SARS-CoV-2 antibody prevalence by industry, workplace characteristics, and workplace infection prevention and control measures, North Carolina, 2021 to 2022.

Background: The COVID-19 pandemic has disproportionately affected workers in certain industries and occupations, and the workplace can be a high risk setting for SARS-CoV-2 transmission. In this study, we measured SARS-CoV-2 antibody prevalence and identified work-related risk factors in a population primarily working at industrial livestock operations. Methods: We used a multiplex salivary SARS-CoV-2 IgG antibody assay to determine infection-induced antibody prevalence among 236 adult (≥18 years) North Carolina residents between February 2021 and August 2022. We used the National Institute for Occupational Safety and Health Industry and Occupation Computerized Coding System (NIOCCS) to classify employed participants' industry and compared infection-induced IgG prevalence by participant industry and with the North Carolina general population. We also combined antibody results with reported SARS-CoV-2 molecular test positivity and vaccination history to identify evidence of prior infection. We used logistic regression to estimate odds ratios of prior infection by potential work-related risk factors, adjusting for industry and date. Results: Most participants (55%) were infection-induced IgG positive, including 71% of animal slaughtering and processing industry workers, which is 1.5 to 4.3 times higher compared to the North Carolina general population, as well as higher than molecularly-confirmed cases and the only other serology study we identified of animal slaughtering and processing workers. Considering questionnaire results in addition to antibodies, the proportion of participants with evidence of prior infection increased slightly, to 61%, including 75% of animal slaughtering and processing workers. Participants with more than 1000 compared to 10 or fewer coworkers at their jobsite had higher odds of prior infection (adjusted odds ratio [aOR] 4.5, 95% confidence interval [CI] 1.0 to 21.0). Conclusions: This study contributes evidence of the severe and disproportionate impacts of COVID-19 on animal processing and essential workers and workers in large congregate settings. We also demonstrate the utility of combining non-invasive biomarker and questionnaire data for the study of workplace exposures.

Evaluation of Calibration Approaches for Indoor Deployments of PurpleAir Monitors.

Low-cost air quality monitors are growing in popularity among both researchers and community members to understand variability in pollutant concentrations. Several studies have produced calibration approaches for these sensors for ambient air. These calibrations have been shown to depend primarily on relative humidity, particle size distribution, and particle composition, which may be different in indoor environments. However, despite the fact that most people spend the majority of their time indoors, little is known about the accuracy of commonly used devices indoors. This stems from the fact that calibration data for sensors operating in indoor environments are rare. In this study, we sought to evaluate the accuracy of the raw data from PurpleAir fine particulate matter monitors and for published calibration approaches that vary in complexity, ranging from simply applying linear corrections to those requiring co-locating a filter sample for correction with a gravimetric concentration during a baseline visit. Our data includes PurpleAir devices that were co-located in each home with a gravimetric sample for 1-week periods (265 samples from 151 homes). Weekly-averaged gravimetric concentrations ranged between the limit of detection (3 µg/m3) and 330 µg/m3. We found a strong correlation between the PurpleAir monitor and the gravimetric concentration (R>0.91) using internal calibrations provided by the manufacturer. However, the PurpleAir data substantially overestimated indoor concentrations compared to the gravimetric concentration (mean bias error ≥ 23.6 µg/m3 using internal calibrations provided by the manufacturer). Calibrations based on ambient air data maintained high correlations (R ≥ 0.92) and substantially reduced bias (e.g. mean bias error = 10.1 µg/m3 using a US-wide calibration approach). Using a gravimetric sample from a baseline visit to calibrate data for later visits led to an improvement over the internal calibrations, but performed worse than the simpler calibration approaches based on ambient air pollution data. Furthermore, calibrations based on ambient air pollution data performed best when weekly-averaged concentrations did not exceed 30 µg/m3, likely because the majority of the data used to train these models were below this concentration.

Characterizing metals in particulate pollution in communities at the fenceline of heavy industry: combining mobile monitoring and size-resolved filter measurements.

