J.R. Bell and P.N. Gillatt
Dr Bell is Head of Food Science Division II in the United Kingdom Ministry of Agriculture, Fisheries and Food and Chairman of the Codex Committee on Fats and Oils (CCFO). Mr Gillatt is the Scientific Secretary to CCFO.
Authenticity of vegetable oils
Useful purity criteria
Codex standards to combat fraud
Future developments
Conclusion
References
Ensuring the authenticity of food has been a problem for millennia. Wherever there is a commodity that commands a premium in the market and has either high value or high-volume sales, some people may be tempted to profit from illegal activity. Food fraud usually involves misleading the purchaser as to the true nature, substance or quality of the goods demanded; thus, food standards and labelling are breached. The offence can take the form of adulteration, which generally involves the dilution of a commodity with less expensive materials. A cheaper food may also be represented as if it were a food of greater value.
Food authenticity problems can create enormous harm in the marketplace. While public health problems are uncommon, consumers are defrauded when premium prices are demanded for goods that are of lower value. Future sales of a product may be damaged when consumers believe that the "quality" product is not worth the extra cost. In addition, market competition is distorted since an honest trader cannot easily match the unrealistic prices that perpetrators of fraud can charge.
Food analysis has improved dramatically and more overt types of adulteration or misrepresentation are now unlikely to escape detection. However, more subtle practices emerge as the unscrupulous seek to ensure future dishonest profits. When a new method is devised, a database must be developed that shows the typical values for a particular parameter when authentic samples are analysed. The food in question is compared to this database to see if it is what it is claimed to be.
Often legislation is deemed necessary to ensure that consumers and fair trade are adequately protected. Such laws have often enshrined a "database", indicating the compositional requirements for a food labelled in a given way. Where goods are traded between countries, there may be a need for international action such as law or standards defining foodstuffs.
The vegetable-oil sector encompasses products of differing value and sales volume. In theory, fraud is attractive and international standards in this area are needed. Standards facilitate trade by establishing a baseline for product quality that is internationally agreed. They enable disputes to be settled and help to protect the purchaser. With the recent General Agreement on Tariffs and Trade (GATT) Agreement on Technical Barriers to Trade, the significance of internationally agreed food standards will increase substantially.
In some cases, the parameters of standards for oil are straightforward and concern the product's quality rather than its character. Contents of moisture, impurities and free fatty acids and peroxide value are examples of such parameters. The limits established need to reflect whether the oil is crude, partly refined or fully refined. Consideration must be given to the concentrations of trace and heavy metals, as the first can give rise to rapid lipid oxidation even if present in relatively small quantities and the second may be of public health concern.
The purpose of this paper, however, is to show how the standards of the Codex Alimentarius Commission (CAC) for fats and oils can help to verify food authenticity. It is therefore concerned with those parameters that can characterize an oil. Vegetable oils and olive oils are considered in particular, since they provide an illustration of the problems that can be encountered and the evolution of standards to tackle them. For vegetable oils, adulteration may comprise the dilution of expensive oils with cheaper oils. In the case of olive oil, the problem may be the addition of lower-value grades of olive oil to those (such as extra-virgin olive oil) that command a premium in the marketplace. In each case, changes in Codex standards over time and expected future developments are discussed.
In past incidents of oil adulteration, palm stearin and olein (fractions of palm oil) have been blended together or with palm oil to give a product of unknown and variable quality; groundnut oils have contained undeclared small quantities (5 percent) of cheaper oils such as soybean oil; rapeseed oil has been added, undeclared, to the more expensive soybean oil and maize (corn) oil; and cottonseed oil has been diluted with palm olein. Historically, detecting such fraud has been difficult because of the small database establishing appropriate purity criteria for authentic edible oils and fats.
Olive oils
Olive oils are distinct from other vegetable oils because they may be consumed without extensive refining. Indeed, refining the product is considered detrimental since the attractive organoleptic properties of oil are diminished in the process. Virgin olive oils attract a higher price than refined olive oils because of their pleasant, rather delicate flavour and aroma and limited production volume. As a result, olive oils are subject to two types of deliberate adulteration. The first is the blending of virgin olive oils with olive oils of lower grade (e.g. refined olive oil or olive-pomace oil). The second is the less subtle mixing of olive oil with liquid vegetable oils. There has been considerable concern regarding the deliberate mislabelling of olive oils since a researcher revealed that most of the olive oils studied were not of the grade claimed (Firestone et al., 1985). As a consequence of such findings the International Olive Oil Council (IOOC) (1993a) and the Codex Alimentarius Commission (1993) have produced standards for virgin and refined olive and olive-pomace oils and certain blends of these products.
