COVID-19 remains an ongoing threat to public health, and reliable, continuous disease monitoring programs are essential for preventing future surges of infection. However, without mandated COVID-19 testing, accurate data of confirmed cases are unavailable. Instead, COVID-19 viruses may be tracked via wastewater samples from sewage manholes in areas of high social connectivity, where captured viral RNA data are biomarkers useful for monitoring and predicting community-level COVID-19 prevalence through machine learning techniques. We construct a prediction model of high sensitivity and specificity to provide evidence of significant underreporting of COVID-19 cases for the time period following the lifting of testing mandates.

Introduction: Researchers, whether working in wet-labs, dry-labs, clinical settings, or field environments, encounter various hazards. However, there has been limited study on the health and safety of academic researchers. This study aimed to investigate hazardous occupational exposures and safety among researchers in academic settings at a large U.S. Midwestern public research university. Method: A cross-sectional survey was administered to university researchers, including students, postdoctoral fellows, staff, and faculty members. The survey covered five domains: demographics, psychological stress, workplace safety perceptions, hazardous exposures, and job-related morbidity and injuries. This analysis reports on the demographic, hazardous exposure, and job-related morbidity and injury data collected. The survey instrument used was validated through an expert review (n = 6) and pretested by volunteers (n = 13). Results: A total of 1,202 participants completed the survey. The majority were wet-lab researchers (42.7%), female (68.8%), and research staff (39.6%), with over half having ≤ 5 years of experience at the institution. The most frequently reported hazard categories were ergonomic issues, including prolonged sitting or standing (64%), eye strain (54.2%), and repetitive motion (32%), as well as psychosocial hazards such as excessive workload (21.6%), ineffective communication or lack of support (20.8%), and isolation (20.4%). Approximately 12.5% of participants reported work-related incidents, including injuries, property damage, and near misses, and another 5.4% reported work-related illnesses, with the predominant cause attributed to physical hazards. Conclusions: The findings highlight the necessity for stress management programs, ergonomic interventions, and ongoing safety training among university researchers to address these challenges. Future research should aim to develop and assess interventions that fill these gaps, creating a safer and more supportive environment for researchers. Practical Applications: The results offer research institutions an opportunity to implement these initiatives to enhance the well-being and productivity of their research workforce.

Environmental regulation of single pollutants likely underprotects communities disproportionately burdened by multiple, overlapping environmental hazards. We investigate whether historically redlined neighborhoods across the US are exposed to cumulative environmental inequities today. We overlaid 1930s-40s Home Owners' Loan Corporation maps of 202 US cities from the Mapping Inequality project onto EPA EJScreen data to analyze whether a collective, simultaneous, IQR increase in 12 environmental hazards (various air pollutants and toxic facility proximities) was significantly associated with higher odds of a neighborhood having been historically redlined (D-grade) using a boosted regression tree model. Controlling for neighborhood socioeconomic status, a collective, simultaneous, IQR increase in 12 environmental hazards was associated with 1.30 (95%CI: 1.17-1.46) times higher odds of a neighborhood having been historically redlined. Proximity to hazardous waste and wastewater discharge sites, traffic volume, and diesel particulate matter were the most pervasive environmental hazards in historically redlined neighborhoods. Cumulative environmental inequities were largest in the Western US by region and in Oklahoma City, Cincinnati, and Detroit by city. We find that historically redlined neighborhoods may be disproportionately impacted by cumulative environmental impacts. Environmental regulation of single pollutants may not be sufficiently protective of historically marginalized communities, which may help us understand why contemporary environmental health disparities persist today.

