Food fraud vulnerability assessments and product authentication in Dutch and Chinese milk supply chains

Recent food fraud incidents have highlighted the importance of fighting food fraud and thereby ensuring food integrity. Milk is reported as one of the most commonly adulterated foods in the world, hence there is a strong demand for the assurance of its integrity. Both technical and managerial approaches can help to combat food fraud and ensure the integrity of the milk supply chain. The Dutch and Chinese milk supply chains are different with regard to their structure, history of development, and historical evidence of fraud. Hence, this thesis aimed to elucidate and compare fraud vulnerabilities in the Dutch and Chinese milk supply chains, their underlying factors, and the relationship with fraud prevalence.Fraud vulnerabilities in the Dutch and Chinese supply chains were examined, and the similarities and differences between the tier groups were identified (Chapters 2 and 3). In general, both supply chains were rated as low to medium vulnerable to food fraud by its actors. The variation in fraud vulnerability among different tier groups, i.e. farmers, processors, and retailers, was relatively small in the Netherlands. When comparing Dutch farms of different management types, the organic farms appeared more vulnerable than their non-organic counterparts (Chapter 2). Unlike the Dutch milk supply chain, the difference in fraud vulnerability level between Chinese farmers and processors was larger. The farmers were the weaker link within the Chinese milk supply chain (Chapter 3). On average, Dutch and Chinese farmers reported a similar level of fraud vulnerability, but the Chinese farmers varied much more in their levels. The Dutch processors appeared slightly more vulnerable on average than their Chinese counterparts, but variation in both groups was similar.Following the profiling of fraud vulnerability of the milk supply chains, analytical methods for detection of milk adulteration were developed. Considering the diversity in milk composition, the authentication approaches for the Dutch and Chinese commercial milks were developed separately, but along the same principles. The latter one was then used for checking the prevalence of suspicion for milk fraud in the Chinese market. The milk compositional parameters measured by routine analysis, which is based on Fourier transform infrared (FTIR) spectroscopy, were used for the development of the milk authentication approaches. Both univariate and multivariate analyses were conducted, and their performance was compared. Using the developed approach, most of the milk samples (more than 90%) that were artificially adulterated with protein-rich, nitrogen-rich, and carbohydrate adulterants could be successfully identified as adulterated. Comparing the univariate and multivariate approaches, the univariate approach performed better in detection of common milk adulteration, missing only one adulterated sample among the protein-rich, nitrogen-rich, and carbohydrate adulterations (Chapter 4). In addition to the routine analysis, a novel, rapid and non-destructive technique, based on an ultrasound echo-pulse system, was explored for the characterisation of dairy products and detection of milk dilution (Chapter 5). The results show that the ultrasonic technique could be used for compositional characterisation of dairy products, and has the potential to detect diluted milk.Using the milk authentication method developed in Chapter 4, milk fraud occurrence in the Chinese market was determined, and its relationship with the fraud vulnerability data of the Chinese milk supply chain was elucidated (Chapter 6). Twelve out of 52 tested samples were suspected to be adulterated, of which the protein, fat, or lactose contents were below their respective thresholds. Proportionally more suspect samples were discovered among the products from central-northern and eastern areas in China. Chapter 3 previously also showed higher vulnerability levels for a number of fraud factors in these areas, which indicates an interesting relationship between fraud vulnerability and prevalence.In conclusion, this thesis highlighted the complexity of food fraud vulnerability in milk supply chains in the Netherlands and China. It also revealed the relationship between vulnerabilities and fraud prevalence as well as the importance of adequate fraud detection systems. It is recommended that a proper guidance for milk fraud mitigation should concern continuous assessment of fraud vulnerabilities in businesses and across the supply chain in combination with availability of efficient and affordable authentication technology