|
K. Tu, K. Donham, R. Ziegenhorn,
Journal of Agricultural Safety and Health
Beginning in the mid 1970s and continuing to the present, research has been devoted to understanding human exposures and health consequences of working in swine confinement environments (Donham, 1990; Donham et al., 1977; Kiekhaefer et al., 1995; Thorne et al., 1992). The results indicate swine confinement workers experience a number of health problems. A notable problem area is the range of respiratory conditions resulting from exposure to gases and dusts while working inside these facilities (Donham, 1993). However, very little research has been conducted on exposures to external emissions.
Research on exterior conditions has primarily targeted the reduction and elimination of odor emissions from swine operations. This research has concentrated 0 identifying compounds producing- odors (Merkel et al., 1969; O'Neill and Phillips 1992; Ritter, 1989), mechanisms for measuring odor (Hobbs, 199D-; Longhurst, 1995; Mannebeck, 1995; Sweeten, 1988), and the development of control technologies (Fullhage, 1995; Voermans, 1995; Yokoyama, 1995). In addition, considerable research has been devoted to the uptake of ammonia from animal manure and the environmental consequences of its redeposition as rain in Europe (ApSimon and Kruse-Plass, 1991; Le--, 1990). However, little work has been devoted to understanding odor-related complaints and health problems among, residents living near large livestock operations.
Emerging research (Schiffman, 1995; Schiffman et al., 1995) has investigated relationships between the psychological health of neighbors and swine-generated odors. This research indicates deleterious psychological health effects such as mood disorders result from a combination of physical agents and physiological responses to swine odor. It also suggests changing, social conditions in rural neighborhoods may be a factor affecting responses. Other research (Thu and Durrenberger, 1994) supports Schiffman's suggestion that rural social issues play a role.
This study addresses a gap in research through a control approach to assessing interrelated issues of health, quality of life, and mental health of residents living, in the vicinity of a large-scale swine confinement facility. The primary purpose of the study was to test a methodology for assessing, neighbor health and quality of life issues, provide preliminary data to identify salient neighbor health and life quality problems, and generate hypotheses for further research.
Methodology
This study is based on a comparative control methodology. Data on physical health status,mental health, and quality of life were collected via personal interviews of neighbors of a large-scale swine production facility and from a random sample of rural residents who do not live near any livestock. Results from the two groups were compared to identify salientdifferences.
Survey Instrument
A questionnaire was developed to elicit data via personal interviews on physical health status, mental health, quality of life, and standard sociodemographics. An initial section was designed to collect basic background information, including demographics, employment, residential history, and previous occupational exposures. The second section elicited symptoms indicative of health status. Health status questions were drawn from earlier health assessment studies of swine confinement workers (Donham, 1990). They consisted of an initial set of open-ended questions concerning health problems, frequency ratings of IS symptoms, and a series of health history and current health status questions.
To assess psychological health, mental health questions were developed in consultation with Professor Susan Schiffm'an, a medical psychologist at Duke University. In her research (Schiffman et al., 1995), Schiffman collected data on mood states between swine operation neighbors and controls utilizing a standardized mood profile scale (McNair and Lorr, 1992). To complement her findings, we included psychological scales to collect data on depression (Zung, 1965) and anxiety (state-trait anxiety inventory from Steer et al., 1993).
A third section included open-ended questions to solicit qualitative information on neighborhood social issues. For the case sample, questions were designed to elicit information on issues such as how well and how long neighbors knew the owners and operators of the swine facility and the nature of their relationship. Both case and control participants responded to a question on the characteristics of a "good neighbor".
Sample Selection and Procedures
A large-scale swine confinement operation was selected as the study site based on its scale and because we knew certain neighbors had expressed environmental and health concerns. The selected swine operation is one of the largest in Iowa, with approximately 4,000 sows in a farrowing operation consisting of six confinement units, an office building, and a two-state outdoor waste lagoon about five acres in size. The entire operation is situated on an estimated,35 acres of land.