Exposures to metals from industrial emissions can pose important health risks. The Chester-Trainer-Marcus Hook area of southeastern Pennsylvania is home to multiple petrochemical plants, a refinery, and a waste incinerator, most abutting socio-economically disadvantaged residential communities. Existing information on fenceline community exposures is based on monitoring data with low temporal and spatial resolution and EPA models that incorporate industry self-reporting. During a 3 week sampling campaign in September 2021, size-resolved particulate matter (PM) metals concentrations were obtained at a fixed site in Chester and on-line mobile aerosol measurements were conducted around Chester-Trainer-Marcus Hook. Fixed-site arsenic, lead, antimony, cobalt, and manganese concentrations in total PM were higher (p < 0.001) than EPA model estimates, and arsenic, lead, and cadmium were predominantly observed in fine PM (<2.5 µm), the PM fraction which can penetrate deeply into the lungs. Hazard index analysis suggests adverse effects are not expected from exposures at the observed levels; however, additional chemical exposures, PM size fraction, and non-chemical stressors should be considered in future studies for accurate assessment of risk. Fixed-site MOUDI and nearby mobile aerosol measurements were moderately correlated (r ≥ 0.5) for aluminum, potassium and selenium. Source apportionment analyses suggested the presence of four major emissions sources (sea salt, mineral dust, general combustion, and non-exhaust vehicle emissions) in the study area. Elevated levels of combustion-related elements of health concern (e.g., arsenic, cadmium, antimony, and vanadium) were observed near the waste incinerator and other industrial facilities by mobile monitoring, as well as in residential-zoned areas in Chester. These results suggest potential co-exposures to harmful atmospheric metal/metalloids in communities surrounding the Chester-Trainer-Marcus Hook industrial area at levels that may exceed previous estimates from EPA modeling.

Laboratory comparison of field portable X-ray fluorescence spectrometer (FP-XRF) and inductively coupled plasma mass spectrometry (ICP-MS) for determination of airborne metals in stainless steel welding fume.

Welding fume is a common exposure in occupational settings. Gravimetric analysis for total particulate matter is common; however, the cost of laboratory analyses limits the availability of quantitative exposure assessment for welding fume metal constituents in occupational settings. We investigated whether a field portable X-ray fluorescence spectrometer (FP-XRF) could provide accurate estimates of personal exposures to metals common in welding fume (chromium, copper, manganese, nickel, vanadium, and zinc). The FP-XRF requires less training and is easier to deploy in many settings than traditional wet laboratory analyses. Filters were analyzed both by FP-XRF and inductively coupled plasma mass spectrometry (ICP-MS). We estimated the FP-XRF limit of detection for each metal and developed a correction factor accounting for the non-uniform deposition pattern on filter samples collected with an Institute of Medicine (IOM) inhalable particulate matter sampler. Strong linear correlation was observed for all metals (0.72

Indoor Pollution and Lung Function Decline in Current and Former Smokers: SPIROMICS AIR.

Rationale: Indoor pollutants have been associated with chronic obstructive pulmonary disease morbidity, but it is unclear whether they contribute to disease progression. Objectives: We aimed to determine whether indoor particulate matter (PM) and nitrogen dioxide (NO2) are associated with lung function decline among current and former smokers. Methods: Of the 2,382 subjects with a history of smoking in SPIROMICS AIR, 1,208 participants had complete information to estimate indoor PM and NO2, using individual-based prediction models, in relation to measured spirometry at two or more clinic visits. We used a three-way interaction model between time, pollutant, and smoking status and assessed the indoor pollutant-associated difference in FEV1 decline separately using a generalized linear mixed model. Measurements and Main Results: Participants had an average rate of FEV1 decline of 60.3 ml/yr for those currently smoking compared with 35.2 ml/yr for those who quit. The association of indoor PM with FEV1 decline differed by smoking status. Among former smokers, every 10 µg/m3 increase in estimated indoor PM was associated with an additional 10 ml/yr decline in FEV1 (P = 0.044). Among current smokers, FEV1 decline did not differ by indoor PM. The results of indoor NO2 suggest trends similar to those for PM ⩽2.5 µm in aerodynamic diameter. Conclusions: Former smokers with chronic obstructive pulmonary disease who live in homes with high estimated PM have accelerated lung function loss, and those in homes with low PM have lung function loss similar to normal aging. In-home PM exposure may contribute to variability in lung function decline in people who quit smoking and may be a modifiable exposure.

Identifying optimal co-location calibration periods for low-cost sensors.

