Concentrated animal feeding operations contribute to local, regional, and global air pollution.1 Pollutants of local importance include odor,2,3 hydrogen sulfide (H2S),4 endotoxin,5 particulate matter (PM),6,7 and ammonia (NH3).8,9
Several cross-sectional studies have examined the health of people living near concentrated hog operations on the basis of residential proximity to classify exposure. In a population-based survey, neighbors of hog operations reported more episodes of headache, runny nose, sore throat, coughing, diarrhea, and burning eyes compared with demographically similar persons who did not live near a hog operation.10 Among children, indicators of asthma have been related to measures of residential11 and school12,13 exposures to pollution from hog operations. In an area of Germany with a high density of concentrated animal feeding operations, reported odor annoyance was associated with prevalence of wheeze without a cold, and physician-diagnosed asthma and allergic rhinitis. Additionally, the number of operations within 500 meters of participants' homes was associated with increased odds of wheezing without a cold, and with diminished lung function.14 These symptoms overlap with conditions reported in studies of occupational exposures of animal-confinement-house workers, including decreased lung function,15–17 chronic cough,17 excess phlegm production, chest tightness,18 scratchy throat, eyes and mucous membrane irritation, shortness of breath,16 and wheezing.18
Community Health Effects of Industrial Hog Operations was a longitudinal, community-driven, participatory study of air pollution, health, and quality of life among persons living near hog operations. We have previously described associations between air pollution and hog odor,19 air pollution and measures of stress and negative mood,20 and factors associated with data quality and completeness.21 Here we report relationships between measures of air pollution, symptoms, and lung function, focusing on physical symptoms that have been of interest in cross-sectional studies.22
Between September 2003 and September 2005, residents of 16 eastern North Carolina communities collected health data for approximately 2 weeks while pollutant concentrations were monitored continuously. Communities participated sequentially using the same set of air-monitoring devices.
Nonsmoking volunteers aged at least 18 years residing within 1.5 miles of at least one hog operation were recruited through community-based organizations. The lead community organization for this study was the Concerned Citizens of Tillery.23 Participants attended a 3-hour training session at which they gave informed consent and practiced completing all data-collection activities. The study design has been described in detail elsewhere.23
The Institutional Review Board of the University of North Carolina at Chapel Hill reviewed and approved study activities annually.
Participants spent 10 minutes outdoors at preselected morning and evening times approximately 12 hours apart. While outside, they rated, on a scale of 0 (none) to 8 (very strong), the strength of the hog odor they recalled having smelled during each of the 12 preceding hours. Participants then returned indoors and rated hog odor present during the 10 minutes outside on the same 9-point scale.
We analyzed 2 hog odor variables. Twelve-hour mean odor is the average of the hourly odor levels reported for each of the 12 hours before the morning or evening data collection time. Twice-daily odor is the odor during the 10 minutes outdoors.
Continuous air pollution monitors, mounted on a trailer that was centrally located in each community, recorded concentrations of hydrogen sulfide (H2S), semi-volatile particulate matter less than 10 micrometers in diameter (semivolatile PM10), particulate matter less than 10 micrometers in diameter that excluded the volatile fraction (PM10), coarse PM (PM2.5–10), fine PM (PM2.5), and endotoxin. An MDA Scientific Single Point Monitor (Honeywell Analytics Inc North America, Lincolnshire, IL) recorded H2S concentrations every 15 minutes in parts per billion (ppb). Hourly concentrations of PM10 and semivolatile PM10 were measured in micrograms per cubic meter (μg/m3), using a Tapered Element Oscillating Microbalance Series 1400a Ambient Particulate Monitor with a Series 8500 Filter Dynamics Measurement System (Thermo Fisher Scientific, Waltham, MA). In the first 12 of 16 communities, a Dichotomous Partisol-Plus 2025-D Sequential Air Sampler (Thermo Fisher Scientific, Waltham, MA) was used to collect 12-hour samples of PM2.5–10 and PM2.5 (μg/m3) on filters that were assayed for endotoxin in endotoxin units per milligram (EU/mg). Endotoxin levels from PM2.5–10 filters were quantified by kinetic chromogenic Limulus amebocyte lysate assay24,25; PM2.5–10 filters contained approximately 60% of the endotoxin in the PM10.