Palm oil is a semi-solid oil at ambient temperatures in northern Europe. This is explained by its fatty acid composition (FAC): the oil contains large quantities (40 to 47 percent) of palmitic acid (C 16:0) and similar amounts of oleic acid (C18:1). The limited presence of polyunsaturated fatty acids permits close packing of the component triglycérides which results in a relatively high slip melting point.
Palm oil can be adulterated with other palm products, and in these cases the determination of fatty acid, sterol and tocopherol compositions is of limited use. However, plotting iodine value against slip melting point produces three distinct groupings for palm oil and its fractions palm stearin and olein, thereby providing a method of identifying impure oils (Tan and Flingoh, 1981).
Palm stearin has high concentrations of tripalmitin; palm oil contains moderate levels but palm olein has almost none. Accordingly, by determining the triglycéride carbon number, the presence of palm oil or palm stearin in palm olein can be detected (Turrell and Whitehead, 1990).
Double-zero ("00") rapeseed oils, which have become more common since the 1980s, are low in erucic acid, come from seeds low in glucosinolates and have a higher linolenic acid (C18:3) content (10 to 14 percent) than the previously predominant single-zero ("0") rapeseed oils (approximately 8 to 10 percent linolenic acid). The "00" oils are therefore significantly more prone to oxidation and consequently might be considered of lower quality.
Rapeseed (low erucic acid) and soybean oils have similar FAC; however, the differences are sufficient to permit identification of blending. For example, low-erucic acid rapeseed oil contains 52 to 67 percent oleic acid and 16 to 25 percent linoleic acid. Soybean oil contains approximately 18 to 26 percent oleic acid and 50 to 57 percent linoleic acid.
Perhaps the most straightforward method of distinguishing the two oils is by their sterol composition. For example, soybean oil has much higher concentrations of stigmasterol than rapeseed oil. More important, rapeseed oil has high levels of brassicasterol, while soybean oil is almost devoid of this compound.
It has been established that the FAC at the 2-position of the triglycérides is significantly different for these oils. In particular, linolenic acid enrichment factor (EF) is approximately 1,75 for rapeseed oil and 0.9 for soybean oil (Rossell, 1991).
Cottonseed, groundnut and sunflower-seed oils are characterized by their lack of linolenic acid (C 18:3) and by the fact that the sum of their oleic and linoleic acid concentrations is approximately 70 to 80 percent of the total. The presence of more than 0.5 percent linolenic acid indicates adulteration, possibly by soybean or rapeseed oil, both of which trade at lower prices than these oils.
Most oils do not contain tocotrienols, the most notable exception being palm oil. When there are tocotrienols in a cottonseed oil the presence of palm oil or a palm oil fraction is indicated. Groundnut and rapeseed oils generally do not contain d -tocopherol, while soybean oil is rich in this compound. If, following measurement of FAC, a groundnut oil appears to contain an adulterant, the determination of d -tocopherol will verify whether or not it is soybean oil.
The mean ratio of a - to d -tocopherol in sunflower-seed oil is approximately 200. In cottonseed, groundnut and maize oils the corresponding ratio is less than 1. Therefore, one might suspect a sunflower-seed oil of containing one of these oils if this ratio is significantly reduced below 200. The use of FAC alone would not permit this detection.
Palm kernel and coconut oils are rich in lauric acid and the short-chain fatty acids C6:0, C8:0 and C 10:0. There are only small differences in the FAC of the two products, so identifying the oil from its FAC is most difficult. However, analysis of the triglycéride carbon number provides a method of categorical differentiation. The carbon number is renormalized to take account of only those triglycérides with carbon numbers between 32 and 42. When the sum of the concentrations of the triglycérides with carbon numbers of 34 and 40 is plotted against the sum of those with carbon numbers of 36 and 38, it is quite apparent whether the oil is coconut or palm kernel oil (King and Zilka, 1986; King, Zilka and Turrell, 1985).
