Characteristics of the study participants are shown in Table 1. Poultry processing workers were slightly older than workers in the comparison population (mean age, 35 vs 32 years, respectively; P < 0.01) and a larger proportion of participants in the poultry processing population reported their country of birth as Guatemala compared with the predominantly Mexican origin reported by the comparison population. The percentage of participants who reported work-related inhalation exposures was higher in the comparison population (74%) than in the poultry processing population (51%).

Table 2 shows the prevalence of reported nasal and respiratory symptoms and conditions. Overall, the lifetime prevalences of symptoms and conditions, as well as the prevalences in the last 12 months, were each higher in the comparison population than in the poultry processing population (eg, ever wheezing or whistling in the chest: 11% vs 8%, respectively). None of the differences, including the difference in prevalence of wheezing or whistling in the chest in the last month, current use of medication for breathing problems, and reporting that breathing problems worsen at work, was statistically different at the α = 0.05 level.

Among the 402 poultry processing workers, packing (26%) and cutting (22%) were the most frequently reported poultry processing activities. Twenty-one percent reported performing two or more of the tasks as part of their poultry processing jobs (Table 3). Overall, the prevalences of nasal symptoms (prevalence, 25%; OR, 0.91; 95% CI, 0.59–1.40) and current asthma (prevalence, 9%; OR, 0.96; 95% CI, 0.57–1.62) were each lower among poultry processing workers than in the comparison population. Including sex and other inhalation exposures in models of the association between poultry processing and nasal symptoms modestly attenuated the effect estimates (including sex: OR, 0.91; 95% CI, 0.59–1.41; including other inhalation exposures: OR, 0.97; 95% CI, 0.62–1.52; including sex and other inhalation exposures: OR, 1.00; 95% CI, 0.63–1.57). Adding country of birth to the final model of the association between poultry processing and current asthma resulted in a negligible change in the adjusted OR (0.95; 95% CI, 0.55–1.63).

The small numbers of symptomatic participants limited our ability to evaluate thoroughly associations between individual poultry processing activities and nasal and respiratory symptoms. Nevertheless, the highest prevalences were observed among poultry processing workers who reported receiving, hanging, killing, and plucking (nasal symptoms: 35%); cutting and evisceration (nasal symptoms: 32%; current asthma: 12%); trimming (nasal symptoms: 33%); and chilling and packing (nasal symptoms: 34%; current asthma: 11%). Of these activities, multivariate analyses generated consistently elevated point estimates for cutting and evisceration (nasal symptoms: 1.10; current asthma: 1.50); trimming (nasal symptoms: 1.41; current asthma: 1.07); and chilling and packing (nasal symptoms: 1.57; current asthma: 1.23). Point estimates below unity were consistently observed for deboning (nasal symptoms: 0.61; current asthma: 0.64) and sanitation (nasal symptoms: 0.82; current asthma: 0.46) activities.

Among participants who completed spirometry testing, unadjusted mean values of FEV₁ and FVC were both higher in the comparison population (men: FEV₁, 3612 mL, FVC, 4382 mL; women: FEV₁, 2771 mL, FVC, 3281 mL) than in the population of poultry processing workers (men: FEV₁, 3337 mL, FVC, 4087 mL; women: FEV₁, 2612 mL, FVC, 3072 mL) (Table 4). Adjusted absolute differences in mean FEV₁ and FVC were lower among poultry processing workers overall and in nearly all of the poultry processing activity groups–most notably among all men (FEV₁, − 89 mL; FVC, − 84 mL) and men performing sanitation activities (FEV₁, − 192 mL; FVC, − 206 mL). Differences in the percentages of FEV₁ and FVC predicted values and in the FEV₁:FVC ratio were modest across all categories, as were differences in FEV₁ and FVC with increasing numbers of poultry processing activities reported.

This large observational study provided us with a unique opportunity to report the prevalences of a wide range of respiratory health outcomes and measurements of lung function among working Latino men and women in rural North Carolina. In these data, employment in poultry processing was not associated with nasal symptoms or current asthma, though we observed lower prevalences of both in the poultry processing cohort than in the comparison population. In contrast, measurements of FEV₁ and FVC were modestly lower among female and male poultry processing workers and across categories of poultry processing job activities. The magnitudes of the differences were larger among men overall, and particularly among those who reported working in sanitation. Together, these findings suggest that despite the low prevalence of respiratory symptoms reported, poultry processing work may affect lung function.

The remarkably low prevalences of nasal and respiratory symptoms in the poultry processing cohort suggest that our study may be affected by a respiratory-specific healthy worker effect. If individuals who are eligible to work in poultry processing but who have allergies or respiratory health conditions that may be exacerbated by the potential exposures do not seek employment in poultry processing or have left poultry processing work, then the individuals employed in the facilities, and thus available for recruitment into this study, may be healthier than those employed elsewhere. Because of the wide range of manual labor jobs held by members of the comparison population, we do not expect the respiratory health or susceptibility profile of the comparison group to have been appreciably or systematically altered by the same phenomenon. This hypothesis is supported by the lower percentages of current smoking and lifetime histories of allergies, asthma, and wheezing in the poultry processing group than in the comparison group. Indeed, the associations between poultry work and respiratory symptoms generated in our study do not point to poultry processing work as a risk factor for adverse respiratory health.