The 27 neighbors living within two miles were identified from plat maps as potential participants. Each household was sent a letter of introduction, a project summary, an invitation to participate, and a stamped return postcard. Of the 27 households contacted, 18 returned the postcard indicating an interest in participating (67% participation rate). Follow-up phone calls were made to each of the 18 interested households to schedule personal health assessment interviews. Of the 18 interested households, 10 households met the selection criterion of living closer to the large-scale swine operation than other livestock operations. Nine of these with 19 participants completed all aspects of the study. Multiple dwellers within a single household were interviewed independently from each other.
A control sample of rural residents not living near any livestock operation was selected. Countv level data from the 1992 Agricultural Census were used to locate areas of minimal livestock production. A county different from the case sample site was selected and all rural zip code areas within the countv were checked to identify areas with the lowest population of livestock. All rural residents (n = 188) within the selected zip code area who owned a telephone were selected from a telephone database. Letters of introduction were sent to all residents, including a project summary, an invitation to participate, and a stamped return postcard. Included in the letter was an additional screening caveat that prospective participants must not live within a mile of any type of livestock operation greater than 50 head.
Of the 188 letters sent, 14 were returned undeliverable by the Post Office, 24 postcards were returned declining participation, and I postcards were returned indicating they met the selection criteria and were interested in participating. All interested participants were contacted by phone to schedule interviews in their homes at their convenience. We requested that as many members of the household as possible participate. A total of 21 interviews were conducted in I households. However, data from two households in which three interviews were conducted had to be eliminated because of a failure to meet our selection criteria. Consequently, the control sample consisted of 18 personal interviews across nine rural households. Neither the control nor case sample participants were provided financial or other incentives to participate.
The principle author and a co-author were the primary interviewers. Both are trained in qualitative and quantitative data collection methods utilizing ethnographic and personal interview techniques from social anthropology and the social sciences (Weller and Romney, 1988). The interviewers have 12 years combined experience in data collection specific to agriculture.
All data from the interviews were coded and entered into a Paradox database. Quantitative analyses were performed using, a SAS statistical package*. Qualitative data were analyzed based on a combination of results from the quantitative analysis and interviewer notes on the questionnaires.
Results
As evidenced in table 1, there was little difference in gender, marital status, age or educational level between the two samples. In addition, all respondents were white and there was a comparable proportion of farmers and nonfarmers in our sample populations. It is unlikely that the findings are biased by demographic differences between the sample and control populations.
Physical Health Symptoms
Results of the frequency of physical symptoms are presented in figure 1. The study population reported higher frequencies of 14 out of the 18 symptoms than the control population. There was no connection between the frequency of reported physical symptoms and distance from the swine facility. Results indicate a pattern of four interconnected clusters of symptoms that include respiratory problems, nausea and weakness, headaches and plugged ears, and irritation of eyes, nose, and throat. This constellation of symptoms matched those reported bv participants in response to an open-ended question posed earlier in the interview. Skin rash, muscle aches. and fever was reported more frequently among the control group, while hearing-, problems were reported at an identical frequency by both groups.
Table 2 presents the results of analyses assessing the significance in differences between the reported symptoms from neighbors of the swine facility and the control population. The constellation of 14 symptoms reported more frequently by the study croup showed composite mean frequency scores of 21 for the study population and 15 for the control. The first line of table 2 labeled "All Symptoms" presents the the results of a Wilcoxon Test (Chi Sq = 2.3; P 0.13) indicating this difference warrants attention but is not conclusive.
More significant is the trend among- clusters of symptoms. Within the range of symptoms reported more frequently by the study sample, four clusters of related symptoms deserve particular attention. These clusters of symptoms have been recognized previously in swine facility workers (Donham, 1995). They represent toxic or inflammatory effects on different segments of the respiratory tract.
The first cluster is a combination of five symptoms indicative of inflammation of the bronchi and bronchioles, or chronic bronchitis and hyperactive airways: sputum, cough, breath shortness, wheezing,, and chest tightness. A variety of standardized survey instruments include this cluster of symptoms: the American Thoracic Society, the British Medical Research Council, and the Agricultural Dust Exposure Assessment. A one-tailed t-test was conducted to determine whether the study population reported experiencing this combination of symptoms more frequently than the control sample. As presented in Cluster 1 of table 2, results indicate that residents living in the vicinity of the large-scale operation do report experiencing significantly higher rates of symptoms associated with chronic bronchitis and hyperactive airways T = 2.12; P = 0.02; 26.7 decrees of freedom). This type of bronchitis is almost invariably associated with environmental exposures, e.c,., air pollution, chronic agricultural dust exposure, and long-term cigarette smoking.