Purity criteria for olive oils
Because of the differences between olive and other vegetable oils, parameters that serve as quality indices only in the case of a particular vegetable oil can be used to help distinguish the different quality grades of olive oil.
Free fatty acid content (FFA). The FFA of an olive oil can be used as a measure of its quality. In refined vegetable oils, the lower the FFA the more acceptable the oil to the human palate. Virgin olive oils cannot be categorized in this way because their characteristic flavour involves a sharp component that results from the FFA present. Therefore, a higher FFA is acceptable in virgin olive oil than would be permitted in refined vegetable oils. IOOC and CAC standards set a limit of 1 percent for FFA in extra-virgin olive oils.
Peroxide value (PV). The PV is an indication of the amount of hydroperoxides present in an oil. These compounds arise from lipid oxidation; therefore, the PV, expressed as milliequivalent oxygen per kilogram oil (meq/kg), is a measure of oil quality. The PV is greatly reduced by the refining process used for most vegetable oils. Virgin olive oils are not exposed to such processes and the PVs permitted in these products are considerably higher. The IOOC and CAC standards permit extra-virgin olive oils to have PVs of up to 20 meq/kg, while pure olive oils, which by definition are blends of virgin and refined olive oils, must have PVs below 10 meq/kg.
Specific extinction (SE). This is a simple and rapid indicator for establishing whether oils labelled as virgin contain refined oils. The SE of an oil will increase on refining because the usual 1:4 méthylène interrupted distribution of double bonds found in linoleic and linolenic acids is altered in part to form a conjugated 1:3 distribution. If the oil has an SE greater than 0.25 at 270 nm, it is considered either to be not virgin or to contain oxidized fatty acids. If an oil is considered suspect, it is treated with alumina, which removes oxidation products, and the SE is determined again. If, after treatment with alumina, the SE is greater than 0.10, the oil is considered to be adulterated with refined oils.
Fatty acid composition. The use of FAC to establish the purity of an olive oil has been criticized because of the very large variations permitted in standards for certain fatty acid ranges (see Table 1). Analysis of the FAC at the 2-position of the triglycéride, a technique used to indicate esterification, is particularly relevant for olive oils. For example, in poor-quality olive oils, fatty acids may be removed from the glycerol by hydrolysis. These can be recovered during oil refining and recombined with glycerol to produce an oil of the same overall FAC but having a different distribution of fatty acids within the triglycérides. In such a case, the esterified oils have a considerably higher concentration of palmitic (C 16:0) and stearic (C 18:0) acids at the 2-position (see Table 2).
This phenomenon has been exploited in setting standards to prevent the addition of esterified oils to virgin oils. For example, the CAC and IOOC standards state that the sum of palmitic and stearic acids at the 2-position should not exceed 1,5 percent in virgin oil or 1.8 percent in refined oils.
TABLE 1 - Partial fatty acid composition of olive oil - Teneur partielle en acides gras de l'huile d'olive - Composición parcial de los ácidos grasos del aceite de oliva
Fatty acid |
Permitted range |
C16:0 |
7.5-20 |
C18:1 |
55-83 |
C18:2 |
3.5-21 |
TABLE 2 - Typical profiles of fatty acids at the triglyceride 2-position for olive and olive-derived oils - Profils typiques des acides gras à la position triglycéride 2 pour les huiles d'olive et les produits dérivés - Perfiles típicos de los ácidos grasos en la posición 2 del triglicérido, en los aceites de oliva y derivados
Fatty acid |
Virgin |
Esterified |
Pomace |
C16:0 |
0.7-1.6 |
9.0 |
2.0 |
C16:1 |
0.5 |
1.1 |
- |
C18:0 |
Trace-0.3 |
2.6 |
0.6 |
C18:1 |
85.5-89.1 |
75.6 |
- |
C18:2 |
1.3-11.7 |
11.6 |
- |
Sterol composition. This particularly useful parameter identifies adulteration with liquid vegetable oils, which tend to contain considerably higher levels of desmethylsterols than olive oils. Early determinations of sterol composition by gas liquid chromatography (GLC) involved the use of packed columns. The stationary phases that were available at that stage had limited resolving power and only six sterols could be identified from olive oils: cholesterol, brassicasterol, campesterol, stigmasterol, b -sitosterol and D -7-stigmastenol. As improved stationary phases were introduced it became possible to separate D -5-avenasterol from b -sitosterol and D -7-avenasterol from D -7-stigmastenol. Subsequent important chromatographic developments now allow the separation of 16 desmethylsterols from olive oils.
