (NO2) in an ethnically diverse, older population.

METHODS

Plasma metabolomic profiles of 107 participants of the Washington Heights and Inwood Community Aging Project in New York City, collected from 1995-2015, including non-Hispanic white, Caribbean Hispanic, and non-Hispanic Black older adults were used. We estimated the association between each metabolic feature and predicted annual mean exposure to the air pollutants using three approaches: 1) A metabolome wide association study framework; 2) Feature selection using elastic net regression; and 3) A multivariate approach using partial-least squares discriminant analysis.

RESULTS

79 features associated with exposure to PM2.5 but none associated with PM10 or NO2. PM2.5 exposure was associated with altered amino acid metabolism, energy production, and oxidative stress response, pathways also associated with Alzheimer's disease. Three metabolites were associated with PM2.5 exposure through all three approaches: cysteinylglycine disulfide, a diglyceride, and a dicarboxylic acid. The relationship between several features and PM2.5 exposure was modified by diet and metabolic diseases.

CONCLUSIONS

These relationships uncover the mechanisms through which PM2.5 exposure can lead to altered metabolic outcomes in an older population.

Matching on Generalized Propensity Scores with Continuous Exposures

In the context of a binary treatment, matching is a well-established approach in causal inference. However, in the context of a continuous treatment or exposure, matching is still underdeveloped. We propose an innovative matching approach to estimate an average causal exposure-response function under the setting of continuous exposures that relies on the generalized propensity score (GPS). Our approach maintains the following attractive features of matching: a) clear separation between the design and the analysis; b) robustness to model misspecification or to the presence of extreme values of the estimated GPS; c) straightforward assessments of covariate balance. We first introduce an assumption of identifiability, called local weak unconfoundedness. Under this assumption and mild smoothness conditions, we provide theoretical guarantees that our proposed matching estimator attains point-wise consistency and asymptotic normality. In simulations, our proposed matching approach outperforms existing methods under settings with model misspecification or in the presence of extreme values of the estimated GPS. We apply our proposed method to estimate the average causal exposure-response function between long-term PM2.5 exposure and all-cause mortality among 68.5 million Medicare enrollees, 2000-2016. We found strong evidence of a harmful effect of long-term PM2.5 exposure on mortality. Code for the proposed matching approach is provided in the CausalGPS R package, which is available on CRAN and provides a computationally efficient implementation.

Quantifying diurnal changes in NO2 due to COVID-19 stay-at-home orders in New York City

Introduction

Policy responses to the COVID-19 pandemic, such as the NY on Pause stay-at-home order (March 22 - June 8, 2020), substantially reduced traffic and traffic-related air pollution (TRAP) in New York City (NYC). We evaluated the magnitude of TRAP decreases and examined the role of modifying factors such as weekend/weekday, road proximity, location, and time-of-day.

Methods

Hourly nitrogen dioxide (NO₂) concentrations from January 1, 2018 through June 8, 2020 were obtained from the Environmental Protection Agency's Air Quality System for all six hourly monitors in the NYC area. We used an interrupted time series design to determine the impact of NY on Pause on NO₂ concentrations, using a mixed effects model with random intercepts for monitor location, adjusted for meteorology and long-term trends. We evaluated effect modification through stratification.

Results

NO₂ concentrations decreased during NY on Pause by 19% (-3.2 ppb, 95% confidence interval [CI]: -3.5, -3.0), on average, compared to pre-Pause time trends. We found no evidence for modification by weekend/weekday, but greater decreases in NO₂ at non-roadside monitors and weak evidence for modification by location. For time-of-day, we found the largest decreases for 5 am (27%, -4.5 ppb, 95% CI: -5.7, -3.3) through 7 am (24%, -4.0 ppb, 95% CI: -5.2, -2.8), followed by 6 pm and 7 pm (22%, -3.7 ppb, 95% CI: -4.8, -2.6 and 22%, -4.8, -2.5, respectively), while the smallest decreases occurred at 11 pm and 1 am (both: 11%, -1.9 ppb, 95% CI: -3.1, -0.7).

Conclusion

NY on Pause's impact on TRAP varied greatly diurnally. Decreases during early morning and evening time periods are likely due to decreases in traffic. Our results may be useful for planning traffic policies that vary by time of day, such as congestion tolling policies.

Heat stress illness outcomes and annual indices of outdoor heat at U.S. Army installations

This study characterized associations between annually scaled thermal indices and annual heat stress illness (HSI) morbidity outcomes, including heat stroke and heat exhaustion, among active-duty soldiers at ten Continental U.S. (CONUS) Army installations from 1991 to 2018. We fit negative binomial models for 3 types of HSI morbidity outcomes and annual indices for temperature, heat index, and wet-bulb globe temperature (WBGT), adjusting for installation-level effects and long-term trends in the negative binomial regression models using block-bootstrap resampling. Ambulatory (out-patient) and reportable event HSI outcomes displayed predominately positive association patterns with the assessed annual indices of heat, whereas hospitalization associations were mostly null. For example, a one-degree Fahrenheit (°F) (or 0.55°C) increase in mean temperature between May and September was associated with a 1.16 (95% confidence interval [CI]: 1.11, 1.29) times greater rate of ambulatory encounters. The annual-scaled rate ratios and their uncertainties may be applied to climate projections for a wide range of thermal indices to estimate future military and civilian HSI burdens and impacts to medical resources.

Long-term Traffic-related Air Pollutant Exposure and Amyotrophic Lateral Sclerosis Diagnosis in Denmark: A Bayesian Hierarchical Analysis

BACKGROUND

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease. Limited evidence suggests ALS diagnosis may be associated with air pollution exposure and specifically traffic-related pollutants.

METHODS

In this population-based case-control study, we used 3,937 ALS cases from the Danish National Patient Register diagnosed during 1989-2013 and matched on age, sex, year of birth, and vital status to 19,333 population-based controls free of ALS at index date. We used validated predictions of elemental carbon (EC), nitrogen oxides (NO x ), carbon monoxide (CO), and fine particles (PM 2.5 ) to assign 1-, 5-, and 10-year average exposures pre-ALS diagnosis at study participants' present and historical residential addresses. We used an adjusted Bayesian hierarchical conditional logistic model to estimate individual pollutant associations and joint and average associations for traffic-related pollutants (EC, NO x , CO).

