News

A Spanish translation of our 2023 Annual Summary has been published here.  Thanks to Merry Rivas for this first step in making our resources more accessible to members who do not have English as a first language.  More language translations are set to follow.

Urea is an adulteration risk in milk, particularly in areas of the world with less systematic or industrialised supply chains.  It is already a natural component of milk but is also cheaply and widely available.  Adding urea to milk increases the nitrogen content, hence increasing the apparent protein, enabling the milk to be diluted with water.  It is not a sophisticated fraud.

Researchers at Baba Mastnath University have developed and published (open access) a simple onsite sensor to measure urea in milk that can be used in the field.  It is based on urease enzymes on a nylon membrane attached to an ammonium ion-selective electrode.  The biosensor gave a rapid 20-second response time at pH 5.5, detecting urea concentrations between 0.001 and 0.80 mM.  The authors validated with recovery experiments from milk spiked with urea (recoveries were above 97%) and went on to characterise the typical natural urea background concentration in milks from different regions of Northern India.  They found that their sensor compared favourably with other urea potentiometric biosensors and with other laboratory-based test methods.

Food Fraud Prevention – Types of Products

Welcome! In support of the Food Authenticity Network (FAN) activity, this new blog series reviews key topics related to food fraud prevention. Watch here for updates that explore the definitions of food fraud terms and concepts.

This blog post builds on our previous review of the definition and scope of food fraud and the subsequent blog post on the types of fraud. We will continue the discussion by examining the types of fraud. The next blog post will review the application of quality management and risk management to expand the focus from detection to prevention and risk to vulnerability.

Early food fraud research and publications focused on what food fraud is or how big the problem is. The research evolved into several paths: incident reviews, detection or authentication development, criminology, and strategic management. Some of the strategic management research included our peer-reviewed, scholarly, SCOPUS-listed publication on Defining the Public Health Threat of Food Fraud, Introducing the Food Fraud Initial Screening model (FFIS), Introducing the Food Fraud Prevention Cycle (FFPC), and Defining the types of counterfeiters, counterfeiting, and offender organizations. Together, the research projects revealed that criminals will attack in just about any way imaginable and most quickly and easily. Together, the research projects emphasized that criminals will  attack by ANY fraud act against ANY product. Thus, to holistically reduce food fraud, we need to focus on ALL types of fraud and for ALL products. We can either complain about this very broad scope or be practical and expand our collective focus on all types of fraud and for all products.

Here, the ‘products’ are not individual commodities such as olive oil, seafood, or spices, but are supply chain inventory types of products such as raw materials, ingredients, work-in-process, or finished goods (see MSU Introduction to Supply Chain Management/ SCM303).

The broad focus on ‘all hazards’ – or for food fraud prevention, for ‘all vulnerabilities’ – is consistent with food safety and HACCP. For example (emphasis added): “HACCP is a management system in which food safety is addressed through the analysis and control of biological, chemical, and physical hazards from raw material production, procurement, and handling, to manufacturing, distribution, and consumption of the finished product” (FDA 2017).”

A food fraud incident can occur in any type of product, so all are within the scope of a food fraud prevention strategy.

While a manufacturer or producer has the most control of THEIR raw materials and incoming goods, their customers are worried about fraud at any point along THEIR entire supply chain – or all products.

This blog post will review the food fraud types of products.

Food Fraud & Definition (From various sources including GFSI and SSAFE with definitions from adapted from Supply Chain Management textbooks):

• Raw Materials/ Commodities: A component of a food, feed or packaging that has not undergone processing (GFSI).
• Incoming Goods/ Ingredients: A component that is being received including food, or feed that has undergone processing (GFSI).
• Incoming Goods/ Packaging: A component that is being received including packaging that has undergone processing (GFSI).
• Work-in-process-manufacturing: product that is actively being transformed from ingredients to finished goods.
• Work-In-process-inventory: product that is actively being transformed but is being held idle while waiting for an additional step to complete the transition finished goods.
• Finished goods in inventory: product that has completed a transformation and is ready to deliver to a customer but it is being held in storage.
• Finished goods in the marketplace: product that has completed a transformation and is being held in a location or format that is ready for a customer to procure.
• Distributors, Wholesalers, and Resellers: firms that sell or deliver merchandise to retail stores or other types of customers.
• Returned goods and reverse logistics: the process of moving finished goods that have been distributed to the marketplace back to the origin or a location to receive, dispose, or rework product.
• Waste disposed, used packaging, and off-specification products: products that have been partially consumed or otherwise determined to be used or unacceptable for further use.

