Behavior-Based Risk Assessment of Indirect Pesticide Exposure Among Older Adults Residing in Agricultural Families in Ubon Ratchathani Province, Thailand

This study utilized a survey research design to evaluate indirect pesticide exposure prevention strategies among older adults living in agricultural families. The data were gathered via a structured assessment questionnaire designed in alignment with the OHSAS 18001 framework, concentrating on household-level activities that affect exposure risk. The survey was administered at the residences of older adults residing in the Hua Ruea Subdistrict of Ubon Ratchathani Province, Thailand, from October 2025 to February 2026.

2.1 Population and sample

The target demographic consisted of older adults aged 60 and above, who were living in agricultural families that were frequently utilizing pesticides in the Hua Ruea Subdistrict of Ubon Ratchathani Province, Thailand. Local administrative records indicated that the studied region had 1,777 older adults across 16 villages, all of which were situated within an agricultural community marked by extensive pesticide use. In each chosen household, the eldest qualifying older adult was appointed as the respondent to ensure uniformity and to reduce intra-household variability.

$\begin{gathered}n=\frac{N Z^2 \alpha / 2[p(1-p)]}{\left[e^2(N-1)\right]+\left[Z^2 \alpha / 2[p(1-p)]\right]} \\ =\frac{1777(1.96)^2[0.589(1-0.589)]}{\left[0.059^2(1777-1)\right]+\left[(1.96)^2[0.59(1-0.59)]\right]}=232\end{gathered}$

The sample size was calculated utilizing the formula for calculating a population percentage, employing n = 232, N = 1777, Zα/2 = standard normal value at confidence level 95% (α = 0.05, Z = 1.96), an estimated proportion of high-level pesticide related behaviors (p = 0.589) derived from existing literature [18], and a margin of error established at 10% of p. Upon applying the finite population correction, the minimum requisite sample size was determined to be 232 participants. To address potential non-response and fieldwork discrepancies, an additional 5% was incorporated, resulting in a final target sample of roughly 250 participants.

Table 1. Number of participants in each village

The stratified random sampling within each village was done according to Table 1 by calculating the population proportion in 16 villages; thereafter, the simple random sampling was applied to collect the data. The participation criteria consisted of the following: (1) being an adult aged 60 years or older, (2) being the eldest of the older adults in the family, (3) being a resident for a minimum of six months, and (4) being capable of communicating and providing informed consent. Individuals exhibiting significant cognitive deficits, communication obstacles, or acute medical conditions that prevented participation were excluded.

2.2 Research tools

Three principal study instruments were utilized to gather data and evaluate indirect pesticide-exposure risks in older adults. All instruments were created in accordance with recognized occupational health and safety standards and conformed to the OHSAS 18001 risk-assessment framework.

1. The General Information Questionnaire: A systematic questionnaire was employed to gather the demographic and health-related data from participants. The tool of 8 items featured multiple-choice questions regarding sex, age, household size, and chronic illnesses, etc.

2. The Hazard Identification Checklist: The checklist functioned as the principal instrument for delineating behavioral routes that enable the transmission of pesticides into residential settings. Each participant was assessed using a 20-item household behavior checklist. Responses indicating compliance with the recommended preventive behavior were classified as safe practices and were not subjected to further risk assessment. Responses indicating non-compliance (unsafe practices) identified from the checklist were considered potential hazards. Each identified hazard was subsequently evaluated using a risk matrix based on the likelihood of occurrence and the severity of potential consequences (Table 2). Risk levels were determined according to the criteria presented in Table 3.

Table 2. The likelihood and severity matrix for hazard evaluation

Table 3. The risk level categorization for indirect pesticide exposure

3. The OHSAS 18001-based Risk Assessment Form: A universal risk assessment form was modified from the Department of Industrial Works (Thailand) to evaluate the risk severity and to prioritize hazardous behaviors, in accordance with the OHSAS 18001 requirements. The instrument evaluated two dimensions: (1) the likelihood of occurrence (four levels: rare to frequent) and (2) the severity of the consequences (four levels: low to extremely high). Risk scores were calculated by multiplying the likelihood and the severity (Table 2). Then the scores were classified into four categories: low (1–2), moderate (3–6), high (8–9), and extremely high (12–16) (Table 3). This facilitated the accurate measurement of the household pesticide exposure hazards and the identification of behaviors requiring prompt intervention.

2.3 Quality assessment of the research tool

The checklist and questionnaire's content validity was assessed by three specialists in occupational health, environmental health, and pesticide safety. Modifications were implemented based on the experts’ suggestions. Furthermore, the Index of Item Objective Congruence (IOC) values varied from 0.67 to 1.00, signifying acceptable validity.

2.4 Data collection and analysis

Data collection was conducted via individual household visits, which were conducted in cooperation with village health volunteers (VHVs), who enabled community access and aided in finding qualifying families. Before the fieldwork commenced, all research assistants underwent uniform training to guarantee the consistent implementation of the survey instruments and compliance with the ethical protocols.

The hazard identification checklist was finalized by conducting participant interviews and by direct observations of households, which allowed the study team to verify the actual practices that were related to pesticide management, contamination routes, and household environmental conditions. Subsequent to the hazard identification process, the OHSAS 18001-compliant risk assessment form was utilized to assess both the likelihood and the severity of the identified hazards. The research team thoroughly evaluated each behavioral component identified as a potential risk by using the established criteria to generate risk scores.

2.5 Data analysis

The data were examined utilizing descriptive statistical methods. Normally distributed variables were presented with means and standard deviations, while non-normally distributed variables were summarized with medians, interquartile ranges, minimums, and maximums.