Low-cost sensors are often co-located with reference instruments to assess their performance and establish calibration equations, but limited discussion has focused on whether the duration of this calibration period can be optimized. We placed a multipollutant monitor that contained sensors that measure particulate matter smaller than 2.5 µm (PM2.5), carbon monoxide (CO), nitrogen dioxide (NO2), ozone (O3), and nitric oxide (NO) at a reference field site for one year. We developed calibration equations using randomly selected co-location subsets spanning 1 to 180 consecutive days out of the 1-year period and compared the potential root mean square errors (RMSE) and Pearson correlation coefficients (r). The co-located calibration period required to obtain consistent results varied by sensor type, and several factors increased the co-location duration required for accurate calibration, including the response of a sensor to environmental factors, such as temperature or relative humidity (RH), or cross-sensitivities to other pollutants. Using measurements from Baltimore, MD, where a broad range of environmental conditions may be observed over a given year, we found diminishing improvements in the median RMSE for calibration periods longer than about six weeks for all the sensors. The best performing calibration periods were the ones that contained a range of environmental conditions similar to those encountered during the evaluation period (i.e., all other days of the year not used in the calibration). With optimal, varying conditions it was possible to obtain an accurate calibration in as little as one week for all sensors, suggesting that co-location can be minimized if the period is strategically selected and monitored so that the calibration period is representative of the desired measurement setting.

Laboratory and field evaluation of a low-cost methane sensor and key environmental factors for sensor calibration.

Low-cost sensors enable finer-scale spatiotemporal measurements within the existing methane (CH4) monitoring infrastructure and could help cities mitigate CH4 emissions to meet their climate goals. While initial studies of low-cost CH4 sensors have shown potential for effective CH4 measurement at ambient concentrations, sensor deployment remains limited due to questions about interferences and calibration across environments and seasons. This study evaluates sensor performance across seasons with specific attention paid to the sensor's understudied carbon monoxide (CO) interferences and environmental dependencies through long-term ambient co-location in an urban environment. The sensor was first evaluated in a laboratory using chamber calibration and co-location experiments, and then in the field through two 8 week co-locations with a reference CH4 instrument. In the laboratory, the sensor was sensitive to CH4 concentrations below ambient background concentrations. Different sensor units responded similarly to changing CH4, CO, temperature, and humidity conditions but required individual calibrations to account for differences in sensor response factors. When deployed in-field, co-located with a reference instrument near Baltimore, MD, the sensor captured diurnal trends in hourly CH4 concentration after corrections for temperature, absolute humidity, CO concentration, and hour of day. Variable performance was observed across seasons with the sensor performing well (R 2 = 0.65; percent bias 3.12%; RMSE 0.10 ppm) in the winter validation period and less accurately (R 2 = 0.12; percent bias 3.01%; RMSE 0.08 ppm) in the summer validation period where there was less dynamic range in CH4 concentrations. The results highlight the utility of sensor deployment in more variable ambient CH4 conditions and demonstrate the importance of accounting for temperature and humidity dependencies as well as co-located CO concentrations with low-cost CH4 measurements. We show this can be addressed via Multiple Linear Regression (MLR) models accounting for key covariates to enable urban measurements in areas with CH4 enhancement. Together with individualized calibration prior to deployment, the sensor shows promise for use in low-cost sensor networks and represents a valuable supplement to existing monitoring strategies to identify CH4 hotspots.

Association of Occupational Exposures and Chronic Obstructive Pulmonary Disease Morbidity.

OBJECTIVE: The aim of the study is to determine whether aggregate measures of occupational exposures are associated with chronic obstructive pulmonary disease (COPD) outcomes in the Subpopulations and Intermediate Outcome Measures in COPD study cohort. METHODS: Individuals were assigned to six predetermined exposure hazard categories based on self-reported employment history. Multivariable regression, adjusted for age, sex, race, current smoking status, and smoking pack-years determined the association of such exposures to odds of COPD and morbidity measures. We compared these with the results of a single summary question regarding occupational exposure. RESULTS: A total of 2772 individuals were included. Some exposure estimates, including "gases and vapors" and "dust and fumes" exposures resulted in associations with effect estimates over two times the estimated effect size when compared with a single summary question. CONCLUSIONS: Use of occupational hazard categories can identify important associations with COPD morbidity while use of single-point measures may underestimate important differences in health risks.

A cross sectional pilot study to assess the role of phthalates on respiratory morbidity among patients with chronic obstructive pulmonary disease.

