Food Science & Technology · Packaging, Safety, Quality & Regulation · also Food Adulteration, Food Authenticity, Food Fraud Prevention, Detection of Adulterants
Detecting and preventing economically motivated adulteration: common adulterants and their health effects, rapid household and laboratory tests, authenticity methods (isotope ratios, DNA, spectroscopic fingerprinting), vulnerability assessment and supply-chain controls.
In the exams: ICAR AICE 35.a Unit 4 (food adulteration); core of food safety officer and food analyst examinations
Food fraud is the deliberate deception of buyers about food for economic gain. It ranges from adding water to milk and dyeing chilli powder to selling horse meat as beef, farmed fish as wild, cheap rice as premium aromatic rice, sugar syrup as honey and conventional produce as organic. Most fraud is not meant to harm anyone, but because fraudsters do not consider safety, some of the worst food poisonings in history have been fraud: melamine in milk powder, industrial oil and toxic seeds in edible oils, methanol in spirits, lead chromate in turmeric. Fraud also damages honest producers, misleads consumers about what they pay for, undermines trade in high-value products and can hide allergens that kill.
This manual covers the field in the order a food analyst or food safety officer needs it. It starts with definitions: adulteration, misbranding and food fraud, and how fraud differs from food safety and food defence. It describes the common adulterants of milk, oils, spices, honey, beverages and grains, and their health effects. It then moves from detection to prevention: rapid screening tests, laboratory confirmation, stable isotope ratio analysis, DNA-based species identification and barcoding, spectroscopic fingerprinting with chemometrics, and authentication of geographic origin and organic status. It ends with food fraud vulnerability assessment and mitigation, and with the traceability and transparency technologies used in supply chains.
Food safety officer and food analyst examinations in many countries make adulteration and its detection a core topic, and doctoral fellowship examinations in food processing include food adulteration as a named unit. Typical questions ask about the adulterants of particular foods and the tests that detect them, health effects, the principles of isotope and DNA methods, chemometric terms, and the elements of a vulnerability assessment. The manual covers each of these, with worked calculations where examinations use them.
Food laws in most jurisdictions use two older legal concepts. Adulteration, broadly, means that a food is unsafe or not of the nature, substance or quality demanded: it contains a poisonous or deleterious substance, filth or decomposed material; a valuable component has been removed or a cheaper one substituted; damage or inferiority has been concealed; something has been added to increase bulk or weight or make it appear better than it is; or it was prepared under insanitary conditions. Misbranding means that the label is false or misleading, for example about the name, ingredients, quantity, origin or nutritional content. Many laws also define 'substandard' food that does not meet a compositional standard but is not unsafe.
These concepts cover accidents and negligence as well as deliberate acts. A food can be adulterated by unintended contamination with a pathogen or a pesticide. Food fraud, by contrast, is defined by intent. A widely used definition describes food fraud as a collective term for the deliberate and intentional substitution, addition, tampering or misrepresentation of food, food ingredients or food packaging, or false or misleading statements about a product, for economic gain. In the United States the related term economically motivated adulteration is used. The European Union uses four criteria to identify fraud: a violation of food law, intention, economic gain and deception of customers. Codex has developed guidance on the prevention and control of food fraud to help national systems.
Fraud takes several forms. Dilution mixes a high-value liquid with a cheaper one, such as water in milk or juice, or a cheap oil in olive oil. Substitution replaces an ingredient or species with a cheaper one: hazelnut oil for olive oil, pangasius for cod, cow milk in buffalo mozzarella, non-premium rice in premium aromatic rice. Concealment hides poor quality, for example by colouring old fish gills or bleaching damaged grain. Unapproved enhancement adds a substance to improve apparent quality, such as melamine to raise apparent protein or Sudan dyes to brighten chilli. Mislabelling makes false claims of origin, organic status, species, weight or expiry. Counterfeiting copies a brand or packaging. Grey-market diversion and theft move genuine goods outside authorised channels, often with altered dates.
Food fraud sits beside three other concepts. Food safety deals with unintentional hazards; food quality with unintentional failures to meet specifications; food defence with intentional acts meant to cause harm, such as tampering for ideological or malicious reasons. Fraud is intentional but aims at gain, not harm. The distinction matters because each needs different controls: hazard analysis for safety, specifications and quality control for quality, threat assessment for defence and vulnerability assessment for fraud. Fraud is hard to find by routine testing, because fraudsters design adulteration to pass the standard tests, so prevention focuses on understanding where and why fraud is attractive.
Milk is one of the most frequently adulterated foods worldwide because it is valuable, liquid and sold in small lots. The commonest fraud is adding water, which lowers solids and raises volume. To hide the dilution, adulterators add substances that restore density or solids, such as sugar, starch, salt, urea or ammonium sulphate, or skim milk powder; neutralisers such as sodium carbonate or bicarbonate to mask souring; preservatives such as formalin or hydrogen peroxide to delay spoilage; detergents and vegetable fat to make 'synthetic' milk; and nitrogen-rich compounds such as urea or melamine to inflate the protein measured by nitrogen methods. Milk from a cheaper species, usually cow, may be added to buffalo, goat or sheep milk and products. Milk fat may be replaced by cheaper vegetable or animal fats in ghee, butter and cheese.
Edible oils are adulterated by mixing high-value oils with cheaper ones and by passing off refined or low-grade oils as premium. Olive oil is a classic target: extra virgin oil may be blended with refined olive oil, pomace oil, or seed oils such as sunflower, soybean or hazelnut oil. Other frauds include mineral oil in edible oil; castor oil; used and reprocessed frying oil; and, historically, toxic contaminants: mustard oil mixed with oil from argemone seeds, which caused outbreaks of epidemic dropsy in South Asia, and cooking oil contaminated with tri-ortho-cresyl phosphate, which caused mass paralysis in North Africa in 1959. Cottonseed oil can be detected by its cyclopropenoid fatty acids, and sesame oil by sesamolin.
Spices are valuable, sold ground and coloured, so they invite adulteration. Chilli powder has been coloured with Sudan dyes, industrial azo dyes found widely in Europe in the early 2000s, and bulked with brick powder or sawdust; turmeric with lead chromate, a bright yellow pigment linked to lead poisoning in South Asia, and with metanil yellow; black pepper with dried papaya seeds; saffron, the most expensive spice, with dyed corn silk, safflower petals or used stigmas; cumin and other spices with ground nut shells, which brought undeclared peanut allergen in some cases; oregano with olive and other leaves. Salt and sugar may be cut with chalk or other powders.
Honey is among the most adulterated foods. It is diluted with sugar syrups from cane or maize, which are C4 plants, or from rice, beet or wheat, which are C3 plants and harder to detect; it is harvested immature and dried artificially; it is mislabelled for botanical or geographical origin; and it is ultrafiltered to remove pollen and so hide its origin.
Beverages: fruit juices are diluted with water, sugar and acids or with cheaper juices such as apple or grape in pomegranate or berry juice; tea may contain exhausted leaves re-coloured, or foreign leaves; coffee may contain chicory, roasted cereals or husks; wine may be sweetened with added sugar or contain unapproved additives, as in the diethylene glycol scandal in Central Europe in 1985; counterfeit spirits made with methanol cause deaths each year in several regions. Grains and pulses: premium aromatic rice is mixed with ordinary rice, durum wheat pasta with common wheat, pulses with cheaper or toxic seeds such as Lathyrus sativus, and grain with stones, sand or dyes. Meat and fish species substitution, such as horse meat found in beef products in Europe in 2013, completes the picture.
Most fraud only cheats consumers, but some adulterants are toxic, and some frauds have caused mass poisonings. Melamine, added to raw milk and milk powders to raise apparent protein, caused kidney stones and kidney failure in infants in East Asia in 2008, when hundreds of thousands of children were affected and several died; melamine combines with cyanuric acid to form insoluble crystals in the kidneys. Codex later set maximum levels for melamine of 1 mg/kg in powdered infant formula and 2.5 mg/kg in other foods.
Argemone oil contains sanguinarine and related alkaloids that cause epidemic dropsy: swelling of the legs, skin changes, heart failure and glaucoma. Lead chromate in turmeric causes lead poisoning, which harms the developing brain, causes anaemia and raises blood pressure. Sudan dyes are classed as genotoxic carcinogens and are not permitted in food at any level. Metanil yellow, a non-permitted azo dye, has been linked to toxic effects in animal studies. Methanol in counterfeit spirits is metabolised to formaldehyde and formic acid, causing metabolic acidosis, blindness and death. Diethylene glycol causes kidney failure. Tri-ortho-cresyl phosphate causes delayed paralysis. Seeds of Lathyrus sativus, if eaten as a staple for months, cause neurolathyrism, an irreversible spastic paralysis of the legs, through the neurotoxic amino acid β-ODAP. Formalin in milk or fish is toxic and irritant; hydrogen peroxide and detergents irritate the gut.
Many frauds cause harm indirectly. Undeclared substitution can introduce allergens: ground peanut or almond shells in spices, milk proteins in products labelled dairy-free, or fish species substitution, can cause severe reactions in allergic people who trusted the label. Species substitution can also introduce toxins or hazards: escolar sold as other white fish causes oily diarrhoea, pufferfish mislabelled as another species has caused tetrodotoxin poisoning, and fish that has been mishandled and relabelled can carry histamine.
Dilution reduces nutritional value: watered milk deprives children of protein and calcium; diluted or substituted fortified foods fail to deliver their nutrients. Unhygienic practices often accompany fraud: water added to milk may be contaminated, and recycled oils and repackaged foods may carry pathogens or chemical degradation products. Mislabelled organic or origin products are rarely harmful in themselves but defraud consumers who pay more.
The burden of fraud falls most heavily where testing and enforcement are weakest and where poor consumers buy loose, unbranded foods. Public-health surveillance, monitoring of outbreaks for unusual causes, and testing of high-risk foods are therefore part of fraud control. When an unusual illness pattern appears, such as many infants with kidney stones or a cluster of oedema cases, investigators should consider adulteration early.
Rapid tests screen samples quickly and cheaply at collection centres, markets, checkposts or in the home. Many are classical colour reactions. Starch in milk, paneer or other products gives a blue colour with iodine. Added sucrose in milk is shown by the resorcinol (Seliwanoff) reaction, giving a red colour on heating with acid. Neutralisers such as sodium carbonate in milk turn rosalic acid solution rose-red. Formalin in milk forms a violet ring at the junction with concentrated sulphuric acid in the presence of ferric chloride. Hydrogen peroxide is detected by a blue colour with potassium iodide and starch. Urea is detected by the yellow colour with p-dimethylaminobenzaldehyde.
For oils, the nitric acid test gives an orange-red colour with argemone oil; Halphen's test, with sulphur in carbon disulphide and amyl alcohol, gives a pink to red colour with cottonseed oil; the Baudouin test, with furfural and hydrochloric acid, gives a red colour with sesame oil; and mineral oil makes a saponified sample turbid when diluted with water (Holde's test). For spices and pulses, metanil yellow turns magenta with hydrochloric acid; brick powder sinks in water while genuine chilli powder floats and colours the water; papaya seeds float in alcohol or water while black pepper sinks; iron filings in tea are picked up with a magnet; exhausted tea leaves release colour quickly into cold water.
Physical tests include the lactometer, which measures milk density and, with fat content, allows solids-not-fat to be estimated by formula; the cryoscope, which measures freezing point; and refractometers for soluble solids in juices and honey.
Instrumental rapid methods are increasingly used. Infrared milk analysers at collection centres measure fat, protein, lactose and solids in under a minute and can flag abnormal spectra suggestive of adulteration. Test strips and kits based on enzymes or antibodies detect urea, melamine, antibiotics, aflatoxin and other targets. Portable near-infrared and Raman spectrometers, and smartphone-based colour readers, screen powders, oils and spices for unusual patterns. Lateral-flow immunoassays detect species proteins and allergens. Loop-mediated isothermal amplification allows DNA tests to be done outside the laboratory.
Rapid tests have limits. Many colour tests have detection limits too high to catch small additions; some give false positives with permitted ingredients or false negatives in coloured or fatty matrices; and adulterators adapt to them, choosing substances that the usual tests do not detect. A rapid test result is therefore a screening result: a positive should lead to a confirmatory laboratory test before legal action, and a negative does not prove authenticity.
Their value lies in coverage. Testing every lot of milk at a village collection centre, or many samples at a market in a day, creates a deterrent that occasional laboratory testing cannot. They also educate consumers, and public campaigns showing simple household tests raise awareness of fraud. Programmes should record results, verify kits with known samples and send a share of samples, especially positives, to accredited laboratories.
Laboratory confirmation identifies and quantifies the adulterant or the deviation from authentic composition using validated methods. Chromatography is central. Gas chromatography with flame ionisation detection measures fatty acid profiles of oils and fats, revealing foreign oils and non-milk fats; sterol and triacylglycerol profiles add specificity, for example to detect seed oils or refined oils in olive oil. Liquid chromatography with diode-array or mass spectrometric detection identifies dyes such as Sudan and metanil yellow, melamine and cyanuric acid, and other additives; tandem mass spectrometry gives very low detection limits and confirms identity by specific ion transitions. Ion and anion-exchange chromatography profiles sugars and oligosaccharides in honey and juices.
Elemental analysis by atomic absorption spectrometry or inductively coupled plasma mass spectrometry measures lead, chromium and other elements, as in turmeric adulterated with lead chromate. Cryoscopy measures the freezing point of milk precisely; because the freezing point is controlled mainly by lactose and salts, which are fairly constant, added water raises it towards 0 °C. The percentage of added water can be estimated as (normal freezing point depression − sample depression) ÷ normal depression × 100; if authentic milk depresses the freezing point by 0.525 °C and a sample by 0.480 °C, about 8.6% water has been added. Nitrogen methods such as Kjeldahl cannot distinguish protein from melamine or urea, so true protein is confirmed by methods that measure protein nitrogen specifically or by amino acid analysis. Electrophoresis, isoelectric focusing and ELISA identify species-specific proteins, and nuclear magnetic resonance gives compositional fingerprints.
For results to stand up in enforcement or trade disputes, laboratories must produce defensible evidence. Official sampling follows prescribed procedures: samples are taken in the presence of the owner, divided into portions, sealed and labelled, with one portion for the official laboratory, one for the owner and one retained for referee analysis, and a documented chain of custody. Laboratories are accredited to ISO/IEC 17025 for the methods they use, and use reference methods from Codex, ISO, AOAC International or national standards where they exist, or fully validated in-house methods.
Validation establishes selectivity, limit of detection and quantification, linearity, accuracy (by certified reference materials or recovery), precision and robustness. Measurement uncertainty must be reported, and decision rules defined, so that a result just above a limit is interpreted consistently. Proficiency testing, in which laboratories analyse the same samples and compare results, verifies competence.
Authenticity testing has a special difficulty: many frauds do not add a foreign substance but shift composition within or slightly outside natural variation. The laboratory must then compare the sample with a database of authentic samples of known origin, variety and season, and the conclusion is a probability that the sample is consistent or inconsistent with the claim, not a simple pass or fail. Databases must be large, representative and maintained, which is one reason authenticity testing is concentrated in a limited number of specialised laboratories.
Many elements have stable isotopes: carbon (¹²C and ¹³C), hydrogen (¹H and ²H), oxygen (¹⁶O and ¹⁸O), nitrogen (¹⁴N and ¹⁵N), sulphur and strontium. Their ratios vary slightly in nature because physical, chemical and biological processes discriminate between light and heavy isotopes. Stable isotope ratio analysis measures these ratios with an isotope ratio mass spectrometer, usually after converting the sample to simple gases by combustion or pyrolysis in an elemental analyser, or after separating individual compounds by gas or liquid chromatography.
Ratios are reported in delta notation, in parts per thousand (per mil, ‰), relative to international standards: δ = (Rsample/Rstandard − 1) × 1,000, where R is the ratio of heavy to light isotope. Carbon is referred to Vienna Pee Dee Belemnite, hydrogen and oxygen to Vienna Standard Mean Ocean Water, and nitrogen to atmospheric nitrogen. A negative δ means the sample has less of the heavy isotope than the standard.
Carbon isotopes distinguish photosynthetic pathways. Most plants, including wheat, rice, sugar beet, grapes and fruit, use the C3 pathway and have δ¹³C of about −22 to −33 ‰, typically near −27 ‰. Maize, sugar cane, sorghum and millet use the C4 pathway and have δ¹³C of about −10 to −16 ‰. Plants with crassulacean acid metabolism, such as pineapple and vanilla, are intermediate. A C3 product adulterated with C4 sugar therefore shifts towards less negative δ¹³C. Hydrogen and oxygen isotopes in plant water reflect local precipitation, which varies with latitude, distance from the sea, altitude and temperature, so they carry geographic information and reveal added water. Nitrogen isotopes reflect fertiliser type and diet, and strontium isotopes reflect bedrock geology.
The best-known application is detecting cane or corn syrup in honey. Because honey is mostly from C3 flowers, C4 sugar raises its δ¹³C. The official internal standard method compares the δ¹³C of the whole honey with that of its own protein, which is not affected by added sugar: a difference of more than about 1 ‰ indicates C4 sugar. The percentage of C4 sugar is estimated as (δprotein − δhoney)/(δprotein − δC4) × 100, with δC4 taken as about −9.7 ‰. If δprotein is −25.0 ‰ and δhoney −23.5 ‰, the C4 sugar is (−1.5)/(−15.3) × 100 ≈ 9.8%. Syrups from C3 plants such as rice or beet escape this test and require liquid chromatography-isotope ratio mass spectrometry of individual sugars, NMR profiling or marker compounds.
Simple two-source mixing models estimate proportions: if cane sugar is −11 ‰ and beet sugar −26 ‰, a sugar measuring −20 ‰ contains (−20 + 26)/(−11 + 26) = 40% cane sugar. Site-specific natural isotope fractionation measured by deuterium NMR, which reveals the deuterium content at particular positions in ethanol, detects added beet or cane sugar in wine (chaptalisation beyond legal limits) and distinguishes natural from synthetic flavour compounds such as vanillin. The δ¹⁸O of water in wine and juice reveals added water. Isotopes of the main elements together with strontium and trace elements authenticate geographic origin of wine, cheese, meat, coffee and rice. Interpretation needs authentic reference data, because isotope values also vary with climate and season.
DNA methods identify the species of animal, fish or plant material in a food. DNA is present in almost every cell, carries species-specific sequences, and is more stable than proteins to many processes, so it can be found in cooked, dried and moderately processed foods. Heavy processing such as sterilisation, prolonged heating at low pH and refining fragments DNA, so methods target short sequences; highly refined products such as oils, sugars and gelatin may contain too little DNA for reliable testing.
The polymerase chain reaction amplifies a chosen DNA region millions of times. Species-specific PCR uses primers that amplify only the target species; real-time PCR measures amplification as it happens with fluorescent probes, allowing sensitive detection and semi-quantitative estimates. Quantifying adulteration by DNA is difficult, because DNA content per gram differs between tissues and species and is altered by processing, so results are usually given as detection above a threshold or as DNA proportions rather than weight per cent. Digital PCR, which partitions a sample into thousands of reactions and counts positives, improves quantification. Mitochondrial targets, present in many copies per cell, give high sensitivity.
PCR methods confirmed horse and pig DNA in beef products during the European meat scandal of 2013, and are used routinely to check meat, fish, milk species in cheese, durum wheat in pasta, premium rice varieties with microsatellite markers, and genetically modified material.
DNA barcoding identifies species by sequencing a standard, short gene region and comparing it with a reference library. For animals, including fish, the standard barcode is a region of about 650 base pairs of the mitochondrial cytochrome c oxidase subunit I gene (COI); for land plants, the plastid genes rbcL and matK are the core barcodes, often supplemented with the nuclear ITS region; ITS is also used for fungi. Reference sequences linked to identified voucher specimens are held in public libraries such as the Barcode of Life Data System. Mini-barcodes, shorter fragments of the same regions, work on processed products.
Barcoding has exposed widespread fish mislabelling, with surveys in several continents finding that a significant share of samples, sometimes a third or more in particular markets and categories, were not the species named; common substitutions include cheaper white fish sold as cod or grouper, farmed salmon as wild, and escolar as tuna. It is also used for herbal products, spices and tea.
Sanger sequencing reads one DNA sequence and suits single-species samples; mixtures give unreadable traces. Next-generation sequencing, or DNA metabarcoding, reads thousands of sequences at once and identifies all species in a mixture, such as meat products or herbal blends, though quantitative interpretation is limited. Limitations include closely related species with nearly identical barcodes, hybrids, errors and gaps in reference libraries, and contamination in the laboratory, which requires strict separation of pre- and post-amplification work and negative controls.
Targeted methods look for a known adulterant. Fraudsters, however, keep changing adulterants, so laboratories increasingly use untargeted fingerprinting: a technique records a rich signal from the whole sample, and statistics decide whether that signal matches authentic material. Near-infrared spectroscopy measures overtones and combinations of vibrations of C–H, O–H and N–H bonds; it is fast, needs little or no sample preparation and can be portable, but its broad, overlapping bands require multivariate analysis. Mid-infrared (Fourier-transform infrared) spectroscopy gives sharper bands from fundamental vibrations and is used for oils, honey, milk and powders. Raman spectroscopy complements infrared, works through glass and plastic and suits water-containing samples and pigments. Fluorescence spectroscopy is sensitive to compounds such as chlorophyll, tocopherols and oxidation products in oils. Nuclear magnetic resonance gives highly reproducible fingerprints of many compounds at once and is used for honey, wine, juice and oils. Hyperspectral imaging combines spectroscopy with imaging to map adulterants in powders or meat.
Fingerprints are also obtained by mass spectrometry, including high-resolution and ambient ionisation techniques, and by chromatographic profiles. These approaches together are sometimes called food metabolomics or 'foodomics'.
Chemometrics is the use of statistical and mathematical methods to extract information from chemical data. Spectra are first pre-processed to remove unwanted variation: scatter corrections such as standard normal variate or multiplicative scatter correction, derivatives to separate overlapping bands, smoothing and scaling.
Principal component analysis is an unsupervised method that reduces many correlated variables to a few principal components that capture most of the variation; plotting samples on these components reveals groups and outliers without using any knowledge of which samples are authentic. Supervised classification methods use known classes to build models: linear discriminant analysis, partial least squares discriminant analysis, support vector machines, random forests and neural networks assign a sample to one of several classes. Class-modelling methods such as soft independent modelling of class analogy build a model of the authentic class alone and ask whether a new sample fits it, which suits authenticity testing, where the possible adulterants are unknown. Partial least squares regression predicts a quantity, such as the percentage of an adulterant.
Models must be validated with samples not used to build them, by cross-validation and, better, an independent external test set, and their performance reported as sensitivity, specificity and error rates. If a model detects 46 of 50 adulterated samples and correctly accepts 95 of 100 authentic ones, its sensitivity is 92% and its specificity 95%. Overfitting, when a model describes noise in a small training set, is the most common failure. Models are only as good as the authentic reference samples behind them, which must cover the natural variation in variety, origin, season and processing.
Many foods command a premium because of where they come from. Geographical indications protect names linked to a place and its know-how, under the World Trade Organization's agreement on intellectual property (articles 22 to 24) and national or regional schemes such as protected designations of origin and protected geographical indications. Examples from many regions include sparkling wines, hard cheeses and hams of particular European regions, high-altitude teas and aromatic rices of South Asia, single-origin coffees of East Africa, the Americas and the Pacific, and specialty honeys and fruits worldwide. Origin labelling is also required for many foods by country-of-origin rules. These claims are attractive targets for fraud.
Origin is authenticated by measuring properties that reflect the environment of production. Hydrogen and oxygen isotopes reflect local water; strontium isotope ratios reflect underlying rock; sulphur and nitrogen isotopes reflect geology, fertiliser and proximity to the sea; and multi-element profiles, including trace and rare-earth elements measured by inductively coupled plasma mass spectrometry, reflect soil. NMR and other metabolomic fingerprints add information on variety and processing. Results are compared with reference databases of authentic samples collected over several seasons; official databases exist for some products, such as wine in the European Union. Origin can usually be confirmed or excluded at the level of a region, rarely of a single farm, and only where reference data exist.
Organic production is defined by standards, such as the Codex guidelines for organically produced foods (CXG 32-1999) and national and regional regulations, and verified by certification: inspection of farms and processors, records of inputs and sales, and mass balance checks. The price premium makes relabelling conventional produce as organic a common fraud, sometimes on a large scale in international trade in grains and oilseeds.
No single laboratory test proves that a food is organic, because organic is a production method, not a composition. Laboratory tests support certification in several ways. Residues of synthetic pesticides above trace levels indicate a likely breach, although their absence does not prove organic production. Nitrogen isotopes help for crops: synthetic nitrogen fertilisers made from atmospheric nitrogen have δ¹⁵N close to 0 ‰, whereas manure and compost are enriched, usually several per mil higher, so organic vegetables tend to have higher δ¹⁵N. Overlap is large, legumes fix atmospheric nitrogen, and some organic fertilisers have low δ¹⁵N, so the test can flag suspicious lots but not prove fraud alone. For animal products, fatty acid profiles and isotopes reflect diet, for example pasture feeding.
The strongest controls against organic fraud are documentary: supply-chain mass balance, which compares the quantities of organic product bought and sold by each operator; electronic certificates of inspection for imports; unannounced inspections; and cross-checks between certification bodies. If a trader sells 100 tonnes of 'organic' grain but can show purchases of only 80 tonnes, the 20 tonnes, a quarter more than was bought, must be explained.
Because fraud cannot be tested out of a supply chain, food businesses are expected to assess where they are vulnerable and to put mitigation in place. The Global Food Safety Initiative made a documented food fraud vulnerability assessment and mitigation plan a requirement of its benchmarked certification schemes from 2018, and standards such as FSSC 22000, BRCGS, SQF and IFS include it. In some jurisdictions, regulations on preventive controls require that economically motivated adulteration be considered in hazard analysis when it could cause a safety hazard.
A vulnerability assessment looks at each raw material, ingredient and finished product and asks how attractive and how easy fraud would be. Factors include opportunity, such as ease of adulteration with available substances, the complexity and opacity of the supply chain, the number of intermediaries, access to the product, and weaknesses in detection; motivation, such as high price, price spikes or shortages caused by poor harvests, economic pressure on suppliers, a history of fraud for that commodity and the culture of the sector; and existing controls, such as supplier relationships, testing, audits and traceability. Information comes from fraud incident databases, official alert systems, market price data, industry intelligence and the company's own testing.
The result is a ranking, often expressed by scoring likelihood and consequence on simple scales and multiplying them, so that a material scoring 4 for likelihood and 5 for impact scores 20 and is treated as high priority. Free tools from industry and standards bodies guide the assessment. The assessment is reviewed regularly and whenever conditions change, such as a crop failure that makes a commodity scarce and expensive.
Mitigation reduces opportunity and raises the chance of detection. Supplier measures include approving suppliers after assessment, preferring direct relationships and shorter chains, buying from producers rather than traders where possible, unannounced audits, and contracts with specifications, authenticity clauses and rights to test. Purchasing whole products rather than ground or processed forms, for example whole spices or whole fish, makes substitution harder. Testing is targeted at the most vulnerable materials with methods able to detect likely frauds, including untargeted screening; results are trended. Mass balance and yield checks reveal unexplained quantities.
Physical measures include tamper-evident seals and closures, secure storage and transport, and authentication features on packaging such as holograms, covert inks, taggants or unique codes, which protect brands against counterfeiting. Organisational measures include training staff to recognise fraud indicators, whistle-blowing channels, horizon scanning of prices and alerts, and a food safety culture in which suspicions are reported.
Food fraud prevention differs from food defence, which uses threat assessment to protect against intentional harm by disgruntled employees, activists or terrorists, and relies on access control, personnel screening and security. Many companies run the two programmes together because both deal with intentional acts and share tools such as access control and tamper evidence. Governments contribute through intelligence-led enforcement, specialised food crime units in some countries, international networks for sharing information on fraud, and penalties that make fraud unprofitable.
Traceability is the ability to follow the movement of a food through specified stages of production, processing and distribution. The Codex principles for traceability or product tracing (CXG 60-2006) treat it as a tool of a food inspection and certification system, and many food laws require at least 'one step back, one step forward' traceability: each business must know from whom it received each input and to whom it supplied each product, with records available to authorities. Internal traceability links incoming lots to outgoing lots through the process.
The foundation is identification: every lot or batch has a unique code on the product and its documents, and records link codes across steps. Global standards from GS1 provide identifiers for products, locations and logistic units and a standard for sharing event data, which allows different companies' systems to exchange traceability information. Barcodes, two-dimensional codes such as QR codes and radio-frequency identification tags carry these identifiers.
Good traceability allows fast, precise recalls, limits the scope of incidents, supports origin and organic claims, and deters fraud, because fraudulent material is harder to insert into a chain where every lot is recorded and reconciled. But traceability only proves where documents say a product went. If false information is entered at the start, or if a lot is diluted or swapped between recorded steps, records alone will not reveal it, so traceability must be combined with mass balance checks and analytical verification.
Several technologies aim to make supply chains more transparent. Distributed ledgers, often called blockchain, record transactions in a shared, append-only database that participants can verify and that is hard to alter retrospectively. They can link records from farm to retail and let buyers or consumers scan a code to see a product's history. Their limitation is that the ledger guarantees only that records were not changed after entry, not that they were true when entered; 'garbage in, garbage out' applies. They work best combined with sensors, independent verification and analytical testing at key points.
Internet-of-things sensors record temperature, location and handling in real time, supporting cold-chain integrity and showing where goods have been. Satellite imagery and remote sensing verify that claimed farms exist and produce the crops claimed. Digital certificates replace paper documents that can be forged. Anti-counterfeiting technologies on packaging include serialised codes checked against a central database, holograms, colour-shifting inks, microtext, chemical taggants and edible DNA or isotopic tags added to products.
Consumers increasingly expect transparency about origin, production methods and sustainability. Technologies that publish supply-chain information can build trust, but only if the information is verified. For a food technologist, the practical tasks are designing lot coding and records that link to process data, reconciling quantities at each step, choosing technologies proportionate to risk and cost, and making sure that analytical checks confirm what the records claim.
A student uses this subject to connect food chemistry and analysis with real cases. They learn which foods are adulterated and why, how a colour test or a freezing point measurement reveals it, and how isotope, DNA and spectroscopic methods work. In examinations they meet questions on adulterants and tests, calculations of added water and C4 sugar, and the concepts of vulnerability assessment and traceability. In practicals they run classical tests and learn their limits. They also learn to read a fraud case critically, separating what was proved from what was alleged.
A quality assurance manager or food safety officer with five years' experience writes and updates the food fraud vulnerability assessment for each material, chooses which suppliers and ingredients to test and with what methods, interprets results from contract laboratories, organises mass balance checks and supplier audits, and responds to alerts about frauds in the market. An enforcement officer plans sampling of high-risk foods, takes legal samples correctly and prepares evidence for prosecution. Both roles require clear reports that explain to non-specialists what a result means and what it does not prove.
A senior technical or regulatory leader designs company-wide or national fraud prevention programmes, decides on investment in testing capacity and traceability systems, builds intelligence sharing with other companies and authorities, and handles major incidents, recalls and public communication. They set purchasing policies that reduce fraud opportunity, such as shorter supply chains and buying whole commodities, and they balance cost against risk across hundreds of materials. They brief boards and ministers on emerging threats, commission new detection methods when a commodity becomes vulnerable, and build partnerships with producers in origin regions so that authentic supply is easier to buy than fraudulent supply.
A senior researcher develops and validates new authenticity methods, builds and curates databases of authentic samples, advances untargeted methods and chemometric models, and studies the economics and criminology of food fraud to predict where it will appear next. They advise standard-setting and enforcement bodies, testify as expert witnesses, and ensure that methods are robust enough for legal use rather than only for publication. They also train analysts and inspectors, contribute to proficiency tests, and help regulators decide when a new method is mature enough to support enforcement.
This manual does not reproduce full test procedures, reagent preparations or method validation data; analysts must use the official method texts. Detection limits of rapid tests vary with kit, matrix and operator, and many classical tests are not specific. Isotope ranges, DNA barcode regions and chemometric performance figures are representative and depend on reference data. Legal definitions, standards and penalties differ by jurisdiction and are described here only in general terms. Figures for historical incidents are those commonly reported and may differ between sources.
Original practice questions written to the syllabus and pattern of the exams named in each question's examTags. They are not past papers and do not reproduce any board's questions.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.