RESULTS

For a standard deviation (SD) increase in 5-year average concentrations, EC (SD = 0.42 µg/m 3 ) had a high probability of individual association with increased odds of ALS (11.5%; 95% credible interval [CrI] = -1.0%, 25.6%; 96.3% posterior probability of positive association), with negative associations for NO x (SD = 20 µg/m 3 ) (-4.6%; 95% CrI = 18.1%, 8.9%; 27.8% posterior probability of positive association), CO (SD = 106 µg/m 3 ) (-3.2%; 95% CrI = 14.4%, 10.0%; 26.7% posterior probability of positive association), and a null association for nonelemental carbon fine particles (non-EC PM 2.5 ) (SD = 2.37 µg/m 3 ) (0.7%; 95% CrI = 9.2%, 12.4%). We found no association between ALS and joint or average traffic pollution concentrations.

CONCLUSIONS

This study found high probability of a positive association between ALS diagnosis and EC concentration. Further work is needed to understand the role of traffic-related air pollution in ALS pathogenesis.

Principal Component Pursuit for Pattern Identification in Environmental Mixtures

BACKGROUND

Environmental health researchers often aim to identify sources or behaviors that give rise to potentially harmful environmental exposures.

OBJECTIVE

We adapted principal component pursuit (PCP)-a robust and well-established technique for dimensionality reduction in computer vision and signal processing-to identify patterns in environmental mixtures. PCP decomposes the exposure mixture into a low-rank matrix containing consistent patterns of exposure across pollutants and a sparse matrix isolating unique or extreme exposure events.

METHODS

We adapted PCP to accommodate nonnegative data, missing data, and values below a given limit of detection (LOD). We simulated data to represent environmental mixtures of two sizes with increasing proportions <LOD and three noise structures. We applied PCP-LOD to evaluate its performance in comparison with principal component analysis (PCA). We next applied principal component pursuit with limit of detection (PCP-LOD) to an exposure mixture of 21 persistent organic pollutants (POPs) measured in 1,000 U.S. adults from the 2001-2002 National Health and Nutrition Examination Survey (NHANES). We applied singular value decomposition to the estimated low-rank matrix to characterize the patterns.

RESULTS

PCP-LOD recovered the true number of patterns through cross-validation for all simulations; based on an a priori specified criterion, PCA recovered the true number of patterns in 32% of simulations. PCP-LOD achieved lower relative predictive error than PCA for all simulated data sets with up to 50% of the data <LOD. When 75% of values were <LOD, PCP-LOD outperformed PCA only when noise was low. In the POP mixture, PCP-LOD identified a rank-three underlying structure and separated 6% of values as extreme events. One pattern represented comprehensive exposure to all POPs. The other patterns grouped chemicals based on known structure and toxicity.

DISCUSSION

PCP-LOD serves as a useful tool to express multidimensional exposures as consistent patterns that, if found to be related to adverse health, are amenable to targeted public health messaging. https://doi.org/10.1289/EHP10479.

Long-Term Fine Particulate Matter Concentrations and Prevalence of Severe Acute Respiratory Syndrome Coronavirus 2: Differential Relationships by Socioeconomic Status Among Pregnant Individuals in New York City

We aimed to determine whether long-term ambient concentrations of fine particulate matter (particulate matter with an aerodynamic diameter less than or equal to 2.5 μm (PM2.5)) were associated with increased risk of testing positive for coronavirus disease 2019 (COVID-19) among pregnant individuals who were universally screened at delivery and whether socioeconomic status (SES) modified this relationship. We used obstetrical data collected from New-York Presbyterian Hospital/Columbia University Irving Medical Center in New York, New York, between March and December 2020, including data on Medicaid use (a proxy for low SES) and COVID-19 test results. We linked estimated 2018-2019 PM2.5 concentrations (300-m resolution) with census-tract-level population density, household size, income, and mobility (as measured by mobile-device use) on the basis of residential address. Analyses included 3,318 individuals; 5% tested positive for COVID-19 at delivery, 8% tested positive during pregnancy, and 48% used Medicaid. Average long-term PM2.5 concentrations were 7.4 (standard deviation, 0.8) μg/m3. In adjusted multilevel logistic regression models, we saw no association between PM2.5 and ever testing positive for COVID-19; however, odds were elevated among those using Medicaid (per 1-μg/m3 increase, odds ratio = 1.6, 95% confidence interval: 1.0, 2.5). Further, while only 22% of those testing positive showed symptoms, 69% of symptomatic individuals used Medicaid. SES, including unmeasured occupational exposures or increased susceptibility to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) due to concurrent social and environmental exposures, may explain the increased odds of testing positive for COVID-19 being confined to vulnerable pregnant individuals using Medicaid.

The associations of prenatal exposure to dioxins and polychlorinated biphenyls with neurodevelopment at 6 Months of age: Multi-pollutant approaches

BACKGROUND

Prenatal exposure to persistent organic pollutants, including polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), dioxin-like polychlorinated biphenyls (DL-PCBs), and nondioxin-like PCBs (NDL-PCBs), has been hypothesized to have a detrimental impact on neurodevelopment. However, the association of prenatal exposure to a dioxin and PCB mixture with neurodevelopment remains largely inconclusive partly because these chemical levels are correlated.

OBJECTIVES

We aimed to elucidate the association of in utero exposure to a mixture of dioxins and PCBs with neurodevelopment measured at 6 months of age by applying multipollutant methods.

METHODS

A total of 514 pregnant women were recruited between July 2002 and October 2005 in the Sapporo cohort, Hokkaido Study on Environment and Children's Health. The concentrations of individual dioxin and PCB isomers were assessed in maternal peripheral blood during pregnancy. The mental and psychomotor development of the study participants' infants was evaluated using the Bayley Scales of Infant Development-2nd Edition (n = 259). To determine both the joint and individual associations of prenatal exposure to a dioxin and PCB mixture with infant neurodevelopment, Bayesian kernel machine regression (BKMR) and quantile-based g-computation were employed.

RESULTS

Suggestive inverse associations were observed between in utero exposure to a dioxin and PCB mixture and infant psychomotor development in both the BKMR and quantile g-computation models. In contrast, we found no association of a dioxin and PCB mixture with mental development. When group-specific posterior inclusion probabilities were estimated, BKMR suggested prenatal exposure to mono-ortho PCBs as the more important contributing factors to early psychomotor development compared with the other dioxin or PCB groups. No evidence of nonlinear exposure-outcome relationships or interactions among the chemical mixtures was detected.

