Lipid Peroxidation Products Are Elevated in Fish Oil

Nutrient Metabolism

Lipid Peroxidation Products Are Elevated in Fish Oil Diets Even in the Presence of Added Antioxidants1 MICHAEL J. GONZALEZ, J. IAN GRAY, RACHEL A. SCHEMMEL, LEROY DÃœGAN, JR. AND CLIFFORD W. WELSCH*2

Departments of Food Science and Human Nutrition and *Pharmacology and Toxicology, Michigan State University, East Lansing, MI 48824

INDEXING KEY WORDS:

â€¢fish oil â€¢lipid peroxidation â€¢antioxidants â€¢BALB/c mice

Polyunsaturated fats can quickly autoxidize at am bient or subambient temperatures (Chan 1987). In the food industry, this problem has been reduced by the addition of antioxidants and by vacuum packaging of foods containing these fats (Buford-Coulter 1988). In recognition of this problem, the American Institute of Nutrition has recommended the addition of a syn thetic antioxidant at a level of 0.02% (by weight of oil) to experimental animal diets containing polyunsaturated fats (AIN 1980). However, diets containing fish oil are more susceptible to autoxidation than diets containing other polyunsaturated fats (e.g., corn 0022-3166/92

$3.00 Â©1992 American

Institute

of Nutrition.

'Supported by research grants AICR-90BW05 and NIH CA42876 to C. W. Welsch and by a predoctoral minigrant of the Department of Food Science and Human Nutrition of Michigan State University, an Equal Opportunity Fellowship from Michigan State University and fellowships from the Government of Puerto Rico and the University of Puerto Rico to M. J. Gonzalez. ^To whom correspondence and reprint requests should be ad dressed. Abbreviations used: BHT, butylated hydroxytoluene; PV, peroxide value; TEA, thiobarbituric acid; TEARS, thiobarbituric acid-reactive substances,- TBHQ, tertiary butylhydroxyquinone.

Received 27 March 1992. Accepted 26 June 1992.

2190

Downloaded from jn.nutrition.org by guest on May 29, 2013

oil) because of their high concentration of long-chain polyunsaturated fatty acids, e.g., eicosapentaenoic acid [20:5(n-3|] and docosahexaenoic acid [22:6(n-3)]. Pritsche and Johnston (1988) reported an extremely rapid autoxidation of diets with added fish oil as measured by the peroxide value (PV).3 Their report, plus an earlier publication by Rasheed et al. (1963), which also reported a high autoxidation in diets con taining fish oil, are two of only a few publications considering the important issue of autoxidation in diets containing fish oil. It is our opinion that in most studies dealing with the role of dietary fish oil in physiological or patho logical processes, the process of dietary autoxidation has been largely ignored. In a recent report (Gonzalez et al. 1991), our laboratory has provided evidence that the inhibitory effect of a fish oil diet on tumorigenic processes may be mediated primarily by a lipid perox idation process. Such evidence has provided the basis for our belief that the importance of autoxidation or lipid peroxidation has been underestimated by those researchers interested in the health impact of fish oil diets. With this in mind, we felt it necessary to further evaluate the issue of autoxidation of diets containing fish oil and the accumulation of oxidation products in organs of mice fed a diet rich in fish oil under normal feeding circumstances.

ABSTRACT Purified corn and fish oil diets with different types and concentrations of antioxidants were evaluated for oxidation products. In addition, a determi nation of different organ and carcass oxidation product levels was performed. Peroxide value and thiobarbituric acid assays were performed on the diets immediately after mixing (0 h) and 24, 48 and 72 h after being fed to mice. The AIN-recommended level of antioxidant ad dition (butylated hydroxytoluene, 0.02 g/100 g oil) and even the addition of 100 times this level (2 g/100 g oil), although decreasing the level of oxidation products, failed to totally prevent oxidative deterioration in diets high in fish oil. Furthermore, other antioxidants added in excess to the fish oil diets also failed to completely suppress oxidative deterioration of the diets and, in ad dition, when fed daily to mice for a period of 4 wk, caused an accumulation of lipid peroxidation products in certain organs (e.g., heart, skeletal muscle, mammary glands) and in the carcass. These results provide evi dence that in the preparation of fish oil diets, the ad dition of antioxidants at the AIN-recommended level, or even levels substantially higher, does not completely suppress oxidative deterioration of experimental diets. J. Nutr. 122: 2190-2195, 1992.

