Soil contamination poses significant human health risks, particularly for agricultural workers with frequent direct contact. Current exposure assessment methods are inadequate, lacking the detail to capture the variability of soil contact in agricultural settings. Existing activity pattern data, while useful for estimating exposure extent, primarily focus on macro-activities (broad activity categories) and micro-activities (small-scale actions like hand-to-mouth). This study addresses the gap by introducing meso-activities – specific tasks within macro-activities – as an intermediate level of activity pattern data to provide greater resolution in soil exposure estimation. The purpose is to develop a framework that organizes factors modifying soil exposure within each meso-activity, improving the accuracy of exposure assessments and guiding future research. The study's importance lies in the refinement of exposure estimation methods for agricultural workers, which can directly lead to better protection of their health.

Previous research highlights the significant health risks associated with exposure to soil contaminants like pesticides, metals, and PFAS among agricultural workers. However, a lack of precise methodology hinders accurate characterization of this exposure. Studies modeling soil ingestion rates suggest higher rates than those previously estimated. While activity pattern data offer a useful tool for exposure assessment, their application has been limited in the agricultural context. Existing data often focus on macro-activities (broad categories of activity), failing to account for variations in behavior influencing exposure within those activities. Similarly, while micro-activity data (characterizing body part contact with surfaces) are crucial for estimating ingestion and dermal exposures, particularly for children, data for adult agricultural workers are scarce. Therefore, this study proposes meso-activity as a bridging concept between macro and micro-activities, recognizing the need for task-specific nuance in assessing occupational soil exposure.

This qualitative study employed purposive sampling to recruit 16 fruit and vegetable growers in Maryland. In-depth interviews (IDIs) explored typical workdays, farm operations, and specific tasks (meso-activities). Interviews lasted 21-92 minutes, were audio-recorded, transcribed verbatim, and analyzed using NVivo software. A framework approach to qualitative data analysis involved deductive coding (based on the interview guide) followed by inductive coding to identify emergent themes. Six meso-activities (bed preparation, planting, irrigation, harvesting, pest management, and produce handling) emerged as key tasks. Ten factors influencing soil exposure were identified and categorized into four classes within the EAT-R framework: Environmental (natural and social/built factors), Activity (crop type, growing practices, ergonomic positioning), Timing (season, day of week, time of day), and Receptor (biological and behavioral factors of the grower). The study then demonstrated the framework's application by integrating the qualitative factors into quantitative models for ingestion and dermal exposure, adapted from US EPA methods. Sensitivity analyses were conducted, iteratively adding EAT-R factors to examine their influence on exposure estimates, comparing results to traditional models.

The interviews revealed six key meso-activities where agricultural workers interact with soil, along with ten factors influencing the frequency, duration, and intensity of soil contact. These factors were effectively organized within the EAT-R framework. Environmental factors included climate, weather, farm size, technology/facilities, and workforce. Activity factors focused on crop type, growing practices (including the use of tools and mechanization), and the grower's ergonomic positioning. Timing factors considered season, day of week, and time of day. Receptor factors included biological characteristics (age, sex) and behavioral factors (PPE use, handwashing practices, and other habits like smoking or sampling produce). Sensitivity analyses showed that integrating EAT-R factors into traditional exposure models significantly improved the accuracy of estimates. Traditional models underestimated or overestimated exposure (depending on the nature of work). The inclusion of meso-activity factors revealed task-specific variations in exposure. Adding timing factors (seasonality) showcased substantial seasonal variation in soil exposure. Environmental factors (soil moisture impacting ingestion and adherence rates) further refined estimates. Finally, incorporating receptor factors (body weight, attire preferences) highlighted significant inter-individual differences in exposure. These findings illustrate the complexity of agricultural soil exposure and the importance of considering individual meso-activities and their modifying factors.

This study addresses a critical gap in assessing agricultural worker exposure to soil contaminants. The novel meso-activity framework provides a more nuanced and accurate approach than traditional methods relying solely on macro-activities. By integrating qualitative data into quantitative exposure models, the study demonstrates the significant impact of task-specific factors, seasonality, environmental conditions, and individual grower characteristics on exposure. The EAT-R framework is particularly valuable in identifying key factors for intervention strategies and future research. The improved accuracy of exposure estimates contributes to more effective risk assessment and development of targeted interventions to protect agricultural workers' health. The framework's applicability extends beyond fruit and vegetable growers in Maryland and can inform exposure assessments for various other populations with soil contact, including children and construction workers, though adaptations may be needed to reflect the distinct meso-activities in those contexts.

This research introduces a novel meso-activity framework for estimating soil exposure in agriculture. This framework integrates qualitative data on influencing factors (EAT-R) into quantitative models, leading to more accurate exposure estimates. The findings underscore the importance of considering task-specific variations, seasonality, environmental conditions, and individual characteristics. Future research should focus on quantifying the influence of each EAT-R factor, building a database of meso-activity-associated exposure factors, and exploring applications to other populations and exposure scenarios.

This study's qualitative nature limits the ability to definitively quantify the impact of each factor on exposure. The sample size (16 growers) may not be fully representative of all agricultural workers. The generalizability to other agricultural contexts (e.g., different crops, climates) remains to be investigated. Quantitative values assigned to qualitative factors in the sensitivity analyses were based on professional judgment due to limited existing data; further empirical validation is needed.