The types of food products are intentionally broad – holistic and all-encompassing- to frustrate the criminal against action of any kind.

Watch out for the next blog, which will review the application of quality management and risk management to expand the focus from detection to prevention and risk to vulnerability,

If you have any questions on this blog, we’d love to hear from you in the comments box below.

References:

• Spink, J, and Moyer, DC, (2011) Defining the Public Health Threat of Food Fraud, Journal of Food Science, Volume 75 (Number 9), p. 57-63, URL: https://ift.onlinelibrary.wiley.com/doi/full/10.1111/j.1750-3841.2011.02417.x
• Spink, John, Moyer, Douglas C, & Speier-Pero, Cheri (2016). Introducing the Food Fraud Initial Screening Model (FFIS), Food Control, Volume 69, November 2016, Pages 306-314, URL: http://www.sciencedirect.com/science/article/pii/S0956713516301219 .
• Spink, John; Zhang, Guangtao; Chen, Weina & Spier-Pero, Cheri; (2019). Introducing the Food Fraud Prevention Cycle (FFPC): A Dynamic Information Management and Strategic Roadmap, Food Control, 105 (November 2019), 233-241, URL: https://www.sciencedirect.com/science/article/pii/S0956713519302567
• Spink, J., Moyer, DC, Park, H., and Heinonen, J. (2013/2017) Defining the Types of Counterfeiters, Counterfeiting, Offender Organizations, Crime Science Journal, Volume 2, Number 8, pp. 1-10, URL: http://onlinelibrary.wiley.com/doi/10.1111/1541-4337.12033/abstract

This article (open access) reviews papers published 2018-2023 that use analysis of phenols, coupled with chemometrics, in order to classify extra virgin olive oil (EVOO).

The authors conclude that the best classification systems with most potential for wider use were achieved by databases that combined phenols with other parameters. Tthe application of untargeted metabolomics for discovering authenticity markers for different cultivars and origins is still relatively scarce.

They also stressed that factors such as climate, cultivar location and agronomic practices, extraction, and processing conditions, as well as storage may affect oil composition including sterols and phenols profiles. Another confounding aspect is that most of the commercial oils are sold as blends, where a contribution of both the cultivars used and the geographical origin is expected.

They concluded that more investigations carried out on a higher number of samples are needed to strengthen the use of this analytical approach both for  geographic traceability and for botanical origin traceability. This review highlighted the fact that most papers did not take into consideration a significant number of samples, and that sometimes there is an unbalanced relationship between the number of samples of a specific cultivar, or a specific geographical origin. Among other points of weakness, research studies did not report the correct legal classification as indicated on the label or certified by official analytical methods/organization.

EVOO samples analyzed in the reviewed papers (2018–2023) were collected or produced in different ways.  The authors therefore recommended that future databases should be built from samples obtained with uniform, harmonized and reproducible extraction conditions and using validated and standardized analytical procedures coupled with appropriate chemometric approach. Such comprehensive information could be useful for the production of a database of the qualitative and quantitative profiles of phenols in PDO, PGI, and monovarietal and blend EVOO samples, obtained from different cultivars or different regions.

This review article (purchase required) covers recent developments in loop-mediated isothermal amplification (LAMP) devices, particularly when coupled with microfluidic chips, for applications as point-of-use tests for meat and seafood species authentication.

While PCR-based methods remain the gold standard for assessment of the species authenticity, the authors consider that there is an urgent need for alternative testing platforms that are rapid, accurate, simple, and portable. Owing to its ease of use, low cost, and rapidity, LAMP is becoming increasingly used method in food analysis. The authors outline how the features of LAMP have been leveraged for species authentication testing of meat and seafood products. LAMP detection is simple and rapid. To make it truly instrument-free it needs an end-point visual detection, and so the authors review the principles of various end-point colorimetry methods. They also summarise different strategies to either suppress the nonspecific amplification or to avoid the results of nonspecific amplification.