We calculated the mean concentrations of H2S, PM10, and semivolatile PM10 in the 1- and 12-hour periods that preceded the time at which participants sat outdoors for 10 minutes. Concentrations of PM2.5, PM2.5–10, and endotoxin were measured on 12-hour filters that typically did not correspond to exposure periods of interest. Thus, we estimated these exposures with a time-weighted average of the concentrations from filters exposed during the 12 hours prior to sitting outdoors. All exposure variables were coded continuously.
Given the short follow-up and focus on transient exposures, we analyzed symptoms that could appear and resolve during follow-up.
After sitting outside their homes for 10 minutes and then returning inside, participants noted whether they experienced cough or irritation of the skin, eyes, nose, or throat while outside (Table 1). Symptoms of acute irritation, reported as present or absent, were analyzed in relation to odor levels reported for the same 10 minutes and in relation to averages of PM and H2S in the hour prior to the time participants returned indoors. After returning indoors, participants rated the extent to which they experienced any of 19 acute physical symptoms in the preceding 12 hours on a scale of 0 (not at all) to 8 (extreme).
We considered the following 12-hour symptoms: respiratory (runny nose, mucus or phlegm, sore throat, cough, wheezing, difficulty breathing, chest tightness), irritation (burning eyes, itching eyes, nasal), gastrointestinal (nausea, diarrhea, poor appetite), neurologic (headache, dizziness), and other (aching joints, difficulty hearing, fever, and backache). Reports of most physical symptoms were uncommon, so we dichotomized them as absent versus present based on the distribution of responses for each symptom such that at least 85% of responses were coded as 0 and no more than 15% were coded as 1. Runny nose, mucus or phlegm, headache, cough, burning eyes, aching joints, nasal irritation, and itching eyes were dichotomized such that a response of 0 or 1 on the original scale was coded as 0 and a response of 2–8 was coded as 1. For the remaining symptoms, a response of 0 on the original scale was coded as 0 and 1–8 was coded as 1.
Participants used an AirWatch personal respiratory monitor (iMetrikus, Inc., Sunnyvale, CA) to measure forced expiratory volume in the first second (FEV1) and peak expiratory flow rate (PEF) during each data collection session. The AirWatch internally recorded each of 3 attempts and flagged any that were made with problematic technique. The highest error-free FEV1 and PEF measurements from each session (sometimes there were none) were included in the analysis as continuous outcome variables.
In this longitudinal study of transient exposures and outcomes, each participant served as her or his own control. The analytic goal was to make valid within-participant comparisons to determine whether increases in air pollutant concentrations or odor ratings were associated with physical symptoms and lung function. Estimates of associations were constructed using conditional fixed-effects linear and logistic regression models. In these models, the within-person correlation due to repeated measures is accounted for by treating each person as a stratum within the model.26 This approach has good control of measured and unmeasured time-invariant individual level confounders. These models account for the longitudinal nature of the data by modeling differences between individuals' time-specific characteristics and their mean value over the entire period of follow-up.
Time of day was integral to the study design because community members collected data at morning and evening times that were approximately 12 hours apart. Physical symptoms, lung function, and hog odors exhibit diurnal variation,19 and thus we made an a priori decision to adjust for potential confounding due to time of day by including a term for morning versus evening in all models. There was little variance in community effects; therefore we did not include the community level in our models.
Because of the large number of exposure and outcome variables, we did not restrict analyses to participant records with complete data for all variables. Each analysis excludes only those observations that were missing data for the exposure and outcome being analyzed.
