Household Air Pollution Interventions to Improve Health in Low- and Middle-Income Countries: An Official American Thoracic Society Research Statement.

Background: An estimated 3 billion people, largely in low- and middle-income countries, rely on unclean fuels for cooking, heating, and lighting to meet household energy needs. The resulting exposure to household air pollution (HAP) is a leading cause of pneumonia, chronic lung disease, and other adverse health effects. In the last decade, randomized controlled trials of clean cooking interventions to reduce HAP have been conducted. We aim to provide guidance on how to interpret the findings of these trials and how they should inform policy makers and practitioners.Methods: We assembled a multidisciplinary working group of international researchers, public health practitioners, and policymakers with expertise in household air pollution from within academia, the American Thoracic Society, funders, nongovernmental organizations, and global organizations, including the World Bank and the World Health Organization. We performed a literature search, convened four sessions via web conference, and developed consensus conclusions and recommendations via the Delphi method.Results: The committee reached consensus on 14 conclusions and recommendations. Although some trials using cleaner-burning biomass stoves or cleaner-cooking fuels have reduced HAP exposure, the committee was divided (with 55% saying no and 45% saying yes) on whether the studied interventions improved measured health outcomes.Conclusions: HAP is associated with adverse health effects in observational studies. However, it remains unclear which household energy interventions reduce exposure, improve health, can be scaled, and are sustainable. Researchers should engage with policy makers and practitioners working to scale cleaner energy solutions to understand and address their information needs.

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.

AirPen: A Wearable Monitor for Characterizing Exposures to Particulate Matter and Volatile Organic Compounds.

Exposure to air pollution is a leading risk factor for disease and premature death, but technologies for assessing personal exposure to particulate and gaseous air pollutants, including the timing and location of such exposures, are limited. We developed a small, quiet, wearable monitor, called the AirPen, to quantify personal exposures to fine particulate matter (PM2.5) and volatile organic compounds (VOCs). The AirPen combines physical sample collection (PM onto a filter and VOCs onto a sorbent tube) with a suite of low-cost sensors (for PM, VOCs, temperature, pressure, humidity, light intensity, location, and motion). We validated the AirPen against conventional personal sampling equipment in the laboratory and then conducted a field study to measure at-work and away-from-work exposures to PM2.5 and VOCs among employees at an agricultural facility in Colorado, USA. The resultant sampling and sensor data indicated that personal exposures to benzene, toluene, ethylbenzene, and xylenes were dominated by a specific workplace location. These results illustrate how the AirPen can be used to advance our understanding of personal exposure to air pollution as a function of time, location, source, and activity, even in the absence of detailed activity diary data.

Large Particle Emissions from Human Vocalization and Playing of Wind Instruments.

Humans emit large salivary particles when talking, singing, and playing musical instruments, which have implications for respiratory disease transmission. Yet little work has been done to characterize the emission rates and size distributions of such particles. This work characterized large particle (dp > 35 µm in aerodynamic diameter) emissions from 70 volunteers of varying age and sex while vocalizing and playing wind instruments. Mitigation efficacies for face masks (while singing) and bell covers (while playing instruments) were also examined. Geometric mean particle count emission rates varied from 3.8 min-1 (geometric standard deviation [GSD] = 3.1) for brass instruments playing to 95.1 min-1 (GSD = 3.8) for talking. On average, talking produced the highest emission rates for large particles, in terms of both number and mass, followed by singing and then instrument playing. Neither age, sex, CO2 emissions, nor loudness (average dBA) were significant predictors of large particle emissions, contrary to previous findings for smaller particle sizes (i.e., for dp < 35 µm). Size distributions were similar between talking and singing (count median diameter = 53.0 µm, GSD = 1.69). Bell covers did not affect large particle emissions from most wind instruments, but face masks reduced large particle count emissions for singing by 92.5% (95% CI: 97.9%, 73.7%).

Application of an Ultra-Low-Cost Passive Sampler for Light-Absorbing Carbon in Mongolia.