LIPID PEROXIDATION

PRODUCTS

IN FISH OIL DIETS

2191

MATERIALS AND METHODS 1DietTABLE

Diets. Purified diets (Table 1) were formulated ac cording to a slight modification of the guidelines de veloped by AIN (1977). Dry components of the diet (Table 1) were mixed for 15 min and stored at -20''C until used for diet preparation. The fish (menhaden) oil was used upon arrival, with the remaining oil stored in small dark containers sealed with nitrogen gas. Mean PV of the menhaden oil when received was 7.25. After 7 d it was 7.76 (-20'C], and after 14 d the PV was 10.45 (-20Â°C). The menhaden oil diet was

d (no significant increase in diet peroxidation was observed during the 7-d storage period). Thiobarbituric acid and peroxide value analyses of diets. Diets were fed (with free access) to mice to expose diets to actual experimental conditions. One hundred twenty mature female mice (BALB/c), from our breeding colony, were housed in 24 cages (five mice/cage). Mice were housed in a temperature-con trolled (24Â°C)and light-controlled (14 h light/d) room. The protocol complied with the NIH Guide for the Care and Use of Laboratory Animals. Triplicate samples of each diet were obtained for analysis im mediately after mixing (0 h), and six diet samples of each diet group were analyzed for each time period, i.e., 24, 48 and 72 h after being fed. Lipid peroxidation was determined by two methods: the PV (mmol peroxides/kg oil) was deter mined by the method of the American Oil Chemist's Society (1973), and thiobarbituric acid-reactive sub stances (TBARS) (ug TBARS/g sample) were measured using a modified thiobarbituric acid (TBA) assay as previously described (Gonzalez et al. 1991). Modifi cation of the TBA assay was as follows: TBHQ (0.2% of total fat, 2 g/kg sample) was added to 10-g diet samples before homogenization at 4Â°C to prevent further oxidation during the assay. The homogenates were treated with 4 mol/L HC1. A distillate was col-

diet20.00 g Casein3DL-MethionineDextroseSucrose

20.170.3532.1816.09

AIN Mineral mix4 AIN Vitamin mix4 Cellulose5Amountg/100 'All ingredients

4.13 1.18 5.90

(dry components)

were obtained from U.S. Bi

ochemical (Cleveland, OH) except sucrose (ICN Biochemicals, Costa Mesa, CA). 2High fat diets (20% wt/wt) contained 20% corn oil or 19% fish oil (menhaden) + 1% corn oil. Menhaden oil contains 0.03 g/kg of all-rac-a-tocopherol. Corn oil (tocopherol stripped) was obtained from U.S. Biochemical. Fish oil (menhaden) was obtained from Zapata Haynie (Reedville, VA). 3Vitamin free, high nitrogen (14.5%). 4AIN (1977). 5Celufil, non-nutritive bulk.

lected and combined with equal amounts of freshly prepared TBA solution. Samples were heated in boiling water, cooled and the absorbance read on a Spectronic 2000 spectrophotometer (Bausch &. Lomb, Rochester, NY). When diet samples were obtained, they were immediately analyzed in duplicate and triplicate for PV and TBARS, respectively. Organ and whole-body analysis. Lipid peroxi dation was assessed in liver, heart, kidney, mammary glands and skeletal muscle (15 animals/group) and in whole-body carcass (10 animals/group) by a modified TBA assay (Gonzalez et al. 1991). Mice were fed one of the following diets daily (non-consumed diet dis carded daily) for 4 wk: 20% corn oil, 19% fish oil + 1% corn oil or 19% fish oil +1% corn oil with added antioxidants, i.e., TBHQ (2.0%, wt/wt oil) and all-raca-tocopherol acetate (8 g/kg diet). Animals used for whole-body carcass TBARS were killed by carbon di oxide inhalation and stored intact in a plastic zippersealed bag at -20Â°C and analyzed for TBARS the following day. Organs for TBARS analysis were col lected and stored in 9 g/L NaCl at -20Â°C. These were analyzed for TBARS the following day. Statistical analysis. Results were statistically ana lyzed by one-way ANOVA and the Newman-Keuls multiple comparison test (Wilkinson 1989). Sig nificance was set at P < 0.01.