BACKGROUND:Outdoor workers are at high risk of hazardous heat exposures and associated adverse health outcomes. Questions remain as to how many workers are exposed. OBJECTIVES:To estimate the yearly incidence of outdoor workers in the United States (US) exposed to hazardous heat from 2010 to 2019 and assess whether marginalized populations of workers are disproportionately exposed. METHODS:We simulated the census tract-level, daily proportion of workers exposed to hazardous heat using multiple datasets containing (i) employment counts, (ii) work characteristics that influence susceptibility to heat exposure (e.g., metabolic rate), and (iii) county-level data on daily-average wet bulb globe temperature. Occupations were classified according to the US SOC structure. Overexposure was determined using ACGIH screening criteria for unacclimatized workers. Daily exposure proportions were summarized as annual, average exposure rates per 100 workers. At the census tract level, annual exposure rates and sociodemographic estimates were merged together to ecologically assess exposure disparities using generalized additive models. Estimates are publicly available alongside an interactive map ( https://sph-umich.shinyapps.io/work-heat-2010-2019/ ). RESULTS:Estimated daily incidence ranged from 0 to 7.3% of workers. Nationwide, there were an estimated 1.46 daily exposures per 100 workers from 2010 to 2019, equivalent to 5.44 billion worker-days. Nationwide exposure rates did not vary substantially year-to-year. Construction and Extraction (9.77 per 100 workers), Installation, Maintenance, and Repair (8.05 per 100 workers), and Farming, Fishing, and Forestry occupations (7.51 per 100 workers) had the highest estimated exposure rates. Regionally, estimated rates were highest in the South (3.26 per 100 workers), particularly in Florida and Texas. Low-income individuals, individuals without a high school diploma, foreign-born populations, and racial and ethnic minority individuals (particularly Hispanic) were disproportionately exposed. SIGNIFICANCE:The frequency of occupational heat exposure, its disproportionate burden on marginalized workers, and the increasing impact of climate change suggests a critical and urgent need for occupational regulation and surveillance of heat exposure. IMPACT:This study characterizes potentially hazardous heat exposures among outdoor workers for every census tract in the contiguous United States during 2010-2019. Current epidemiological and health impact studies of heat-related disease and mortality do not typically consider how occupational heat exposure may exacerbate adverse heat-related health outcomes in communities experiencing high heat. Our approach could reduce exposure misclassification bias in these studies. Policy-wise, these spatiotemporal estimates can support regulatory enforcement and consultation of occupational heat exposure by better targeting particular communities with high heat exposure. These estimates can also inform surveillance efforts and healthcare resources of state and local health departments.

The contribution of occupational exposures to the extent of cumulative environmental impacts, and their implications for environmental justice (EJ), have not been investigated. We (a) characterized communities with cumulatively high occupational and environmental exposures, (b) examined whether marginalized, historically redlined neighborhoods were disproportionately affected by these exposures, and (c) evaluated the implications of failing to consider workplace exposures in EJ screening tools in Michigan. At the census tract-level, we combined occupational exposure estimates of six common workplace hazards, environmental exposures from EJScreen and the National Transportation Noise Map, demographic information from the American Community Survey, and redlining information from the 1930s Home Owners' Loan Corporation maps to test the first two objectives using supervised and unsupervised statistical methods. The last objective incorporated the occupational indicators into the Michigan-specific EJ screening tool (MiEJScreen) to test the third objective. Among 2,772 Michigan census tracts, 738 (27%) had cumulatively high occupational and environmental exposures, primarily in urban areas. Tracts with >90% (compared to <10%) of racial and ethnic minority individuals had 2.31 (95% CI: 1.78-3.03) times higher odds of cumulatively high exposures. A simultaneous increase to the 90th percentile (relative to the 50th) in all 13 occupational and environmental exposures was associated with 2.47 (95% CI: 1.20-5.36) times higher odds of a tract having been historically redlined. Not incorporating occupational exposures into the MiEJScreen would overlook 90 census tracts with cumulatively high environmental and occupational impacts, affecting around 255,000 individuals. Ignoring occupational exposures in cumulative environmental impact assessments may overlook important EJ hotspots.

