Review of the regulatory management of food allergens

3 Label information

Consideration of this issue has been broken down into two components. The first component relates to the clarity of information presented on the food label and, in particular, whether the source of the allergenic ingredient should be declared on the label. For example, should ‘milk’ be declared as the source of the ingredient ‘casein’ on the label. The second component related to the usefulness and accuracy of ‘may contain’ and similar precautionary labelling statements which refer to the possible inadvertent presence of the allergenic substance in the product. A useful source of information for the review is the recently completed consumer survey on allergen labelling (FSANZ, 2009), which provides an insight into consumer views and behaviour in relation to allergen labelling.

3.1 Source of allergenic ingredient

3.1.1 Australia and New Zealand

Clause 4 of Standard 1.2.3 – Mandatory Warnings and Advisory Statements and Declarations of the Code refers to the mandatory declaration on food labels of certain substances in food. The Table to clause 4 lists those substances that must be declared.

These are: cereals containing gluten and their products, namely, wheat, rye, barley, oats and spelt and their hybridised strains; crustacea and their products; egg and egg products; fish and fish products; milk and milk products; peanuts and soybeans and their products; added sulphites in concentrations of 10mg/kg or more; and tree nuts and sesame seeds and their products. Although not explicitly stated in the text to clause 4, the substances listed in the Table to clause 4 are major food allergens, likely to cause adverse and potentially severe, reactions in some consumers. Clause 4 specifies which substances ‘and their products’ must be declared on a food label. However, it does not regulate the terminology to be used to identify these allergenic foods and their products.
Clause 4 of Standard 1.2.4 – Labelling of Ingredients, includes additional conditions relating to the declaration of some of the substances identified in Standard 1.2.3. Where the cereal is wheat, rye, barley, oats or spelt or their hybridised strains, then the specific name of the cereal must be declared. Where the source of the vegetable oil is peanut, soy bean or sesame, the specific source name must be declared. The specific names of the crustacea and nut present in a product must also be declared.
As a result of the current regulations covering the declaration of the major food allergens, the source of the ingredient declared on the label may not always be clearly stated, for example, ovalbumin, which is a product derived from egg. In addition, current regulations do not adequately meet consumers’ needs for information on certain ingredients that can be derived from a number of different sources, of which not all are allergenic.

An example is lecithin, which can potentially be derived from soy or rice, and maltodextrin, which can potentially be derived from wheat among other sources. In such circumstances, clearly stating the source of the ingredient may provide useful information regarding the substance to consumers who need to avoid the allergen in question in order to prevent life threatening adverse reactions. It may also help consumers avoid unnecessary restrictions in their food selection, where the source of the ingredient is non-allergenic.

3.1.2 International regulations

The Codex Alimentarius
The Codex General Standard for the Labelling of Pre-packaged Foods (CODEX STAN 1-1985) states that ‘the following foods and ingredients are known to cause hypersensitivity and shall always be declared: cereals containing gluten; i.e., wheat, rye, barley, oats, spelt or their hybridized strains and products of these; crustacea and products of these; eggs and egg products; fish and fish products; peanuts, soybeans and products of these; milk and milk products (lactose included); tree nuts and nut products; and sulphite in concentrations of 10 mg/kg or more.’ However, there is no prescriptive requirement for how to name the food source from which the allergenic ingredient is derived.
The European Union
The European Union (EU) directive (EU Directive 2003/89/EC amending 2000/13/EC) makes it mandatory for the food industry to list 12 potential food allergens on the product labels regardless of the quantity in the finished product. The regulations state: ‘The list of allergenic substances should include those foodstuffs, ingredients and other substances recognised as causing hypersensitivity’. Again, there is no specific requirement as to the terminology to be used to declare the substance.
The United States of America
In the United States, the Food Allergen Labelling and Consumer Protection Act 2004 (FALCPA) of the US Food and Drug Administration (USFDA) mandates manufacturer disclosure of the most common allergens (milk, egg, wheat, soy, peanut, tree nuts, fish and crustacean shellfish) in plain English, using the common or usual name, in the ingredient list or in a separate allergens summary statement. The name of the food source from which the allergenic substance is derived is required e.g. ‘milk casein’. In addition, FALCPA mandates the disclosure of the type of tree nut, fish or crustacean shellfish.