Low-cost, long-term measures of air pollution concentrations are often needed for epidemiological studies and policy analyses of household air pollution. The Washington passive sampler (WPS), an ultra-low-cost method for measuring the long-term average levels of light-absorbing carbon (LAC) air pollution, uses digital images to measure the changes in the reflectance of a passively exposed paper filter. A prior publication on WPS reported high precision and reproducibility. Here, we deployed three methods to each of 10 households in Ulaanbaatar, Mongolia: one PurpleAir for PM2.5; two ultrasonic personal aerosol samplers (UPAS) with quartz filters for the thermal-optical analysis of elemental carbon (EC); and two WPS for LAC. We compared multiple rounds of 4-week-average measurements. The analyses calibrating the LAC to the elemental carbon measurement suggest that 1 µg of EC/m3 corresponds to 62 PI/month (R2 = 0.83). The EC-LAC calibration curve indicates an accuracy (root-mean-square error) of 3.1 µg of EC/m3, or ~21% of the average elemental carbon concentration. The RMSE values observed here for the WPS are comparable to the reported accuracy levels for other methods, including reference methods. Based on the precision and accuracy results shown here, as well as the increased simplicity of deployment, the WPS may merit further consideration for studying air quality in homes that use solid fuels.

Effects of household and participant characteristics on personal exposure and kitchen concentration of fine particulate matter and black carbon in rural Honduras.

Traditional cooking with solid fuels (biomass, animal dung, charcoals, coal) creates household air pollution that leads to millions of premature deaths and disability worldwide each year. Exposure to household air pollution is highest in low- and middle-income countries. Using data from a stepped-wedge randomized controlled trial of a cookstove intervention among 230 households in Honduras, we analyzed the impact of household and personal variables on repeated 24-h measurements of fine particulate matter (PM2.5) and black carbon (BC) exposure. Six measurements were collected approximately six-months apart over the course of the three-year study. Multivariable mixed models explained 37% of variation in personal PM2.5 exposure and 49% of variation in kitchen PM2.5 concentrations. Additionally, multivariable models explained 37% and 47% of variation in personal and kitchen BC concentrations, respectively. Stove type, season, presence of electricity, primary stove location, kitchen enclosure type, stove use time, and presence of kerosene for lighting were all associated with differences in geometric mean exposures. Stove type explained the most variability of the included variables. In future studies of household air pollution, tracking the cooking behaviors and daily activities of participants, including outdoor exposures, may explain exposure variation beyond the household and personal variables considered here.

Acute differences in blood lipids and inflammatory biomarkers following controlled exposures to cookstove air pollution in the STOVES study.

Household air pollution is a leading risk factor for morbidity and premature mortality. Numerous cookstoves have been developed to reduce household air pollution, but it is unclear whether such cookstoves meaningfully improve health. In a controlled exposure study with a crossover design, we assessed the effect of pollution emitted from multiple cookstoves on acute differences in blood lipids and inflammatory biomarkers. Participants (n = 48) were assigned to treatment sequences of exposure to air pollution emitted from five cookstoves and a filtered-air control. Blood lipids and inflammatory biomarkers were measured before and 0, 3, and 24 hours after treatments. Many of the measured outcomes had inconsistent results. However, compared to control, intercellular adhesion molecule-1 was higher 3 hours after all treatments, and C-reactive protein and serum amyloid-A were higher 24 hours after the highest treatment. Our results suggest that short-term exposure to cookstove air pollution can increase inflammatory biomarkers within 24 hours.

Airborne Transmission of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2): What We Know.

We examine airborne transmission of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) potential using a source-to-dose framework beginning with generation of virus-containing droplets and aerosols and ending with virus deposition in the respiratory tract of susceptible individuals. By addressing 4 critical questions, we identify both gaps in addressing 4 critical questions with answers having policy implications.

Quantitative Protection Factors for Common Masks and Face Coverings.