Despite these advances, many standards still require the concentration of b -sitosterol to be greater than 93 percent of the total sterols. Internationally, it has been agreed that this limit should apply to "apparent b -sitosterol", referring to the sum of the concentrations of b -sitosterol, D -5 avenasterol, D -5,23-stigmastadienol, D -5,24-stigmastadienol, chlerosterol and sitostanol.
Although an important aspect of sterol composition is the apparent b -sitosterol composition, the CAC and IOOC have implemented some or all of the following criteria (calculated as percentage of total sterols): cholesterol, 0,5 percent maximum; brassicasterol, 0.2 percent maximum; campesterol, 4 percent maximum; D -7-stigmastenol, 0,5 percent maximum. In addition, the concentration of stigmasterol should be less than that of campesterol.
Over the years, the Codex standards for fats and oils have been gradually modified to enhance their usefulness in tackling authenticity problems. The pace of the changes is inevitably affected by the availability of data of sufficient quality for inclusion in the database. One significant source of data used to establish purity criteria for edible oils and fats of major importance has been research in the United Kingdom funded by the Ministry of Agriculture, Fisheries and Food (MAFF) and the Federation of Oils, Seeds and Fats Associations (FOSFA) at the Leatherhead Food Research Association.
To produce meaningful data, it is essential that sufficient samples be collected from representative geographical origins and that the oils be pure. In the MAFF/FOSFA work, over 600 authentic commercial samples of vegetable oilseeds of known origin and history, generally of ten different geographical origins, were studied for each of 11 vegetable oils. The oil from these seeds was extracted in the laboratory, except in the case of palm oil, which was obtained from palm plantations because the parent fruit is perishable and cannot be transported. The extracted oils were analysed to determine their overall FAC. FAC at the 2-position of the triglyceride, sterol and tocopherol composition, triglyceride carbon number and iodine value, slip melting point and solid fat content as appropriate.
Prior to 1981, FAC data were not included in Codex standards because data of sufficient quality were not available. In 1981, standards were adopted that included FAC ranges as mandatory compositional criteria. The MAFF/FOSFA work provided the basis for later revisions to these ranges.
In general, as more data became available, it was possible to propose fatty acid ranges much narrower and consequently more specific than those adopted in 1981. Table 3 gives examples of FAC of oils that were adopted by the Codex Alimentarius Commission (CAC) in 1981 and ranges for the same oils proposed at Step 4 at the Codex Committee on Fats and Oils (CCFO) meeting held in September 1993.
Further MAFF/FOSFA information has enabled the British secretariat of CCFO to propose the inclusion of data relating to sterol and tocopherol/tocotrienol (tocol) composition in the new draft standards. The combination of data on fatty acids with those on sterol and tocol composition provides a powerful method of identifying oils and oil blends.
TABLE 3 - Codex standards for fatty acid composition of oils - Normes Codex pour la teneur en acides gras des huiles végétales - Normas del Codex para la composición en ácidos grasos de los aceites
Fatty acid |
Soybean oil |
Groundnut oil |
Cottonseed oil |
Sunflower-seed oil |
||||
1981 |
1993 |
1981 |
1993 |
1981 |
1993 |
1981 |
1993 |
|
C14:0 |
< 0.5 |
< 0.2 |
< 0.6 |
< 0.1 |
0.4-2 |
0.6-1 |
< 0.5 |
< 0.2 |
C16:0 |
7-14 |
8-13.3 |
6-16 |
8.3-14 |
17-31 |
21.4-26.4 |
3-10 |
5.6-7.6 |
C16:1 |
< 0.5 |
< 0.2 |
< 1 |
< 0.2 |
0.5-2 |
0-1.2 |
< 1 |
< 0.3 |
C18:0 |
1.4-5.5 |
2.4-5.4 |
1.3-6.5 |
1.9-4.4 |
1-4 |
2.1-3.3 |
1-10 |
2.7-6.5 |
C18:1 |
19-30 |
17.7-26.1 |
35-72 |