Racial and ethnic inequities in environmental noise exist in the US, partially attributable to historical structural racism. However, previous studies have not considered the totality of people’s exposures. Since people spend most of their waking time at work, there is a need to consider cumulative exposure to noise both in and out of the workplace to understand who is most at risk of noise pollution-related adverse health outcomes. To (1) investigate whether racial and ethnic minority communities are disproportionately burdened by transportation- and workplace-related noise pollution, and (2) assess whether structural racism through historically redlined neighborhoods with sustained mortgage discrimination partially contribute to the hypothesized inequity. We characterized the prevalence of workplace noise and transportation noise exposure by census tract across the US. We analyzed the census tract-level association between racial and ethnic composition and the population exposed to both transportation- and workplace-related noise pollution in the 2010s using geospatial models. We then assessed census tract-level associations with transportation and workplace noise pollution using historical redlining in the 1930s as the primary covariate, stratified by mortgage discrimination in the 1990s using a similar geospatial model, controlling for census tract-level indicators of low socioeconomic status. Higher percentages of racial and ethnic minority individuals, particularly Hispanic/Latino and non-Hispanic Black Americans, were associated with significantly higher odds of exposure to both transportation and workplace noise (odds ratio = 8.59, 95% CI: 7.38–10.0, when comparing within-metropolitan area, highest to lowest quintile percentages). These disparities are particularly profound in urban areas. Urban tracts which experienced residential segregation in the 1930s, even without sustained mortgage discrimination in the 1990s, have a significantly higher percentage of individuals exposed to both transportation and workplace noise today compared to those without historical segregation (1.55%, 95% CI: 1.37–1.74). This inequity is even higher among historically segregated tracts that experienced sustained mortgage discrimination (1.83%, 95% CI: 1.66–2.01). These findings can advance environmental justice initiatives by informing regulatory action to protect communities of color from noise pollution both environmentally and during work. Our study provides evidence that neighborhoods with a higher proportion of racial and ethnic minority individuals are cumulatively burdened by noise pollution both during work and from transportation sources in their home communities. This suggests that not incorporating workplace exposures when assessing environmental impacts may overlook the most burdened communities. Future environmental justice efforts and policies should consider assessing workplace exposures to reduce environmental health disparities more effectively.

Solid waste workers encounter a number of occupational hazards that are likely to induce stress. Thus, there are likely to be psychosocial factors that also contribute to their overall perceptions of organizational health. However, attitudes regarding the aforementioned among solid waste workers’ have not been assessed. This descriptive, cross-sectional pilot study operationalized the INPUTS Survey to determine workers’ perceptions of organizational health and other psychosocial factors of work. Percentage and mean responses to each INPUTS domain are presented in accordance with their survey manual. Pearson’s chi-squared tests were run on count data; Fisher’s exact tests were run for count data with fewer than five samples. ANOVAs were run on the continuous items. Due to a relatively low sample size (N = 68), two-sided p values < 0.1 were considered statistically significant. Most solid waste worker participants reported high decision authority, that they perceived their management to prioritize workplace health and safety, and had high job satisfaction. However, perceptions of support for health outside of the realm of occupational safety and health was lower. Addressing traditional occupational health hazards continues to take precedence in this industry, with less of a focus on how the social determinants of health may impact workplace health.

Ignoring workplace exposures that occur beyond the local residential context in place-based risk indices like the CDC's Social Vulnerability Index (SVI) likely misclassifies community exposure by under-counting risks and obscuring true drivers of racial/ethnic health disparities. To investigate this hypothesis, we developed several place-based indicators of occupational exposure and examined their relationships with race/ethnicity, SVI, and health inequities. We used publicly available job exposure matrices and employment estimates from the United States (US) Census to create and map six indicators of occupational hazards for every census tract in the US. We characterized census tracts with high workplace-low SVI scores. We used natural cubic splines to examine tract level associations between the percentage of racial/ethnic minorities (individuals who are not non-Hispanic White) and the occupational indicators. Lastly, we stratified each census tract into high/low occupational noise, chemical pollutant, and disease/infection exposure to examine racial/ethnic health disparities to diabetes, asthma, and high blood pressure, respectively, as a consequence of occupational exposure inequities. Our results show that racial/ethnic minority communities, particularly those that are also low-income, experience a disproportionate burden of workplace exposures that may be contributing to racial/ethnic health disparities. When composite risk measures, such as SVI, are calculated using only information from the local residential neighborhood, they may systematically under-count occupational risks experienced by the most vulnerable communities. There is a need to consider the role of occupational justice on nationwide, racial/ethnic health disparities.