There was a median of 9 hog operations within 2 miles of participating communities, and the median number of hogs within that radius was approximately 42,000 (Table 1). Study participants ranged in age from 19 to 90; their mean age was 53. Over half of participants were women, and most participants described themselves as black. Overall, the study population was healthy, with zero participants reporting emphysema and 12 reporting asthma or chronic bronchitis (Table 1). The participants provided 2949 journal entries. There were approximately 2600 responses about irritation symptoms following the 10-minutes outdoors, 2900 responses about physical symptoms experienced in the last 12 hours, and 1900 error-free measurements of lung function (eAppendix 1, http://links.lww.com/EDE/A453).
Average ambient air pollutant values are presented in eAppendix 2 (http://links.lww.com/EDE/A453). There were approximately 2700 values of H2S and 2000 values of semivolatile PM10 and PM10; the smaller numbers of observations for the latter 2 pollutants were due to equipment malfunction in hot and humid weather. There were approximately 1750 values for PM2.5–10, PM2.5, and endotoxin in the 12 communities where these pollutants were measured. Overall means, minimum and maximum community means, and between-community variation (as a % of total) are reported in eAppendix 2 (http://links.lww.com/EDE/A453). Two negative minimum community means for semivolatile PM10 occurred due to measurement error in the microbalance estimates of mass close to zero. More than half of the total variation in air pollutant measurements occurred between communities for 12-hour odor and 12-hour semivolatile PM10. For the other measured pollutants, the majority of the variation occurred within the communities over time. This was particularly true for 1-hour and 12-hour H2S and 1-hour and 12-hour PM10, for which the between-community variances were approximately 4%, 6%, 6%, and 15%, respectively.
Associations of acute irritation symptoms with twice-daily (10-minute) odor reports and 1-hour average pollution levels are presented in Table 2. Irritation symptoms were elevated in association with odor and H2S, and most coefficients were substantially greater than their standard errors. Estimates of associations between 1-hour PM10 and irritation symptoms were near zero for nasal and throat irritation, and cough, whereas associations were positive for eye and skin irritation. Coefficients for semivolatile PM10 were both positive and negative and similar in magnitude or smaller than their standard errors.
Estimates of associations of 12-hour average odor, H2S, PM10, and semivolatile PM10 with lung function measures and 12-hour symptom variables are presented in Table 3. Point estimates for PEF and FEV1 are negative except for the coefficient for PM10 and PEF. T values indicate that the negative coefficients are less than or equal in value to their standard errors, the largest being for the association between odor and FEV1.
Point estimates of associations of respiratory symptoms with odor and H2S were positive except for the coefficient between H2S and chest tightness (Table 3). The log odds of having experienced 4 of the 7 respiratory symptoms were positive for PM10 and semivolatile PM10. However, most of these estimates were close to zero, with the exception of difficulty breathing and 12-hour mean semivolatile PM10. Additionally, sore throat symptom reports were negatively associated with increases in PM10.
We observed positive associations (with high χ2 values) of irritation symptoms in the past 12 hours with 12-hour mean odor and with 12-hour mean H2S (Table 3). Twelve-hour irritation symptoms were not associated with 12-hour mean PM10 or semivolatile PM10 (Table 3). Overall, we found little association between gastrointestinal symptoms and 12-hour mean odor, H2S, PM10, or semivolatile PM10, with the exception of a positive association between PM10 and poor appetite. We found little evidence of associations between neurologic symptoms and 12-hour mean odor, H2S, PM10, or semivolatile PM10. Point estimates for the symptoms in the “other” category varied in magnitude and direction. Eleven of the 16 point estimates were negative, although most had very small χ2 values. The highest χ2 values were for the relationships of aching joints and difficulty hearing with 12-hour mean semivolatile PM10, although the estimates were in opposite directions (−0.93 ± 0.47 and 1.78 ± 0.65, respectively).
Twelve-hour average concentrations of PM2.5–10, PM2.5, and endotoxin were modeled as predictors of lung function and 12-hour symptoms in the 12 communities with results from the sequential air sampler (n = 70 participants, Table 4). T values for beta coefficients from linear conditional fixed effects models were small except for the association between PM2.5 and FEV1; FEV1 decreased 0.04 ± 0.02 L per 10 μg/m3 increase in 12 hour mean PM2.5.