36.4-67.1 |
13-44 |
14.7-21.7 |
14-65 |
14-39.4 |
C18:2 |
44-62 |
49.8-57.1 |
13-45 |
14-43 |
33-59 |
46.7-58.2 |
20-75 |
48.3-74 |
C18:3 |
4-11 |
5.5-9.5 |
< 1 |
< 0.1 |
0.1-2.1 |
0-0.4 |
0-0.7 |
0-0.2 |
C20:0 |
<1 |
0.1-0.6 |
1-3 |
1.1-1.7 |
0-0.7 |
0.2-0.5 |
0-1.5 |
0.2-0.4 |
C20:1 |
<1 |
<0.3 |
0.5-2.1 |
0.7-1.7 |
0-0.5 |
0-0.1 |
0-0.5 |
0-0.2 |
C22:0 |
< 0.5 |
0.3-0.7 |
1-5 |
2.1-4.4 |
0-0.5 |
0-0.6 |
0-1 |
0.5-1.3 |
C22:1 |
- |
< 0.3 |
< 2 |
< 0.3 |
0-0.5 |
0-0.3 |
0-0.5 |
0-0.2 |
C22:2 |
- |
- |
- |
- |
- |
- |
- |
0-0.3 |
024:0 |
- |
< 0.4 |
0.5-3 |
1.1-2.2 |
0-0.5 |
0-0.1 |
0-0.5 |
0.2-0.3 |
C24:1 |
- |
- |
- |
< 0.3 |
- |
- |
< 0.5 |
- |
Sources: Codex Alimentarius Commission, 1983,1993.
Olive oils
In processing, many crude vegetable oils undergo a bleaching step which involves heating the oil to approximately 103°C under vacuum with the addition of a bleaching earth (often acid activated). The objective is to remove pigments and produce an oil of light-yellow appearance. The bleaching process can be used on olive oil, usually in refining of oils of lower value. However, this process produces other changes within the oil. For example, it causes the dehydroxylation of sterols to produce steroidal hydrocarbons such as stigmasta-3, 5-diene. If an oil has undergone high-temperature deodorization as well, the formation of dehydroxylated sterols will be even greater. However, it is unlikely that these compounds will be produced to any great extent in virgin olive oils; thus their presence is indicative of mixing with either refined olive oil or other vegetable oils. The latest IOOC trade standard for olive and olive-pomace oils sets limits for these compounds (see Table 4), and CCFO will consider this relatively new development in due course.
The standards for oils and fats were developed and recognized throughout the world relatively recently. The current standards will be improved, with better databases of oil composition, as technology develops. For example, 13C/12C stable-isotope mass spectrometry seems set to become an important tool in determining adulteration of maize oil products.
Historically, determining the authenticity of maize oil with traditional methods was problematic because its FAC overlaps with that of several other vegetable oils. In addition, the concentration of sterols in maize oils is very great in comparison with that of other vegetable oils, so that the sterol composition of any blend will comprise predominantly those from maize oil. It is possible to form blends of oils whose characteristics according to traditional analysis are very similar to those of pure maize oil.
The determination of the 13C/12C stable-isotope ratio (SIR) enables the identification of blends of maize oil with other vegetable oils (see Table 5). The sample oil is burned to form carbon dioxide which is purified by GLC and then analysed by mass spectrometry. The results are presented not as an absolute abundance of each carbon isotope but rather as a ratio of the heavy 13C isotope to 12C, measured as parts per thousand (ppt) with respect to an international standard, carcon dioxide produced from calcite of Pee Dee formation belemnite (PDB) with phosphoric acid.
TABLE 4
Levels of stigmasta-3, 5-diene permitted in olive oils by the International Olive Oil Council - Niveaux de stigmasta-3, 5-diene autorisés par le Conseil oléicole international pour les huiles d'olive - Niveles de estigmasta-3, 5-dieno de los aceite de oliva permitidos por el Consejo Oleícola Internacional
Olive oil grade |
Maximum permissible level of stigmasta-3, 5-diene |
Edible virgin olive oils |
0.1 |
Lampante virgin olive oil |
0.5 |
Refined olive oil |
50 |
Olive oila |
50 |
Crude olive-pomace oil |
0.5 |
Refined olive-pomace oil |
120 |
Olive-pomace oilb |
120 |
a Denotes blend of virgin olive oil and refined olive oil.
b Denotes blend of virgin olive oil and refined olive-pomace oil.