Objectives: This study assessed the relationship between occupational noise exposure and the incidence of workplace fatal injury (FI) and nonfatal injury (NFI) in the United States from 2006 to 2020. It also examined whether distinct occupational and industrial clusters based on noise exposure characteristics demonstrated varying risks for FI and NFI. Methods: An ecological study design was utilized, employing data from the U.S. Bureau of Labor Statistics for FI and NFI and demographic data, the U.S. Census Bureau for occupation/industry classification code lists, and the U.S./Canada Occupational Noise Job Exposure Matrix for noise measurements. We examined four noise metrics as predictors of FI and NFI rates: mean Time-Weighted Average (TWA), maximum TWA, standard deviation of TWA, and percentage of work shifts exceeding 85 or 90 dBA for 619 occupation-years and 591 industry-years. Kmeans clustering was used to identify clusters of noise exposure characteristics. Mixed-effects negative binomial regression examined the relationship between the noise characteristics and FI/NFI rates separately for occupation and industry. Results: Among occupations, we found significant associations between increased FI rates and higher mean TWA (IRR: 1.06, 95% CI: 1.01-1.12) and maximum TWA (IRR: 1.10, 95% CI: 1.07-1.14), as well as TWA exceedance (IRR: 1.04, 95% CI: 1.01-1.07). Increased rates of NFI were found to be significantly associated with maximum TWA (IRR: 1.06, 95% CI: 1.04-1.09) and TWA exceedance (IRR: 1.03, 95% CI: 1.01-1.05). In addition, occupations with both higher exposure variability (IRR with FI rate: 1.49, 95% CI: 1.23-1.80; IRR with NFI rate: 1.40, 95% CI: 1.14-1.73) and higher level of sustained exposure (IRR with FI rate: 1.27, 95% CI: 1.12-1.44; IRR with NFI rate: 1.21, 95% CI: 1.05-1.39) were associated with higher rates of FI and NFI compared to occupations with low noise exposure. Among industries, significant associations between increased NFI rates and higher mean TWA (IRR: 1.05, 95% CI: 1.02-1.08) and maximum TWA (IRR: 1.06, 95% CI: 1.04-1.08) were observed. Unlike the occupation-specific analysis, industries with higher exposure variability and higher sustained exposures did not display significantly higher FI/NFI rates compared to industries with low exposure. Conclusions: The results suggest that occupational noise exposure may be an independent risk factor for workplace FIs/NFIs, particularly for workplaces with highly variable noise exposures. The study highlights the importance of comprehensive occupational noise assessments.

In the United States, the majority of waste workers work with solid waste. In solid waste operations, collection, sorting, and disposal can lead to elevated biohazard exposures (e.g., bioaerosols, bloodborne and other pathogens, human and animal excreta). This cross-sectional pilot study aimed to characterize solid waste worker perception of biohazard exposures, as well as worker preparedness and available resources (e.g., access to personal protective equipment, level of training) to address potential biohazard exposures. Three sites were surveyed: (1) a family-owned, small-scale waste disposal facility, (2) a county-level, recycling-only facility, and (3) an industrial-sized, large-scale facility that contains a hauling and landfill division. Survey items characterized occupational biohazards, resources to mitigate and manage those biohazards, and worker perceptions of biohazard exposures. Descriptive statistics were generated. The majority of workers did not report regularly coming into contact with blood, feces, and bodily fluids (79%). As such, less than one-fifth were extremely concerned about potential illness from biological exposures (19%). Yet, most workers surveyed (71%) reported an accidental laceration/cut that would potentially expose workers to biohazards. This study highlights the need for additional research on knowledge of exposure pathways and perceptions of the severity of exposure among this occupational group.

Background:Effort-reward imbalance (ERI) and overcommitment at work have been associated poorer mental health. However, nonlinear and nonadditive effects have not been investigated previously. Methods:The association between effort, reward, and overcommitment with odds of poorer mental health was examined among a sample of 68 formal United States waste workers (87% male). Traditional, logistic regression and Bayesian Kernel machine regression (BKMR) modeling was conducted. Models controlled for age, education level, race, gender, union status, and physical health status. Results:The traditional, logistic regression found only overcommitment was significantly associated with poorer mental health (IQR increase: OR = 6.7; 95% CI: 1.7 to 25.5) when controlling for effort and reward (or ERI alone). Results from the BKMR showed that a simultaneous IQR increase in higher effort, lower reward, and higher overcommitment was associated with 6.6 (95% CI: 1.7 to 33.4) times significantly higher odds of poorer mental health. An IQR increase in overcommitment was associated with 5.6 (95% CI: 1.6 to 24.9) times significantly higher odds of poorer mental health when controlling for effort and reward. Higher effort and lower reward at work may not always be associated with poorer mental health but rather they may have an inverse, U-shaped relationship with mental health. No interaction between effort, reward, or overcommitment was observed. Conclusion:When taking into the consideration the relationship between effort, reward, and overcommitment, overcommitment may be most indicative of poorer mental health. Organizations should assess their workers' perceptions of overcommitment to target potential areas of improvement to enhance mental health outcomes.