The performance of masks, whether intended to protect the community from exhaled infectious aerosol or to protect the wearer from inhaled infectious aerosol, depends on factors such as filtration efficiency, particle size distribution, leakage, and ventilation rate. These factors depend on the activities and facial features of the mask wearer so that the mask performance for real-world applications is difficult to predict. The present work shows how protection factor, a quantity often used to describe mask performance, can be estimated without involving human volunteers. By constraining these factors to known values, mask protection factors can be compared fairly and efficiently following a series of filtration efficiency measurements performed in the laboratory. Protection factors and mask emissions for exhalation and inhalation were evaluated for masks of seven types currently in use around the world and for a hypothetical mask with 99% efficiency on all particles. The performance of reusable masks made from cotton fabric was limited by the size of the native cotton fibers. Masks that utilized finer fibers, particularly electret fibers with relatively small diameters, showed excellent performance with moderate flow resistance. Results from this work, in addition to simple guidance for mask fit and usage, can facilitate risk communication and decision-making efforts during the COVID-19 pandemic.

Design and Testing of a Low-Cost Sensor and Sampling Platform for Indoor Air Quality.

Americans spend most of their time indoors at home, but comprehensive characterization of in-home air pollution is limited by the cost and size of reference-quality monitors. We assembled small "Home Health Boxes" (HHBs) to measure indoor PM2.5, PM10, CO2, CO, NO2, and O3 concentrations using filter samplers and low-cost sensors. Nine HHBs were collocated with reference monitors in the kitchen of an occupied home in Fort Collins, Colorado, USA for 168 h while wildfire smoke impacted local air quality. When HHB data were interpreted using gas sensor manufacturers' calibrations, HHBs and reference monitors (a) categorized the level of each gaseous pollutant similarly (as either low, elevated, or high relative to air quality standards) and (b) both indicated that gas cooking burners were the dominant source of CO and NO2 pollution; however, HHB and reference O3 data were not correlated. When HHB gas sensor data were interpreted using linear mixed calibration models derived via collocation with reference monitors, root-mean-square error decreased for CO2 (from 408 to 58 ppm), CO (645 to 572 ppb), NO2 (22 to 14 ppb), and O3 (21 to 7 ppb); additionally, correlation between HHB and reference O3 data improved (Pearson's r increased from 0.02 to 0.75). Mean 168-h PM2.5 and PM10 concentrations derived from nine filter samples were 19.4 µg m-3 (6.1% relative standard deviation [RSD]) and 40.1 µg m-3 (7.6% RSD). The 168-h PM2.5 concentration was overestimated by PMS5003 sensors (median sensor/filter ratio = 1.7) and underestimated slightly by SPS30 sensors (median sensor/filter ratio = 0.91).

Janus Electrochemical Paper-Based Analytical Devices for Metals Detection in Aerosol Samples.

Exposure to trace metals in airborne particulate matter (PM) has been linked to various adverse health effects. Quantifying metals in PM is important; however, current analytical tools tend to be bulky and expensive. A need therefore exists for more rapid, low-cost, portable tools for multiplexed determination of metals in PM. Electrochemical paper-based devices (ePADs) have been used for detecting metals in PM but require different devices and methods for different metals, making the systems more complicated than desired. Recently reported Janus ePADs offer a solution to this problem by allowing for multiple electrochemical experiments from a single sample. Here, we sought to determine if a Janus ePAD containing four independent channels and working electrodes could be used for simultaneous detection of multiple metals in PM. Online sample pretreatment in each channel during sample delivery yielded optimal conditions for each experiment. The design allows the device to conduct square-wave anodic stripping voltammetry (SWASV) and square-wave cathodic stripping voltammetry (SWCSV) for simultaneous detection of Cd, Pb, Cu, Fe, and Ni from a single sample. Two detection zones each with shared reference and counter electrodes were used for SWASV and SWCSV, respectively. The proposed sensors reached LODs down to 0.5, 0.5, 1.0, 0.5, and 1.0 µg L-1, for Cd(II), Pb(II), Cu(II), Fe(II), and Ni(II), respectively. The linear working ranges were 0.5-400.0 µg L-1 for Cd(II), Pb(II), and Fe(II), 1.0-400.0 µg L-1 for Cu(II), and 0.5-200.0 µg L-1 for Ni(II). The devices were applied for Cd(II), Pb(II), Cu(II), Fe(II), and Ni(II) determination in PM samples, and the results agreed with those using traditional ICP-MS analyses at 95% confidence.

