Residential exposure to outdoor air pollution from livestock operations and perceived annoyance among citizens.

Epidemiological studies have shown that residential exposure to livestock odors can affect the health and wellbeing of rural citizens. However, exposure-response models for this relationship have not been developed. One of the main challenges is to identify a compound that can be used as proxy for livestock odor exposure. In this paper we developed models that describe the relationship between long-term averaged outdoor residential ammonia (NH(3)) exposures and livestock odor annoyance experienced by rural residents, and investigated person-related variables associated with annoyance responses. We used emission-based atmospheric dispersion modeling data to estimate household-specific outdoor concentrations and survey data to characterize the study subjects. Binomial and multinomial logistic regressions were used for model development. Residential NH(3) exposure was positively associated with moderate, high and extreme odor annoyance (adjusted odds ratio=10.59; 95% confidence interval: 1.35-83.13, for each unit increase in Log(e)NH(3) exposure). Specific characteristics of the exposed subjects (i.e., age, time per week spent at home, presence of children at home and job) act as co-determinants of odor annoyance responses. Predictive models showed classification accuracies of 67-72%. The results suggest that NH(3) exposure in the residential outdoor environment can be used as a predictor of livestock odor annoyance in population studies.