Associations between symptoms and pollutants measured by the sequential sampler in the 12 communities with these measurements are also presented in Table 4. Most χ2 values were small, indicating that these exposure measures were poor predictors of symptoms. High χ2 values were observed for associations between PM2.5–10 and 3 symptoms, PM2.5 and 5 symptoms, and endotoxin and 3 symptoms. PM2.5–10 was negatively associated with chest tightness and nausea and positively associated with aching joints. Symptoms showed more consistently positive associations with PM2.5 (wheezing, difficulty breathing, burning eyes, nasal irritation, backache) and endotoxin (sore throat, chest tightness, nausea).
The models reported in Tables 2–4 were also fit using random effects mixed models and produced very similar results.
Concerns about air pollution from animal production facilities have grown with the global industrialization of food animal production.1,10–14,27,28 Concentrated hog feeding operations release air pollutants from confinement buildings, manure holding pits, and land-application of animal wastes.1,29,30 Although cross-sectional studies have documented relationships of proximity to hog operations with physical symptoms10–14,31,32 and with reduced FEV1,14 they have lacked air pollution measures and most have depended solely on participant recall of symptoms over time periods of 6–12 months. The present study contributes to the literature by linking twice-daily symptom reports and lung function measurements of people residing near hog operations with physical measures of ambient air pollutant concentrations.
Several limitations should be considered in interpreting the results of this study. First, although we have repeated measures for each participant, the number of people in the study is small. The small sample size contributes to imprecision of measures of association and also limits our ability to quantify variability in measures of association between subgroups.
Several factors may limit the external validity of the study findings. The 16 study communities are not a random sample of eastern North Carolina, and we are not able to evaluate the extent to which the characteristics of air pollutants or the volunteers in the study are representative of other populations living near industrial hog operations. Furthermore, participants were nonsmoking volunteers, mostly free of chronic respiratory diseases. Associations between hog operation pollutants and health outcomes may be different among smokers and people with asthma or other conditions that increase responsiveness to pollutants. About three-fourths of the study participants reported growing up around livestock, which has been associated with lower levels of atopy in some studies.33–35 We did not measure atopy; however, 43% of participants who grew up on a farm reported hay fever compared with 19% of those who did not, suggesting that early exposure to livestock may not have resulted in reduced responsiveness to pollutants in this population.
The air-monitoring equipment for this study was large and difficult to conceal. In some communities, participants reported reductions in hog odor and spraying of hog waste during the study compared with time periods before and after the equipment was in their neighborhoods. Changes in waste management practices could have lowered exposure levels during the study, and consequently our ability to detect effects. In addition, exposure variability within communities could not be quantified by the stationary, centrally located monitors.
Finally, lung function data were of lower quality and were less complete than other outcome data.21 Lung function assessment depends upon proper technique and is ideally conducted by a laboratory technician.36 In this study, participants were trained to make 3 measurements to the best of their ability each time they collected data. Given the community-based setting, we did not feel it was appropriate to apply American Thoracic Society/European Respiratory Society standards to these measurements.37 Instead, we analyzed only error-free readings, further reducing sample size and the precision of estimates of association. Therefore, it is of interest that a 10 μg/m3 increase in PM2.5 (measured only in 12 of the 16 communities) was associated with a 0.04 ± 0.02 L decrease in FEV1 (T = −2.12).
Despite these limitations, most exposure-outcome relationships were in the predicted direction; most of those not in the predicted direction were weak. We observed unexpected negative associations between PM10 and sore throat, PM2.5–10 and nausea and chest tightness, and semivolatile PM10 and aching joints. We are not aware of any biologic mechanisms whereby these air pollutants or unmeasured copollutants could protect against development of these symptoms. Although the study design and analytic methods preclude confounding by time-invariant characteristics of participants, these negative associations could reflect uncontrolled time-related confounding, measurement error, or both.
