1 Introduction
Marine oil products are valued for their high eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) content. These fatty acids have been shown to exert positive health effects, such as reducing inflammation (Allaire et al., 2016) and lowering blood triglycerides (Zulyniak et al., 2016). However, the highly unsaturated structure of these fatty acids makes them susceptible to lipid oxidation, a deteriorative process that leads to off-flavors and off-odors. If consumed, the oxidized products may have negative health consequences (Esterbauer, 1993). For regulatory and quality control purposes, it is common practice to monitor the level of oxidation in marine oils. This is typically achieved by measuring multiple oxidation markers, such as peroxides and carbonyls. Peroxide value, as a primary oxidation indicator, can be measured using the American Oil Chemists’ Society (AOCS) titration-based method (Cd 8b-90). Carbonyls, generated during hydroperoxide decomposition, are often used as secondary oxidation markers. Volatile carbonyls, such as hexanal, are commonly quantified using gas chromatography methods. Non-volatile carbonyls can be measured using the 2,4-Dinitrophenylhydrazine (DNPH) assay, thep -Anisidine Value (pAV), or the thiobarbituric acid reactive substances (TBARS) test.
Concentrations of non-volatile carbonyls, primarily unsaturated aldehydes such as 2-alkenals and 2,4-alkadienals, are often assessed using AOCS Official Method for pAV (Cd 18-90). This is a spectrophotometric method based on the absorbance of light by an imine chromophore that forms from the reaction of aldehydes with thep -anisidine reagent. Flavoring agents added to marine oils interfere with the test, as aldehydes in flavors also react with thep -ansidine reagent and form additional imine chromophores, leading to inaccurate results (The Global Organization for EPA and DHA (GOED), 2018; Jackowski et al., 2015; Semb, 2012) and incorrect conclusions about the shelf-life and quality of the oil products. Unexpectedly high pAV have been reported in previous studies in flavored marine oil products (Albert et al., 2015; Jackowski et al., 2015).
To compensate for the increased pAV caused by flavors, GOED (2018) recommends a protocol to assess the pAV of flavored marine oils. There are four major steps: 1) measure pAV of an unflavored base oil (pAV); 2) determine the maximum allowable pAV increase in this oil (ΔpAV); 3) add flavor to the base oil and measure the pAV of this flavored oil (pAV*); 4) determine the maximum allowable pAV of the flavored oil (pAV*max), which constitutes a useful reference value for the shelf-life testing. ΔpAV is calculated by subtracting the pAV of the fresh unflavored base oil from the maximum acceptable pAV suggested by GOED. pAV*max is calculated by adding ΔpAV to pAV*. This recommendation is based on the assumption that the flavors themselves do not degrade or in any way change influence on the measured pAV over time (GOED, 2018). If this assumption is violated, estimates of the oxidation level will be inaccurate. However, this is very likely to occur; several common flavor compounds, such as citral (Djordjevic et al., 2008; Liang et al., 2004; Schieberle & Grosch, 1988) and vanillin (Mourtzinos et al., 2009), have been shown to degrade under a variety of conditions.
In this study, we tested the null hypothesis that flavors do not change their contribution to the measured pAV over the course of oxidation. To do this, we first evaluated 14 flavors to identify those with the greatest contribution to the measured pAV and would thus be most likely to change in observable ways over the course of oxidation. Following this, we performed a series of accelerated stability studies to compare oil samples to which flavor had been added either before or after oxidation. The pAV of all samples was evaluated at several sampling points during oxidation. We anticipated finding differences between the measured pAV of the two sample types, which would suggest the GOED recommendation may not apply to these flavored marine oils, and that an alternative protocol is needed to evaluate the aldehydes that result from lipid oxidation.