Share this article on:

Air Pollution, Lung Function, and Physical Symptoms in Communities Near Concentrated Swine Feeding Operations

Schinasi, Leaha; Horton, Rachel Averya; Guidry, Virginia T.a; Wing, Stevea; Marshall, Stephen W.a; Morland, Kimberly B.b

doi: 10.1097/EDE.0b013e3182093c8b
Air Pollution: Original Article

Background: Concentrated animal feeding operations emit air pollutants that may affect health. We examined associations of reported hog odor and of monitored air pollutants with physical symptoms and lung function in people living within 1.5 miles of hog operations.

Methods: Between September 2003 and September 2005, we measured hydrogen sulfide (H2S), endotoxin, and particulate matter (PM10, PM2.5, and PM2.5–10) for approximately 2-week periods in each of 16 eastern North Carolina communities. During the same time periods, 101 adults sat outside their homes twice a day for 10 minutes, reported hog odor and physical symptoms, and measured their lung function. Conditional fixed-effects logistic and linear regression models were used to derive estimates of associations.

Results: The log odds (±1 standard error) of acute eye irritation following 10 minutes outdoors increased by 0.53 (±0.06) for every unit increase in odor, by 0.15 (±0.06) per 1 ppb of H2S, and by 0.36 (±0.11) per 10 μg/m3 of PM10. Odor and H2S were also associated with irritation and respiratory symptoms in the previous 12 hours. The log odds of difficulty breathing increased by 0.50 (±0.15) per unit of odor. A 10 μg/m3 increase in mean 12-hour PM2.5 was associated with increased log odds of wheezing (0.84 ± 0.29) and declines in forced expiratory volume in 1 second (−0.04 ± 0.02 L). A 10 EU/mg increase in endotoxin was associated with increased log odds of sore throat (0.10 ± 0.05), chest tightness (0.09 ± 0.04), and nausea (0.10 ± 0.05).

Conclusions: Pollutants measured near hog operations are related to acute physical symptoms in a longitudinal study using analyses that preclude confounding by time-invariant characteristics of individuals.


From the aDepartment of Epidemiology, University of North Carolina at Chapel Hill, Chapel Hill, NC; and bDepartment of Community and Preventive Medicine, Mount Sinai School of Medicine, New York, NY.

Submitted 2 April 2010; accepted 18 October 2010; posted 11 January 2011.

Supported by National Institute of Environmental Health Sciences (grant R01 ES011359); Biostatistics for Research in Environmental Health Training Grant of the National Institute of Environmental Health Sciences (grant 5-T32-ES07018); and United States Environmental Protection Agency cooperative agreement CR829522.

Editors' note: A commentary on this article appears on page 216.

Supplemental digital content is available through direct URL citations in the HTML and PDF versions of this article (

Correspondence: Steve Wing, Department of Epidemiology, CB 7435, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599–7435. E-mail:

© 2011 Lippincott Williams & Wilkins, Inc.