RESULTS At 0 h, the PV of the corn oil diet was lower than that of any of the fish oil diets except for the fish oil


prepared as follows: previously mixed dry ingredients (800 g) were combined with 190 g of menhaden oil and 10 g of corn oil (19% fish oil +1% corn oil). Corn oil was added to fish oil to increase the amount of linoleic acid, a fatty acid essential for normal growth and development. The use of corn oil, a polyunsaturated vegetable fat [-60% linoleic acid, 18:2 (n-6)}, as a control comparison for fish oil was based on its fre quent use in experimental animal dietary studies. These diets, which contained 20% fat (200 g/kg diet), are considered high fat diets. The antioxidant tertiary butylhydroxyquinone (TBHQ) (Eastman Chemical Products, Kingsport, TN) or butylated hydroxytoluene (BHT) (Sigma Chemical, St. Louis, MO) was immediately added to the fish oil at a level of 0.02% (BHT) (wt/wt oil) or 2.0% BHT or 2.0% TBHQ Â±8 g/kg all-rac-a-tocopherol acetate (U.S. Biochemical, Cleveland, OH). Sufficient diet for one day's feeding was packed into small plastic zipper-sealed bags. Each bag was flushed with nitrogen gas and stored at -20Â°C for no longer than 7

compositionIngredient'Fat2

2192

GONZALEZ

ET AL.

TABLE 2 Effect of supplemented dietary antioxidants on peroxide values of fish (menhaden) oil diets1 Time Diet0

h24

h48 hmmol/kg

h72

oil20% oil19%Corn Fish19% 1%oil + oilcorn 0.02% + Fish19% 1%oil oilcorn BHT2+ + 2% Fish19% 1%oil oilcorn BHT+ 2% + Fish19% 1%oil oilcorn TBHQ3+ Fishoil + 1%corn oil+ 2% TBHQ + vitamin

0.17aÂ± 0.34aÂ± 0.88aÂ± 1.46Â«Â± 1.25dÂ± 2.28dÂ± 2.89fÂ± 7.27*+ 0.88CÂ± 0.88CÂ± 1.46eÂ± 2.09e+ 0.58bÂ± 0.58bÂ± 2.48dÂ± 4.42dÂ± 0.29aÂ± 1.15bÂ± 0.58CÂ± 1.20CÂ± E41.3313.576.674.012.071.50Â± 0.29Â»8.6745.3325.6717.0814.0811.67Â± 0.88a11.6771.2039.6730.6722.0117.05Â± 1.16b19.67262.6778.2048.6638.6730.15Â± 0.50b

'Values are means Â±SEM,n - 6. Statistical comparisons made at each time point among diets determined by ANOVA and Neuman-Keuls multiple comparison test. Values with different superscript 2BHT - butylated hydroxytoluene. 3TBHQ - tertiary butyl-hydroxyquinone. 4all-rac-ot-Tocopherol acetate, 8 g/kg diet.

letters in the same column are significantly

increased in all diet groups over time throughout the entire experimental period (0, 24, 48 and 72 h). At 0 h, the dietary TEARS levels were lowest in the corn oil diet and highest in the fish oil diet not supplemented with antioxidants (Table 3). The ad dition of antioxidants (BHT or TBHQ) to the fish oil diet decreased the levels of TEARS when compared with levels in the non-supplemented fish oil diet. Increasing the level of BHT from 0.02 to 2.0% reduced TEARS at all time points. Tertiary butyl hydroxyquinone (2%) was slightly more effective than BHT (2%) in reducing TEARS. The addition of vitamin E along with TBHQ further reduced, albeit slightly, TEARS. However, at 0 and 24 h, the levels of TEARS were higher in fish oil diets at all levels of antioxidant and supplementation than that observed

TABLE

3

Effect of supplemented dietary antioxidants on levels of thiobarbituric acid-reactive substances (TBARS) in fish (menhaden) oil diets1 TimeDiet0 h24h48h72hug

diet20% oil19%Corn Fish19% +oil oil1%corn BHT2+ 0.02% Fish19% +oil oil1%corn BHT+ 2% Fish19% +oil oil1%corn TBHQ3+ 2% Fish19% +oil oil1%com Fishoil +1% corn oil+ 2% TBHQ + vitamin