In the hazard identification checklist, each of the 20 behavioral components was designated as a binary variable (1 = Safe practice observed, 0 = Unsafe practice observed). Frequencies and percentages were computed to determine the prevalence of both safe and risky household behaviors contributing to indirect pesticide exposure among older adults. Risk assessments were conducted by the research team based on both participant interviews and direct household observations. To minimize subjectivity, all identified hazards were reviewed and discussed among the investigators, and final risk scores were assigned through consensus according to the predefined scoring criteria. In addition, the risk assessment instrument and scoring framework were reviewed by experts in occupational health, environmental health, and pesticide safety during the content validation process. Risk scoring was performed using a standardized assessment guide adapted from the Department of Industrial Works (Thailand) and the OHSAS 18001 risk assessment framework. The guide specified operational definitions for each likelihood level (rare, unlikely, occasional, and frequent) and severity level (low, moderate, high, and extremely high).

The behavioral items were subsequently grouped into five indirect pesticide exposure pathways according to the primary contamination route represented by each behavior. This classification was used to facilitate visualization and interpretation of exposure pathways in the Sankey diagram.

2.6 Ethical aspects

The protection of the rights of the volunteers who participated in the project was taken into consideration. On September 11, 2025, the Human Research Ethics Committee of Ubon Ratchathani University gave its approval to this study, which was given the project number: UBU-REC-177/2568.

This study found that older adults living in agricultural families continue to face a risk of indirect pesticide exposure, despite a high incidence of having previously participated in chemical safety training. Despite the fact that compliance with numerous recommended preventive measures had exceeded 90%, several risk behaviors remained, especially those associated with contaminated clothing, proximity to post-spraying activities, and the sharing of household objects. The findings distinctly demonstrated a persistent knowledge-practice gap, a phenomenon that has been consistently documented in recent studies, indicating that information-based training alone is inadequate when seeking to achieve sustained behavioral changes in rural and semi-urban agricultural communities [14, 19, 20].

The Sankey diagram (Figure 1) illustrates that hygiene and sanitation, along with direct contact pathways, had been the primary contributors to home pesticide risk factors, each representing 26.8% of all detected risks. This was succeeded by food and water contamination and storage-related pathways. This pattern robustly corroborates the existing evidence that the take-home transfer of pesticides through contaminated clothing, skin contact, shared fabrics, and insufficiently segregated living spaces constitutes the primary mechanism by which pesticides infiltrate domestic environments [21, 22].

Risk categorization of the domestic activities (Figure 2) demonstrated a significant distinction between the structural restrictions and the behavioral deficiencies. Behaviors requiring physical or infrastructural modifications, such as the absence of designated storage areas for pesticide equipment, public drinking vessels, and inadequate wastewater disposal systems, have led to disproportionately raised and extraordinarily high-risk assessments. These findings corresponded with increasing regional and international evidence that environmental and housing limitations in rural areas can significantly restrict the practicality of safe pesticide handling guidelines, even among informed households [13, 18]. Conversely, the activities that occur more frequently, such as drying clothing indoors, sharing personal goods, and engaging in post-spraying interactions that are in proximity to the farmers, were predominantly situated within the moderate risk category. Yet, these had significantly contributed to overall exposure. This underscored that moderate risk behaviors, when performed daily, may contribute to chronic low-level exposure, which is particularly concerning for older adults with age-related physiological susceptibilities [9, 10].

The bubble chart analysis (Figure 3) further demonstrated an inverse correlation between the frequency of conduct and the severity of risk. Infrequent behaviors, such as inadequate equipment storage or utilizing domestic containers for pesticide formulation, correlated with elevated severity scores, which signified underlying, yet significant vulnerabilities that may be neglected in standard safety trainings. In contrast, there are high-frequency behaviors of moderate intensity, which include the procedures of doing the laundry and having interpersonal interactions immediately after farmers have sprayed. These behaviors indicate entrenched home patterns that are resistant to alteration. Comparable trends have been observed in agricultural communities in Africa, South Asia, and Southeast Asia, where households possessing sufficient pesticide knowledge still encounter difficulties in altering their culturally ingrained practices concerning clothing management, domestic hygiene, and familial interactions after the application of pesticides [16, 20].

Overall, over fifty percent of the detected risks were categorized as being moderate risk, while more than twenty percent were classified as high or extremely high, indicating that household-level exposure pathways represent the primary source of indirect pesticide risk for older adults. The findings aligned with the international literature, which identified contaminated clothes, communal living environments, and inadequately designated storage locations as significant reservoirs of pesticide residues that impact vulnerable communities [21, 22].

A limitation of this study is that the OHSAS 18001 based risk matrix evaluates behavioral and environmental risk factors rather than actual pesticide exposure doses. Therefore, the risk classifications should be interpreted as relative indicators for prioritizing preventive interventions and not as direct measures of toxicological risk. Future studies should combine behavioral assessment with environmental sampling, biomonitoring, or pesticide-exposure models to improve exposure characterization. Furthermore, the risk classification relied on expert judgment using a structured risk matrix rather than direct exposure measurements. Although standardized scoring criteria and consensus review procedures were applied, formal inter-rater reliability testing was not performed. Future studies should evaluate inter-rater agreement and incorporate quantitative exposure assessments to improve the robustness of risk estimation. Moreover, because this study employed a cross-sectional design and relied on behavioral assessments and risk scoring, the findings should not be interpreted as evidence of actual pesticide exposure levels or causal health effects. The identified hazards represent potential exposure pathways that warrant further investigation using environmental monitoring and biomarker-based exposure assessments.