A recent audit of manufactured products (Pieretti, 2009) for use of allergen labelling statements identified only a very small percentage of products with FALCPA violations involving the use of non-food source terms, for example, ‘whey’ without the term milk, ‘durum flour’ without the term ‘wheat’.

From the results of this audit, it appears that regulating for the declaration of the source of allergen has been a successful means of ensuring that food manufacturers provide adequate information on the label about allergenic ingredients. Given the high level of compliance in the US (and given current labelling practices in Australia which indicate that manufacturers are already meeting these requirements, as discussed further on), results may also suggest that should such a regulatory approach be implemented by FSANZ, it may not pose undue difficulties for manufacturers to implement.

Pieretti (2009) identified several labelling ambiguities which may present areas of potential confusion for the individual with food allergy. One of the main issues, that has already been outlined in 3.1.1 above, involved the declaration of ingredients that may have allergenic or non-allergenic sources. Given that non-allergenic sources are not required to be qualified in Australia and New Zealand this requires a relatively sophisticated understanding of the labelling requirements by consumers.

3.1.3 Outcome of stakeholder consultation

In March 2008, FSANZ released an issues paper for targeted stakeholder consultation. The paper outlined the issues that FSANZ intended to cover in the review, sought comment on these and requested information to inform the review. Submissions were received from approximately 20 key stakeholders representing jurisdictions, the food industry, consumer support groups and health professionals. Responses to the issues paper revealed general support for the need for clarification of terms in the Code. In particular, there was support for the source of ingredients to be declared, with stakeholders proposing that this would involve the replacement of terms such as ovalbumin with egg and casein/ whey with milk. The comment was also made that they understood that the original intent of the mandatory declaration of allergens under clause 4 of Standard 1.2.3 was for common names to be used. However, the drafting of the current standard had left this aspect open to interpretation.

There was also support for the need to clarify the source of ingredients that may have been derived from alternative sources, not all of which are allergenic, for example lecithin (as previously mentioned) and thickener, (which can be derived from wheat, or from other non-allergenic sources such as maize).

3.1.4 Consumer research

Two surveys examining consumers’ perspectives in relation to allergen labelling have been undertaken by FSANZ as part of the FSANZ Evaluation Strategy.

The first survey, a benchmark study conducted in 2003, collected baseline data on consumers’ views and behaviours towards the then newly introduced food allergen labelling requirements (FSANZ, 2004). A follow-on survey was conducted in 2008, largely replicating the measures from the benchmark survey (FSANZ, 2009). The aim of the follow-on study was to provide an indication of the current situation as well as a comparison to practices in 2003 when the changes to allergen labelling provisions were only just being implemented.
The benchmark survey found that many people with food allergies, when presented with a list of substances, did not recognise all of the terms used to describe the allergenic substances. For example, a large percentage of those with a wheat allergy did not recognise that thickener, semolina, couscous, cornflour, starch, icing sugar mix, textured vegetable protein and maltodextrin could all potentially contain wheat.
The follow-on survey noted that the rate of recognition of many ingredients and products has increased in 2008 compared with 2003. However, a number of ingredients are poorly recognised when the source is not included. For example, among milk allergic individuals 81% identified lactose, 76% identified butterfat, 73% identified casein and 71% identified whey as words that indicated the presence of ingredients of concern to them. Although the risk to the consumer varies according to their sensitivity and the allergenic content of the ingredient, and therefore personal judgement and experience play a major role in decision making, clearly some consumers are unable to identify ingredients of concern.