Source: IOOC, 1993b.
TABLE 5 - Carbon stable-isotope ratio for maize oil and other oils - Ratio isotopique du carbone stable pour l'huile de maïs et les autres huiles - Relación del isótopo de carbono estable para el aceite de maíz y otros aceites
Oil |
Mean D 13C |
Range D 13C |
No. of samples |
Maize |
-14.95 |
-13.71 to -16.36 |
42 |
Other vegetable |
-28.99 |
-25.38 to -32.39 |
68 |
Cereal |
-31.26 |
-30.38 to -32.39 |
4 |
Fish |
-26.66 |
-25.37 to -27.95 |
4 |
Animal |
-30.28 |
-27.56 to -32.08 |
5 |
It is estimated that the determination of the 13C/12C SIR and the calculated iodine value (determined from the FAC) will permit the identification of maize oil adulterated with 5 to 10 percent other vegetable oil (Lee. Gillatt and Rossell, 1994). This is a powerful way to identify and thus prevent maize oil adulteration, and criteria relating to this technique may be expected to appear in the appropriate Codex standards in the future.
Recently, two instrumental methods, one based on chemical theory and the other on computer technology - i. e pyrolysis mass spectrometry and artificial neural networks - have been coupled to give a rapid assessment of adulteration of olive oils with seed oils (Goodacre, Kell and Bianchi, 1993). Although at present the system appears complex and expensive, it may one day form part of an internationally recognized standard to help prevent olive oil adulteration.
International standards for vegetable oils are evolving. They provide a valuable internationally accepted database which, apart from being useful for assessment of product quality, is vital for verifying oil authenticity. Clearly, Codex standards will assume increasing significance under the recently concluded GATT agreement; it is therefore incumbent upon Member Governments to seek to ensure that the standards are up to date and accurate to the extent possible.
Codex Alimentarius Commission. 1983. Codex standards for edible fats and oils. Supplement 1 to Codex Alimentarius Volume XI, Rome, FAO/WHO.
Codex Alimentarius Commission. 1993. Report of the Fourteenth Session of the Codex Committee on Fats and Oils, London, 27 September - 1 October 1993, Alinorm 95/17. Rome, FAO/WHO.
Firestone, D., Summers, J.L, Reina, R.J. & Adams, W.S. 1985. Detection of adulterated and misbranded olive oil products. J. Am. Oil Chem. Soc., 62: 1558-1562.
Goodacre, R., Kell, D.B. & Bianchi, G. 1993. Rapid assessment of the adulteration of virgin olive oils by other seed oils using pyrolysis mass spectrometry and artificial neural networks. J. Sci. Food Agric., 63: 297-307.
International Olive Oil Council (IOOC). 1993a. International trade standard applying to olive oils and olive-pomace oils. COI/T. 15/NC, No. 1/Rev. 6, Madrid.
IOOC. 1993b. Resolution RES-2/65-IV/93. Madrid.
King, B. & Zilka, S.A. 1986. Authenticity of edible vegetable oils and fats. Part XI. Palm kernel oil. Research Report No. 559, Leatherhead, Surrey, UK, Leatherhead Food Research Association.
King, B., Zilka, S.A. & Turrell, J.A. 1985. Authenticity of edible vegetable oils and fats. Part X. Coconut oil. Research Report No. 531. Leatherhead, Surrey, UK, Leatherhead Food Research Association.
Lee, K., Gillatt, P.N. & Rossell, J.B. 1994. Authenticity of edible oils and fats. Part XX. Determination of maize oil purity by stable carbon isotope ratio analysis (SCIRA). Research Report No. 719, Leatherhead, Surrey, UK, Leatherhead Food Research Association.
Rossell, J.B. 1991. Purity criteria for edible vegetable oils and fats. Fat Sci. Technol., 93(4): 526-531.
Tan, B.K. & Flingoh, C.H. 1981. Oleins and stearins from Malaysian palm oil: chemical and physical characteristics. PORIM (Palm Oil Res. Inst. Malaysia) Technol., No. 4. 6 pp.
Turrell, J.A. & Whitehead, P.A. 1990. Authenticity of edible vegetable oils and fats. Part XVI. Analysis of additional samples of palm, soyabean and rapeseed oils, Research Report No. 665. Leatherhead, Surrey, UK, Leatherhead Food Research Association.