Acute differences in pulse wave velocity, augmentation index, and central pulse pressure following controlled exposures to cookstove air pollution in the Subclinical Tests of Volunteers Exposed to Smoke (SToVES) study.

Household air pollution emitted from solid-fuel cookstoves used for domestic cooking is a leading risk factor for morbidity and premature mortality globally. There have been attempts to design and distribute lower emission cookstoves, yet it is unclear if they meaningfully improve health. Using a crossover design, we assessed differences in central aortic hemodynamics and arterial stiffness following controlled exposures to air pollution emitted from five different cookstove technologies compared to a filtered air control. Forty-eight young, healthy participants were assigned to six 2-h controlled treatments of pollution from five different cookstoves and a filtered air control. Each treatment had a target concentration for fine particulate matter: filtered air control = 0 µg/m3, liquefied petroleum gas = 10 µg/m3, gasifier = 35 µg/m3, fan rocket = 100 µg/m3, rocket elbow = 250 µg/m3, three stone fire = 500 µg/m3. Pulse wave velocity (PWV), central augmentation index (AIx), and central pulse pressure (CPP) were measured before and at three time points after each treatment (0, 3, and 24 h). Linear mixed models were used to assess differences in the outcomes for each cookstove treatment compared to control. PWV and CPP were marginally higher 24 h after all cookstove treatments compared to control. For example, PWV was 0.15 m/s higher (95% confidence interval: -0.02, 0.31) and CPP was 0.6 mmHg higher (95% confidence interval: -0.8, 2.1) 24 h after the three stone fire treatment compared to control. The magnitude of the differences compared to control was similar across all cookstove treatments. PWV and CPP had no consistent trends at the other post-treatment time points (0 and 3 h). No consistent trends were observed for AIx at any post-treatment time point. Our findings suggest higher levels of PWV and CPP within 24 h after 2-h controlled treatments of pollution from five different cookstove technologies. The similar magnitude of the differences following each cookstove treatment compared to control may indicate that acute exposures from even the cleanest cookstove technologies can adversely impact these subclinical markers of cardiovascular health, although differences were small and may not be clinically meaningful.

C-reactive protein from dried blood spots: Application to household air pollution field studies.

Household air pollution (HAP) is estimated to be an important risk factor for cardiovascular disease, but little clinical evidence exists and collecting biomarkers of disease risk is difficult in low-resource settings. Among 54 Nicaraguan women with woodburning cookstoves, we evaluated cross-sectional associations between 48-hour measures of HAP (eg, fine particulate matter, PM2.5 ) and C-reactive protein (CRP) via dried blood spots; secondary analyses included seven additional biomarkers of systemic injury and inflammation. We conducted sub-studies to calculate the intraclass correlation coefficient (ICC) in biomarkers collected over four consecutive days in Nicaragua and to assess the validity of measuring biomarkers in dried blood by calculating the correlation with paired venous-drawn samples in Colorado. Measures of HAP were associated with CRP (eg, a 25% increase in indoor PM2.5 was associated with a 7.4% increase in CRP [95% confidence interval: 0.7, 14.5]). Most of the variability in CRP concentrations over the 4-day period was between-person (ICC: 0.88), and CRP concentrations were highly correlated between paired dried blood and venous-drawn serum (Spearman ρ = .96). Results for secondary biomarkers were primarily consistent with null associations, and the sub-study ICCs and correlations were lower. Assessing CRP via dried blood spots provides a feasible approach to elucidate the association between HAP and cardiovascular disease risk.

Comparing regional stove-usage patterns and using those patterns to model indoor air quality impacts.

