Food fraud
Dangerously well-camouflaged: How modern analytics identifies food fraud
Every weekly shop is a series of decisions: forest honey or blossom honey? Simple rapeseed oil or high-grade olive oil? Pasture milk from grass-fed cows or the cheap discount variety from cows in stalls? Aside from taste and ecological factors, it is mainly price that plays a role here. And illegal schemes are never far away when money is involved. The seemingly harmless supermarket shelf can become the scene of a crime. We are talking here about food fraud. This occurs, for instance, when manufacturers claim that a food product contains very high-quality ingredients, but in reality they are just cheap substitutes. A typical example is the use of chemically synthesised vanilla that is listed on a product as “natural vanilla flavouring” – and has a price tag to match.
What is food fraud?
The European Commission has set out criteria to describe food fraud [1]. It involves …
- … the violation of European food law …
- … by intentionally deceiving or misleading consumers, …
- … resulting in an economic advantage or gain for the food fraudster.
Food fraud is by no means a niche problem. According to estimates, it affects around ten million kilograms of food in Europe each year [2]. The European Union estimates the global financial damage to be roughly 30 billion euros per year, putting it on the same scale as economic damage from the illegal heroin trade [3].
Food fraud can take many different forms [1]. These are just the most common ones:
- Dilution – Has the food been thinned out through the addition of other ingredients? Example: Water content of fruit juice is too high
- Substitution – Have individual ingredients been (partially) replaced by lower-quality ingredients? Example: Use of cistus leaves in oregano
- Mislabelling – Have false claims been made about the quality or origin of the food? Example: Conventional tomatoes marked as organic
- Unapproved enhancement – Have unknown or undeclared substances been added to the food to fake its quality attributes? Example: Treatment of tuna with nitrites for a lasting red colour
Is it food fraud?
One aspect of food fraud is to deliberately change product characteristics such as the appearance or smell to fake higher quality. However, there are also tricks that are perfectly legal, which food producers use to influence consumers. For example, you can come across a “light” version of food fraud when you buy oranges. They are usually in plastic nets in supermarkets, which are woven from orange fibres for good reason. Due to an optical illusion (colour assimilation or the “confetti illusion”), the orange colour of the net makes the fruit inside it look riper and distracts us from seeing a green tinge. This type of overlaid pattern is known as a Munker net, named after colour researcher Hans Munker. Whether this can be considered food fraud is up to the individual to decide. At any rate, it is more environmentally-friendly to buy oranges without a plastic net [11].
Example 1: Dangerous protein content in milk powder
The main reason for food fraud is generally to maximise profit. Yet in some cases, misdeclarations and substitute ingredients can also endanger consumers’ health. One particularly dramatic example of this was the melamine scandal uncovered in 2008. A Chinese milk powder producer had substituted the powder with melamine to achieve a higher protein content in the end product. In food analytics, proteins are typically determined using the Kjeldahl method based on the nitrogen content. However, Kjeldahl only determines the total nitrogen concentration of a sample, and not whether the nitrogen actually comes from proteins. Amino acids, the building blocks of protein, mostly have a nitrogen content of less than 20 per cent by mass, up to a maximum of 32 per cent by mass for arginine. As melamine, a small molecule with six N atoms, is very rich in nitrogen (67 per cent by mass), comparatively small additions were sufficient to fake a high protein content in the milk powder.
As the milk powders prepared in this way were also sold as infant formula, numerous cases of kidney stones and kidney failure among infants and young children occurred in 2007, caused by the melamine. Six children died from the effects [4].
This is a tragic example that illustrates one of the major challenges in food control: methods test exactly what they are supposed to, and a method is only used if there is something suspect. For example, food screenings identify typical dangerous substances that are known contaminants, for example mycotoxins or certain bacteria, such as salmonella in products containing raw eggs. In the Chinese infant formula case, there had been no suspicions in the past that a nitrogen-rich, non-protein substance had been added – let alone melamine – which is why the milk powder was not tested or analysed for such substances.
Example 2: The 2013 horse meat scandal
Another striking example which attracted a huge amount of media attention was the “horse meat scandal”, which came to light in 2013. Various retail groups and food companies had diluted beef products with cheaper horse meat, generating additional profit of 550,000 euros, according to the food safety organisation Foodwatch. In just a few months, around 750,000 tonnes of horse meat ended up in circulation in lasagne and other beef products [5].