The US FDA have published their annual surveillance test results for imported honey authenticity.

They found 3 of 107 samples to be adulterated with syrups or sweeteners.  These were from the Dominican Republic, Mexico and Yemen.  This compares with 14 of 144 samples reported as adulterated in the previous year (2021/2022).

The authors of this study (purchase required) have developed and validated a point-of-use test for bovine DNA in dairy products.

It is based on direct PCR within a closed-loop system (a single tube, using a portable PCR machine), which eliminates the need to extract DNA from samples.  Visualisation is by SYBR Green I detection of PCR amplicons giving a fluorescent marker that can be read by the naked eye. This was enhanced with the use of a self-designed portable LED lightbox, which mitigated colour interference, enabling detection in coloured products such as yoghurts without any sample preparation.  Analysis time was 50 minutes.

Validation studies revealed 100% specificity when tested with 13 other species, over 99% accuracy with 208 cow’s milk and dairy products in both liquid and solid forms, 1×10⁻⁵ ng limit of detection, and 100% accuracy with blind testing.

The authors then conducted a small market survey of “milk replacement” dairy products on sale in Thailand.  They found that  10.7%, contained bovine DNA, an indicator of potential dilutions or substitution with cows’ milk.

Photo by Tijana Drndarski on Unsplash

This comprehensive review article (open access) lists over 60 published chemometric applications for authenticating edible oils and fats using FT-IR.  FT-IR spectra are driven by the oils’ triglyceride composition.

The authors categorise the applications into four sections

• Adulteration with used oil (“gutter oil”)
• Adulteration of vegetable fat with animal fats or vice-versa (e.g. lard added to palm oil, or margarine added to butter)
• Substitution of one vegetable oil with another
• Geographic origin (mainly of premium olive oils).

The authors conclude that FT-IR is a valuable technique in the armoury, and often has advantages over alternative methods (e.g. DSC, GC-MS, HPLC) in terms of speed, simplicity and robustness.

Photo by Roberta Sorge on Unsplash

This study (purchase required, free to IFST members) sought to characterise premium smoked meats from different regions of China by differences in their key aroma compounds. The researchers looked at products made from different meats (chicken, pork, mutton, duck, goose) and from different regional origins.  They analysed these reference samples for 40 key aroma compounds using chromatographic/mass spectrometric techniques.

They found that 15 of the 40 key aroma compounds were co-occurring in all reference samples. The highest contributor among them was 2-methoxy-phenol. Norharman was more abundant in smoked pork.  N^ε-carboxymethyl lysine and N^ε-carboxyethyl lysine were higher in smoked pork and chicken.  Phenolics were higher in smoked pork and mutton, aldehydes in smoked duck and goose and nitrogenous compounds mainly in smoked chicken. Products originating from the southwest of China showed higher levels of key phenolics such as 2-methoxy-phenol.

The authors identified 12 potential markers for distinguishing between different regions and types of smoked meat.

Photo by Steven Weeks on Unsplash

This paper (open access) reviews the use of chemometrics and machine learning for building food authenticity classification databases.  It highlights best practice.  It concentrates on one-class classification models (“Is this sample X or is this sample NOT X?”) and on the complication, common within authenticity applications, that adulterants which should classify the sample as “NOT X” may be present in very low proportions.

The paper describes a generic and structured approach to building a classification database.  It concludes with 10 “golden rules”

1 Before choosing a data analysis method, it is important to understand what question needs to be answered.

2. Authentication should be developed using an one-class classification (OCC) approach. Discrimination, such as PLS-DA, may be used in exceptional cases where an exhaustive list of classes is available.

3.The OCC model must be developed using a representative training set collected from the target class, a further representative set for model optimization and, at the end, a test set for model validation. The key to success is good sampling.

4.Proper data preprocessing is essential because all data sets contain both useful and unwanted information.