In addition, our findings were generally consistent with prior studies of airborne emissions from industrial hog operations. For example, in a controlled experiment, 48 healthy adult volunteers (mean age = 26) reported eye irritation and nausea more frequently when exposed to diluted swine air than when exposed to clean air.38 Radon et al14 found evidence of decreased FEV1 and increased wheezing in association with the number of concentrated animal feeding operations near participants' residences, and increased reports of asthma and nasal allergies in association with reported annoyance with odor. Mirabelli et al12 observed a 23% higher prevalence of wheezing among children who attended schools where staff reported livestock odor inside school buildings twice or more per month, compared with schools where no livestock odor was reported. In a cross-sectional study of rural Iowa children, living on a farm that raised swine and added antibiotics to animal feed was associated with asthma-related outcomes.11 Finally, endotoxin exposures have been associated with increased respiratory and systemic symptoms and decreased lung function,39 and working in hog operations has also been associated with respiratory symptoms, reduced lung function, and organic dust toxic syndrome.15,16,40,41
Interestingly, in contrast to some other studies, we did not observe an association between hog operation air pollutants and headaches.10,38,42 It is possible that headaches are more prevalent among individuals living near hog operations, but that the incidence of headaches does not covary with odor and pollutants on the short-time scale used in our study. Although an acute association with headache was observed in a chamber study,38 that exposure was diluted air from a swine confinement building, and the experimental subjects were naive volunteers who did not live near hog operations.
Among the pollutants we measured, H2S (which is produced by anaerobic decomposition of sulfur-containing organic matter in hog waste pits1) provides a fairly specific measure of hog operation pollution in these rural areas where there are few other industrial sources of H2S. In contrast to H2S, PM is a ubiquitous air pollutant with many sources and has been previously associated with lower lung function, heart rate variability, and mortality.43–46 In addition to solid particle sources, constituents of PM may form indirectly in the atmosphere through reactions of precursor gases such as NH3 47 to form soluble substances such as ammonium nitrate.48 These particles may be semivolatile, in equilibrium between gas and particle phases,49 and may have different effects than nonvolatile fractions of PM. Therefore observed associations between PM, symptoms, and lung function could be due to PM emitted by hog operations, PM from other sources, or both. We were specifically interested in PM2.5–10 because of the possibility that hog dander, feed, dried feces, endotoxin, and other microbial matter would be present in the coarse fraction.30 However, of all the pollution measures, PM2.5–10 showed the smallest and least precise associations with symptoms and lung function.
This longitudinal study contributes to evidence obtained from cross-sectional research that suggests that air pollutants near hog operations cause acute physical symptoms, particularly upper respiratory symptoms and irritation of the nose and eyes. Despite limitations of measurements of exposure and outcome, the temporal nature of the analysis eliminates confounding from time-invariant factors and strengthens the evidence. Adjustment for time of day helps reduce any time-related confounding that could be introduced by diurnal covariation in symptoms and air pollutants. Variability in pollutants within morning and evening periods is large enough so that overadjustment is not a concern.
Industrial hog operations in North Carolina are disproportionately located in low-income communities of color10,29 where there is more potential for exposure to outdoor air pollutants due to older homes that are not air tight and have no air conditioning. Many residents also lack the financial resources to travel and choose activities that could help them avoid high pollution. Exposure to air pollution from hog operations is an environmental injustice in rural areas hosting facilities that supply pork to populations spared the burdens of its production.
John Creason, Doris Taylor Davis, Brenda Denzler, Jerry Godwin, Gary R. Grant, Steve Hutton, Amy Lowman, Naeema Muhammad, Susan Schiffman, James W. Scott, Jessica Thompson, Jerry Watkins, and Susanne Wolf played key roles in field work and study support. John Creason, Karin Foarde, James Raymer, and Susan Schiffman helped to design the study. For their support and encouragement throughout all stages of this research, we are especially indebted to the study participants, community members, and community-based organizations, including the Concerned Citizens of Tillery, the Alliance for a Responsible Swine Industry, and others who must remain unnamed to protect confidentiality. We are indebted to the study participants for their hard work and commitment to collection of data.
