Living in the 70’s – Detecting adulteration by seed oils
admin | August 19, 2012Remember the 1970’s when a person could be convicted of murder on the basis of a strand of hair left at the crime scene. If it was the same colour and was the same thickness to that of the accused, then that would do. We look back on that now and most of us would immediately recognise the huge potential for incorrect prosecution. DNA methods have reduced the potential error rates from 1 in 100 to 1 in hundreds of millions. Trying to convict someone today solely on the basis of the physical characteristics of a hair sample is now unthinkable.
But when it comes to detecting olive fraud, that is exactly what is happening right now.
Here are some examples of the “conviction by hair strand” approach that we currently apply to detecting fraud in olive oil, and how it is supposed to work.
A whole bunch of chemistry standards have been approved by the International Oil Council in an attempt to identify olive oils that have been adulterated with another cheaper seed oil such as sunflower, canola and soybean. However, the problem is that most edible oils are made of most of the same sort of stuff as olive oil. For example, olive oil contains a lot of the fatty acid oleic acid (typically around 70-75%). However, other oils such as canola contain credible amounts of this same fatty acid, and sunflowers have been bred to produce oil that has oleic levels equivalent to or greater than olive oil. So if a fraudster added 30% high oleic sunflower oil or 20% canola oil to a typical EVOO it will create a fraudulent oil with an oleic content that still is well within the range normally expected of real EVOO (officially 55-83%).
So the answer? Introduce more chemical measures of course! Edible oils contain substances called sterols. Some are found exclusively in one type of oil, but the majority are found in all oils but the amounts vary depending on the type of oil. For example, the sterol Brassicasterol is found only in canola oil in more than trace amounts. Conversely, the sterol campesterol while being the predominant sterol in sunflower oil, is also found in reasonable quantities in olive oil.
So while the oleic acid level of an olive oil will not change significantly when adulterated with canola oil, the resulting fraudulent oil will contain high levels of brassicasterol – a dead sure sign that it had been adulterated. It’s not just a smoking gun, the culprit has been caught standing over the body with the smoking gun in hand, with the act caught on DVD quality CCTV.
On the other hand, if the olive oil was adulterated with sunflower oil it would register an elevated level of campesterol. But elevated compared with what? Real olive oil naturally contains campesterol as well, and being a natural product, that level varies from oil to oil (mostly dependent on the variety of olive). Think about it for a moment. if you added sunflower oil to an olive oil with a low campesterol level, then it may actually have a total campesterol level lower than some 100% olive oils. As the international convention stands, oils with more than 4% campesterol are deemed not to be extra virgin, and I guess by inference they have been deemed to have had sunflower oil added (as campesterol has no negative sensory impact and it is a healthy cholesterol lowering sterol). But make no mistake, there are thousands of legitimately made EVOO’s made each year throughout the world that have more than 4% campesterol. The 4% level has been arbitrarily set by EU olive oil politicians and as such those thousands of oils have (analogously) been convicted on the basis of the width of a hair found at a crime scene.
So what is the solution? Arguments regarding what is an appropriate level of this or that when it comes to catching out fake olive oils will continue forever, but every level that is chosen will necessarily be arbitrary. If the level is lowered, it will be harder to adulterate with sunflower oil and get away with it, but it will also mean that even more hard working legitimate producers of olive oil will be labelled as cheats. If the level is increased, the number of false positives for cheaters will decline but it will be easier to adulterate oils and get away with it. Science has little to do with it, so long as campesterol is found in oils other than olive (i.e. forever).
But there is a solution. DNA testing. And it’s not a pipe dream.
To date, DNA testing has been considered by many (mostly chemists surprisingly) not to be a viable alternative to solving the intractable problem of using chemical markers of adulteration that are found in olive oil itself.
The reason is that very little of the original plant DNA remains in oil once it is refined (as is the case with all oils used for adulteration). But remain it does. Olive oil may contain a lot of things but sunflower, hazelnut or soybean DNA aren’t on the list.
Methods have been developed to extract DNA fragments from edible oils and the efficiency of procedures that are used to amplify these fragments have increased by orders of magnitude in recent years. These improvements have meant that lower volumes of oil need to be extracted in order to obtain a sufficient amount of DNA for accurate identification (Agrimonti et al. 2011). Very recently, Zhang et al.(2012) detected adulteration of olive oil by sunflower, sesame and soybean oils as low as 10%. I’m not an expert in this area, but their methods seemed orthodox using a couple of commercially available primers and standard amplification techniques. The authors concluded that:
“Despite all the advantages of DNA method, DNA collection from oil turns into a bottleneck for a widespread application in market supervision. Therefore for all DNA methods sample preparation is always a critical point to be improved”.
So there is quite a bit of work to be done. However, consider this. It took over 2,000 man-years of effort and over 6 billion dollars to sequence the human genome using technology available 10-15 years ago. Using soon to be released technology from Oxford University (you can read about it here doi:10.1038/nature.2012.10051), it is estimated that 96% of the human genome could be sequenced for a few thousand dollars over a weekend, and that simpler organisms such as yeasts and bacteria could be sequenced in a few minutes by a smaller one off use $800 device that connects to a USB slot on a laptop. That is the rate of advancement in this field. So in this context it would appear that the problem of cost efficient ‘sample preparation’ of edible oils before DNA analysis is far from insurmountable.
The advances in this field are extraordinary, and while we continue to spend money on trying to tweek possible chemistry based solutions (and even more time arguing amongst ourselves at international forums as to what is an appropriate level of this or that), a definitive genomic based solution is probably just around the corner.
So surely it’s time to move towards DNA evidence to convict real adulterators and make false convictions by the physical appearances of a hair or two a thing of the past.
Sources: (past the doi number into the search box at www.doi.org)
Agrimonti et al. (2011) doi:10.1016/j.tifs.2011.02.002
Zhang et al. (2012) doi:10.1016/j.foodcont.2012.03.027