5.The quality of a model is assessed using two main figures of merit (FoM); sensitivity, which is the rate of true acceptance of samples from the target class, and specificity, which is the rate of true rejection of samples not belonging to the target class.

6.In probabilistic models such as SIMCA, 100% sensitivity cannot be expected.  Their variability due to the limited number of samples should be taken into account.

7.There are two ways to optimize an OCC model: rigorous, which is based only on the target data, and compliant, which also uses data belonging to other classes.

8.In either case, OCC optimization is performed by comparing the results of the training and test data, and the optimal complexity (e.g., the number of PCs) is chosen at a FoM convergence.

9.When choosing an alternative set from a non-target class for compliant optimization, the class that needs to be chosen must be as similar as possible to the target one and must reflect the real authentication problem, otherwise the specificity estimate may be overoptimistic or non-realistic.

10.Once the model is properly optimized, its actual performance needs to be verified on the test set, which must be completely extraneous to the previous training and optimization phases. FoM calculated on the test set are the ones that must be considered to define model prediction ability.

(image taken from the paper)

In this survey (open access literature) the researchers purchased 46 processed edible insect products from nine EU e-commerce platforms.  They tested the authenticity of the labelled speciest by genomic metabarcoding. A 200 bp region from 16S rRNA gene was used as molecular target. Sequences were taxonomically assigned through BLAST analysis against the GenBank database. Procedural blanks and positive controls were included in the analysis, and threshold values were established to filter the final data.

They found that the mislabeling rate (i. e. the mismatch between the species declared on the label and the species identified by metabarcoding) was, on average, 33% although it varied depending on the e-commerce platform and the insect species.  A. domesticus was particularly involved. The use of species not authorized under Novel Foods legislation (e. g. Gryllus locorojo), and/or the partial replacement of high value species with lower value species was highlighted. The presence of insect pests was also detected.

The authors conclude that metabarcoding is an effective tool for edible insect product species authentication. This study can provide useful data on the main authenticity issues involving the EU edible insect product market, providing evidence that both official controls and FBO’s self-controls should be strengthened.

The second monthly report of EU alerts relating to “suspicions” (i.e. where there is no food safety risk, so official notifications that would not be captured on the RASFF system) has been published here.  The data need careful interpretation, as they include (and are heavily weighted by) non-fraud issues such as pesticide residue limit non-compliance.  There is still valuable insight to be gained on fraud risks, such as lists of imports from unapproved establishments; meat-containing dried noodles from factories that do not hold a licence for making products of animal origin appear to be a particular watch-out.

Image from the report.

The authors of this paper (open access) have reviewed the recent uses of internet-of-things (IoT), artificial intelligence and machine learning (AI-ML) and blockchain technology (BT) in the wine industry supply chain.

They propose a future framework  They consider that this IoT-AI(ML)-BCT framework integrates IoT, AI, and BCT would revolutionize the wine industry’s logistics and quality control processes. It considers that all actors of the supply chain are part of the system: Grape producers use ML algorithms to predict crop yields, monitor soil conditions, and forecast weather patterns, aiding vineyard management and optimizing production; wineries use IoT devices for tracking wine barrels throughout the supply chain, ensuring proper storage conditions and minimizing spoilage risks. Meanwhile, BCT provides transparency and immutability to supply chain data, enabling consumers to trace the journey of each bottle from vineyard to shelf, guaranteeing authenticity and quality.

The authors recognise that there are many practical factors to overcome in order to roll out the proposed framework, including

• Conducting case studies or empirical studies to validate the proposed framework in real-world settings.
• Exploring the regulatory landscape surrounding the adoption of BCT, IoT, and AI in the wine industry, considering factors such as data privacy, security, and compliance with industry standards and regulations.
• Investigating the cultural and economic implications of integrating these technologies into the wine supply chain, including stakeholder perceptions, adoption barriers, and economic feasibility.
• Continuously monitoring technological advancements and emerging trends in BCT, IoT and AI to ensure the proposed framework remains relevant and up-to-date.

They conclude that deep learning and time series algorithms in more complex wine supply chain management scenarios with demands derived from actual data, would be exciting to examine.