1.National Research Council Ad Hoc Committee on Air Emissions from Animal Feeding Operations. Air Emissions from Animal Feeding Operations: Current Knowledge, Future Needs
. Washington, DC: The National Academies Press; 2003.
2.Schiffman SS. Livestock odors: implications for human health and well-being. J Anim Sci
3.Schiffman SS, Walker JM, Dalton P, et al. Potential health effects of odor from animal operations, wastewater treatment, and recycling of byproducts. J Agromedicine
4.Hoff SJ, Bundy DS, Nelson MA, et al. Emissions of ammonia, hydrogen sulfide, and odor before, during, and after slurry removal from a deep-pit swine finisher. J Air Waste Manag Assoc
5.Thorne PS, Ansley AC, Perry SS. Concentrations of bioaerosols, odors, and hydrogen sulfide inside and downwind from two types of swine livestock operations. J Occup Environ Hyg
6.Walker JT, Robarge WP, Shendrikar A, Kimball H. Inorganic PM2.5 at a U.S. agricultural site. Environ Pollut
7.Goetz S, Aneja VP, Zhang Y. Measurement, analysis, and modeling of fine particulate matter in eastern North Carolina. J Air Waste Manag Assoc
8.Costanza JK, Marcinko SE, Goewert AE, Mitchell CE. Potential geographic distribution of atmospheric nitrogen deposition from intensive livestock production in North Carolina, USA. Sci Total Environ
9.Sherlock RR, Sommer SG, Khan RZ, et al. Ammonia, methane, and nitrous oxide emission from pig slurry applied to a pasture in New Zealand. J Environ Qual
10.Wing S, Wolf S. Intensive livestock operations, health, and quality of life among eastern North Carolina residents. Environ Health Perspect
11.Merchant JA, Naleway AL, Svendsen ER, et al. Asthma and farm exposures in a cohort of rural Iowa children. Environ Health Perspect
12.Mirabelli MC, Wing S, Marshall SW, Wilcosky TC. Asthma symptoms among adolescents who attend public schools that are located near confined swine feeding operations. Pediatrics
13.Sigurdarson ST, Kline JN. School proximity to concentrated animal feeding operations and prevalence of asthma in students. Chest
14.Radon K, Schulze A, Ehrenstein V, van Strien RT, Praml G, Nowak D. Environmental exposure to confined animal feeding operations and respiratory health of neighboring residents. Epidemiology
15.Eduard W, Pearce N, Douwes J. Chronic bronchitis, COPD, and lung function in farmers: the role of biological agents. Chest
16.Senthilselvan A, Chenard L, Ulmer K, Gibson-Burlinguette N, Leuschen C, Dosman JA. Excess respiratory symptoms in full-time male and female workers in large-scale swine operations. Chest
17.Zejda JE, Hurst TS, Rhodes CS, Barber EM, McDuffie HH, Dosman JA. Respiratory health of swine producers. Focus on young workers. Chest
18.Andersen CI, Von Essen SG, Smith LM, Spencer J, Jolie R, Donham KJ. Respiratory symptoms and airway obstruction in swine veterinarians: a persistent problem. Am J Ind Med
19.Wing S, Horton RA, Marshall SW, et al. Air pollution and odor in communities near industrial swine operations. Environ Health Perspect
20.Horton RA, Wing S, Marshall SW, Brownley KA. Malodor as a trigger of stress and negative mood in neighbors of industrial hog operations. Am J Public Health
21.Schinasi L, Horton RA, Wing S. Data completeness and quality in a community-based and participatory epidemiologic study. Prog Community Health Partnersh
22.Tajik M, Muhammad N, Lowman A, Thu K, Wing S, Grant G. Impact of odor from industrial hog operations on daily living activities. New Solut
23.Wing S, Horton RA, Muhammad N, Grant GR, Tajik M, Thu K. Integrating epidemiology, education, and organizing for environmental justice: community health effects of industrial hog operations. Am J Public Health
24.Menetrez MY, Foarde KK, Esch RK, et al. An evaluation of indoor and outdoor biological particulate matter. Atmos Environ
25.Foarde K, Berry M. Comparison of biocontaminant levels associated with hard vs. carpet floors in nonproblem schools: results of a year long study. J Expo Anal Environ Epidemiol
26.Allison PD. Fixed Effects Regression Methods for Longitudinal Data Using SAS
. Cary, NC: SAS Institute; 2005.