BACKGROUND: Chronic Obstructive Pulmonary Disease (COPD) affects ∼16 million U.S. adults. Phthalates, synthetic chemicals in consumer products, may adversely impact pulmonary function and airway inflammation; however, their role on COPD morbidity remains unknown. OBJECTIVE: We examined associations between phthalate exposures and respiratory morbidity among 40 COPD patients who were former smokers. METHODS: We quantified 11 phthalate biomarkers in urine samples collected at baseline in a 9-month prospective cohort study in Baltimore, Maryland. COPD baseline morbidity measures included: health status and quality of life measures (CAT: COPD Assessment Test, CCQ: Clinical COPD Questionnaire, SGRQ: St. George's Respiratory Questionnaire; mMRC: Modified Medical Research Council Dyspnea Scale), and lung function. Information on prospective exacerbation data was monitored monthly during the 9-month longitudinal follow-up period. To examine associations between morbidity measures and phthalate exposures, we used multivariable linear and Poisson regression models for continuous and count outcomes, respectively, adjusting for age, sex, race/ethnicity, education, and smoking pack-years. RESULTS: Higher mono-n-butyl phthalate (MBP) concentrations were associated with increased CAT(ß, 2.41; 95%CI, 0.31-4.51), mMRC (ß, 0.33; 95%CI 0.11-0.55), and SGRQ (ß, 7.43; 95%CI 2.70-12.2) scores at baseline. Monobenzyl phthalate (MBzP) was also positively associated with CCQ and SGRQ scores at baseline. Higher concentrations of the molar sum of Di (2-ethylhexyl) phthalate (DEHP) were associated with increased incidence of exacerbations during the follow-up period (incidence rate ratio, IRR = 1.73; 95%CI 1.11, 2.70 and IRR = 1.94; 95%CI 1.22, 3.07, for moderate and severe exacerbations, respectively). MEP concentrations were inversely associated with incidence of exacerbations during the follow-up period. CONCLUSIONS: We found that exposure to select phthalates was associated with respiratory morbidity among COPD patients. Findings warrant further examination in larger studies given widespread phthalate exposures and potential implications for COPD patients should relationships observed be causal.

Effects of a household air pollution intervention using liquefied petroleum gas stoves, continuous fuel distribution and behavioural messaging on dietary and sodium intake of adult women in Puno, Peru: a randomised controlled trial.

OBJECTIVE: Household air pollution (HAP) is a widespread environmental exposure worldwide. While several cleaner fuel interventions have been implemented to reduce personal exposures to HAP, it is unclear if cooking with cleaner fuels also affects the choice of meals and dietary intake. DESIGN: Individually randomised, open-label controlled trial of a HAP intervention. We aimed to determine the effect of a HAP intervention on dietary and Na intake. Intervention participants received a liquefied petroleum gas (LPG) stove, continuous fuel delivery and behavioural messaging during 1 year whereas control participants continued with usual cooking practices that involved the use of biomass-burning stoves. Dietary outcomes included energy, energy-adjusted macronutrients and Na intake at baseline, 6 months and 12 months post-randomisation using 24-h dietary recalls and 24-h urine. We used t-tests to estimate differences between arms in the post-randomisation period. SETTING: Rural settings in Puno, Peru. PARTICIPANTS: One hundred women aged 25-64 years. RESULTS: At baseline, control and intervention participants were similar in age (47·4 v. 49·5 years) and had similar daily energy (8894·3 kJ v. 8295·5 kJ), carbohydrate (370·8 g v. 373·3 g) and Na intake (4·9 g v. 4·8 g). One year after randomisation, we did not find differences in average energy intake (9292·4 kJ v. 8788·3 kJ; P = 0·22) or Na intake (4·5 g v. 4·6 g; P = 0·79) between control and intervention participants. CONCLUSIONS: Our HAP intervention consisting of an LPG stove, continuous fuel distribution and behavioural messaging did not affect dietary and Na intake in rural Peru.

Sustained use of liquefied petroleum gas following one year of free fuel and behavioral support in Puno, Peru.