(graphic from the paper)

One of our Partners, Tenet, has published an article that explores the threat caused by Supply Chain Fraud, how it manifests and what you can do to protect your business and customers.  

Some of the most common fraud threats that supply chains are open to are:

1. Invoice Fraud
2. Product Substitution
3. Kickbacks and Bribery
4. Data Manipulation
5. Ghost Employees and Vendors
6. Collusion and collaboration
7. Sanctions violations
8. Contract and Misrepresentation Fraud.

Read the article here.

The EC Knowledge Centre for Food Fraud and Quality (the Joint Research Centre, “JRC”) have published their monthly collation of global food fraud media reports for February 2024. 

Thanks, as always, for FAN member Bruno Sechet for formatting these into this infographic, which you can download here.  If you would like to join the JRCs mailing list to sign up for these monthly summaries then the link is here.  You can also follow Bruno's LinkedIn feed here.

Other interesting articles signposted by the JRC:

• The United Kingdom Food Standards Agency (FSA) has updated its guidance for the private sector to prevent food crime.
  Food.gov.uk     Food-safety
• DG SANTE ACN food fraud monthly reports
  The Commission has published its first monthly report on agri-food fraud suspicions. The report compiles information gathered
  from the Alert and Cooperation Network (ACN), which facilitates the exchange of information between Member States on agrifood controls. The report includes cases of cross-border non-compliance, which ACN members have identified and shared as
  suspected fraud.
  DG SANTE    Food Authenticity Network    Food Safety News
• UK FSA Surveillance Sampling Programme
  A project commissioned by the UK FSA led to the sampling of retail food products from food business operators (e.g. national
  supermarkets and small independent retailers) in October 2022. A total of 1 215 food samples from 28 different food
  commodity types were collected and tested. 81% of the samples tested for compliance were satisfactory.
  FSA    Food Authenticity Network

Paw San rice, also known as “Myanmar pearl rice”, is considered the highest quality rice in Myanmar. Methods for its authentication are one of the worldwide research priorities for specific regional foods identified by the Joint FAO/IAEA Centre for Nuclear Techniques and these new methods (here – free link for 50 days) is one of the outputs of that programme.

Shwe Bo District is one of the most popular rice growing areas in the Sagaing region of Myanmar which produces the most valued and highly priced Paw San rice (Shwe Bo Paw San). The verification of the geographical origin of Paw San rice is not readily undertaken in the rice supply chain because the existing analytical approaches are time-consuming and expensive.

In this 4-year study, two rapid screening techniques, Fourier-transform near-infrared (FT-NIR) spectroscopy and headspace-gas chromatography-ion mobility spectrometry (HS-GC-IMS), coupled with chemometric modelling, were applied and compared for the regional differentiation of Paw San rice. In addition, low-level fusion of the FT-NIR and HS-GC-IMS data was performed and its effect on the discriminative power of the chemometric models was assessed. Extensive model validation, including the validation using independent samples from a different production year, was performed. Furthermore, the effect of the sample preparation technique (grinding versus no sample preparation) on the performance of the discriminative model, obtained with FT-NIR spectral data, was assessed. The study discusses the suitability of FT-NIR spectroscopy, HS-GC-IMS and the combination of both approaches for rapid determination of the geographical origin of Paw San rice.

The results demonstrated the excellent potential of the FT-NIR spectroscopy as well as HS-GC-IMS for the differentiation of Paw San rice cultivated in two distinct geographical regions. The OPLS-DA model, built using FT-NIR data of rice from 3 production years, achieved 96.67% total correct classification rate of an independent dataset from the 4th production year. The DD-SIMCA model, built using FT-NIR data of ground rice, also demonstrated the highest performance: 94% sensitivity and 97% specificity. This study has demonstrated that FT-NIR spectroscopy can be used as an accessible, rapid and cost-effective screening tool to discriminate between Paw San rice cultivated in the Shwe Bo and Ayeyarwady regions of Myanmar.

More than 1000 Volatile Organic Compounds (VOCs) have been identified in roasted coffee. They have been shown to be indicative of the roasting process, the type (Aribica or Rustica) and the specific variety and origin of the coffee bean.

