Effects of duration and temperature of heating of sunflower oil on ruminal linoleic acid biohydrogenation in vitro

Feeding cows with heat treated oilseeds results in an increase of milk conjugated linoleic acids (CLA) content compared to feeding raw seeds. Heating leads to a lipid peroxidation, which could reduce the amount of fatty acids (FA) subjected to biohydrogenation and produce peroxides which could act directly on this reaction. The effects of heating duration and temperature, on the biohydrogenation were examined. Sunflower oil was heated at 150°C during 0, 3, 14, or 22 h, or during 3 h at 70, 100, 130, or 160°C. Peroxides in oils were determined by the peroxide index (PI), and fatty acids in samples incubated during 6 h were analysed by gas chromatography. Increasing heating duration was more efficient than increasing heating temperature for peroxides production: PI = 15.9, 41.9, 93.0, and 162.8 mEq/kg for 0, 3, 14, and 22 h of heating, respectively, and PI = 40.9, 35.9, 84.0, and 88.2 mEq/kg for 70, 100, 130, and 160 °C of heating, respectively. For each mEq of PI/kg, FA content decreased by 0.11% (P=0.003; R2=0.177). After incubation, the percents of cis9,trans11-CLA (C18:2c9t11) and trans-vaccenic acid (C18:1t11) were significantly correlated with the heating duration: C18:2c9t11 (% of FA) = -0.040 DURATION (h) + 1.783 (P=0.000; R2=0.830) and C18:1t11 (% of FA) = -0.743 DURATION (h) + 3.711 (P=0.001; R2=0.396). The percent of C18:2c9t11 was significantly correlated with the heating temperature: C18:2c9t11 (% of FA) = -0.007 TEMPERATURE (°C) + 2.457 (P=0.006; R2=0.296), but this relationship was lower than for heating duration, and there was no effect of temperature on C18:1t11 percent, possibly because high heating temperature generated less peroxides than long heating duration. Across the two studied parameters, the relationship between the percent of C18:2c9t11 and the level of peroxides was significant: C18:2c9t11 (% of FA) = -0.007 PI (mEq/kg) + 1.998 (P=0.000; R2=0.494). This 0.35% relative decrease for one mEq of PI/kg, suggested that the effect was mainly due to the peroxides, not only to the decrease of FA content. In conclusion, increasing temperature and mainly heating duration of oil led to a decrease of C18:2c9t11 production in the rumen probably due to the lipid peroxidation. Increasing heating duration also decreased C18:1t11 production. These results suggest that peroxides cannot explain the increase of milk CLA noticed in the milk from cows eating heated oilseeds, and that a good quality fat, without lipids oxidation products, is necessary to increase CLA and C18:1t11 production in the rumen