OBJECTIVES:To elucidate whether occupational exposure to soft paper dust increases the incidence of cancer.METHODS:We studied 7988 workers in Swedish soft paper mills from 1960 to 2008, of whom 3233 (2 187 men and 1046 women) had more than 10 years of employment. They were divided into high exposure (>5 mg/m3 for >1 year) or lower exposure to soft paper dust based on a validated job-exposure matrix. They were followed from 1960 to 2019, and person-years at risk were stratified according to gender, age, and calendar-year. The expected numbers of incident tumors were calculated using the Swedish population as the reference, and standardized incidence ratios (SIR) with 95% confidence intervals (95% CI) were assessed.RESULTS:Among high-exposure workers with more than 10 years of employment, there was an increased incidence of colon cancer (SIR 1.66, 95% CI 1.20-2.31), small intestine cancer (SIR 3.27, 95% CI 1.36-7.86), and thyroid gland cancer (SIR 2.68, 95% CI 1.11-6.43), as well as lung cancer (SIR 1.56, 95% CI 1.12-2.19). Among the lower-exposed workers there was an increased incidence of connective tissue tumors (sarcomas) (SIR 2.26, 95% CI 1.13-4.51) and pleural mesothelioma (SIR 3.29, 95% CI 1.37-7.91).CONCLUSION:Workers in soft paper mills with high exposure to soft paper dust have an increased incidence of large and small intestine tumors. Whether the increased risk is caused by paper dust exposure or some unknown associated factors is unclear. The increased incidence of pleural mesothelioma is probably linked to asbestos exposure. The reason for increased incidence of sarcomas is unknown.

OBJECTIVE:To elucidate whether occupational noise exposure increases the mortality from ischemic heart disease (IHD) and stroke, and if exposure to paper dust modified the risks. METHODS:We studied 6686 workers from soft paper mills, with occupational noise exposure, < 85 dBA, 85-90 dBA and > 90 dBA, and high (> 5 mg/m3) exposure to paper dust. Person-years 1960-2019 were stratified according to gender, age, and calendar-year. Expected numbers of deaths were calculated using the Swedish population as the reference and standardized mortality ratios (SMR) with 95% confidence intervals (95% CI) were assessed. RESULTS:SMR for IHD was 1.12 (95% CI 0.88-1.41) for noise < 85 dBA, 1.18 (95% CI 0.90-1.55) for 85-90 dBA, and 1.27 (95% CI 1.10-1.47) among workers exposed > 90 dBA. Joint exposure to high noise exposure and high exposure to paper dust resulted in slightly higher IHD mortality (SMR 1.39, 95% CI 1.15-1.67). SMR for ischemic stroke was 0.90 (95% CI 0.37-2.15) for noise < 85 dBA, 1.08 (95% CI 0.45-2.59) for 85-90 dBA, and 1.48 (95% CI 0.99-2.00) among workers exposed > 90 dBA. High noise exposure and high exposure to paper dust resulted in higher ischemic stroke mortality (SMR 1.83, 95% CI 1.12-2.98). CONCLUSION:Noise levels > 90 dBA was associated with increased IHD mortality. Combined exposures of noise and paper dust may further increase the risks. Our results do not provide support for a causal relationship for ischemic stroke. Residual confounding from smoking has to be considered. Workers need to be protected from occupational noise levels exceeding 90 dBA.