In this paper (purchase required) the authors used an untargeted strategy to process SPME-GC-MS data coupled with chemometrics to identify VOCs volatile as possible markers to discriminate Arabica coffee and its main adulterants (corn, barley, soybean, rice, coffee husks, and Robusta coffee). They reported that Principal Component Analysis (PCA) showed the difference between roasted ground coffee and adulterants, while the Hierarchical Clustering of Principal Components (HCPC) and heat map showed a trend of adulterants separation. The partial Least-Squares Discriminant Analysis (PLS-DA) approach confirmed the PCA results. 24 VOCs were putatively identified, and 11 VOCs considered candidates as markers to detect coffee fraud, found exclusively in one type of adulterant: coffee husks, soybean, and rice.

Photo by Nathan Dumlao on Unsplash

 We are delighted to be collaborating with the Global Alliance (GA) on Food Crime to bring a dedicated page for the GA on our website.

The GA is a coalition of international leaders who have agreed to work together on the prevention, detection and disruption of food crime. The GA initially agreed to have a small number of founder participants, consisting of food regulatory and enforcement organisations from Australia, Canada, New Zealand, the UK and the USA, but are looking to involve any country that is willing and able to contribute to the aims and objectives of the GA moving forward. 

The current Chair of the GA is Ron McNaughton, Head of the Scottish Food Crime and Incidents Unit at Food Standards Scotland, who said “Its fantastic that the Global Alliance now has a page on the Food Authenticity Network’s website. This gives us the ability to share information on the work of the GA in one place. This will be particularly important in terms of outlining progress towards achieving our strategic objectives, so many thanks to the Network for providing this great opportunity.”

Visit the Global Alliance page for more information and updates.

This dissertation from Florida State University describes designing and validating a species-specific PCR-lateral flow assay for Atlantic white shrimp (Litopenaeus setiferus) utilizing a miniaturized and cost-effective PCR instrument.  The selectivity was tested against 68 shrimp, prawn, and fish samples from 14 seafood species. L. setiferus was simultaneously amplified by the multiplex assay to give three visual bands, which distinguished it from other species having either one or two bands on the lateral flow stick.

The researcher also developed a specific red snapper assay, validating a rhPCR lateral flow assay where the thermotolerant RNase H2 enzyme was included in the PCR reaction to activate the red snapper-specific rhPCR primer. Amplicons generated in the duplex rhPCR reaction were detected using dual target lateral flow strips. The standardized assay was validated with 108 barcoded fish samples from 16 finfish species. Samples identified as Lutjanus campechanus or L. purpureus by DNA barcoding formed three distinct bands, while other fish species formed only two bands on the lateral flow strips. A minimum of 0.37 ng/μL crude DNA was needed to obtain a visible band on the lateral flow dip stick.

Both assays showed 100% specificity and took 90–120 minutes for completion. The researcher concludes that this confirms the suitability of PCR and rhPCR-lateral flow assays as an economical on-site tool for species authentication in the seafood industry.

Photo by Michal Mrozek on Unsplash

AcciMap is a systems-based technique used to analyse accidents that occurred within complex socio-technical systems. It has been used to drill into the causes of serious accidents in fields such as transportation or recreation, and also for applications such as identifying risk factors for child labour in supply chains.  More recently it has been used for food safety incidents and foodborne disease outbreaks.

This pre-publication paper is the first report on this methodology being used to analyse contributory factors in food fraud incidents.  The authors applied the model to the milk adulteration scandal (melamine) that emerged in China in 2008.  They coupled AcciMap with a classical Food Fraud Vulnerability Assessment approach (systemic weaknesses associated with opportunities, motivation and control measures).  Their model identified forty-eight contributory factors of influence grouped across six sociotechnical levels across the Chinese dairy system from government to equipment and surroundings. Lack of vertical integration (processes and communication) contributed to the failure. The authors conclude that, when viewed in a broader perspective, the melamine milk scandal can be linked to a series of human error and organisational issues associated with government bodies, the dairy supply chain, individual organisations and management decisions and individual actions of staff or processes.

Image from the publication