CONCLUSIONS

Applying the two complementary statistical methods for chemical mixture analysis, we demonstrated limited evidence of inverse associations of prenatal exposure to dioxins and PCBs with infant psychomotor development.

Smoking During Pregnancy and Risk of Attention-deficit/Hyperactivity Disorder in the Third Generation

BACKGROUND

Animal experiments indicate that environmental factors, such as cigarette smoke, can have multigenerational effects through the germline. However, there are little data on multigenerational effects of smoking in humans. We examined the associations between grandmothers' smoking while pregnant and risk of attention-deficit/hyperactivity disorder (ADHD) in her grandchildren.

METHODS

Our study population included 53,653 Nurses' Health Study II (NHS-II) participants (generation 1 [G1]), their mothers (generation 0 [G0]), and their 120,467 live-born children (generation 2 [G2]). In secondary analyses, we used data from 23,844 mothers of the nurses who were participants in the Nurses' Mothers' Cohort Study (NMCS), a substudy of NHS-II.

RESULTS

The prevalence of G0 smoking during the pregnancy with the G1 nurse was 25%. ADHD was diagnosed in 9,049 (7.5%) of the grandchildren (G2). Grand-maternal smoking during pregnancy was associated with increased odds of ADHD among the grandchildren (adjusted odds ratio [aOR] = 1.2; 95% confidence interval [CI] = 1.1, 1.2), independent of G1 smoking during pregnancy. In the Nurses' Mothers' Cohort Study, odds of ADHD increased with increasing cigarettes smoked per day by the grandmother (1-14 cigarettes: aOR = 1.1; 95% CI = 1.0, 1.2; 15+: aOR = 1.2; 95% CI = 1.0, 1.3), compared with nonsmoking grandmothers.

CONCLUSIONS

Grandmother smoking during pregnancy is associated with an increased risk of ADHD among the grandchildren.

Can weather help explain 'why now?': The potential role of hourly temperature as a stroke trigger

BACKGROUND

While evidence suggests that daily ambient temperature exposure influences stroke risk, little is known about the potential triggering role of ultra short-term temperature.

METHODS

We examined the association between hourly temperature and ischemic and hemorrhagic stroke, separately, and identified any relevant lags of exposure among adult New York State residents from 2000 to 2015. Cases were identified via ICD-9 codes from the New York Department of Health Statewide Planning and Reearch Cooperative System. We estimated ambient temperature up to 36 h prior to estimated stroke onset based on patient residential ZIP Code. We applied a time-stratified case-crossover study design; control periods were matched to case periods by year, month, day of week, and hour of day. Additionally, we assessed effect modification by leading stroke risk factors hypertension and atrial fibrillation.

RESULTS

We observed 578,181 ischemic and 164,755 hemorrhagic strokes. Among ischemic and hemorrhagic strokes respectively, the mean (standard deviation; SD) patient age was 71.8 (14.6) and 66.8 (17.4) years, with 55% and 49% female. Temperature ranged from -29.5 °C to 39.2 °C, with mean (SD) 10.9 °C (10.3 °C). We found linear relationships for both stroke types. Higher temperature was associated with ischemic stroke over the 7 h following exposure; a 10 °C increase over 7 h was associated with 5.1% (95% Confidence Interval [CI]: 3.8, 6.4%) increase in hourly stroke rate. In contrast, temperature was negatively associated with hemorrhagic stroke over 5 h, with a 5-h cumulative association of -6.2% (95% CI: 8.6, -3.7%). We observed suggestive evidence of a larger association with hemorrhagic stroke among patients with hypertension and a smaller association with ischemic stroke among those with atrial fibrillation.

CONCLUSION

Hourly temperature was positively associated with ischemic stroke and negatively associated with hemorrhagic stroke. Our results suggest that ultra short-term weather influences stroke risk and hypertension may confer vulnerability.

Air Pollution in American Indian Versus Non-American Indian Communities, 2000-2018

Objectives. To compare fine particulate matter (PM2.5) concentrations in American Indian (AI)-populated with those in non-AI-populated counties over time (2000-2018) in the contiguous United States. Methods. We used a multicriteria approach to classify counties as AI- or non--AI-populated. We ran linear mixed effects models to estimate the difference in countywide annual PM2.5 concentrations from well-validated prediction models and monitoring sites (modeled and measured PM2.5, respectively) in AI- versus non-AI-populated counties. Results. On average, adjusted modeled PM2.5 concentrations in AI-populated counties were 0.38 micrograms per cubic meter (95% confidence interval [CI] = 0.23, 0.54) lower than in non-AI-populated counties. However, this difference was not constant over time: in 2000, modeled concentrations in AI-populated counties were 1.46 micrograms per cubic meter (95% CI = 1.25, 1.68) lower, and by 2018, they were 0.66 micrograms per cubic meter (95% CI = 0.45, 0.87) higher. Over the study period, adjusted modeled PM2.5 mean concentrations decreased by 2.13 micrograms per cubic meter in AI-populated counties versus 4.26 micrograms per cubic meter in non-AI-populated counties. Results were similar for measured PM2.5. Conclusions. This study highlights disparities in PM2.5 trends between AI- and non-AI-populated counties over time, underscoring the need to strengthen air pollution regulations and prevention implementation in tribal territories and areas where AI populations live. (Am J Public Health. 2022;112(4): 615-623. https://doi.org/10.2105/AJPH.2021.306650).

Association of Tropical Cyclones With County-Level Mortality in the US

Importance

Tropical cyclones have a devastating effect on society, but a comprehensive assessment of their association with cause-specific mortality over multiple years of study is lacking.

Objective

To comprehensively evaluate the association of county-level tropical cyclone exposure and death rates from various causes in the US.

Design, Setting, and Participants

A retrospective observational study using a Bayesian conditional quasi-Poisson model to examine how tropical cyclones were associated with monthly death rates. Data from 33.6 million deaths in the US were collected from the National Center for Health Statistics over 31 years (1988-2018), including residents of the 1206 counties in the US that experienced at least 1 tropical cyclone during the study period.

Exposures

Tropical cyclone days per county-month, defined as number of days in a month with a sustained maximal wind speed 34 knots or greater.

Main Outcomes and Measures

Monthly cause-specific county-level death rates by 6 underlying causes of death: cancers, cardiovascular diseases, infectious and parasitic diseases, injuries, neuropsychiatric conditions, and respiratory diseases. The model yielded information about the association between each additional cyclone day per month and monthly county-level mortality compared with the same county-month in different years, up to 6 months after tropical cyclones, and how these estimated associations varied by age, sex, and social vulnerability. The unit of analysis was county-month.