In contrast, the consistently lower measures of FEV₁ and FVC in the poultry processing cohort suggest that, despite the low prevalence of symptoms reported, poultry processing may affect lung function. Because the differences in mean lung function measurements did not reach statistical significance among men or women, these data should be interpreted with caution. If the degree of lung function impairment observed in the poultry processing cohort was insufficient to trigger symptoms or other functional consequences, then this modest decline may indicate unrecognized respiratory disease. If participants experienced symptoms but did not report them in our survey, then our lung function findings may reveal systematic differences in the ways members of the two study populations reported symptoms. In both populations, median educational attainment was equivalent to a primary school education. Compared with other jobs available to Latino immigrants with limited education, jobs in poultry processing come with some noteworthy advantages. For example, working conditions were reported more favorably in the poultry processing cohort, where only 1% (vs 5% in the comparison population) reported having lost their job in the last year; 41% (vs 32%) reported having modified their work stations or tasks to make them safer or more comfortable; 87% (vs 60%) agreed that workers receive safety instruction when hired; 64% (vs 12%) reported having health insurance; and 25% (vs 4%) reported that the insurance was paid by their employer. If poultry processing jobs are considered more desirable than other employment opportunities available in the communities, then these jobs may be held by healthier workers, including men and women with less respiratory impairment and fewer functional consequences of the corresponding symptoms or conditions. Alternatively, individuals in these jobs may have better access to health care and fewer untreated symptoms or they may be reluctant to report symptoms that may be attributed to a relatively desirable job.

Few data about the respiratory health of poultry processing workers or rural Latino immigrants exist with which to compare our findings. Cross-shift increases in the prevalence of coughing (35% preshift vs 52% postshift), shortness of breath (0% vs 9%), nasal irritation (9% vs 22%), and runny nose (4% vs 17%) as well as decreases in FEV₁ (− 4.1% predicted) and FVC ((−3.1% predicted) have been reported in a small cohort of poultry slaughterhouse workers,⁵ highlighting the importance of poultry processing exposures and the impact of the timing of symptom and lung function data collection. Thirty-day period prevalences of coughing or wheezing (11%) and shortness of breath (7%) have been reported in a poultry working cohort that included chicken catchers and processing workers.¹⁸ In our data, 6% of the poultry processing population and 5% of the comparison population reported wheezing or whistling in the chest in the last month; these prevalences were both notably lower than that reported among Latino farmworkers (16%),²⁶ whose symptoms may be attributed to the wide range of outdoor exposures such as agricultural pesticides, allergens, and organic dusts encountered in farm work.^3,27–29

In this study, assessment of potential inhalation exposures in the poultry processing plants was based on self-reported job activities. Our understanding of specific work tasks, as described by poultry processing workers in this study and by other accounts in the literature,^14,30,31 enabled us to develop plausible hypotheses about the relationships between poultry processing workplace exposures to respiratory irritants and respiratory symptoms. For example, workers who receive and handle live birds or carcasses likely encounter allergen, bacteria, and dust exposures. Air quality measurements from the breathing zones of slaughterhouse shacklers in Sweden indicated exposures to time-weighted average concentrations of total dust ranging from 0.4 to 15.3 mg/m³ (mean, 6.3 mg/m³), with higher levels of airborne bacteria in the hanging and evisceration departments than in the packaging areas.⁵ As the birds are killed and continue along the assembly line, worker exposures to cold temperatures as well as vapors, gases, dusts, and fumes arising from disinfectants and other cleaning agents, machining fluids, and packaging materials may be expected throughout the plants. In these data, we observed lower metrics of lung function that support one task-specific hypothesis: The use of cleaning agents in the poultry processing affects lung function. This finding supports and extends a growing body of evidence that occupational exposure to cleaning agents with irritant properties affects lung function.^32,33 If our use of self-reported work activities or groupings of activities incorrectly grouped workers with varying degrees of exposure, the resulting misclassification likely limited the ability of our analysis to detect other potential effects. Developing and implementing methods to improve the assessment of exposures, including concentrations, durations, mixtures, and the use of respiratory protection and ventilation, would improve our understanding of the degree to which workers encounter inhalation hazards in various poultry processing jobs.

These data illustrate the importance and feasibility of including spirometry as one component of a large occupational health study. In fact, on the basis of the respiratory symptoms alone, the data may be viewed as suggesting that poultry work is associated with a lower burden of respiratory symptoms, whereas in combination with the available lung function measurements they suggest that poultry processing work may affect lung function. By presenting sex-stratified data, we also show lower lung function measurements, as a percentage of predicted values, among women than among men. This finding would be expected if women in the study experienced other inhalation exposures, such as those encountered in cooking, cleaning, gardening, and other avocational activities that were not accounted for in our data. Sex-based differences in these exposures would result in differential misclassification affecting our analysis of nasal and respiratory symptoms and may explain the lower lung function measures observed among women.

There are few large population-based surveys focused on occupational exposures and respiratory health of the Latino work-force in the United States. In this study, working Latino adults were successfully recruited and enrolled into a research study focused on the health of workers in an industry in which investigators and other public health personnel have limited workplace access. The generally low prevalence of respiratory health symptoms in the poultry processing cohort suggests the influence of the healthy worker effect. If poultry processing jobs are more desirable than other employment opportunities available, then those who leave poultry processing work due to their respiratory health may be at a unique employment disadvantage because of their health. These findings justify efforts to evaluate and monitor the health of new employees and to reduce exposures to inhalation hazards in poultry processing.