The industry guideline on allergen labelling, discussed below, recommends the use of clear and easy to understand terms in association with the ingredient.

Overall, there were improvements in the area of label clarity in 2008 compared with 2003. However, results indicate potential risk to allergic consumers who do not recognise ingredients of concern. Less than half of the respondents agreed that it was easy to understand and use food labels.
In particular, respondents reported that the use of many or different names for the same ingredient was a problem for them, making it difficult to select appropriate food products and avoid the allergens of concern. This was reflected in their suggested improvements to labelling, where comments centred around the need to be more specific as to the types and source of ingredients such as vegetable oil, emulsifiers, thickeners etc. Similar issues were identified in a recent European study examining food allergic consumers’ preferences for labelling practices. The study, which included 40 participants, reported that the ingredient list was considered incomplete and the information not sufficiently specific (Voordouw et al., 2009).
It should be noted that despite the call for further clarity and detail in the ingredient information provided on labels, there does not appear to be any evidence of this as a cause of allergic reactions. In response to an open-ended question, 5% of respondents reporting a serious reaction since the allergy was first identified said this was due to unlabelled or incorrectly labelled food. Consumer comments did not indicate any specific deficiencies in relation to the clarity of information as being a cause of a repeat allergic reaction. However in response to a separate question, lack of clarity of information about the allergenic ingredient was not reported as being a cause of a repeat allergic reaction.

3.1.5 Food industry initiatives

In 2005, the Allergen Bureau¹ was established as an industry funded resource providing information and tools to improve awareness and skills in relation to allergens in the food manufacturing environment. The Australian Food and Grocery Council (AFGC) published the ‘Food industry guide to allergen management and labelling (revised edition)’ in 2007, providing guidelines on the management of allergens in the food processing environment and recommended labelling formats (AFGC, 2007). The aim of the industry guide is to promote the declaration of allergen information on food labels in a clear and consistent manner to enable food allergic consumers and their carers to easily determine the suitability of particular foods. The guide recommends that:

• 
  all allergen information should be grouped together to be easily identified and not hidden amongst other labelling information
  
• 
  product description and representation should provide an accurate expectation of the product and should not be misleading
  
• 
  allergens must be declared using plain English terms consistent with the Code
  
• 
  the print size should be big enough to be easily read, preferably at a minimum 1.5mm with sans serif font, and the font colour should contrast distinctly from the background.
  

• 
  the use of lower or upper case will depend on the overall presentation of labelling information.
  

The recommended labelling format consists of an ingredient list declaring allergenic ingredients and their derivatives in bold, an allergen summary statement, and a precautionary statement. The ingredient/ component should be qualified according to the allergenic foods listed in the Table to clause 4 of Standard 1.2.3, either in the ingredient list itself or in the allergen summary statement. This is to ensure that the allergenic ingredients are clear to the sensitive consumer, through use of the source name. Implementation of the guide by food manufacturers is voluntary.

3.1.6 Label monitoring surveys

Label survey information that gives an indication of the state of allergen labelling is useful to complement the information gathered in consumer surveys on how consumers are perceiving allergen labelling, its usefulness and limitations. Such information is available from various sources including label monitoring surveys conducted by FSANZ and the food industry.

3.1.7 FSANZ label monitoring surveys

Food label monitoring surveys have been undertaken by FSANZ since 2002. These surveys provide an indication of current allergen labelling practices, the results of which can complement the consumer information gathered.

