Document resource
Objective Agricultural studies mostly focus on farmers’ exposures, but their family members often conduct similar agricultural tasks. To quantify the endotoxin exposure for spouses of farmers in the Agricultural Health Study (AHS), we modified an algorithm previously developed for AHS farmers in the Biomarkers of Exposure and Effect in Agriculture (BEEA) sub-study.Methods The BEEA algorithm quantified endotoxin exposure for 13 bioaerosol-related agricultural tasks queried via questionnaire. It multiplied task frequency by task-specific intensity derived from meta-analyses of published endotoxin measurements. The AHS enrollment questionnaire completed by 32,345 spouses (>99% female) included 5 of these tasks, three with frequency information (‘number of times’: poultry contact, swine contact, milking cows), and two with yes/no information (grinding feed, harvesting grain). We used the median hours/time reported for each task in BEEA to convert ‘number of times’ to hours and the median task hours to derive hours for yes/no tasks. We used the ratio of the 75th/25th percentile of exposure estimates as a measure of contrast. We calculated Spearman correlations between the task-specific and total estimates.Results Overall, 39% of spouses reported at least one task, with 12% reporting two or more. Swine contact was the most common (27%); milking cows was the least common (3%). We estimated a 90-fold contrast amongst endotoxin-exposed spouses, with 38-, 10-, and 7-fold contrast amongst those exposed to poultry, swine, and milk cows, respectively. The task-specific estimates had low correlation (<0.25) with each other. The ‘total’ (summed) exposure had high correlation with swine contact, moderate correlation with grinding feed and poultry contact, and low correlation with the remaining tasks.Conclusions Although we could not account for all bioaerosol-related tasks, we identified substantial endotoxin exposure variability in the AHS spouses. These estimates will be used to examine exposure-response associations in future etiologic analyses with cancer and other outcomes.