A second cluster of related symptoms was examined that included: nausea, weakness, dizziness, and fainting,. Previous research among- swine workers reveal this group of symptoms is fairly common (Donham, 1993). A one-tailed t-test was again conducted to determine whether the study population reported experiencing this combination of symptoms more frequently than the control sample. As presented in Cluster 2 of table 2, results indicate that residents living, in the vicinity of the large-scale operation do report experiencing significantly higher rates of nausea, weakness, dizziness, and fainting (T = 1.83; P = 0.04; 24.5 decrees of freedom). Research among swine confinement workers suggests that long-term exposure to less than acutely toxic levels of endotoxin and hydro-en sulfide merit investigation in conjunction with these symptoms (Aucer et al., 1994).
A third combination of symptoms, headaches and plu2-ed ears, is another frequently observed among, swine confinement workers. Once again, a one-tailed t-test was conducted to determine whether the study population reported experiencing this combination of symptoms more frequently than the control sample. As presented in Cluster 3 of table 2, results indicate that residents living in the vicinity of large-scale swine operation reported experiencing higher rates of headaches and plugged, though the difference is marginally less significant than the first two clusters (T = 1.67; P = 0.06; 24.5 decrees of freedom). The physiological explanation for these symptoms among swine confinement workers is that they are often associated with chronic sinusitis. Symptoms of chronic sinusitis are seen in nearly a quarter of active swine producers (Donham, 1993).
A final cluster of symptoms was examined that included: burning eyes, runny nose, and scratchy throat. The one-tailed t-test was replicated to compare the study and control sample. As presented in Cluster 4 of table 2, results indicate that the higher rates of these reported symptoms among neighboring residents of the large-scale operation warrant notice but the difference is less clear (T = 1. 18; P = 0. 12; 33 degrees of freedom). Among interior swine confinement workers, these symptoms are associated with a condition called mucous membrane irritation. Irritant cases and particulates inside swine confinement buildings are thought to affect the mucous membranes of the eyes and upper airways, resulting in the symptoms reported.
Differences in reported physical health symptoms between the study and control population are present. More notable than individual symptoms or clusters of symptoms, is the overall trend of interrelated symptom clusters reported more frequently among, neighbors of the swine facility than the control sample. The constellation of symptoms reported in excess by neighbors is consistent with, but less severe and frequent, compared to symptoms of workers in swine confinement facilities. A companion article to this article reveals that ammonia, dust, and endotoxin are present in the air downwind from large swine facilities. However, these levels are much lower than those previously associated with any known illness (Reynolds et al., in press). This raises the question as to whether low levels may be associated with reported symptoms.
Research in North Carolina (Schiffman et al., 1995) reported that persons living, near large-scale swine operations exhibited significantly higher rates of mood disorders than did matched control participants as measured by a Profile of Mood States (POMS) scale. Neighbors living, near large swine facilities experienced higher rates of tension, anger, fatigue, and confusion. Schiffman discusses how molecules responsible for odors can potentially result in physical responses linked to mood alterations. She also suggests that odor may play a role in suppressing immune system responses via physical connections between the olfactory and immune systems. The psychological scales we used measured depression and anxiety as a comparative supplement to Schiffman's research.
The depression scale is based on the work of Zung, (1964) and is derived from established research utilizing factor analyses to derive the most common set of underlying characteristics that predict depression in a clinical setting-. Participants in our pi 'lot study were administered 20 questions from the Self-Rating Depression Scale (SDS) derived from this clinical work. The comparative results of mean scores of individual items are presented in figure 2.
Little difference in depressive symptoms exists between the study and control populations. Following- Zung's (1964) methodology a depression index was created by totaling, the raw scores of participants and dividing them by the total possible score. The composite mean depression index for case study participants totaled 0.37 compared with 0.40 for the controls and were not significantly different (Chi Sq = 0.35; P = 0.55). These scores compare with a mean depression index of 0.74 in Zung,'s clinically admitted population of depressed patients. Zung's control, or 44 normal" population, scored 0.33. Thus our study population is well within the range of Zung's control population, exhibiting very little depressive symptomology.