Results

There were 33 619 393 deaths in total (16 691 681 females and 16 927 712 males; 8 587 033 aged 0-64 years and 25 032 360 aged 65 years or older) from the 6 causes recorded in 1206 US counties. There was a median of 2 tropical cyclone days experienced in total in included US counties. Each additional cyclone day was associated with increased death rates in the month following the cyclone for injuries (3.7% [95% credible interval {CrI}, 2.5%-4.9%]; 2.0 [95% CrI, 1.3-2.7] additional deaths per 1 000 000 for 2018 monthly age-standardized median rate [DPM]; 54.3 to 56.3 DPM), infectious and parasitic diseases (1.8% [95% CrI, 0.1%-3.6%]; 0.2 [95% CrI, 0.0-0.4] additional DPM; 11.7 to 11.9 DPM), respiratory diseases (1.3% [95% CrI, 0.2%-2.4%]; 0.6 [95% CrI, 0.1-1.1] additional DPM; 44.9 to 45.5 DPM), cardiovascular diseases (1.2% [95% CrI, 0.6%-1.7%]; 1.5 [95% CrI, 0.8-2.2] additional DPM; 129.6 to 131.1 DPM), neuropsychiatric conditions (1.2% [95% CrI, 0.1%-2.4%]; 0.6 [95% CrI, 0.1-1.2] additional DPM; 52.1 to 52.7 DPM), with no change for cancers (-0.3% [95% CrI, -0.9% to 0.3%]; -0.3 [95% CrI, -0.9 to 0.3] additional DPM; 100.4 to 100.1 DPM).

Conclusions and Relevance

Among US counties that experienced at least 1 tropical cyclone from 1988-2018, each additional cyclone day per month was associated with modestly higher death rates in the months following the cyclone for several causes of death, including injuries, infectious and parasitic diseases, cardiovascular diseases, neuropsychiatric conditions, and respiratory diseases.

A Bayesian Multi-Outcome Analysis of Fine Particulate Matter and Cardiorespiratory Hospitalizations

BACKGROUND

Short-term fine particulate matter (PM2.5) exposure is positively associated with acute cardiovascular and respiratory events. Understanding whether this association varies across specific cardiovascular and respiratory conditions has important biologic, clinical, and public health implications.

METHODS

We conducted a time-stratified case-crossover study of hospitalizations from 2000 through 2014 among United States Medicare beneficiaries aged 65+. The outcomes were hospitalizations with any of 57 cardiovascular and 32 respiratory discharge diagnoses. We estimated associations with two-day moving average PM2.5 as a piecewise linear term with a knot at PM2.5 = 25 g/m3. We used Multi-Outcome Regression with Tree-structured Shrinkage (MOReTreeS) to identify de novo groups of related diseases such that PM2.5 associations are: (1) similar within outcome groups; but (2) different between outcome groups. We adjusted for temperature, humidity, and individual-level characteristics. We introduce an R package, moretrees.

RESULTS

Our dataset included 16,007,293 cardiovascular and 8,690,837 respiratory hospitalizations. Of 57 cardiovascular diseases, 51 were grouped and positively associated with PM2.5. We observed a stronger positive association for heart failure, which formed a separate group. We observed negative associations for groups containing the outcomes other aneurysm and intracranial hemorrhage. Of 32 respiratory outcomes, 31 were grouped and were positively associated with PM2.5. Influenza formed a separate group with a negative association.

CONCLUSIONS

We used a new statistical approach, MOReTreeS, to uncover variation in the association between short-term PM2.5 exposure and hospitalizations for cardiovascular and respiratory causes controlling for patient characteristics, time trends, and environmental confounders.

Prenatal maternal phthalate exposures and trajectories of childhood adiposity from four to twelve years

BACKGROUND/AIM

Adiposity trajectories reflect dynamic process of growth and may predict later life health better than individual measures. Prenatal phthalate exposures may program later childhood adiposity, but findings from studies examining these associations are conflicting. We investigated associations between phthalate biomarker concentrations during pregnancy with child adiposity trajectories.

METHODS

We followed 514 mother-child pairs from the Mexico City PROGRESS cohort from pregnancy through twelve years. We measured concentrations of nine phthalate biomarkers in 2nd and 3rd trimester maternal urine samples to create a pregnancy average using the geometric mean. We measured child BMI z-score, fat mass index (FMI), and waist-to-height ratio (WHtR) at three study visits between four and 12 years of age. We identified adiposity trajectories using multivariate latent class growth modeling, considering BMI z-score, FMI, and WHtR as joint indicators of latent adiposity. We estimated associations of phthalates biomarkers with class membership using multinomial logistic regression. We used quantile g-computation to estimate the potential effect of the total phthalate mixture and assessed effect modification by sex.

RESULTS

We identified three trajectories of child adiposity, a "low-stable", a "low-high", and a "high-high" group. A doubling of the sum of di (2-ethylhexyl) phthalate metabolites (ΣDEHP), was associated with 1.53 (1.08, 2.19) greater odds of being in the "high-high" trajectory in comparison to the "low-stable" group, whereas a doubling in di-isononyl phthalate metabolites (ΣDiNP) was associated with 1.43 (1.02, 2.02) greater odds of being in the "low-high" trajectory and mono (carboxy-isononyl) phthalate (MCNP) was associated with 0.66 (0.45, 97) lower odds of being in the "low-high" trajectory. No sex-specific associations or mixture associations were observed.

CONCLUSIONS

Prenatal concentrations of urinary DEHP metabolites, DiNP metabolites, and MCNP, a di-isodecyl phthalate metabolite, were associated with trajectories of child adiposity. The total phthalate mixture was not associated with early life child adiposity.

Mixed metals exposure and cognitive function in Bangladeshi adolescents

BACKGROUND

Over 57 million people in Bangladesh have been chronically exposed to arsenic-contaminated drinking water. They also face environmental exposure to elevated levels of cadmium (Cd), manganese (Mn), and lead (Pb), all of which have been previously observed in environmental and biological samples for this population. These metals have been linked to adverse neurocognitive outcomes in adults and children, though their effects on adolescents are not yet fully characterized. Additionally, previous studies have linked selenium (Se) to protective effects against the toxicity of these other metals.