Background: Existing efforts to promote cleaner fuels have not achieved exclusive use. We investigated whether receiving 12 months of free liquefied petroleum gas (LPG) and behavioral support could motivate continued purchase and use. Methods: The Cardiopulmonary outcomes and Household Air Pollution (CHAP) trial enrolled 180 women. Half were randomly assigned to an intervention group, which included free LPG delivered in months 1-12 followed by a post-intervention period in which they no longer received free fuel (months 13-24). For the purposes of comparison, we also include months 1-12 of data from control participants. We tracked stove use with temperature monitors, surveys, and observations, and conducted in-depth interviews with 19 participants from the intervention group at the end of their post-intervention period. Results: Participants from the intervention group used their LPG stove for 85.4 % of monitored days and 63.2 % of cooking minutes during the post-intervention months (13-24) when they were not receiving free fuel from the trial. They used a traditional stove (fogón) on 45.1 % of days post-intervention, which is significantly lower than fogón use by control participants during the intervention period (72.2 % of days). In months 13-24 post-intervention, participants from the intervention group purchased on average 12.3 kg and spent 34.1 soles (10.3 USD) per month on LPG. Continued LPG use was higher among participants who said they could afford two tanks of LPG per month, did not cook for animals, and removed their traditional stove. Women described that becoming accustomed to LPG, support and training from the project, consistent LPG supply, choice between LPG providers, and access to delivery services facilitated sustained LPG use. However, high cost was a major barrier to exclusive use. Conclusion: A 12-month period of intensive LPG support achieved a high level of sustained LPG use post-intervention, but other strategies are needed to sustain exclusive use.

Principal stratification analysis to determine health benefit of indoor air pollution reduction in a randomized environmental intervention in COPD: Results from the CLEAN AIR study.

BACKGROUND: Indoor air quality represents a modifiable exposure to Chronic Obstructive Pulmonary Disease (COPD) health. In a randomized controlled trial (CLEAN AIR study), air cleaner assignment had causal effect in improving COPD outcomes. It is unclear, however, what is the treatment effect among those for whom intervention reduced air pollution and whether it was reduction in fine particulate matter (PM2.5) or nitrogen dioxide (NO2) that contributed to such improvement. Because pollution is a posttreatment variable, treatment effect cannot be assessed while controlling for pollution using intention-to-treat (ITT) analysis. OBJECTIVE: Using principal stratification method, we assess indoor pollutants as the intermediate variable, and determine the causal effect of reducing indoor air pollution on COPD health. METHOD: In randomized controlled trial, former smokers with COPD received either active or placebo HEPA air cleaners and were followed for 6 months. Saint George's Respiratory Questionnaire (SGRQ) was the primary outcome and secondary measures included SGRQ subscales, COPD assessment test (CAT), dyspnea (mMRC), and breathlessness, cough, and sputum scale (BCSS). Indoor PM2.5 and NO2 were measured. Principal stratification analysis was performed to assess the treatment effect while controlling for pollution reduction. RESULTS: Among those showing at least 40 % PM2.5 reduction through air cleaners, the intervention showed improvement in respiratory symptoms for the active (vs. placebo), and the size of treatment effect shown for this subgroup was larger than that for the overall sample. In this subgroup, those with active air cleaners (vs. placebo) showed 7.7 points better SGRQ (95%CI: -14.3, -1.1), better CAT (ß = -5.5; 95%CI: -9.8, -1.2), mMRC (ß = -0.6; 95%CI: -1.1, -0.1), and BCSS (ß = -1.8; 95%CI: -3.0, -0.5). Among those showing at least 40 % NO2 reduction through air cleaners, there was no intervention difference in outcomes. CONCLUSION: Air cleaners caused clinically significant improvement in respiratory health for individuals with COPD through reduction in indoor PM2.5. TRIAL REGISTRATION: ClinicalTrials.gov: NCT02236858.

Ambient Air Pollution Exposure and Sleep Quality in COPD.

Rationale: Ambient air pollution exposure is associated with respiratory morbidity among individuals with chronic obstructive pulmonary disease (COPD), particularly among those with concomitant obesity. Although people with COPD report high incidence of poor sleep quality, no studies have evaluated the association between air pollution exposure, obesity, and sleep disturbances in COPD. Methods: We analyzed data collected from current and former smokers with COPD enrolled in the Subpopulations and Intermediate Outcome Measures in COPD -Air Pollution ancillary study (SPIROMICS AIR). Socio-demographics and anthropometric measurements were collected, and 1-year mean historical ambient particulate matter (PM2.5) and ozone concentrations at participants' residences were estimated by cohort-specific spatiotemporal modeling. Sleep quality was assessed with the Pittsburgh Sleep Quality Index (PSQI), and regression models were constructed to determine the association of 1-year PM2.5 (1Yr-PM2.5) and 1-year ozone (1Yr-ozone) with the PSQI score, and whether obesity modified the association. Results: In 1308 participants (age: 65.8±7.8 years, 42% women), results of regression analyses suggest that each 10µg/m3 increase in 1Yr-PM2.5 was associated with a 2.1-point increase in PSQI (P=0.03). Obesity modified the association between 1Yr-PM2.5 and PSQI (P=0.03). In obese and overweight participants, a 10µg/m3 increase in 1Yr-PM2.5 was associated with a higher PSQI (4.0 points, P<0.01, and 3.4 points, P<0.01, respectively); but no association in lean-normal weight participants (P=0.51). There was no association between 1 Yr-ozone and PSQI. Conclusions: Overweight and obese individuals with COPD appear to be susceptible to the effects of ambient PM2.5 on sleep quality. In COPD, weight and ambient PM2.5 may be modifiable risk factors to improve sleep quality.

