Fraudulent Olive Oil Is Tainting the European Market, and This Test Can Prove It
Sofia Drakopoulou, a researcher at the National and Kapodistrian University of Athens, discussed the test at RAFA 2024.
Complete the form below to unlock access to ALL audio articles.
The price of olive oil isn’t what it used to be. Recent droughts and floods across Europe have devastated thousands of once-reliable groves, leaving farmers with fewer crops and consumers with higher prices.
Looking to capitalize on the crisis, fraudsters have since polluted the European market with tainted, adulterated olive oil.
In the first quarter of 2018, just 15 such cases of contaminated olive oil were recorded by the European Union. In the first quarter of 2024, 50 such cases were reported. These are only the incidents that have been raised by member states to the EU; experts believe the true scale of fraud is considerably higher.
So, how can watered-down oil be better detected? Technology Networks attended RAFA 2024 to hear about one promising technique, direct analysis in real time–high-resolution mass spectrometry (DART-HRMS), from Sofia K. Drakopoulou, a postdoctoral researcher at the National and Kapodistrian University of Athens.
A review of EVOO
“If we’d like to talk about numbers, the adulteration costs approximately 8-12 billion euros per year,” Drakopoulou told the RAFA audience. “That’s a huge amount of money.”
When it comes to extra virgin olive oil (EVOO), this kind of adulteration, says Drakopoulou, is largely achieved via substitutions with cheaper ingredients, such as vegetable oil.
“In the case of extra virgin olive oil, we can see that one of the most common practices is the partial substitution with cheaper alternatives,” she said. “This might mean the substitution with some vegetable or seed oils or even more olive oil of lower quality.”
“Not only is consumer trust at risk but also is [their] health, sometimes due to allergic reactions,” she warned. “So, we need to take action and develop some reliable and holistic methodologies in order to proceed with the authenticity testing.”
To do just that, Drakopoulou and her colleagues at the National and Kapodistrian University of Athens built on previous research which demonstrated that high-resolution mass spectrometry (HRMS) could aptly distinguish between different varieties of EVOO; recognizing the speed offered by direct analysis in real time (DART) methods, the team paired the two techniques together to create their DART-HRMS test.
“We have the fast speed of DART and, of course, we have the reliability of HRMS,” she summarized in her RAFA presentation.
To put their technique to the test, Drakopoulou and her team sourced 10 different types of oil and contaminated them with varying levels of adulterants. These samples were then analyzed by the DART-HRMS technique, which managed to give an accurate snapshot of the oils’ contents.
“We analyzed a total of 10 different oil categories, including extra virgin olive oils, olive oils of lower quality, spent oils – seed oils, vegetable oils – labeled as olive oil,” Drakopoulou said.
“We have conducted an adulteration experiment using nine different adulteration levels in all categories starting from 50 down to 1% of adulteration level,” she added. “And we managed to detect down to 1% of adulteration in the case of vegetable oils.”
And yet, despite this accuracy, Drakopoulou says there’s still room to improve the method, with a little help from machine learning.
“We want to take a step forward,” she said. “We’re thinking what we can do in case of data treatment in order to get the results faster.”
“[We developed] a machine learning-based approach at the university that we use to detect some specific region of interest in the spectrum in the case of EVOO and refined olive oil.”
With the benefit of this machine learning component, Drakopoulou says her DART-HRMS technique can turnover results in rapid time – hundreds of times faster than contemporary methods.
“In the case of time, with chromatography and the omics approaches, we have an analysis time of 20 minutes per sample, using LC [liquid chromatography],” she told the RAFA audience. “In DART-HRMS, we have an analysis time of three seconds.”
“Then the algorithm returns a probability score, the authenticity score,” she added. “That means if it is possible or not this sample [is] authentic. And, of course, we have the predicted label, if it is categorized as EVOO or not.”
Satisfied with the results of her study and her technique’s speed and reliability, Drakopoulou is confident that the machine learning-backed DART-HRMS method could be an indispensable tool to address the growing problem of olive oil fraud across Europe.