OBJECTIVES

To examine the associations between mixed metals exposure and cognitive function in Bangladeshi adolescents.

METHODS

The Metals, Arsenic, & Nutrition in Adolescents study (MANAs) is a cross-sectional study of 572 Bangladeshi adolescents aged 14-16 years, whose parents were enrolled in the Health Effects of Arsenic Longitudinal Study (HEALS). Biosamples were collected from these adolescents for measurement of whole blood metalloid/metal levels of As, Cd, Mn, Pb, and Se. Participants also completed an abbreviated version of The Cambridge Neuropsychological Test Automated Battery (CANTAB), a cognitive function test designed to measure performance across several aspects of executive function. Linear regression was used to examine associations for each metal while controlling for the other metals. Bayesian Kernel Machine Regression (BKMR) assessed the overall mixture effect in addition to confirming the effects of individual metal components observed via linear regression.

RESULTS

Linear regression revealed negative associations for Spatial Working Memory and both As and Mn (As B=-2.40, Mn B=-5.31, p < 0.05). We also observed negative associations between Cd and Spatial Recognition Memory (B=-2.77, p < 0.05), and Pb and Delayed Match to Sample, a measure of visual recognition and memory (B=-3.67, p < 0.05). Finally, we saw a positive association for Se and Spatial Span Length (B=0.92, p < 0.05). BKMR results were largely consistent with the regression analysis, showing meaningful associations for individual metals and CANTAB subtests, but no overall mixture effect. Via BKMR, we observed negative associations between Pb and Delayed Match to Sample, and Cd and Spatial Recognition Memory; this analysis also showed positive associations for Se and the Planning, Reaction Time, and Spatial Span subtests. BKMR posterior inclusion probability consistently reported that Se, the only component of the mixture to show a positive association with cognition, was the most important member of the mixture.

CONCLUSIONS

Overall, we found Se to be positively associated with cognition, while Mn and As were linked to poorer working memory, and Cd and Pb were associated with poorer visual recognition and memory. Our observations are consistent with previous reports on the effects of these metal exposures in adults and children. Our findings also suggest agreement between linear regression and BKMR methods for analyzing metal mixture exposures. Additional studies are needed to evaluate the impact of mixed metals exposure on adverse health and poorer cognition later in life for those exposed during adolescence. Findings also suggest that metal exposure mitigation efforts aimed at adolescents might influence lifelong cognitive outcomes in regions where environmental exposure to metals is endemic.

Powering Research through Innovative Methods for Mixtures in Epidemiology (PRIME) Program: Novel and Expanded Statistical Methods

Humans are exposed to a diverse mixture of chemical and non-chemical exposures across their lifetimes. Well-designed epidemiology studies as well as sophisticated exposure science and related technologies enable the investigation of the health impacts of mixtures. While existing statistical methods can address the most basic questions related to the association between environmental mixtures and health endpoints, there were gaps in our ability to learn from mixtures data in several common epidemiologic scenarios, including high correlation among health and exposure measures in space and/or time, the presence of missing observations, the violation of important modeling assumptions, and the presence of computational challenges incurred by current implementations. To address these and other challenges, NIEHS initiated the Powering Research through Innovative methods for Mixtures in Epidemiology (PRIME) program, to support work on the development and expansion of statistical methods for mixtures. Six independent projects supported by PRIME have been highly productive but their methods have not yet been described collectively in a way that would inform application. We review 37 new methods from PRIME projects and summarize the work across previously published research questions, to inform methods selection and increase awareness of these new methods. We highlight important statistical advancements considering data science strategies, exposure-response estimation, timing of exposures, epidemiological methods, the incorporation of toxicity/chemical information, spatiotemporal data, risk assessment, and model performance, efficiency, and interpretation. Importantly, we link to software to encourage application and testing on other datasets. This review can enable more informed analyses of environmental mixtures. We stress training for early career scientists as well as innovation in statistical methodology as an ongoing need. Ultimately, we direct efforts to the common goal of reducing harmful exposures to improve public health.

Grandmothers' endocrine disruption during pregnancy, low birth weight, and preterm birth in third generation

BACKGROUND

Diethylstilbestrol (DES) is an endocrine-disrupting pharmaceutical prescribed to pregnant women to prevent pregnancy complications between the 1940s and 1970s. Although DES has been shown in animal studies to have multigenerational effects, only two studies have investigated potential multigenerational effects in humans on preterm birth (PTB), and none on low birthweight (LBW)-major determinants of later life health.

METHODS

Nurses' Health Study (NHS) II participants (G1; born 1946-64) reported their mothers' (G0) use of DES while pregnant with them. We used cluster-weighted generalized estimating equations to estimate odds ratios (OR) and 95% confidence intervals (CI) for risk of LBW and PTB among the grandchildren by grandmother use of DES. G1 birthweight and gestational age were considered to explore confounding by indication.

RESULTS

Among 54 334 G0-G1/grandmother-mother pairs, 973 (1.8%) G0 used DES during pregnancy with G1. Of the 128 275 G2 children, 4369 (3.4%) were LBW and 7976 (6.2%) premature. Grandmother (G0) use of DES during pregnancy was associated with an increased risk of G2 LBW [adjusted OR (aOR) = 3.09; 95% CI: 2.57, 3.72], that was reduced when restricted to term births (aOR = 1.59; 95% CI: 1.08, 2.36). The aOR for PTB was 2.88 (95% CI: 2.46, 3.37). Results were essentially unchanged when G1 birthweight and gestational age were included in the model, as well as after adjusting for other potential intermediate variables, such as G2 pregnancy-related factors.

CONCLUSIONS

Grandmother use of DES during pregnancy is associated with an increased risk of LBW, predominantly through an increased risk of PTB. Results when considering G1 birth outcomes suggest this does not result from confounding by indication.