SARS-CoV-2 Antibody Prevalence among Industrial Livestock Operation Workers and Nearby Community Residents, North Carolina, 2021 to 2022.

Industrial livestock operations (ILOs), particularly processing facilities, emerged as centers of coronavirus disease 2019 (COVID-19) outbreaks in spring 2020. Confirmed cases of COVID-19 underestimate true prevalence. To investigate the prevalence of antibodies against SARS-CoV-2, we enrolled 279 participants in North Carolina from February 2021 to July 2022: 90 from households with at least one ILO worker (ILO), 97 from high-ILO intensity areas (ILO neighbors [ILON]), and 92 from metropolitan areas (metro). More metro (55.4%) compared to ILO (51.6%) and ILON participants (48.4%) completed the COVID-19 primary vaccination series; the median completion date was more than 4 months later for ILO compared to ILON and metro participants, although neither difference was statistically significant. Participants provided a saliva swab we analyzed for SARS-CoV-2 IgG using a multiplex immunoassay. The prevalence of infection-induced IgG (positive for nucleocapsid and receptor binding domain) was higher among ILO (63%) than ILON (42.9%) and metro (48.7%) participants (prevalence ratio [PR], 1.38; 95% confidence interval [CI], 1.06 to 1.80; reference category ILON and metro combined). The prevalence of infection-induced IgG was also higher among ILO participants than among an Atlanta health care worker cohort (PR, 2.45; 95% CI, 1.80 to 3.33) and a general population cohort in North Carolina (PRs, 6.37 to 10.67). The infection-induced IgG prevalence increased over the study period. Participants reporting not masking in public in the past 2 weeks had higher infection-induced IgG prevalence (78.6%) than participants reporting masking (49.3%) (PR, 1.59; 95% CI, 1.19 to 2.13). Lower education, more people per bedroom, Hispanic/Latino ethnicity, and more contact with people outside the home were also associated with higher infection-induced IgG prevalence. IMPORTANCE Few studies have measured COVID-19 seroprevalence in North Carolina, especially among rural, Black, and Hispanic/Latino communities that have been heavily affected. Antibody results show high rates of COVID-19 among industrial livestock operation workers and their household members. Antibody results add to evidence of health disparities related to COVID-19 by socioeconomic status and ethnicity. Associations between masking and physical distancing with antibody results also add to evidence of the effectiveness of these prevention strategies. Delays in the timing of receipt of COVID-19 vaccination reinforce the importance of dismantling vaccination barriers, especially for industrial livestock operation workers and their household members.

Infinite hidden Markov models for multiple multivariate time series with missing data.

Exposure to air pollution is associated with increased morbidity and mortality. Recent technological advancements permit the collection of time-resolved personal exposure data. Such data are often incomplete with missing observations and exposures below the limit of detection, which limit their use in health effects studies. In this paper, we develop an infinite hidden Markov model for multiple asynchronous multivariate time series with missing data. Our model is designed to include covariates that can inform transitions among hidden states. We implement beam sampling, a combination of slice sampling and dynamic programming, to sample the hidden states, and a Bayesian multiple imputation algorithm to impute missing data. In simulation studies, our model excels in estimating hidden states and state-specific means and imputing observations that are missing at random or below the limit of detection. We validate our imputation approach on data from the Fort Collins Commuter Study. We show that the estimated hidden states improve imputations for data that are missing at random compared to existing approaches. In a case study of the Fort Collins Commuter Study, we describe the inferential gains obtained from our model including improved imputation of missing data and the ability to identify shared patterns in activity and exposure among repeated sampling days for individuals and among distinct individuals.

Indoor Air Pollution and Impaired Cardiac Autonomic Function in Chronic Obstructive Pulmonary Disease.

