DPPH Radical Scavenging Activity of Cold-Pressed Oils
Table 2 shows the DPPH antioxidant activity of cold-pressed oils expressed in TEAC, for both hydrophilic and lipophilic fractions, as well as for oils not subjected to extraction. The DPPH radical scavenging activity of the studied oils ranged from 0.17 up to 2.32 mM TEAC/kg. The lipophilic fraction of all the studied oils was characterized by much higher activity than its hydrophilic equivalent, which is corroborative with previous studies and reflects more significant amounts of lipophilic antioxidants (tocopherols) than hydrophilic ones (phenolic compounds) present in oils . The ratio LF/HF ranged from 1.31 (MACO) up to 7.60 (SAFO). Apart from SAFO, LINO and FLAO were also characterized by high LF/HF ratios (7.37 and 7.06, respectively). A similar result was obtained, for FLAO only, by Tuberoso et al. . In individual samples of SESO and WALO, high LF/HF ratios were observed once more (9.36 and 7.34, respectively). ROSO was characterized by the largest antioxidant activity (2.32 mM TEAC/kg), which may result from the high content or synergistic activity of antioxidant compounds in this oil. Data on antioxidants occurring in ROSO is extremely sparse, however, it has been found that ROSO contains considerable amounts of carotenoids (46–145 mg/kg) .
Fatty Acid Composition
Table 3 shows the fatty acid composition of the fresh oils studied. The highest content of C18:2n-6 was found in SAFO, LINO and POPO, of C18:3n-3 in FLAO and of C18:1n-9 in MACO and AVO. MACO contained the highest amounts of C16:1n-7 among the studied oils, and HEMO was distinguished by approximately 2 % of C18:3n-6. Fatty acid composition in oils was in agreement with the previous data [12, 20–24]. However, the main fatty acid content in SAFO was found to be outside the limit of the range presented in the literature: 20.6 %o C18:1n-9 (typical content 11–16 %) and 67.3 % of C18:2n-6 (typical content 72–79 %) [25, 26]. Table 4 presents SFA, MUFA, PUFA and TFA content in fresh and stored oils. A large variability of MUFA and PUFA contents between the analyzed brands of SAFO, PUMO, FLAO and HEMO was observed (relative standard deviation was up to 33.3 % for MUFA in HEMO). Such variability of fatty acid composition of these oils can be found in literature [12, 20, 27–31]. ROSO was characterized by the lowest SFA and highest PUFA contents (an especially high percentage of alpha-linolenic acid), suggesting its susceptibility to oxidation. However, no significant changes in fatty acid contents in ROSO, or other studied oils, were found during storage. In the correlation test we also observed the effect of the antiradical activity of oils on inhibition of PUFA deterioration, expressed as percentage of change of PUFA content after 12 months of storage (Table 5). The TFA content was very low in fresh oils (0.1–0.7 %) and only a slight increase could be observed during storage, as the highest value did not exceed 0.13 % at the end of their shelf life. From the correlation test we can conclude that the antioxidants of oils could protect from trans isomerization of fatty acids for up to 6 months of storage.
Oxidative Stability Parameters of Oils
The acid value (AV) measures the content of free fatty acids formed upon the hydrolytic degradation of lipid molecules, thus contributing to the reduction of the shelf life of the oil . The AV of fresh and stored oils are shown in Table 6. The acid value of each of these cold-pressed oils in each of the indicated periods of storage was within the limit of up to 4 mg KOH/g of oil, according to the Codex Alimentarius Commission standard for cold-pressed and virgin oils . Gorjanović et al.  found considerably higher concentrations of acids in the fresh samples of PUMO, as high as 1.75 mg KOH/g. Other authors, too, reported higher levels of this parameter in SESO, WALO, LINO, FLAO and SAFO [22, 34, 35] in comparison with our results. Assessment of the relationship between the antiradical activity of fresh oils and oxidative stability parameters as measured during the shelf life of oils showed a statistically significant positive correlation of AV values in fresh oils with DPPH values in nonfractionated and fractionated oils (Table 5). These effects were constant for 6 months of storage. However, a significant negative correlation between antiradical activity of the lipophilic fraction of oil and percentage of AV change after 3 months of storage suggests that lipophilic antioxidants decelerate aldehyde formation in oils. It was observed both for lipophilic and hydrophilic fractions of oils, and also in nonfractionated oil after 12 months of storage. In non-refined cold-pressed oils, adverse relations were observed between acidity and DDPH values . However, the studied oils could have been subjected to refining methods commercially used in manufacturing cold-pressed oils, such as deacidification. This could result in a decrease in AV as well as antioxidant contents, so this process could influence the observed relationship .