27.Pew Commission on Industrial Farm Animal Production. Putting Meat on the Table: Industrial Farm Animal Production in America
. Baltimore, MD: The Johns Hopkins Bloomberg School of Public Health; 2006.
28.Thorne PS. Environmental health impacts of concentrated animal feeding operations: anticipating hazards—searching for solutions. Environ Health Perspect
29.Wing S, Cole D, Grant G. Environmental injustice in North Carolina's hog industry. Environ Health Perspect
30.Iowa State University and the University of Iowa Study Group. Iowa Concentrated Animal Feeding Operation Air Quality Study.
Iowa City, IA: The University of Iowa College of Public Health, Iowa State University College of Agriculture; 2002.
31.Thu KM. Public health concerns for neighbors of large-scale swine production operations. J Agric Saf Health
32.Villeneuve PJ, Ali A, Challacombe L, Hebert S. Intensive hog farming operations and self-reported health among nearby rural residents in Ottawa, Canada. BMC Public Health
33.Naleway AL. Asthma and atopy in rural children: is farming protective? Clin Med Res
34.Vedanthan PK, Mahesh PA, Vedanthan R, Holla AD, Liu AH. Effect of animal contact and microbial exposures on the prevalence of atopy and asthma in urban vs rural children in India. Ann Allergy Asthma Immunol
35.Von Ehrenstein OS, Von Mutius E, Illi S, Baumann L, Bohm O, von Kries R. Reduced risk of hay fever and asthma among children of farmers. Clin Exp Allergy
36.Miller MR, Crapo R, Hankinson J, et al. General considerations for lung function testing. Eur Respir J
37.Miller MR, Hankinson J, Brusasco V, et al. ATS/ERS Task Force. Standardisation of spirometry. Eur Respir J
38.Schiffman SS, Studwell CE, Landerman LR, Berman K, Sundy JS. Symptomatic effects of exposure to diluted air sampled from a swine confinement atmosphere on healthy human subjects. Environ Health Perspect
39.Heederik D, Sigsgaard T, Thorne PS, et al. Health effects of airborne exposures from concentrated animal feeding operations. Environ Health Perspect
40.Vogelzang PF, van der Gulden JW, Folgering H, van Schayck CP. Organic dust toxic syndrome in swine confinement farming. Am J Ind Med
41.Charavaryamath C, Singh B. Pulmonary effects of exposure to pig barn air. J Occup Med Toxicol
42.Thu K, Donham KJ, Ziegenhorn R, et al. A control study of the physical and mental health of residents living near a large-scale swine operation. J Agric Saf Health
43.Downs SH, Schindler C, Liu LJ, et al. Reduced exposure to PM10 and attenuated age-related decline in lung function. N Engl J Med
44.Lewtas J. Air pollution combustion emissions: characterization of causative agents and mechanisms associated with cancer, reproductive, and cardiovascular effects. Mutat Res
45.Wheeler A, Zanobetti A, Gold DR, Schwartz J, Stone P, Suh HH. The relationship between ambient air pollution and heart rate variability differs for individuals with heart and pulmonary disease. Environ Health Perspect
46.Zeka A, Zanobetti A, Schwartz J. Individual-level modifiers of the effects of particulate matter on daily mortality. Am J Epidemiol
47.Aneja VP, Schlesinger WH, Erisman JW. Effects of agriculture upon the air quality and climate: research, policy, and regulations. Environ Sci Technol
48.Harrison RM, Yin J. Particulate matter in the atmosphere: which particle properties are important for its effects on health? Sci Total Environ
49.Grover BD, Eatough NL, Eatough DJ, et al. Measurement of both nonvolatile and semi-volatile fractions of fine particulate matter in Fresno, CA. Aerosol Sci Technol