Early-Life Exposure to Green Space and Mid-Childhood Cognition in the Project Viva Cohort, Massachusetts

The association between early-life greenness and child cognition is not well understood. Using prospective data from Project Viva (n = 857) from 1999-2010, we examined associations of early-life greenness exposure with mid-childhood cognition. We estimated residential greenness at birth, early childhood (median age 3.1 years), and mid-childhood (7.8 years) using 30-m resolution Landsat satellite imagery (normalized difference vegetation index). In early childhood and mid-childhood, we administered standardized assessments of verbal and nonverbal intelligence, visual-motor abilities, and visual memory. We used natural splines to examine associations of early life-course greenness with mid-childhood cognition, adjusting for age, sex, race, income, neighborhood socioeconomic status, maternal intelligence, and parental education. At lower levels of greenness (greenness <0.6), greenness exposure at early childhood was associated with a 0.48% increase in nonverbal intelligence and 2.64% increase in visual memory in mid-childhood. The association between early-childhood greenness and mid-childhood visual memory was observed after further adjusting for early childhood cognition and across different methodologies, while the association with nonverbal intelligence was not. No other associations between early life-course greenness and mid-childhood cognition were found. Early childhood greenness was nonlinearly associated with higher mid-childhood visual memory. Our findings highlight the importance of nonlinear associations between greenness and cognition.

Risk for preeclampsia following exposure to PM2.5 during pregnancy

BACKGROUND

Previous findings concerning the risk for preeclampsia following exposure to particulate matter are inconclusive.

METHODS

We used data from all singleton pregnancies of women insured by the "Clalit health services" (CHS) maintenance organization in southern Israel that resulted in delivery or perinatal mortality at Soroka Medical Center (SMC). Daily PM_2.5 concentrations were estimated by a hybrid satellite-based model at one-squared kilometer spatial resolution. We used Cox proportional hazard models coupled with distributed lag models to examine the association between the mean exposure to PM_2.5 in every gestational week and the diagnosis of preeclampsia, adjusting for maternal age, parity, year of birth, season of birth and socio-economic status. Hazard Ratios (HR) and 95% Confidence Intervals (CI) were calculated for individual gestational weeks and for cumulative exposure until the 25th gestational week.

RESULTS

A total of 133,197 pregnancies ended at SMC during the study period, of which 68,126 (51.1%) were Jewish and 65,071 (48.9%) were Bedouin. For pregnancies of Jewish women, exposure to PM_2.5 from the 7th until the 14st gestational week was significantly associated with preeclampsia (maximal HR = 1.06; 95%CI: 1.01 - 1.11 during the 10th gestational week per 10 μg/m³ increase in PM_2.5). Cumulative exposure to PM_2.5 during the first 25th gestational weeks was also significantly associated with preeclampsia (HR = 2.08; 95%CI: 1.10 - 3.94 per 10 μg/m³ increase in PM_2.5). We observed no association for pregnancies of Bedouin women.

CONCLUSIONS

Exposure to PM_2.5 between the 7th and the 14st gestational weeks was associated with preeclampsia among Jewish women but not among Bedouin women.

Parkinson's disease aggravation in association with fine particle components in New York State

BACKGROUND

Long-term exposure to fine particulate matter (PM2.5) has been associated with neurodegenerative diseases, including disease aggravation in Parkinson's disease (PD), but associations with specific PM2.5 components have not been evaluated.

OBJECTIVE

To characterize the association between specific PM2.5 components and PD first hospitalization, a surrogate for disease aggravation.

METHODS

We obtained data on hospitalizations from the New York Department of Health Statewide Planning and Research Cooperative System (2000-2014) to calculate annual first PD hospitalization counts in New York State per county. We used well-validated prediction models at 1 km2 resolution to estimate county level population-weighted annual black carbon (BC), organic matter (OM), nitrate, sulfate, sea salt (SS), and soil particle concentrations. We then used a multi-pollutant mixed quasi-Poisson model with county-specific random intercepts to estimate rate ratios (RR) of one-year exposure to each PM2.5 component and PD disease aggravation. We evaluated potential nonlinear exposure-outcome relationships using penalized splines and accounted for potential confounders.

RESULTS

We observed a total of 197,545 PD first hospitalizations in NYS from 2000 to 2014. The annual average count per county was 212 first hospitalizations. The RR (95% confidence interval) for PD aggravation was 1.06 (1.03, 1.10) per one standard deviation (SD) increase in nitrate concentrations and 1.06 (1.04, 1.09) for the corresponding increase in OM concentrations. We also found a nonlinear inverse association between PD aggravation and BC at concentrations above the 96th percentile. We found a marginal association with SS and no association with sulfate or soil exposure.

CONCLUSION

In this study, we detected associations between the PM2.5 components OM and nitrate with PD disease aggravation. Our findings support that PM2.5 adverse effects on PD may vary by particle composition.

A 1-km hourly air-temperature model for 13 northeastern U.S. states using remotely sensed and ground-based measurements

BACKGROUND

Accurate and precise estimates of ambient air temperatures that can capture fine-scale within-day variability are necessary for studies of air temperature and health.

METHOD

We developed statistical models to predict temperature at each hour in each cell of a 927-m square grid across the Northeast and Mid-Atlantic United States from 2003 to 2019, across ~4000 meteorological stations from the Integrated Mesonet, using inputs such as elevation, an inverse-distance-weighted interpolation of temperature, and satellite-based vegetation and land surface temperature. We used a rigorous spatial cross-validation scheme and spatially weighted the errors to estimate how well model predictions would generalize to new cell-days. We assess the within-county association of temperature and social vulnerability in a heat wave as an example application.

RESULTS

We found that a model based on the XGBoost machine-learning algorithm was fast and accurate, obtaining weighted root mean square errors (RMSEs) around 1.6 K, compared to standard deviations around 11.0 K. We found similar accuracy when validating our model on an external dataset from Weather Underground. Assessing predictions from the North American Land Data Assimilation System-2 (NLDAS-2), another hourly model, in the same way, we found it was much less accurate, with RMSEs around 2.5 K. This is likely due to the NLDAS-2 model's coarser spatial resolution, and the dynamic variability of temperature within its grid cells. Finally, we demonstrated the health relevance of our model by showing that our temperature estimates were associated with social vulnerability across the region during a heat wave, whereas the NLDAS-2 showed a much weaker association.

CONCLUSION

Our high spatiotemporal resolution air temperature model provides a strong contribution for future health studies in this region.

Short-term PM2.5 and cardiovascular admissions in NY State: assessing sensitivity to exposure model choice

BACKGROUND

Air pollution health studies have been increasingly using prediction models for exposure assessment even in areas without monitoring stations. To date, most studies have assumed that a single exposure model is correct, but estimated effects may be sensitive to the choice of exposure model.

METHODS

We obtained county-level daily cardiovascular (CVD) admissions from the New York (NY) Statewide Planning and Resources Cooperative System (SPARCS) and four sets of fine particulate matter (PM_2.5) spatio-temporal predictions (2002-2012). We employed overdispersed Poisson models to investigate the relationship between daily PM_2.5 and CVD, adjusting for potential confounders, separately for each state-wide PM_2.5 dataset.