The scandal was not about the lower quality meat in itself, but more about the deception. The misdeclared or undeclared meat was proven by an analysis of species-specific proteins or PCR analysis. In the latter method, DNA traces are reproduced from the sample and identified using a polymerase chain reaction. This enables a distinction to be made between different animal species – Equus (horses, asses and zebras), beef, pork and lamb, in some cases down to the individual breed – using DNA [6].
The detection limit for meat content in processed products is around 0.5%. It is much lower in raw, unprocessed meat, at around 0.1%, as the DNA has not been damaged by the heating process [7]. That means the method is sensitive enough to detect meat that has been mixed in intentionally. According to the Freiburg Chemical and Veterinary Investigation Office (Chemisches und Veterinäruntersuchungsamt Freiburg), a content level of more than five per cent of a particular DNA type in a food is considered to be a separate ingredient – such as horse. A content level of between one and five per cent indicates an undeclared species, and lower than one per cent is deemed to be interference from the production process [6].
Example 3: The true origin of food – organic and regional?
In the above cases of melamine in milk powder and horse meat in lasagne, it is clear what substance needed to be found analytically to uncover the fraud. But what happens when producers do not change the content of the food, and instead fake the origin? For example, how can we check whether milk actually deserves its “organic” label? Or whether our apple juice really comes from apples cultivated locally?
It may seem impossible to prove – but this is where stable isotope analysis comes into play. This method makes it possible to discern amazingly precise differences in food production. For example, to test whether milk is organic, the proportion of carbon isotope 13C is measured. This differs significantly if you compare C4 plants, such as maize, which is used in conventional cattle feed, and C3 plants such as grass and clover in “organic feed”. In accordance with the different proportions of 13C in the feed plants, the same isotope ratios can be found in the milk. This means the claim that milk is organic can be verified analytically [8].
Using the same principle, the 18O oxygen isotope can also be analysed, or to be more precise: the ratio of heavy 18O oxygen isotope to the lighter 16O “standard atom”, known as the δ18O value. This is particularly suitable for checking geographic data, as the value changes to increasingly negative values with distance from the coast.
To investigate the origin of apple juice, the following conclusion has been reached, for example: higher δ18O values of -3.5 to -3.99‰ are achieved exclusively in north-western Bavaria (Lower Franconia). According to the Reference Guideline for Apple Juice (as at January 2016) of the European Fruit Juice Association (AIJN), the average δ18O value of apple juice (fresh juice) is at least -6.5‰; values of -5 to -4‰ relative to Vienna Standard Mean Ocean Water (V-SMOW) are to be expected for Central Europe. If the measured values differ significantly from those expected, this indicates the origin of the apple juice has been misdeclared or there is inadmissible added water [9].
The δ18O value indicates distance from the coast
The isotope value of 18O in water decreases with increasing distance from the coast towards the interior. This is because heavier isotopes in water vapour condense more readily, which means clouds contain less and less 18O the further inland they move. This produces a geographic gradient in groundwater: coastal regions have higher 18O values, while the values are lower inland [12].
The master class: non-target screening
The list of food fraud examples is endless. Authorities continuously face the challenge of exposing the tricks perpetuated by food fraudsters. In the meantime, instrumental analysis also offers the possibility of non-target screening, which is more complex than searching for a specific foreign substance, but has the major advantage of identifying previously unknown discrepancies.
Non-target methods like this create a characteristic chemical fingerprint of a food or animal feed, for example using high performance liquid chromatography (HPLC) or gas chromatography (GC), combined with high-resolution mass spectrometry (HRMS). The measurement data generated is then compared against data from a reference library and so can verify the authenticity of the food sample – or falsify it if there are discrepancies, so that further analyses can be carried out. With holistic analyses like these, we will be able to guarantee the authenticity and safety of our food increasingly efficiently in the future. And that will not only protect consumers’ wallets, but their health too [10].
Sources:
[2] https://vsvbb.de/lebensmittelbetrug/
[3] https://www.mri.bund.de/de/nrz/das-nrz-authent/
[4] https://www.spektrum.de/kolumne/eine-prise-chemie-melamin-skandal-um-babymilch-in-china/2227016
[6] https://www.ua-bw.de/uploaddoc/cvuafr/Nachweis_Pferd_Poster_Lebensmittelchemikertag_2013.pdf
https://www.q-s.de/services/files/newscenter/13%2002%2014%20DNA-Analytik.pdf