While perceptions of risk have been examined in the workplace to understand safety behavior, hazard perception has been overlooked, particularly for chemical, physical, and biological agents.This study sought to establish the prevalence of one type of mismatch in hazard perception, - noise misperception - among miners, to examine whether different types of noisy environments (e.g., continuous, highly variable, etc.) alter workers' misperception of their noise exposures, and to evaluate whether noise misperception is associated with hearing protection device (HPD) use behavior.In this cross-sectional study across 10 surface mines in the USA, 135 normal-hearing participants were surveyed on their perceptions of exposure to noise at work and were monitored for three shifts, each with personal noise dosimetry, to examine which workers had a mismatch in perceived versus true noise exposure by 8-hr, time-weighted average, NIOSH exposure limits (TWANIOSH). Mixed effects logistic regression and probit Bayesian Kernel Machine Regression (BKMR) models examining on the odds of noise misperception associated with four different noise metrics (kurtosis, crest factor, variability, and number of peaks >135 dB) were used to determine which types of noisy environments may influence noise misperception. The relationship between noise misperception and odds of not wearing HPDs during a work shift was further examined.Our findings showed that nearly 1 in 3 workers underestimated their exposure to noise when their true exposure was in fact hazardous (TWANIOSH >= 85 dBA) for at least one shift, and 6% misperceived hazardous exposures for all shifts. Work shifts with highly kurtotic noise distributions (>3) had 3.1 (95% CI: 1.1 to 8.4) times significantly higher odds of resulting in misperceived noise; no other noise metric was significantly associated with noise misperception. BKMR modeling provided further evidence that kurtosis dominates this relationship, with an IQR increase in kurtosis significantly associated with 1.68 (95% CI: 1.13 to 2.50) higher odds of noise misperception. Although not statistically significant, misperception of hazardous noise exposure was associated with 3.2 (95% CI: 0.8 to 12.5) times higher odds of not using earplugs during a work shift.Misperception of noise occurs in the workplace, and likely occurs for other physical, chemical, and biological exposures. This hazard misperception may influence risk perceptions and worker behavior and reduce the effectiveness of behavior-related training. Elimination, substitution, or engineering controls of exposures is the best way to prevent hazard misperceptions and exposure-related diseases.

Globally, noise exposure from occupational and nonoccupational sources is common, and, as a result, noise-induced hearing loss affects tens of millions of people. Occupational noise exposures have been studied and regulated for decades, but nonoccupational sound exposures are not well understood. The nationwide Apple Hearing Study, launched using the Apple research app in November 2019 (Apple Inc., Cupertino, CA), is characterizing the levels at which participants listen to headphone audio content, as well as their listening habits. This paper describes the methods of the study, which collects data from several types of hearing tests and uses the Apple Watch noise app to measure environmental sound levels and cardiovascular metrics. Participants, all of whom have consented to participate and share their data, have already contributed nearly 300 x 10(6) h of sound measurements and 200 000 hearing assessments. The preliminary results indicate that environmental sound levels have been higher, on average, than headphone audio, about 10% of the participants have a diagnosed hearing loss, and nearly 20% of the participants have hearing difficulty. The study's analyses will promote understanding of the overall exposures to sound and associated impacts on hearing and cardiovascular health. This study also demonstrates the feasibility of collecting clinically relevant exposure and health data outside of traditional research settings. (C) 2022 Acoustical Society of America.

Recently, the National Institute for Occupational Safety and Health (NIOSH) released an updated version of the NIOSH Industry and Occupation Computerized Coding System (NIOCCS), which uses supervised machine learning to assign industry and occupational codes based on provided free-text information. However, no efforts have been made to externally verify the quality of assigned industry and job titles when the algorithm is provided with inputs of varying quality. This study sought to evaluate whether the NIOCCS algorithm was sufficiently robust with low-quality inputs and how variable quality could impact subsequent job estimated exposures in a large job-exposure matrix for noise (NoiseJEM). Using free-text industry and job descriptions from >700,000 noise measurements in the NoiseJEM, three files were created and input into NIOCCS: (1) N1, "raw" industries and job titles; (2) N2, "refined" industries and "raw" job titles; and (3) N3, "refined" industries and job titles. Standardized industry and occupation codes were output by NIOCCS. Descriptive statistics of performance metrics (e.g., misclassification/discordance of occupation codes) were evaluated for each input relative to the original NoiseJEM dataset (N0). Across major Standardized Occupational Classifications (SOC), total discordance rates for N1, N2, and N3 compared to N0 were 53.6%, 42.3%, and 5.0%, respectively. The impact of discordance on the major SOC group varied and included both over- and under-estimates of average noise exposure compared to N0. N2 had the most accurate noise exposure estimates (i.e., smallest bias) across major SOC groups compared to N1 and N3. Further refinement of job titles in N3 showed little improvement. Some variation in classification efficacy was seen over time, particularly prior to 1985. Machine learning algorithms can systematically and consistently classify data but are highly dependent on the quality and amount of input data. The greatest benefit for an end-user may come from cleaning industry information before applying this method for job classification. Our results highlight the need for standardized classification methods that remain constant over time.