RESULTS

For all PM_2.5 datasets, we observed positive associations between PM_2.5 and CVD. Across the modeled exposure estimates, effect estimates ranged from 0.23% (95%CI: -0.06, 0.53%) to 0.88% (95%CI: 0.68, 1.08%) per 10 µg/m³ increase in daily PM_2.5. We observed the highest estimates using monitored concentrations 0.96% (95%CI: 0.62, 1.30%) for the subset of counties where these data were available.

CONCLUSIONS

Effect estimates varied by a factor of almost four across methods to model exposures, likely due to varying degrees of exposure measurement error. Nonetheless, we observed a consistently harmful association between PM_2.5 and CVD admissions, regardless of model choice.

Association Between Periconceptional Weight of Maternal Grandmothers and Attention-Deficit/Hyperactivity Disorder in Grandchildren

IMPORTANCE

Neurodevelopmental disorders have been proposed to involve alterations to epigenetic regulation, and epigenetic effects may extend to germline cells to affect later generations. Weight status may affect DNA methylation, and maternal weight before and during pregnancy has been associated with offspring DNA methylation as well as attention-deficit/hyperactivity disorder (ADHD).

OBJECTIVE

To assess whether a woman's weight before and during pregnancy is associated with ADHD in her grandchild.

DESIGN, SETTING, AND PARTICIPANTS

This cohort study analyzed data from 19 835 grandmother-mother dyads and 44 720 grandchildren in the Nurses' Health Study II (NHS-II) cohort (2001-2013), a population-based prospective cohort study. Cluster-weighted generalized estimating equations were modeled to estimate the association of grandmother's prepregnancy body mass index (BMI) and gestational weight gain with grandchild risk of ADHD. Data analyses were conducted from May 2018 to April 2021. Grandmothers reported their height and weight before, and weight gain during, their pregnancy with the NHS-II participants. Mothers self-reported height and weight prior to pregnancy. From those data, grandmother BMI and mother BMI were calculated as weight in kilograms divided by height in meters squared and categorized as underweight (<18.5), healthy/normal (18.5-24.9), overweight (25.0-29.9), or obese (≥30).

MAIN OUTCOMES AND MEASURES

Cases of ADHD identified by maternal report of having a child with a diagnosis of ADHD.

RESULTS

In total, 19 835 grandmothers (97.6% White race/ethnicity; 2113 [10.7%] prepregnancy underweight and 1391 [7.0%] prepregnancy overweight or obese) were included in this cohort study. Of 44 720 grandchildren, 3593 (8%) received a diagnosis of ADHD. Higher odds of ADHD among grandchildren were found for those whose grandmother was underweight compared with healthy weight prior to pregnancy with the NHS-II participant (adjusted odds ratio, 1.25; 95% CI, 1.10-1.42). By contrast, grandmother gestational weight gain was not significantly associated with risk of grandchild ADHD (adjusted odds ratio for <20 lbs [9.1 kg], 1.06; 95% CI, 0.96-1.16; adjusted odds ratio for >29 lbs [13.2 kg], 1.01; 95% CI, 0.91-1.13). Mother prepregnancy BMI showed an association with ADHD among offspring, with a stronger association detected for obese status (adjusted odds ratio, 1.27; 95% CI, 1.07-1.49) than for overweight status (adjusted odds ratio, 1.13; 95% CI, 1.02-1.26) compared with normal weight as a reference group. The positive association between grandmother prepregnancy underweight and ADHD risk among the grandchildren remained unchanged after further adjustment for potential mediators, including maternal prepregnancy BMI.

CONCLUSIONS AND RELEVANCE

The results of this cohort study indicate that grandmother underweight prior to pregnancy is associated with an increased risk of ADHD among grandchildren, independent of grandmother gestational weight gain and independent of maternal prepregnancy weight status.

Air Pollution and Risk of Placental Abruption: A Study of Births in New York City, 2008-2014

We evaluated the associations of exposure to fine particulate matter (particulate matter with an aerodynamic diameter ≤2.5 μm (PM2.5) at concentrations of <12 μg/m3, 12-14 μg/m3, and ≥15 μg/m3) and nitrogen dioxide (at concentrations of <26 parts per billion (ppb), 26-29 ppb, and ≥30 ppb) with placental abruption in a prospective cohort study of 685,908 pregnancies in New York, New York (2008-2014). In copollutant analyses, these associations were examined using distributed-lag nonlinear models based on Cox models. The prevalence of abruption was 0.9% (n = 6,025). Compared with a PM2.5 concentration less than 12 μg/m3, women exposed to PM2.5 levels of ≥15 μg/m3 in the third trimester had a higher rate of abruption (hazard ratio (HR) = 1.68, 95% confidence interval (CI): 1.41, 2.00). Compared with a nitrogen dioxide concentration less than 26 ppb, women exposed to nitrogen dioxide levels of 26-29 ppb (HR = 1.11, 95% CI: 1.02, 1.20) and ≥30 ppb (HR = 1.06, 95% CI: 0.96, 1.24) in the first trimester had higher rates of abruption. Compared with both PM2.5 and nitrogen dioxide levels less than the 95th percentile in the third trimester, rates of abruption were increased with both PM2.5 and nitrogen dioxide ≥95th percentile (HR = 1.44, 95% CI: 1.15, 1.80) and PM2.5 ≥95th percentile and nitrogen dioxide <95th percentile (HR = 1.43 95% CI: 1.23, 1.66). Increased levels of PM2.5 exposure in the third trimester and nitrogen dioxide exposure in the first trimester are associated with elevated rates of placental abruption, suggesting that these exposures may be important triggers of premature placental separation through different pathways.

The association between ambient temperature variability and myocardial infarction in a New York-State-based case-crossover study: An examination of different variability metrics

BACKGROUND

Short-term temperature variability has been consistently associated with mortality, with limited evidence for cardiovascular outcomes. Previous studies have used multiple metrics to measure temperature variability; however, those metrics do not capture hour-to-hour changes in temperature.

OBJECTIVES

We assessed the correlation between sub-daily temperature-change-over-time metrics and previously-used metrics, and estimated associations with myocardial infarction (MI) hospitalizations.