An anxiety scale was administered based on the Beck Anxiety Inventory (BAI) developed by Beck and Steer (Steer et al., 1993). The scale is derived from analyses of in-patients exhibiting a set of symptoms distinct from other mental disorders in a clinical setting. Participants in our pilot study were administered 21 questions from the BAI derived from this clinical work. 'Me comparative results of mean scores of individual items are presented in figure 3.
Little difference in anxiety symptoms exists between the study and control populations. Following the methodology of Steer et al. (1993), an anxiety index was created for each case by totaling the raw scores of participants and dividing it by the total possible score. The composite mean anxiety indexes for case study and control participants were virtually identical: 0.11. These scores compare with a mean anxiety score of 0.29 in Steer and coworkers' population of 250 clinically admitted patients categorized as "moderately anxious". Our study population does not appear to be suffering from anxiety related psychological symptoms. Moreover, no significant differences were found in anxiety between the study participants and the control population.
Conclusion
Evidence indicates that neighbors of the large-scale swine operation in our study reported experiencing increased rates of a number of interrelated symptoms, including headaches, respiratory problems, eye irritation, nausea, weakness, and chest tightness. The pattern of differential symptomology rates between the study and control samples suggest further study is warranted. There is little evidence to suggest that neighbors of the large-scale swine operation suffer hi-her rates of anxiety or depression.
Further study is needed to test the hypothesis that neighbors of large-scale swine operations experience higher rates of physical symptoms comparable to the types of symptoms experienced by interior confinement workers. A larger population-based study is needed that includes neighbors of a cross-section of various sizes and types of swine and other livestock operations. Such a study should continue to use personal interviews as the basis of health assessments. A central issue in these investigations is the reliability and quality of data. Personal interviews by trained and experienced interviewers in the homes of rural residents provide a comfortable setting for participants to discuss issues in a forthright and open manner. A report based on a 1985 National Science Foundation conference on data collection points to natural settings as providing the best opportunity for collecting reliable interview data (Bernard et al., 1986). Validity of data collection is related to a host of factors, including the extent of open exchange between interviewers and persons being interviewed.
Neighbors did not appear to be concocting evidence of health or psychological problems based on any personal or political agenda. Evidence for the credibility of physical symptom reports comes from the psychological profile data. If participants wanted to concoct evidence it would have been easy for them to report high rates of depression and/or anxiety. Such reporting did not occur. Physical assessments of neighbors would provide clarification of these issues.
Permeating all the responses, regardless of whether respondents had specific health problems, was the underlying view that the owner was creating, social and class divisions in the neighborhood and community. Most believed that the construction and presence of the facility violated core rural values of being, a good neighbor". For virtually all respondents, rural "neighborliness" embodies central cultural principles of egalitarian relationships, reciprocal exchange such as helping, or sharing in times of need, mutual respect, and being kept informed. The facility's construction and continuing presence was viewed as eroding these cornerstones of agrarian life. Often discussed outside the strictures of the questionnaire, participants voiced concern about such issues as labor turn-over, social chasms emerging, between neighbors and between children of neighbors, the influence of the facility's owner on local political and economic decision-making, boards, and the ability of residents to have control over their land, homes, families, and quality of life. Clearly the issues confronting rural residents in this study reflect an intertwining of personal, environmental, economic, and social health. Further study should seek to clarify and broaden our understanding of these interrelated issues.
References
Auger P., P. Gourdeau and J. Miller. 1994. Clinical experience with patients suffering from a chronic fatigue-like syndrome and repetitious upper respiratory infection in relation to airborne molds. Am. J. Ind. Med. 25(l):41-42.
Bernard, H. R., P. J. Pelto et al. 1986. The construction of primary data in cultural anthropology. Current Anthropology 27(4):382-396.
DeLind, L. 1995. The state, hog hotels and the "Right to Farm": A curious relationship.
Agric. and human Values 12(3):34-44.
Donham, K. J. 1995. A review - The effects of environmental conditions inside swine housing on worker and pig health. In Manipulating Pig Production V, 203-22l,'cds. D. P Hennessy and P. D. Cranwell. Werribee, Victoria, Australia: Australasian Pig Science Assoc.