Rationale: Indoor air pollution represents a modifiable risk factor for respiratory morbidity in chronic obstructive pulmonary disease (COPD). The effects of indoor air pollution, as well as the impact of interventions to improve indoor air quality, on cardiovascular morbidity in COPD remain unknown. Objectives: To determine the association between indoor particulate matter (PM) and heart rate variability (HRV), a measure of cardiac autonomic function tied to cardiovascular morbidity and mortality, as well as the impact of household air purifiers on HRV. Methods: Former smokers with moderate-severe COPD were recruited from a 6-month randomized controlled trial of a portable air cleaner intervention to undergo paired assessment of both in-home PM and HRV using 24-hour Holter monitoring at up to five time points. Primary outcomes were HRV measures tied to cardiovascular morbidity (standard deviation of normal-to-normal intervals [SDNN] and root mean square of successive differences between normal-to-normal intervals [RMSSD]). Measurements and Results: Eighty-five participants contributed 317 HRV measurements. A twofold increase in household PM ⩽2.5 µm in aerodynamic diameter was associated with decreases in SDNN (ß, -2.98% [95% confidence interval (CI), -5.12 to -0.78]) and RMSSD (ß, -4.57% [95% CI, -10.1 to -1.60]). The greatest effects were observed with ultrafine particles (<100 nm) (RMSSD; ß, -16.4% [95% CI, -22.3 to -10.1]) and among obese participants. Participants randomized to the active air cleaner saw improvements in RMSSD (ß, 25.2% [95% CI, 2.99 to 52.1]), but not SDNN (ß, 2.65% [95% CI, -10.8 to 18.1]), compared with the placebo group. Conclusions: This is the first U.S. study to describe the association between household PM and cardiac autonomic function among individuals with COPD, as well as the potential cardiovascular health benefits of household air cleaners.

Association of School Infrastructure on Health and Achievement Among Children With Asthma.

OBJECTIVE: To determine whether school infrastructure is associated with health and academic outcomes among elementary school children with asthma. METHODS: We conducted a retrospective cohort study of linked medical, academic, and facilities data from a large mid-Atlantic school district of the United States. All K-5 students with asthma who were enrolled under the state's Children's Health Insurance Program were included. We estimated associations of the infrastructure quality of the student's school, as assessed by an engineering firm in Summer 2011 and represented by the Facility Condition Index (FCI), with asthma health outcomes, absenteeism, and standardized test scores in math and reading in the 2 academic years thereafter. RESULTS: A total of 6558 students were identified, the majority non-Hispanic Black, across 130 schools. Most schools (97/130, 75%) were in very poor or worse condition. In cluster-adjusted models accounting for demographics, grade, school-specific area deprivation, and inhaled corticosteroid use, a one standard deviation increase in FCI, corresponding to greater infrastructure deficiency, was associated with higher rates of asthma-related hospitalizations (incidence rate ratio [IRR] 1.16; 95% confidence interval [CI] 1.03, 1.32), more absenteeism (IRR 1.05; 95% CI 1.01, 1.08), and lower scores in math (mean difference [MD] -3.3; 95% CI -5.5, -1.2) and reading (MD -3.0; 95% CI -5.1, -0.9). There were no differences in rates of asthma-related emergency visits or steroid prescriptions. CONCLUSIONS: Children with asthma attending schools with poorer infrastructure had worse health and academic outcomes. Public policy emphasizing reinvestment in school infrastructure may be a potential means of addressing asthma disparities.

A DYNAMIC SPATIAL FILTERING APPROACH TO MITIGATE UNDERESTIMATION BIAS IN FIELD CALIBRATED LOW-COST SENSOR AIR POLLUTION DATA.

Low-cost air pollution sensors, offering hyper-local characterization of pollutant concentrations, are becoming increasingly prevalent in environmental and public health research. However, low-cost air pollution data can be noisy, biased by environmental conditions, and usually need to be field-calibrated by collocating low-cost sensors with reference-grade instruments. We show, theoretically and empirically, that the common procedure of regression-based calibration using collocated data systematically underestimates high air pollution concentrations, which are critical to diagnose from a health perspective. Current calibration practices also often fail to utilize the spatial correlation in pollutant concentrations. We propose a novel spatial filtering approach to collocation-based calibration of low-cost networks that mitigates the underestimation issue by using an inverse regression. The inverse-regression also allows for incorporating spatial correlations by a second-stage model for the true pollutant concentrations using a conditional Gaussian Process. Our approach works with one or more collocated sites in the network and is dynamic, leveraging spatial correlation with the latest available reference data. Through extensive simulations, we demonstrate how the spatial filtering substantially improves estimation of pollutant concentrations, and measures peak concentrations with greater accuracy. We apply the methodology for calibration of a low-cost PM2.5 network in Baltimore, Maryland, and diagnose air pollution peaks that are missed by the regression-calibration.