METHODS

Hour-to-hour change-over-time was measured via three metrics: 24-hr mean absolute hourly first difference, 24-hr maximum absolute hourly first difference, and 24-hr mean hourly first difference. We first assessed the Spearman correlations between these metrics and four previously-used metrics (24-hr standard deviation of hourly temperature, 24-hr diurnal temperature range, 48-hr standard deviation of daily minimal and maximal temperatures, and 48-hr difference of daily mean temperature), using hourly data from the North America Land Data Assimilation System-2 Model. Subsequently, we estimated the association between these metrics and primary MI hospitalization in adult residents of New York State for 2000-2015 using a time-stratified case-crossover design.

RESULTS

The hour-to-hour change-over-time metrics were correlated, but not synonymous, with previously-used metrics. We observed 809,259 MI, 45% of which were among females and the mean (standard deviation) age was 70 (15). An increase from mean to 90th percentile in mean absolute first difference of temperature was associated with a 2.04% (95% Confidence Interval [CI]: 1.30-2.78%) increase in MI rate. An increase from mean to 90th percentile in mean first difference also yielded a positive association (1.86%; 95%CI: 1.09-2.64%). We observed smaller- or similar-in-magnitude positive associations for previously-used metrics.

DISCUSSION

First, short-term hour-to-hour temperature change was positively associated with MI risk. Second, all other variability metrics yielded positive associations with MI, with varying magnitude. In future research on temperature variability, researchers should define their research question, including which aspects of variability they intend to measure, and apply the appropriate metric.

ALTERNATIVE

All metrics of temperature variability, including short-term hour-to-hour temperature changes, were positively associated with MI risk, though the magnitude of effect estimates varied by metric.

Reflection on modern methods: good practices for applied statistical learning in epidemiology

Statistical learning includes methods that extract knowledge from complex data. Statistical learning methods beyond generalized linear models, such as shrinkage methods or kernel smoothing methods, are being increasingly implemented in public health research and epidemiology because they can perform better in instances with complex or high-dimensional data-settings in which traditional statistical methods fail. These novel methods, however, often include random sampling which may induce variability in results. Best practices in data science can help to ensure robustness. As a case study, we included four statistical learning models that have been applied previously to analyze the relationship between environmental mixtures and health outcomes. We ran each model across 100 initializing values for random number generation, or 'seeds', and assessed variability in resulting estimation and inference. All methods exhibited some seed-dependent variability in results. The degree of variability differed across methods and exposure of interest. Any statistical learning method reliant on a random seed will exhibit some degree of seed sensitivity. We recommend that researchers repeat their analysis with various seeds as a sensitivity analysis when implementing these methods to enhance interpretability and robustness of results.

Air pollution, methane super-emitters, and oil and gas wells in Northern California: the relationship with migraine headache prevalence and exacerbation

BACKGROUND

Migraine-an episodic disorder characterized by severe headache that can lead to disability-affects over 1 billion people worldwide. Prior studies have found that short-term exposure to fine particulate matter (PM_2.5), nitrogen dioxide (NO₂), and ozone increases risk of migraine-related emergency department (ED) visits. Our objective was to characterize the association between long-term exposure to sources of harmful emissions and common air pollutants with both migraine headache and, among patients with migraine, headache severity.

METHODS

From the Sutter Health electronic health record database, we identified 89,575 prevalent migraine cases between 2014 and 2018 using a migraine probability algorithm (MPA) score and 270,564 frequency-matched controls. Sutter Health delivers care to 3.5 million patients annually in Northern California. Exposures included 2015 annual average block group-level PM_2.5 and NO₂ concentrations, inverse-distance weighted (IDW) methane emissions from 60 super-emitters located within 10 km of participant residence between 2016 and 2018, and IDW active oil and gas wells in 2015 within 10 km of each participant. We used logistic and negative binomial mixed models to evaluate the association between environmental exposures and (1) migraine case status; and (2) migraine severity (i.e., MPA score > 100, triptan prescriptions, neurology visits, urgent care migraine visits, and ED migraine visits per person-year). Models controlled for age, sex, race/ethnicity, Medicaid use, primary care visits, and block group-level population density and poverty.

RESULTS

In adjusted analyses, for each 5 ppb increase in NO₂, we observed 2% increased odds of migraine case status (95% CI: 1.00, 1.05) and for each 100,000 kg/hour increase in IDW methane emissions, the odds of case status also increased (OR = 1.04, 95% CI: 1.00, 1.08). We found no association between PM_2.5 or oil and gas wells and migraine case status. PM_2.5 was linearly associated with neurology visits, migraine-specific urgent care visits, and MPA score > 100, but not triptans or ED visits. NO₂ was associated with migraine-specific urgent care and ED visits, but not other severity measures. We observed limited or null associations between continuous measures of methane emissions and proximity to oil and gas wells and migraine severity.

CONCLUSIONS

Our findings illustrate the potential role of long-term exposure to multiple ambient air pollutants for prevalent migraine and migraine severity.

Environmental hazards, social inequality, and fetal loss: Implications of live-birth bias for estimation of disparities in birth outcomes

Restricting to live births can induce bias in studies of pregnancy and developmental outcomes, but whether this live-birth bias results in underestimating disparities is unknown. Bias may arise from collider stratification due to an unmeasured common cause of fetal loss and the outcome of interest, or depletion of susceptibles, where exposure differentially causes fetal loss among those with underlying susceptibility.

METHODS

We conducted a simulation study to examine the magnitude of live-birth bias in a population parameterized to resemble one year of conceptions in California (N = 625,000). We simulated exposure to a non-time-varying environmental hazard, risk of spontaneous abortion, and time to live birth using 1000 Monte Carlo simulations. Our outcome of interest was preterm birth. We included a social vulnerability factor to represent social disadvantage, and estimated overall risk differences for exposure and preterm birth using linear probability models and stratified by the social vulnerability factor. We calculated how often confidence intervals included the true point estimate (CI coverage probabilities) to illustrate whether effect estimates differed qualitatively from the truth.

RESULTS

Depletion of susceptibles resulted in a larger magnitude of bias compared with collider stratification, with larger bias among the socially vulnerable group. Coverage probabilities were not adversely affected by bias due to collider stratification. Depletion of susceptibles reduced coverage, especially among the socially vulnerable (coverage among socially vulnerable = 46%, coverage among nonsocially vulnerable = 91% in the most extreme scenario).

CONCLUSIONS

In simulations, hazardous environmental exposures induced live-birth bias and the bias was larger for socially vulnerable women.