Monitoring improved cookstove adoption and usage in developing countries can help anticipate potential health and environmental benefits that may result from household energy interventions. This study explores stove-usage monitor (SUM)-derived usage data from field studies in China (52 stoves, 1422 monitoring days), Honduras (270 stoves, 630 monitoring days), India (19 stoves, 565 monitoring days), and Uganda (38 stoves, 1007 monitoring days). Traditional stove usage was found to be generally similar among four seemingly disparate countries in terms of cooking habits, with average usage of between 171 and 257 minutes per day for the most-used stoves. In Honduras, where survey-based usage data were also collected, there was only modest agreement between sensor data and self-reported user data. For Indian homes, we combined stove-usage data with a single-zone Monte Carlo box model to estimate kitchen-level PM2.5 and CO concentrations under various scenarios of cleaner cookstove adoption. We defined clean cookstove performance based on the International Standards Organization (ISO) voluntary guidelines. Model results showed that even with 75% displacement of traditional stoves with the cleanest available stove (ISO tier-5), World Health Organization 24 hours PM2.5 standards were exceeded in 96.4% of model runs, underscoring the importance of full displacement.

Acute changes in lung function following controlled exposure to cookstove air pollution in the subclinical tests of volunteers exposed to smoke (STOVES) study.

Background: Exposure to household air pollution generated as a result of cooking and heating is a leading contributor to global disease. The effects of cookstove-generated air pollution on adult lung function, however, remain uncertain.Objectives: We investigated acute responses in lung function following controlled exposures to cookstove-generated air pollution.Methods: We recruited 48 healthy adult volunteers to undergo six two-hour treatments: a filtered-air control and emissions from five different stoves with fine particulate matter (PM2.5) targets from 10 to 500 µg/m3. Spirometry was conducted prior to exposure and immediately, and three and 24 h post-exposure. Mixed-effect models were used to estimate differences in post-exposure lung function for stove treatments versus control.Results: Immediately post-exposure, lung function was lower compared to the control for the three highest PM2.5-level stoves. The largest differences were for the fan rocket stove (target 250 µg/m3; forced vital capacity (FVC): -60 mL, 95% confidence interval (95% CI) -135, 15; forced expiratory volume (FEV1): -51 mL, 95% CI -117, 16; mid-expiratory flow (FEF25-75): -116 mL/s, 95% CI -239, 8). At 3 h post-exposure, lung function was lower compared to the control for all stove treatments; effects were of similar magnitude for all stoves. At 24 h post-exposure, results were consistent with a null association for FVC and FEV1; FEF25-75 was lower relative to the control for the gasifier, fan rocket, and three stone fire.Conclusions: Patterns suggesting short-term decreases in lung function follow from exposure to cookstove air pollution even for stove exposures with low PM2.5 levels.

Bayesian Nonparametric Monotone Regression.

In many applications there is interest in estimating the relation between a predictor and an outcome when the relation is known to be monotone or otherwise constrained due to the physical processes involved. We consider one such application-inferring time-resolved aerosol concentration from a low-cost differential pressure sensor. The objective is to estimate a monotone function and make inference on the scaled first derivative of the function. We proposed Bayesian nonparametric monotone regression which uses a Bernstein polynomial basis to construct the regression function and puts a Dirichlet process prior on the regression coefficients. The base measure of the Dirichlet process is a finite mixture of a mass point at zero and a truncated normal. This construction imposes monotonicity while clustering the basis functions. Clustering the basis functions reduces the parameter space and allows the estimated regression function to be linear. With the proposed approach we can make closed-formed inference on the derivative of the estimated function including full quantification of uncertainty. In a simulation study the proposed method performs similar to other monotone regression approaches when the true function is wavy but performs better when the true function is linear. We apply the method to estimate time-resolved aerosol concentration with a newly-developed portable aerosol monitor. The R package bnmr is made available to implement the method.

Exposure to household air pollution from biomass cookstoves and self-reported symptoms among women in rural Honduras.

Household air pollution from combustion of solid fuels is an important risk factor for morbidity and mortality, causing an estimated 2.6 million premature deaths globally in 2016. Self-reported health symptoms are a meaningful measure of quality of life, however, few studies have evaluated symptoms and quantitative measures of exposure to household air pollution. We assessed the cross-sectional association of self-reported symptoms and exposures to household air pollution among women in rural Honduras using stove type (traditional [n = 76]; cleaner-burning Justa [n = 74]) and 24-hour average personal and kitchen fine particulate matter (PM2.5) concentrations. The odds of prevalent symptoms were higher among women using traditional stoves vs Justa stoves (e.g. headache: odds ratio = 2.23; 95% confidence interval = 1.13-4.39). Associations between symptoms and measured PM2.5 were generally consistent with the null. These results add to the evidence suggesting reduced exposures and better health-related quality of life among women using cleaner-burning biomass stoves.

