Margin of Exposure

The MOE is an approach used by risk assessors to assess possible safety concerns arising from the presence of substances in food and feed when it is not appropriate or possible to establish a HBGV. The MOE is a ratio of:

• the dose derived from an experimental or observational dose–response relationship for a critical toxicological effect (called ‘reference point’ or ‘point of departure’), and
• the estimated human exposure to the substance.

Importantly, the output of the MOE approach is not a ‘health-based guidance value’, i.e. it is not a safety threshold below which risk assessors conclude that the intake is safe. When there is evidence of harmful effects but not enough to confirm how much is safe, the MOE tells us if current intakes are likely to be harmful or not: a low MOE represents a greater risk than a higher MOE. Based on this, risk managers can consider if action is needed.

Once the reference point /exposure ratio has been calculated, the result indicates the level of concern associated with the exposure to the substance. The minimum magnitude of the MOE needed for concluding that the actual human exposure is of low toxicological concern differs for the two different categories of substances.

• Substances that are both genotoxic and carcinogenic

EFSA’s Scientific Committee published a scientific opinion in 2005 recommending the use of the MOE approach for the assessment of substances that are both genotoxic and carcinogenic The Scientific Committee stated that an MOE of 10,000 or higher, if it is based on an animal study, would be of low concern from a public health point of view and might reasonably be considered a low priority for risk management actions. The Scientific Committee stated that while such a judgment is ultimately a matter for the risk managers, an MOE of 10,000 or more should not preclude the application of risk management measures to reduce human exposure.

• All other substances (that are not ‘both genotoxic and carcinogenic')

In this category, uncertainty about the effects of these substances does not allow establishing a HBGV - the minimum magnitude of an MOE for these substances is usually 100 or larger, i.e. a ratio that is below 100 would be considered a concern for public health.

The figure of 100 is based on long-standing principles in toxicological risk assessment about how to account for uncertainties and differences among and within species . An uncertainty factor (sometimes also called “safety factor”) of 10 accounts for differences between humans and the animals used in experimental studies, then an additional factor of 10 accounts for the variability among humans. Together this results in a minimum factor of 100-fold for substances that are not genotoxic nor carcinogenic.

The 100-fold factor used for substances that are not genotoxic carcinogens is multiplied by an additional factor of 10 to account for differences in the ability of human cells to repair DNA as this influences the carcinogenic process. Then, to account for any other uncertainties in the assessment of genotoxic carcinogens, an additional factor of 10 is added, resulting in the MOE of 10,000, i.e. 100 x 100.

If there are uncertainties arising from gaps in the toxicological data, these need to be embedded in the MOE to conclude on the potential concerns. Therefore, the factor of 100 needs to be multiplied by an additional factor considered adequate to account for these additional uncertainties (often between 2 and 5, resulting in a minimum magnitude of an MOE of 200 to 500).

Alternatively, it is also good practice to set a different minimum MOE if uncertainties about one or more of the conditions described in the answers to question 2 and question 3 can be reduced. This may happen, if there is data to account for differences between humans and experimental animals, or for differences among humans. For example, experts assessing the safety of sulfites used as food additives were able to set an MOE of 80 because they had data on how humans respond differently to sulfites after ingestion.

Substances that are both genotoxic and carcinogenic must not be added intentionally to food. However, their presence in food and feed, while not desirable, is still possible for different reasons:

• their natural presence, e.g. Allylbenzene (methyleugenol/estragol) in basil or pyrrolizidine alkaloids (retrosine/Intermedine) in honey
• from microbial activities, e.g. mycotoxins, such as ochratoxin A or aflatoxins, produced by mould that may contaminate various crops
• from environmental pollution, e.g. benzo(a)pyrene resulting from incomplete combustion of organic matter
• from the unintended consequences of cooking or other manufacturing processes, e.g. acrylamide naturally forms in starchy food products during high-temperature cooking.

They may also be present in food and feed at very low levels resulting from impurities in substances used in manufacturing processes. EFSA’s Scientific Committee advised in a statement in 2012 that the MOE could be useful in such cases to support risk managers in defining possible actions required to keep exposure to such impurities as low as possible.

No. The ‘margin of safety’ and ‘margin of exposure’ are two different concepts that are sometimes mistakenly used interchangeably. 

The margin of safety (MOS) is the ratio between a safe threshold (i.e. the ‘health-based guidance value’ such as an ADI ) and the actual or estimated exposure.

Due to inconsistencies in the use of these two terms, EFSA's Scientific Committee clarified in a 2025 statement that from then on only the term margin of exposure should be used in EFSA scientific assessments both for human and animal health. The term margin of safety should no longer be used.

Before 2005, the general advice of risk managers regarding unavoidable substances in food that are both genotoxic and carcinogenic was to reduce exposure to such substances to a level that is ‘as low as reasonably achievable’ (known as the ALARA principle) irrespective of the nature of the carcinogen. This ALARA principle does not allow setting priorities for management action according to the potency of the carcinogenic substances under consideration.

The approach developed by EFSA for use of the MOE for substances that are both genotoxic and carcinogenic aimed at informing risk managers about the level of toxicological concern coming from a given human exposure to such substances and allowing for setting priorities for management action. EFSA’s Scientific Committee is made up of highly experienced scientists from across Europe, who have published widely in scientific literature and are well placed to consider new methodologies for risk assessment. The Scientific Committee recommended using the MOE approach for these substances in 2005, and subsequently again for genotoxic/carcinogenic impurities present in food in 2012.

In 2024 EFSA’s Scientific Committee kicked off the revision of its 2005 guidance on the MOE approach with completion expected by the end of 2027.

In November 2005, a joint EFSA/WHO conference reached similar overall conclusions that the MOE approach was a useful and pragmatic option for the risk assessment of substances that are both genotoxic and carcinogenic and had the potential to improve the advice provided to risk managers.

Consequently, expert committees of the United Nations Food and Agriculture Organization (FAO) and World Health Organization (WHO), as well as other UN bodies (e.g. the United Nations Environment Programme) have used the margin of exposure approach. The Joint FAO/WHO Expert Committee on Food Additives (JECFA) routinely uses the MOE approach for assessing the risk of genotoxic and carcinogenic contaminants in foods.