Food label monitoring surveys were undertaken in 2002/2003 (Phase 1) and again in 2005/2006 (Phase 2). Surveys were conducted to assess how food manufacturers were managing key labelling requirements. In any one year of the survey, between 1200-1300 food labels were collected from 14 different food categories in Australia and New Zealand and assessed for consistency with the Code.
Phase 1 was undertaken during a period of transition to the new Code and shortly thereafter. Phase 2 was undertaken when changes to the allergen labelling provisions had been in full effect for several years. Each label was assessed against twelve key labelling elements, including allergen labelling.
In Phase 1, an assessment of allergen labelling covered legibility only. All labels assessed in Phase 1 were assessed as legible and therefore consistent with the requirements of the Code (FSANZ, 2004). In Phase 2, the methodology was altered to include an assessment of whether ingredients were accurately qualified, for example, flour qualified with the cereal type (FSANZ, 2008). In Phase 2, a small number of labels were assessed as inconsistent with this criterion, with the most common inconsistency being failure to qualify the cereal/ flour with the cereal type.
The results of Phase 1 indicate that the most common method of declaring the presence of allergens was in the ingredient list, rather than elsewhere on the label, and using the common/source name rather than a technical term. This apparent trend continued in Phase 2, with virtually all labels declaring the presence of allergens in the ingredient list. Use of common/source names was again high.

3.1.8 FSANZ allergen label monitoring survey

Information on the extent of application of the industry guide to allergen management by the food industry would be useful to complement information gathered by FSANZ through the food label monitoring surveys.

In 2009, FSANZ undertook a label monitoring survey focusing on allergen declaration. The survey analysed 182 labels containing allergen declarations, to assess allergen label information on products in Australia against the allergen labelling provisions in the Code as well as the labelling recommendations set out in the industry guide. As in Phase 2 of the label monitoring survey discussed above, (99%) of these products were assessed as consistent with the Code’s Standards 1.2.3 and 1.2.4. Only two cases of allergen labelling were found to be inconsistent with the Code, due to failure to qualify the flour with the cereal type. Thirty-four percent of labels were assessed as consistent with all of the labelling recommendations set out in the industry guide.
As noted in previous surveys, the most common method of declaring the presence of allergens was in the ingredient list, rather than elsewhere on the label. In some cases the ingredient was declared, for example, ‘casein’, but in direct association with the source (e.g. ‘milk casein’ or ‘casein (milk)’), or else the ingredient was declared in the ingredient list, with the source declared in a separate summary statement (e.g. ‘this product contains milk’).
Less than one in ten labels did not clearly state the source/ common name, either in the ingredient list or in a separate allergy statement. These labels included the terms ‘cheese’, ‘cream’ and ‘butter’ (instead of ‘milk’), ‘gluten’, ‘cereal/ flour’ (instead of the specific cereal such as ‘wheat’), ‘nut’ (instead of ‘tree-nut’ or ‘peanut’), ‘salmon’ (instead of ‘fish’), and ‘shrimp’ (instead of ‘crustacea’).

3.1.9 Allergen Bureau labelling review survey 2009

An allergen labelling survey of 340 packaged retail food products with allergen declarations was carried out by the Allergen Bureau in 2009. This survey followed on from an earlier survey of 213 labels also undertaken by the industry in 2005.

The aim of the second survey was to provide information on how allergen related information is currently being declared on food labels, compared to the recommended labelling formats set out in the industry guide (Allergen Bureau, 2009).
Reflecting the results of FSANZ’s surveys, the most common method of declaring the presence of allergens was in the ingredient list, rather than in a summary statement or as a separate claim elsewhere on the label. As previously stated, the industry guide recommends that the ingredient/ component be qualified according to the allergenic foods listed in the Table to clause 4 of Standard 1.2.3. An assessment of this aspect would provide an indication of the extent of use of the source name (as opposed to ingredient names or technical names). This aspect was not measured directly, but the use of plain English terminology across the two surveys, was around 96% of labels that had allergens declared in the ingredient list. From these results, the authors contend that this aspect is being addressed adequately by the majority of food manufacturers surveyed.
Together with the results of the recent label monitoring surveys conducted by FSANZ, these results support the view that food manufacturers are making efforts to adhere to industry guidelines, in declaring the source of the ingredient, either in the ingredient list or in a separate allergen summary statement. As such, the need for added regulation in this area would not appear to be imperative.