.1993. Respiratory disease hazards to workers in livestock and poultry confinement structures. Seminars in Resp. Med 14(i):49-59.
1990. Health effects from work in swine confinement buildings. Am J. Ind. Med. 17(l):17-25.
Donham, K. J., M. J. Rubino et al. 1977. Potential health hazards of workers in swine confinement buildings. J. Occ. Med 19(6):383-387.
Fullhage, C. 1995. Minimizing odor problems in lagoons/composting of dead animals. In Int. Round Table on Swine Odor Control, 72-77. Ames, Iowa: Iowa State University.
Hobbs, P. 1995. Measurement of swine odor using electronic nose technology. In [lit. Round Table on Swine Odor Control, 36-39. Ames, Iowa: Iowa State University.
Kiekhaefer, M. S., K. J. Donham et al. 1995. Cross seasonal studies of airborne microbial populations and environment in swine buildings: Implications for worker and animal health. Ann. Agric. Environ. Med. 2(2):37-44. ego,, B. J. 1990. Farm and food waste: utilization without pollution. In Proc. of the Sixth Int. Symp. on Agricultural and Food Processing Wastes, xi-xxi. St. Joseph, Mich.: ASAE.
Longhurst, P. 1995. Applications of electronic nose technology/community assessment of odor. In Int. Round Table on Swine Odor Control, 32-35. Ames, Iowa: Iowa State University.
Mannebeck, H. 1995. Olfactory measurement. In Int. Round Table on Swine Odor Control, 2-4. Ames, Iowa: Iowa State University.
McNair, D. M., M. Lorr. et al. 1992. Rev. Manual: Profile of Mood States. San Diego, Calif.:
Education and Testing Service.
Merkel, J. A., T. E. Hazen et al. 1969. Identification of cases in a swine confinement building atmosphere. Transactions of the ASAE 12(3):310-315.
O'Neill, D. H. and V. R. Phillips. 1992. A review of the control of odor nuisance from livestock buildings: Part 3, Properties of the odorous substances which have been identified in livestock wastes or in the air around them. J. Agric. Eng. Res. 34:23-50.
Reynolds, S., K. J. Donham et al. Air quality assessments in the vicinity of swine production facilities. J. of Agromedicine (In press).
Ritter, W. F. 1989. Odour control of livestock wastes: State-of-the art in North America.
J. Agric. Eng. Res. 42:51-62.
Schiffman, S. 1995. Psychological effects of swine odor on humans. In Int. Round Table on Swine Odor Control, 8-1 1. Ames, Iowa: Iowa State University.
Schiffman, S. S., E. A. S. Miller et al. 1995. The effect of environmental odors emanating from commercial swine operations on the mood of nearby residents. Brain Res. Bulletin
37(4):369-375. 1
Steer, R. A., D. J. Rissmiller et al. 1993. Structure of the computer-assisted Beck Anxiety Inventory with psychiatric inpatients. J. Personality. Assessment 60(3):532-542.
Sweeten, J. M. 1988. Odor measurement and control for the swine industry - Recent developments. J. Env. Health 50(5):282-286.
Thorne, P. S., M. S. Kiekhaefer et al. 1992. Comparison of bioaerosol sampling methods in
barns housing swine. Appl. and Environ. Microbiology- 58(8):2543-255 1.
Thu, K. 1995/96. Piggeries and politics: rural development and lowa's multibillion dollar
swine industry. Culture and Agric. 53:19-23.
Thu, K. and E. P. Durrenberger. 1994. North Carolina's hog industry: The rest of the story.
Culture and Agic. 49:20-23.
Voermans, J. 1995. Reducing ammonia and odor emissions from hog barns. In lilt. Round Table on Swine Odor Control, 87-92. Ames, Iowa: Iowa State University.
Weller, S. and A.K. Romney. 1988. Systematic Data Collection. Newbury Park, Calif.: Sage.
Yokoyama, M. 1995. Ozonation to reduce odors generated from swine industry waste slurry.
In Int. Round Table on Swine Odor Control. 92-97. Ames, Iowa: Iowa State University.
Zung, W. W. K. 1965. A self-rating depression scale. Archive Gen. Psychiatry 12:63-70.
|