Black carbon content in airway macrophages is associated with increased severe exacerbations and worse COPD morbidity in SPIROMICS.

BACKGROUND: Airway macrophages (AM), crucial for the immune response in chronic obstructive pulmonary disease (COPD), are exposed to environmental particulate matter (PM), which they retain in their cytoplasm as black carbon (BC). However, whether AM BC accurately reflects environmental PM2.5 exposure, and can serve as a biomarker of COPD outcomes, is unknown. METHODS: We analyzed induced sputum from participants at 7 of 12 sites SPIROMICS sites for AM BC content, which we related to exposures and to lung function and respiratory outcomes. Models were adjusted for batch (first vs. second), age, race (white vs. non-white), income (<$35,000, $35,000~$74,999, ≥$75,000, decline to answer), BMI, and use of long-acting beta-agonist/long-acting muscarinic antagonists, with sensitivity analysis performed with inclusion of urinary cotinine and lung function as covariates. RESULTS: Of 324 participants, 143 were current smokers and 201 had spirometric-confirmed COPD. Modeled indoor fine (< 2.5 µm in aerodynamic diameter) particulate matter (PM2.5) and urinary cotinine were associated with higher AM BC. Other assessed indoor and ambient pollutant exposures were not associated with higher AM BC. Higher AM BC was associated with worse lung function and odds of severe exacerbation, as well as worse functional status, respiratory symptoms and quality of life. CONCLUSION: Indoor PM2.5 and cigarette smoke exposure may lead to increased AM BC deposition. Black carbon content in AMs is associated with worse COPD morbidity in current and former smokers, which remained after sensitivity analysis adjusting for cigarette smoke burden. Airway macrophage BC, which may alter macrophage function, could serve as a predictor of experiencing worse respiratory symptoms and impaired lung function.

Non-linear probabilistic calibration of low-cost environmental air pollution sensor networks for neighborhood level spatiotemporal exposure assessment.

BACKGROUND: Low-cost sensor networks for monitoring air pollution are an effective tool for expanding spatial resolution beyond the capabilities of existing state and federal reference monitoring stations. However, low-cost sensor data commonly exhibit non-linear biases with respect to environmental conditions that cannot be captured by linear models, therefore requiring extensive lab calibration. Further, these calibration models traditionally produce point estimates or uniform variance predictions which limits their downstream in exposure assessment. OBJECTIVE: Build direct field-calibration models using probabilistic gradient boosted decision trees (GBDT) that eliminate the need for resource-intensive lab calibration and that can be used to conduct probabilistic exposure assessments on the neighborhood level. METHODS: Using data from Plantower A003 particulate matter (PM) sensors deployed in Baltimore, MD from November 2018 through November 2019, a fully probabilistic NGBoost GBDT was trained on raw data from sensors co-located with a federal reference monitoring station and compared against linear regression trained on lab calibrated sensor data. The NGBoost predictions were then used in a Monte Carlo interpolation process to generate high spatial resolution probabilistic exposure gradients across Baltimore. RESULTS: We demonstrate that direct field-calibration of the raw PM2.5 sensor data using a probabilistic GBDT has improved point and distribution accuracies compared to the linear model, particularly at reference measurements exceeding 25 µg/m3, and also on monitors not included in the training set. SIGNIFICANCE: We provide a framework for utilizing the GBDT to conduct probabilistic spatial assessments of human exposure with inverse distance weighting that predicts the probability of a given location exceeding an exposure threshold and provides percentiles of exposure. These probabilistic spatial exposure assessments can be scaled by time and space with minimal modifications. Here, we used the probabilistic exposure assessment methodology to create high quality spatial-temporal PM2.5 maps on the neighborhood-scale in Baltimore, MD. IMPACT STATEMENT: We demonstrate how the use of open-source probabilistic machine learning models for in-place sensor calibration outperforms traditional linear models and does not require an initial laboratory calibration step. Further, these probabilistic models can create uniquely probabilistic spatial exposure assessments following a Monte Carlo interpolation process.