Toward determining fat quality parameters of fish oil by means of ¹H NMR spectroscopy

Abstract

The consumption of fish oil-based dietary supplements has increased dramatically in recent years, which is mainly due to their high content of health promoting omega-3 fatty acids. However, these polyunsaturated fatty acids are extremely prone to oxidation. This study investigated the potential of 1 H NMR spectroscopy for the assessment of the oxidative deterioration of fish oils. Nine raw fish oils of different fish species were stored for 3 months at room temperature at different degrees of sunlight exposure as well as under standardized accelerated storage conditions for 6 days (at 40°C and under constant light exposure). Fish oil samples were analyzed by 1 H NMR spectroscopy as well as by traditional methods to determine the fat quality parameters peroxide value (PV), anisidine value (AnV), TOTOX value, and acid value (AV). PLS (partial least squares) regression models were used to predict these fat quality parameters based on the 1 H NMR spectra. The best regression models reached an R 2 of 0.949, 0.962, 0.991, and 0.977 for PV, AnV, TOTOX value, and AV, respectively. In conclusion, 1 H NMR spectroscopy is a promising approach for a fast, reliable, and sustainable assessment of fish oil quality with regard to lipid oxidation. Practical applications:The results indicate a great potential of 1 H NMR spectroscopy in the quality assessment of fish oils. This technology requires little time, work, amount of sample, and solvent and provides extensive information that can be obtained from a single spectrum. Such an approach is significantly more specific and detailed than traditional lipid oxidation parameters.

Assessment of Fish Oil Oxidation Using Low-Field Proton Nuclear Magnetic Resonance Spectroscopy

“…Additionally, parts of triglycerides previously used to measure oxidation were also included. These results again mimic previous high-field studies of fish oil oxidation. , …”

“…Similar peaks have been used in modeling of fish oil and DHA oxidation using high-field NMR. 6,8 The linear model was able to predict the PV more accurately than PLSR, with a higher R 2 value and lower error. The PLSR model included more integral bins than the linear model.…”

“…High-field nuclear magnetic resonance (HF-NMR) has been proposed to measure oxidation in fish oil previously. − Figure shows an example triglyceride molecule labeled with protons commonly measured in NMR oxidation studies. Previous methods have monitored integrals of olefinic, bis-allylic, and terminal methyl group protons, which generally decrease over time as hydrogens are abstracted during initiation of oxidation. , As a reference, aliphatic protons from saturated and monounsaturated fatty acids or the glyceryl protons in the sn-2 or sn-1 and sn-3 positions of triglycerides have been used, as these are expected to remain relatively unchanged during oxidation .…”

“…Peaks corresponding to oxidation products including peroxides, conjugated dienes, aldehydes, and alcohols have been measured . In addition to monitoring specific peaks, binning of spectra and multivariate data analysis methods such as partial least squares regression (PLSR) have been used. , These methods summarize the entire spectra into components that can be used to predict a reference parameter such as the peroxide value or p -anisidine value.…”

Study on the Volatile Oxidation Compounds and Quantitative Prediction of Oxidation Parameters in Walnut (Carya cathayensis Sarg.) Oil

“…Considering this, the R 2 value of 0.99 was quite favorable for quantification model in this research. Giese reported PLSR model was used to predict PV, AnV, TOTOX, and AV in fish oil by 1 H NMR spectra and the best regression models reached an R 2 of 0.949, 0.962, 0.991, and 0.977, respectively. Zhang reported that the PLS models performed better than the PCR models in quantitative prediction of AV of edible oils by UV–vis spectrometry, with R 2 values of 0.9904–0.9977.…”