TBARS/g 0.04aÂ± 0.03bÂ± Â±2.93 Â±6.53 5f0.14e0.04d0.05C0.04b0.7418.951.951.100.910.79Â± 0.45eÂ± 0.67eÂ± Â±1.18 Â±1.68 0.09dÂ± 0.05dÂ± Â±0.70 Â±0.95 0.04CÂ± 0.05eÂ± Â±0.84 Â±0.63 0.03bÂ± 0.04bÂ± Â±0.49 Â±0.67 E40.35 0.05a0.5019.060.970.800.470.37Â± 0.03a Â±0.01a0.33*0.03e0.03d0.03e0.03b0.49 Â±0.03a0.1

'Values are means Â±SEM,n - 6. Statistical comparisons made at each time point among diets determined by ANOVA and Neuman-Keuls multiple comparison test. Values with different superscript letters in the same column are significantly different at P < 0.01. 2BHT - butylated hydroxytoluene. 3TBHQ - tertiary butyl-hydroxyquinone. 4all-nÂ¡c-a-Tocopherol acetate, 8 g/kg diet.


diets containing 2.0% TBHQ Â±vitamin E (Table 2). At 0 h, the PV of the non-supplemented fish oil diet was higher than the PV of the corn oil diet or the PV of any of the fish oil diets supplemented with antiox idants. This trend was observed throughout the other time periods (24, 48, 72 h). Increasing the level of BHT from 0.02 to 2.0% reduced the PV at all time points. Tertiary butylhydroxyquinone (2%) was slightly more effective than BHT (2%) in reducing PV. The addition of vitamin E to the TBHQ-treated group further reduced, albeit slightly, PV. Peroxide value at O h (immediately after diet preparation) was the lowest compared with all other time points (24, 48, 72 h) in both the corn oil and fish oil diets. The rate of PV increase was highest between diet preparation (0 h) and after being fed for a day (24 h). The PV levels

different at P < 0.01.

LIPID PEROXIDATION

PRODUCTS

IN FISH OIL DIETS

2193

TABLE 4 Effect of supplemented

dietary antioxidants on levels of thiobarbituric acid-reactive in organs of mice fed fish (menhaden) oil diets for 4 wk1

DietLiverHeartKidneySkeletal

substances

(TEARS)

glandsCarcassug muscleMammary

tissue20%

TBARS/g

oil19%Corn +19% Fish oil oilcorn +2% Fish oil +vitamin oil TBHQ2 +1%1%corn E30.62

Â±0.98 Â±0.50

0.02a0.55 Â± 0.03C0.35 Â±

Â±0.02b0.63 Â±0.03b0.05C0.03a0.23

Â±0.97 Â±0.56

Â±0.73 Â±0.56

0.03a2.07 Â± 0.07C0.77 Â±

Â±0.03b1.53 Â±0.03b0.04C0.03a0.30 Â±0.03a0.04e0.03b0.38

0.06a2.95 Â± 0.08C2.01 Â± Â±0.05b

Values are means Â±SEM,n - 10 except for heart, kidney and mammary glands, for which 77-6. Statistical comparisons of organs and carcass between diet groups were determined by ANOVA and Neuman-Keuls multiple comparison test. Values with different superscript letters in the same column are significantly different at P < 0.01. 2TBHQ - tertiary butyl-hydroxyquinone. ^all-rac-ot-Tocopherol acetate, 8 g/kg diet.

DISCUSSION The PV of fish oil diets constantly increased with time exposure to air and under normal feeding condi tions. Peroxide formation is likely to occur as sus ceptible polyunsaturated fatty acids are available in the oil (Esterbauer et al. 1986). Increases in PV are catalyzed by free radical attacks at multiple sites. This might explain the observed higher rate of dietary peroxide formation attained during the 0- to 24-h period in which more susceptible sites for free radical attack are likely to be present. The addition of the synthetic antioxidant BHT (0.02%, wt/wt oil) to the fish oil diet significantly reduced lipid peroxidation as compared with the control fish oil diet. The higher concentration of BHT (2.0%) was more effective than

