“Because the MRL is agronomic in origin, it is not a direct line between a food being safe or unsafe for consumption…”
Adriana Torres de Sousa is the Risk Monitoring and Assessment Manager at the General Toxicology Management Division of Anvisa (Brazil’s Health Regulatory Agency). She oversees non-dietary and dietary risk assessment activities, including establishing Maximum Residue Limits (MRLs) for pesticides in food. She also manages toxicological re-evaluations, the Pesticide Residue Analysis Program in Food (PARA)—which monitors pesticide residues in food sold in Brazil—and post-registration pesticide petitions submitted for toxicological evaluation.
Sousa holds a bachelor’s degree and teaching certification in Chemistry, as well as a master’s degree in Inorganic Chemistry from the University of Brasília (UnB). She is a specialist in Toxicology Applied to Health Surveillance from the State University of Londrina (UEL).
Adriana Torres, Risk Monitoring and Assessment Manager at ANVISA.
AgriBrasilis – Is the presence of banned or excessive pesticides in food a recurring issue in Brazil?
Adriana Torres – The detection of pesticides authorized in the country but not approved for a specific crop typically represents the largest percentage of non-conformities identified. In the 2024 cycle, the most recent released, the majority of non-conformities (14.6%) involved the use of pesticides not permitted for the specific crop, while 7.9% showed residue concentrations above the MRL. A single sample may contain more than one type of non-conformity.
The presence of active ingredients banned in Brazil was low, occurring in 0.10% of analyzed samples (paraquat in one soybean sample and carbofuran in two orange samples).
AgriBrasilis – When a food exceeds the residue limit, what does that mean for the consumer?
Adriana Torres – If a food contains residue above the MRL, there is a non-conformity, but the consumer is not necessarily at risk. Because the MRL has agronomic origins, it is not a direct line between safe and unsafe food for consumption. The MRL value is derived from supervised field studies that simulate the required use of the pesticide on the crop.
To determine risks to the consumer, a specific technical-scientific procedure must be carried out: dietary risk assessment. This evaluation considers, in addition to residue results, food consumption data, body weight, toxicological parameters and other factors.
Nevertheless, production that does not comply with approved recommendations constitutes an administrative and sanitary violation and may result in the destruction of food with residues above permitted levels. Non-conformities may also pose health risks to farmworkers due to pesticide application that deviates from—or lacks—authorized use recommendations.
AgriBrasilis – How is enforcement carried out and how are those responsible traced?
Adriana Torres – Enforcement involves shared responsibilities among different federal, state and municipal agencies, as provided by current legislation.
PARA monitoring is advisory in nature and is not classified as an enforcement action. The results guide the actions of regulatory bodies and production chains regarding non-conformities, encouraging the adoption of Good Agricultural Practices (GAPs) and interventions at critical points in the production process.
In this context, food traceability is one of the main tools for managing risks and ensuring the quality of products offered to the public. Joint Normative Instruction ANVISA-MAPA No. 2/2018 organizes and assigns responsibility to all entities in the fruit and vegetable supply chains, with a focus on pesticide residues, helping to identify the origin of non-conforming products and promote more effective action.
Traceability enforcement is conducted by MAPA (Ministry of Agriculture) and local health surveillance agencies, as a decentralized action delegated to states and municipalities.
AgriBrasilis – How does Anvisa define Maximum Residue Limits and measure dietary risk?
Adriana Torres – The MRL is defined based on results from field studies simulating pesticide use on the crop according to the GAP required in the registration process, aiming for the lowest possible residue level that still ensures agricultural efficacy.
To ensure the safety of the MRL, Anvisa conducts a dietary risk assessment prior to pesticide approval, following guidelines from the World Health Organization (WHO) and the Food and Agriculture Organization of the United Nations (FAO).
This procedure evaluates both acute and chronic exposure for each active ingredient. Acute, or short-term, exposure is based on the consumption of large quantities of a specific food over a short period, typically one day or one meal. Chronic, or long-term, exposure estimates lifetime exposure through the daily consumption of various foods containing pesticide residues.
In calculating exposure, the values found in field studies form the calculation basis along with body weight and food consumption data for the Brazilian population, extracted from the POF/IBGE (Household Budget Survey).
Consumption of food containing residues at levels represented by the MRL must not be unsafe for consumers. Therefore, if the calculated exposure exceeds the short-term and long-term toxicological reference parameters, an unacceptable health risk is identified and the MRL is not approved.
AgriBrasilis – What criteria define which foods are monitored?
Adriana Torres – Food selection in PARA prioritizes plant-based foods most consumed by the Brazilian population, according to data from the IBGE Household Budget Survey (POF/IBGE). In addition to consumption, Anvisa considers the history of foods with higher risk indices and non-conformities, as well as product availability in each state.
Thus, 36 foods were selected, representing 80% of national vegetable consumption.
AgriBrasilis – How do data from the Pesticide Residue Analysis Program influence regulatory decisions?
Adriana Torres – PARA data support toxicological re-evaluations, informing regulatory decisions such as use restrictions, bans and others. Examples include the bans on carbofuran and carbendazim.
The data also support the exclusion of uses for crops where risks have been identified based on PARA results—for example, removing citrus from the monograph for the active ingredient methidathion and excluding strawberry and grape crops from the formetanate monograph, among others.
PARA also revealed unauthorized pesticides being used on crops with low registration demand from companies—known as Crops with Insufficient Phytosanitary Support (CSFI). In response, regulatory agencies published Joint Normative Instruction No. 1/2014, establishing guidelines to encourage registration for these crops. Since then, more than 3,000 new limits have been set, regularizing use and promoting less toxic options for health.
READ MORE: