News

In this study (here) researchers used a nearest neighbours Machine Learning algorithm to build a classification model for brandy and cognac based upon Raman spectroscopy.  The test is rapid and non-destructive.  They reported good ability to discriminate spirits by age (differentiating “less than 5 years”, “5-7 years” and “more than 7 years”) and also clear discrimination of Armenian and Dagestanian spirits from those from the Russian Federation (excluding Dagestan).  The authors hypothesise that this is because production standards in Armenia and Dagestan stipulate the exclusive use of local ingredients.

Photo by Nika Benedictova on Unsplash

The UK Food Standards Agency (FSA) has published a 3-year corporate plan, which explains how their 5-year strategy will be turned into concrete actions.

In the next 3 years FSA want to:

• Maintain the current high levels of trust and confidence in the food system and FSA. 
• Maintain food standards, so that food is safe and what it says it is, and consumers can continue to have confidence in their food 
• grow our contribution to and influence on food that is healthier and more sustainable, building on the work we have started since we published our strategy 

Read the Executive Summary and the full Corporate Plan.

In a recent publication (here), researchers from Tecnologico de Monterrey, Mexico, built a multivariate classification model for cereal bars based upon simultaneous ICP-MS analysis of 22 trace elements.  The authors used the model as a relatively simple and quick analytical tool to verify the labelled variety of cereal bars (gluten free vs conventional, or whether the bars were based on yoghurt, chocolate or fruit).  They did not explore the potential of the model to check for brand counterfeiting, which could widen the scope of applicability to any similar type of branded manufactured food.

Photo by Towfiqu barbhuiya on Unsplash

Researchers at Queens University Belfast have developed a multi-variate analysis classification model for salmon by fusing together data from mass spectrometry of lipids (“lipidomics”, using direct analysis with the “ion knife” REIMS sampling source) and elemental analysis of 46 metals and minerals.  The results were published last month in Nature Communications (here).  The classification model could discriminate farmed vs wild salmon.  It could also discriminate between fish from Alaska, Norway, Iceland and Scotland.  The main lipid biomarkers that drove this discrimination were identified.  The differences were probably due to vegetable oil supplementation in fish feed.  An incidental finding from the work, previously unreported, was that many Alaskan and Icelandic wild salmon contain elevated levels of cadmium.

Photo by Oxana Kolodina on Unsplash

AOAC International's Food Authenticity Task Force has developed standard method performance requirements (SMPRs®) for targeted and non-targeted food authenticity methods. SMPR set minimum performance criteria that food authenticity testing methods need to fulfil. 

AOAC SMPRs® can be downloaded on the AOAC Resources page.

As of July 2023, the following SMPRs®, relevant for food authenticity, are available:

• Determination of Biological Spices and Botanicals, and Relevant (Common) Biological Adulterants
• Quantitation of Cannabinoidsin Beverages
• Targeted Testing (TT) of Vanilla Adulterants
• Targeted Testing (TT) of Turmeric Adulterants
• Targeted Testing (TT) of Saffron Adulterants
• Nontargeted Testing (NTT) of Ingredients for Food Authenticity/Fraud Evaluation of Vanilla Powder and Extracts
• Nontargeted Testing (NTT) of Ingredients for Food Authenticity/Fraud Evaluation of Saffron
•  Nontargeted Testing (NTT) of Ingredients for Food Authenticity/Fraud Evaluation of Turmeric Spice Powder
• Quantitation of Cannabinoids in Edible Chocolate.

Information on the AOAC food authenticity SMPRs® has been added to the Quality section of this website.

Ideagen have launched a free guide (here), targeted at food manufacturers, on best practice to enhance supply chain intelligence and traceability.  Ideagen supply digital traceability solutions for regulated industries but this guide is written in a generic manner without reference to specific systems.  It is aimed particularly at smaller businesses and supply chains within developing countries where traceability information may be more difficult to capture.  It includes dealing with challenges such as regulatory differences in different sourcing countries.

We have added a new page to the website that signposts to analytical databases of authentic foods held by laboratories and researchers, and highlights those laboratories offering testing against their database.  It is searchable by food type, analytical technique, or authentic attribute (e.g. "organic").

Scroll down on the home page and look for the "Search for Food Authenticity Databases" button (the link has been provided also). 

We hope you find this new functionality useful and would welcome feedback on how it can be improved.  If your organisation holds an "authentic food" analytical database and you would like it listed on the site then please contact me.

This resource was made possible through the generosity of our funding partners.

A recent study (here – purchase needed) piloted the use of a non-destructive in-line sensor to detected minced beef adulteration at economically-significant levels.  The authors used hyperspectral imaging (HSI) in the 400–1000 nm spectral range in tandem with multivariate analysis.  They evaluated the use of partial least squares regression (PLSR) method, the ensemble Monte Carlo variable selection method (EMCVS), a range of spectral pre-treatments, and combinations of any two of them to predict the amount of minced beef in each prepared sample scanned. They developed a prediction model using data from beef adulterated with minced chicken and turkey meats and validated with data from beef adulterated with pork meat at adulteration levels ranging from 0 to 51% at approximately 3% increments. They reported good prediction results using the EMCVS, on the asymmetric least squares (AsLs) + Standard Normal variate (SNV) pre-treated reflectance spectra, using 23 selected wavelengths. They then tested the model on an independent set of beef samples adulterated with lamb and duck meat at concentrations ranging from 3 to 21% and reported good results with 9 optimum wavelengths.  They reported almost perfect classification for calibration, cross-validation, and prediction using 12 selected spectral bands. They concluded that this method could be used to further develop low-cost portable sensors for the digital sorting of adulterated minced beef and that their study demonstrates the feasibility of generic models to detect minced beef meat adulterated with other types of meat.

Picture by Charlie Harris on Unsplash

For many years the Joint FAO/IAEA Center of Nuclear Techniques in Food and Agriculture has had a remit and active programme to build analytical testing capacity in Low & Middle Income Countries; in effect, test capacity to support food safety.  This specifically includes stable-isotope techniques to check food authenticity.  They offer laboratory training programmes and workshops, both in their laboratory in Austria and at the recipient laboratories.

The IAEA have posted a summary of stable-isotope applications here, along with their remit and scope

• Supporting UN Member States in the use of nuclear and complementary techniques for science-based solutions to improve food safety, food authenticity, and security, as well as sustainable agricultural practices.
• Developing nuclear technologies to improve the safety and quality of food products, tracking the origin of food products and checking their authenticity.
• Supporting UN Member States in improving their laboratory and regulatory capacity to trade safe and high-quality food products and verify their authenticity using stable isotope measurements.
• Gathering best practices and providing guidance on the use of nuclear techniques for the verification of origin of food products.
• Conducting Coordinated Research Projects (CRP), focusing on the use of nuclear and complementary techniques, for instance, the identification of food fraud or products with high-value labelling claims.

In this study (here – purchase required) the isotopic composition of carbon, strontium and the concentrations of seventeen elements were determined in olive oils from Tunisia, Southern France and the South Basque country. The preliminary results overlapped and showed that, taken individually, the isotopic and elemental approaches were not discriminant. A linear discriminant analysis applied to δ¹³C, ⁸⁷Sr/⁸⁶Sr and to the concentrations of 4 selected trace elements (Fe, Mn, V and Cr) allowed classification of the olive oils into 3 groups according to their origin. The authors conclude that this novel classification approach succeeds because of variations in the plant growing environment, the geological background, the mineral composition of the soil and the production process.

Deloitte have worked with McCormick to identify best practice in food fraud risk management.  They have recommended six key focus areas.

• Governance, Leadership & “Tone from the Top”
• Policies & Procedures
• Training & Communication
• Vulnerability Assessment & Supply Chain Illumination
• Monitoring & Horizon Scannning – Risk Sensing & Dupe Killer
• Food Testing, Supplier Inspections & Deeper Dives

The first in a series of Deloitte blogs on the topic can be read here.

This annual report covers all the information shared in 2022 within iRASFF, between members of the Alert and Cooperation Network (ACN) that includes the Rapid Alert System for Food and Feed network (RASFF), the Administrative Assistance and Cooperation network (AAC) or the Agri-Food Fraud Network (FFN).

2022 shows an increase in the number of notifications shared in the AAC and FFN in comparison with previous years. This illustrates the continued commitment of Member States’ competent authorities to detect and report non-compliances, even if without health risk, or when suspected of fraudulent practice. In 2022, a high number of those notifications related to pesticide residues.

The European Commission continued to assist Member States, through both expertise and IT support to facilitate the increasing exchange of information.

Due follow-up was given to suspicious cross-border fraudulent activities, leading to launching specific actions such as the coordinated control plans on the illegal trade of pets, and to deter certain fraudulent practices in honey.

Read full report.

The report has also been added to the tools, guides & reports tab of our food fraud prevention section.

This article by Which? states that recent food fraud revelations show we can't take food safety for granted, and that careful oversight is essential for food safety and security.

The article focuses on the following areas:

• Why food safety laws need saving.
• Lack of border checks and staff cuts make life easy for fraudsters.
• Which foods are most at risk of food fraud?
• The human cost of food fraud.

Using examples, Which? show, why we can't take food safety for granted, and it's vital we take the opportunity to review and strengthen food safety laws, via a transparent process that allows both consumers and legitimate businesses to have confidence in the system.

Read full article.

A recent article (here – purchase required) reviews the application of advanced molecular analytical techniques to authentication of products of animal origin.  It covers some less-published application areas such as vegan authentication and concludes that there are still some analytical gaps, particularly for such applications that require species identification at low concentrations.  They conclude that law enforcement and analytical approaches must evolve in an integrated manner.

Thank you again to FAN member Brunot Sechet of Intelagrim (www.integralim.net) for distilling the monthly JRC food fraud bulletin into this infographic.

The authors of a recent study (here – purchase required) developed a portable and non-destructive NIR model to classify and identify artisan butter cheeses that had been adulterated with soya oil.  They developed the multivariate model using cheeses manufactured under laboratory supervision (both authentic and intentionally adulterated) and also market cheeses that had been confirmed as adulterated by fatty acid analysis.  They concluded that routine nutritional proximates analysis or simple colour tests are insufficient for detecting this type of adulteration but that an NIR multivariate model is effective.

A recent paper (here – open access) reviews the current applications of blockchain technology in the livestock industry; case studies, pilot schemes and perceived challenges.  The authors identify training of livestock farmers and other key people at the bottom of the supply chain as a key challenge.  They conclude that blockchain traceability will have most immediate practical use either for niche high-value livestock or for large commercial companies with large-scale primary production.

New publication using outputs of MedISys-FF finds that:

• Meat and meat products were the most reported fraudulent food products.
• Adulterated food commodities are mainly associated with the expiry date and tampering.

The MedISys-FF tool developed by Bouzembrak and colleagues (Bouzembrak et al., 2018) uses the MedISys portal of the European Media Monitor (EMM), a system that uses text mining to collect media articles worldwide.

Read the full article.

Whistleblowing is an important tool in any fraud risk-management strategy.  Research has been published (here – open access) into attitudes to whistleblowing by employees in the Halal meat industry.  The authors used a social science approach to assess the factors influencing a willingness to whistleblow (attitude, subjective norm, perceived behavioural control, perceived organizational support, religious obligation, knowledge, personal cost reporting, personal responsibility, and seriousness of wrongdoing) based upon a structured questionnaire.  Reassuringly the majority of respondents were positive, although this may be heavily influenced by a shared religious commitment.  The authors make recommendations to encourage a whistleblowing culture.

This study (here – open access) used an electronic nose based on an array of metal oxide semiconductor sensors, of the type routinely used in the beverage industry for quality control, to characterise lemon juice based upon its Volatile Organic Compound (VOC) profile. The authors prepared their own provenant lemon juice and their own samples adulterated with lemon pulp, water, citric acid, sugar and wheat straw. They then used chemometric methods such as principal component analysis (PCA), linear and quadratic analysis (LDA), support vector machines (SVMs), and artificial neural networks (ANNs) to analyze the response patterns of the sensors. Of the total data, 60% (for training), 20% (for validation), and 20% (for testing) were used. All models could classify the adulterated samples with an accuracy of more than 95%. The Nu-SVM linear function method had the highest accuracy among all models. The authors concluded that the use of metal oxide semiconductor sensors combined with chemometric methods can be an effective tool with high efficiency for rapid and nondestructive classification of pure lemon juice and its counterfeits.