UC Davis: Most Restaurant ‘Extra Virgin’ Fails the Taste Test
A UC Davis Olive Center study released on September 18 found that 9 of 15 “extra virgin” olive oils sold to American restaurants and caterers failed the basic taste test for the grade, and two of the 21 oils sampled contained canola oil. Nearly all of them passed the chemical tests the trade relies on. Here is what the researchers found, why the usual lab checks missed it, and what it means when you dip bread in a restaurant.
What the study found
The report, Evaluation of Olive Oil Sold to Restaurants and Foodservice, is by Selina Wang, Edwin Frankel and Dan Flynn of the UC Davis Olive Center. Sodexo, the largest foodservice provider in the United States, helped the team buy 21 samples through the distributors restaurants actually use: 15 labeled “extra virgin” and 6 labeled plain “olive oil.” Most came through Sysco. The samples were tested blind by the Australian National Oils Laboratory, with extra tasting panels in Spain and Italy.
All but one of the 15 extra virgin samples passed the three chemistry checks buyers normally use: free acidity, peroxide value and ultraviolet absorbance. Nine of the 15, or 60 percent, failed the USDA extra virgin sensory standard, which only asks that a trained panel finds no defects and at least some fruitiness. The panels described rancid oils (“stale walnuts,” “crayons”), fusty oils (“olive mill waste pond,” “sweaty gym clothes”) and musty ones. Some were so bad the panels graded them lampante, the grade the USDA calls not fit for human consumption without further processing.
| Test | Result for the 15 “extra virgin” samples |
|---|---|
| Common USDA chemistry (acidity, peroxide, UV) | 14 passed |
| USDA sensory standard (trained panels) | 9 failed (60 percent) |
| DAGs test (Australian standard) | All 9 sensory failures also failed |
| PPP test (Australian standard) | 6 of the 9 sensory failures also failed |
| Purity tests for other oils | 1 sample adulterated with canola (plus 1 of the 6 “olive oil” samples) |
The canola
The adulterated samples were both from the Auguri brand, which the report lists as made by Mediterranean Olive Oil Corp. and bought through Sysco. The “extra virgin” sample was over the USDA limit for linolenic acid by 33 percent, for brassicasterol by 3,000 percent and for campesterol by 186 percent, and its stigmastadiene level, a marker of refined oil, was more than 16,000 percent over the limit. That pattern points to refined canola oil. The Auguri “olive oil” grade looked worse, with a higher share of canola, and a second bottle shipped from a different location showed the same thing. Olive Oil Times reported that sample at about 70 percent canola.
Why the lab tests missed it
Acidity, peroxide and UV tests catch oil that is badly made or badly abused. They are cheap and fast. What they do not catch well is old oil, oil from tired fruit, or good oil stretched with a little refined oil. The UC Davis team ran two newer tests, DAGs (diacylglycerols) and PPP (pyropheophytins), which the Australian Olive Association has adopted and which the International Organization for Standardization recognized in 2009. Those matched the tasters much better.
“Results of this study make it very clear that efforts to control quality will likely fail if based only on the most commonly used chemical analyses,” Dan Flynn said in the university’s announcement. The report asks the USDA to add DAGs and PPP to its extra virgin standard, and asks distributors to add tasting, DAGs and PPP to their own checks. The authors call for a larger study. The same pattern showed up in their supermarket studies of 2010 and 2011, and the USDA standards themselves only took effect in 2010 and are voluntary (see USDA Olive Oil Standards Take Effect Today).
What it means for diners and buyers
Foodservice buys on price per ounce, and the report prints what Sodexo paid: most of the extra virgin samples cost between 13 and 29 cents an ounce. It comes weeks after Consumer Reports found similar problems on supermarket shelves (Consumer Reports Olive Oil Test: Only 9 of 23 Pass as Extra Virgin).
- A restaurant cannot easily tell what it was sold either. The chef sees a label and a price, not a lab sheet.
- If the oil on the table smells of crayons or old nuts, it is rancid, whatever the bottle says. Our older piece on fusty, musty, winey and rancid off-flavours explains what you are smelling.
A lab certificate that says “extra virgin” usually means acidity, peroxide and UV passed on the day the sample was drawn, often at the port or at the bottler. Nobody tastes it again after months in a warm storeroom or a sunny dining room. For the cheapest grades of foodservice oil, the certificate is doing a job, but it is not the job the customer thinks it is.
Sources
- UC Davis: Study suggests raising the bar for olive oil quality control
- UC Davis Olive Center: Evaluation of Olive Oil Sold to Restaurants and Foodservice (PDF)
- Olive Oil Times: New Study Finds Some Foodservice Olive Oil ‘Not Fit for Consumption’