Background and Aim: The Home Owners' Loan Corporation (HOLC) in the 1930s drew maps of cities across the US that labeled neighborhoods by mortgage risk. This historical practice, commonly called "redlining", labeled neighborhoods deemed "hazardous" with the color red. This policy denied mortgage loans to minority persons seeking homes in White/affluent neighborhoods, segregating these areas by race and ethnicity. This legacy of de jure segregation shaped neighborhoods today and influenced their environmental exposures, a form of environmental racism. The aim of this study was to identify the most pervasive exposures associated with redlining. Methods: The Detroit shapefile defined by the HOLC and digitized by the Mapping Inequality project overlaid onto the Environmental Protection Agency's EJSCREEN and the Department of Transportation National Transportation Noise Map was used to determine modern environmental exposures and transportation noise within historical boundaries. Differences in demographic and environmental hazards between redlined (red or D grade) and non-redlined neighborhoods (grades A, B, and C) were assessed using hypothesis testing and a boosted classification tree algorithm. Results: Historically redlined Detroit neighborhoods experience significantly higher environmental hazards than non-redlined neighborhoods from diesel particulate matter (PM), traffic volumes, hazardous road noise, cancer risk from air pollution, and are closer to hazardous waste and Risk Management Plan (RMP) sites. With all factors taken together, boosted regression trees indicated the most pervasive environmental exposures among redlined neighborhoods in Detroit are the proximity to RMP sites, hazardous road noise, diesel PM, and cancer risk from air pollution. Conclusions: Institutional segregation via historical redlining is associated with environmental injustices in Detroit today. Policies targeting transportation-related air and noise pollution, particularly from sources of diesel exhaust, in redlined neighborhoods may ameliorate some of the disproportionate impacts of historical redlining, providing a proof-of-concept to apply to other redlined cities. Keywords: environmental justice; redlining; EJSCREEN; noise, Detroit

BACKGROUND AND AIM: Systematic racial/ethnic segregation via historical redlining has been identified as a potential contributor to environmental injustices in the US today, particularly air pollution. However, no studies have examined how the effect of current-day racial/ethnic segregation has modified this association. This study examined this hypothesized effect modification on current levels of air pollution (i.e., PM2.5 and diesel PM) across the USA. METHODS: Current-day demographics and air pollution levels were estimated for USA census tracts using the EPA EJSCREEN. Census-tract-level HOLC grade was assessed using 2010 Historical Redlining Scores, and current-day segregation from the dissimilarity index. We examined differences in air pollution by HOLC grades using city-adjusted, intra-urban air pollution levels from mixed-effects linear regression models using cities as a random intercept to account for city-to-city differences in air pollution levels. We examined our hypothesized effect by investigating the joint effects of redlining (comparing A- and D-grade census tracts) and current-day segregation (comparing low and moderate/high levels of dissimilarity) on log-transformed, population-weighted average diesel PM and PM2.5 levels. RESULTS: Overall, D-grade air pollution levels are 28% (95% CI: 25% to 31%) and 1.2% (1% to 1.4%) higher than A-grade census tracts for diesel PM and PM2.5, respectively. Moderately/highly segregated, D-grade census tracts have 70% (49% to 95%) times higher diesel PM levels and 5% (0%, 11%) times higher PM2.5 levels compared to lowly segregated, A-grade census tracts. Estimates indicate significant departures from additive (RERI: 24%; 14% to 34%) and multiplicative joint effects (percent change of interaction term: 10%; 3% to 17%) for diesel PM, and similarly for PM2.5 (RERI: 1%, 1% to 2%; percent change of interaction term: 1%, 1% to 2%). CONCLUSIONS: Current-day segregation modifies the association between historical redlining and air pollution in the USA. KEYWORDS: environmental justice; redlining; air pollution; effect modification; RERI