Design and performance of UPAS inlets for respirable and thoracic mass sampling.

The Ultrasonic Personal Aerosol Sampler (UPAS) is a small, lightweight, and quiet sampler that collects airborne particulate matter on a filter for gravimetric or compositional analysis. The objective of this work was to develop UPAS inlets with collection efficiencies that match criteria for respirable or thoracic mass sampling. The two-stage inlet for respirable mass described here utilizes an impaction stage and a cyclone, whereas the one-stage inlet for thoracic mass sampling utilizes a circular slot impactor. Inlet designs are based on particle collection theory used in conjunction with an optimization algorithm to predict initial inlet dimensions; these predictions were the starting points for experiments that finalized dimensions and operating conditions. Both the respirable mass inlet and the thoracic mass inlet described here are interchangeable with the UPAS, and both have efficiencies that match well with their respective standards. With either inlet, the collected sample should be within ±5% of what the standard specifies for aerosols with reasonably broad size distributions.

Variability of aerosol mass and number concentrations during taconite mining operations.

This study characterized concentration metrics of airborne nanoparticles and their time series across major operations of a taconite mine through monitoring respirable and ultrafine particle concentrations at four major processing departments of the mine: crushing, dry milling, wet milling, and pelletizing (United Taconite Mine, Iron Junction, MN, USA). We used three area stations of direct-reading instruments to estimate concentration metrics including PM1 (particles with an aerodynamic diameter <1 µm), respirable dust (particles sampled according to the respirable convention with a 50% sampling efficiency at an aerodynamic diameter of 4 µm), PN (total number concentration of particles), and lung-deposited surface area concentrations (LDSA) of particles smaller than 300 nm, on two different days. Results for each station were compared using bivariate correlation analysis to obtain insight into the spatial distribution, and intra-class correlation coefficients (ICCs) to evaluate the between-day repeatability between the measurements. Comparability of the LDSA concentrations measured by two different devices was also investigated using linear regression. Results revealed that the pelletizing operation produced the highest average LDSA concentration on both days (with a maximum concentration of 199 ± 48 µm2/cm3 in pelletizing, 141 ± 52 µm2/cm3 in crushing, 91 ± 9 µm2/cm3 in dry milling, and 85 ± 7 µm2/cm3 in wet milling). Concentrations in all operations showed a fair to excellent between-day repeatability but they were significantly different within stations of each operation. Measured LDSA concentrations did not show a linear correlation between different instruments, except for crushing.

Effects of Fuel Moisture Content on Emissions from a Rocket-Elbow Cookstove.

Exposure to air pollution from solid-fuel cookstoves is a leading risk factor for premature death; however, the effect of fuel moisture content on air pollutant emissions from solid-fuel cookstoves remains poorly constrained. The objective of this work was to characterize emissions from a rocket-elbow cookstove burning wood at three different moisture levels (5%, 15%, and 25% on a dry mass basis). Emissions of carbon dioxide (CO2), carbon monoxide (CO), methane, fine particulate matter (PM2.5), PM2.5 elemental carbon (EC), PM2.5 organic carbon, formaldehyde, acetaldehyde, benzene, toluene, ethylbenzene, and xylenes were measured. Emission factors (EFs; g·MJdelivered-1) for all pollutants, except CO2 and EC, increased with increasing fuel moisture content: CO EFs increased by 84%, PM2.5 EFs increased by 149%, formaldehyde EFs increased by 216%, and benzene EFs increased by 82%. Both modified combustion efficiency and the temperature at the combustion chamber exit decreased with increasing fuel moisture, suggesting that the energy required to vaporize water in the fuel led to lower temperatures in the combustion chamber and lower gas-phase oxidation rates. These results illustrate that changes in fuel equilibrium moisture content could cause EFs for pollutants such as PM2.5 to vary by a factor of 2 or more across different geographic regions.