Simplicity is a Worthwhile Goal but Does Not Always Provide the Correct Answer

Abstract

Dear Editor, The economic importance of being able to predict how long deep-frying oils can be used, before they have to be discarded, is beyond dispute. It would enable users to choose the most stable frying oils and guide producers in their development of more stable oils. It is therefore with great interest that I studied the article by Liu et al.1 because it claims to provide a novel evaluation index for frying oils that only takes their fatty acid composition into account. Another reason that I studied the article with great interest is that some 20-odd years ago, I wrote a paper2 that also correlates the stability of an oil to its fatty acid composition. The difference between the work by Liu et al. and mine is that they carried out actual frying experiments, whereas I based my work solely on Rancimat Oil Stability Index (OSI) determinations. This difference may be essential since the mechanism that causes the unsaturated fatty acid content of the frying oil to decrease (Liu et al.) may not be the same as the mechanism that causes oil to produce compounds that increase the conductivity of water in the Rancimat apparatus (Dijkstra et al.). Because the paper I wrote2 is not that readily accessible since it dates from before 2000—the year in which Oléagineux, Corps gras, Lipides, the French oils and fats journal (OCL)-back issues became freely accessible via the Internet—I will briefly summarise its findings. These include that the Oil Stability Index (OSI) of oil blends can be expressed as the sum of the antioxidant contents of the blend components divided by the sum of the weighted susceptibilities to oxidation of the fatty acids in the blend. These susceptibilities were determined by measuring the OSI-values of 48 series of binary blends of 20 different commonly used components such as liquid oil and partially hydrogenated oils and fats. Saturated and monounsaturated fatty acids were found to be almost equally insusceptible to oxidation, but linoleic acid and linolenic acid were 130–140 times more susceptible, respectively. Given the magnitudes of these susceptibilities, it is clear that the polyunsaturated fatty acids (PUFA) content of a component or a blend has far more influence on its rate of degradation than its saturated fatty acid or monounsaturated fatty acids (MUFA) content. Apparently, the methylene group in between two double bonds, which provides two bis-allylic hydrogen atoms, is the prime point of attack during an OSI determination; mono-allylic hydrogen atoms as present in MUFAs hardly contribute to oil instability having about the same susceptibility to oxidation as saturated fatty acids. Liu et al.1 on the other hand concluded that “the ratio of palmitic to oleic acid in the oils before deep frying has an important impact on the degradation of a variety of unsaturated fatty acid based oils, and can be used as an evaluation index for predicting the degradation rate of such oils before deep frying.” Now, it could be that this ratio is a roundabout way of expressing the polyunsaturated content of the various oils, but studying the paper by Liu et al. in detail and e-mail exchange with the corresponding and first authors made me conclude that their paper oversimplifies what happens and therefore might well mislead its readers. Hence, this letter and its title. The authors find little variation between the values of A, the range being from 0.20 to 0.26. This is surprising because their oils differ widely in PUFA content: 6.8% for olive oil versus 78.2% for perilla oil, which have A-values of 0.228 and 0.237, respectively, i.e., quite close. How can this be explained? Well, the authors only look at the relative decrease of UFAs and not at their absolute rate of decrease, which is likely to correspond to the rate of formation of polar compounds, a well-accepted criterion for frying oil change. If there are few UFA at the start of the experiment, the absolute rate of formation of polar compounds may be slow but the relative decrease of UFA content can be more or less standard for the simple reason that UFA react the way they do. Accordingly, just looking at the rate at which the relative UFA content decreases is not a correct way of measuring oil deterioration during deep fat frying. The authors used “frozen par-fried French fries obtained from a local food store.” They do not mention in what oil these prefried potatoes had been fried. This is relevant information since it is likely that some oil used in prefrying leaches out into the oil used during the deep frying experiments and that consequently, this leached oil affects the fatty acid composition of the frying oil samples. Therefore, the authors should have measured the fatty acid compositions of the oil retained by the prefries and verified that these were the same for all batches. They should also have determined the amount of oil retained by the prefries since a higher oil content might well lead to increased leaching. In summary, Liu et al.1 aimed at arriving at a simple expression to correlate oil composition with oil deterioration, which is a worthy endeavour, but they oversimplified by not taking the antioxidant content of their oils into account. In addition, they studied the relative decrease in unsaturated fatty acids instead of its absolute decrease and the evaluation index they suggest lacks a chemical justification.