PV defines the content of lipid hydroperoxides in oils formed under conditions of auto- and photo-oxidation. All the oils under study (fresh and stored) were characterized by low mean values of PV (Table 7), and none of them exceeded the recommended limit for cold-pressed oils of 15 mequiv O2/kg . In the majority of the tested oils, increased PV value was observed after 3, 6, and 12 months, but a statistically significant difference was reported only in SESO (up to 191 %) and POPO (up to 145 %) at the end of its shelf life. A wide range of PV in fresh AVO brands was noted (4.42–15.74); moreover in two individual samples of AVO, this parameter was slightly above the recommended level. Nevertheless, the level of hydroperoxides remained unchanged throughout the whole period of AVO storage. The data published so far regarding PV in the oils covered by this study, are scarce and only limited to oils that were not subjected to storage. Gorjanović et al.  reported the PV for PUMO pressed from three Cucurbita pepo varieties was at a lower range (3.44–5.54 mequiv O2/kg), while Wroniak et al.  and Czaplicki et al.  quote values for SESO, FLAO, SAFO and WALO 1.5—three times higher than those presented in this work. In our study, a linear decrease of PV values in nonfractionated and fractionated oils was observed in fresh oils, and also after 3 months of storage, as the DPPH value increased in nonfractionated and fractionated oils (Table 5). Longer shelf life did not result in this correlation in fractionated oils, moreover after 6 months of storage, higher DPPH values in the lipophilic fractions were accompanied by an increase in percentage of PV. No correlation was observed in hydrophilic fractions in either fresh or stored oils. So the use of DPPH protocol to predict hydroperoxide formation is limited to the lipophilic fractions of oils and cannot be a hallmark of oxidative resistance during longer shelf life.
The p-AV reflects the content of secondary products of lipid oxidation, resulting from the decomposition of hydro-peroxides. p-AV along with PV may therefore offer elucidation of the rancidity of oils . The lowest p-AV of fresh oils was found in all SESO brands (range 0.2–0.3) and the values did not change during storage (Table 8). The largest variability of p-AV in fresh oils occurred between brands of ROSO (4.22–11.88) and PUMO (1.48–8.17). The highest peak was found in ROSO, with an average 68 % increase in the sixth months of storage. Nevertheless, this increment was not statistically significant. The health safety of oils in relation to p-AV is difficult to assess because of the lack of an established limit of this parameter in cold-pressed oils. In reviewing the literature, the data on the p-AV of the oils analyzed in this work were very limited. The results of p-AV determination in commercially available cold-pressed oils obtained by Wroniak et al.  showed lower values for SAFO (0.23), LINO (0.36) and FLAO (0.48), and higher ones in PUMO (6.92) and WALO (6.07) than those found in our study. A significant negative correlation between DPPH values in nonfractionated oils under study and p-AV of the stored oils was shown after 12 months of shelf life, which indicates that the secondary oxidative product formation in cold-pressed oils rich in antioxidants is very slow (Table 5) .
The formation of hydroperoxides from PUFA in the early stages of oxidation may result in double bond isomerization. The determination of conjugated dienoic and trienoic fatty acid derivatives (CD and CT) enable definition of the oxidation state of an oil, in addition to PV and p-AV .
The CD content in fresh oils ranged from 1.86 (% E) in MACO up to 4.31 (% E) in PUMO (Table 9). The lowest average CD content of MACO is related to its specific fatty acid composition, which is meager in linoleic acid. In three oil types rich in linoleic acid, POPO, MILO and LINO, the CD content increased significantly after 6 months of storage, (mean increase up to 86 %). Wagner et al.  found that a rapid increase in CD generation in POPO stored for 6 days at 40 °C was the result of mechanical damage to the poppy seeds used for the oil production. This effect was also accompanied by a considerable increase of p-AV.
The CT content of fresh oils ranged from 0.18 to 2.96 % E (Table 7). MACO and POPO (both 0.18 % E) were characterized by the lowest CT level, undoubtedly related to a scarce concentration of α-linolenic acid [24, 28]. However, fresh PUMO, with low levels of this fatty acid (0.3 %), was characterized by the highest average CT value (2.96 % E, equivalent to ca. 0.3 % CT in total fatty acids) indicating that the high CT value of PUMO cannot simply be attributed to α-linolenic acid oxidation . The study shows rather constant CT values of oils, with no signs that high-linolenic oils have a higher susceptibility to isomerization during the whole period of their shelf life. Moreover the correlations between DPPH values and percentage changes in the CD and CT of the studied oils could suggest that antioxidants prevent the formation of conjugated trienoic acid and do not inhibit diene isomerization throughout the whole period of storage (Table 5).
PUFA/MUFA ratio and oil oxidative stability
As the oils rich in PUFA were found to have high antiradical activity, a similar relationship was observed between the PUFA/MUFA ratio and oxidative stability parameters as in the case of DPPH values (Table 5).