the 0.02% level. Tertiary butylhydroxyquinone at the 2.0% level was slightly more effective in reducing peroxide formation. Tertiary butylhydroxyquinone has been reported previously to be the most effective antioxidant for fish oil (Ke et al. 1977). Nevertheless, peroxide formation still occurred in the fish oil diet even when 100 times the AIN recommended level (0.02%, wt/wt of oil) for a synthetic antioxidant (2.0% TBHQ or BHT) was added. Because peroxides manifest a transitory nature, over time conversion to secondary products of oxidation is favored. Secondary products of lipid peroxidation (TEARS) measured by the TBA assay also displayed an in creasing trend over the first 48 h. However, this trend was not apparent at 72 h, when a decline in TEARS values was observed. In contrast to PV, TEARS for mation depends on specific site oxidation accom panied by serial decomposition (Dahle et al. 1962). Further examination of TEARS at shorter wavelengths (455 nm and 495 nm) suggests the possi bility of a subsequent conversion of these secondary products of lipid peroxidation to other related com pounds, most probably aldehydes such as alkanals and alk-2-enals (Kosugi et al. 1987) in the 72-h sample. Such TEARS determined at shorter wavelengths should be expected to increase as secondary reactions proliferate with continuing oxidation. It is also pos sible that the observed reductions in TEARS values at 72 h may be due, at least in part, to possible reactions between aldehydes (i.e., malondialdehyde and/or alkanals) and proteins (Cuppett et al. 1989). The ad dition of synthetic antioxidants to the fish oil diets resulted in a substantial decrease in dietary TEARS, as was observed for PV, i.e., 2.0% TBHQ was most effective, followed by 2.0% BHT, then 0.02% BHT. Once again, and importantly, the addition of syn thetic antioxidants at a level 100 times greater than the AIN recommended level (0.02%, wt/wt oil) did not totally prevent the formation of oxidation


in the corn oil diet. Dietary TEARS values at 72 h were lower than those at 48 h except for the fish oil diet not supplemented with antioxidants. In all of the foregoing, the TEARS were measured at the standard wavelength of 532 nm. However, an increase in dietary TEARS was observed at wavelengths 455 nm and 495 nm at 72 h as compared with 455 nm and 495 nm wavelength readings at 0, 24 and 48 h time points (data not shown). The TEARS concentrations of organs (liver, heart, kidney, mammary gland, skeletal muscle) and whole bodies of mice fed the fish oil diet without antiox idants were consistently higher than that observed in organs and whole bodies of mice fed the corn oil diet (Table 4). The addition of high levels of antioxidants to the fish oil diet reduced TEARS concentrations in all organs and in whole body (carcass). However, TEARS concentrations in heart, skeletal muscle, mammary glands and whole body (carcass) of mice fed fish oil containing antioxidants were still higher than TEARS values of mice fed a corn oil diet.

2194

GONZALEZ

levels of polyunsaturated fatty acids seems to be in sufficient to completely inhibit oxidative deterio ration when such diets have fish oil as their fat base. Thus, a distinct recommendation for fish oil diets compared with other polyunsaturated fat diets seems appropriate. Meanwhile, to minimize oxidative de terioration in experimental diets with high levels of fish oil, a synthetic antioxidant should be added in quantities substantially above the AIN recommenda tions. The PV and TEARS values should be monitored throughout all feeding experiments involving dietary fish oil. Animals must be provided a fresh diet at least every 24 h in clean food jars; the remaining nonconsumed diet should be discarded. In addition, the mixing of diets in an inert atmosphere (e.g., nitrogen or carbon dioxide) should be mandatory. Freshly pre pared diets should be stored in individual zippersealed bags in an amount sufficient for a single day's feeding and kept at -20Â°C until fed. These recommen dations will substantially reduce, albeit not com pletely eliminate, excessive oxidative deterioration of experimental diets incorporating high levels of fish oil. Research laboratories studying the effects of dietary fish oil on physiological and/or pathological processes should be cognizant of these concerns.

ACKNOWLEDGMENT We thank Marlene Green for the preparation manuscript.

of the

LITERATURECITED American Institute of Nutrition (1977| Report of the American Institute of Nutrition ad hoc committee on standards for nutri tional studies, f. Nutr. 107: 1340-1348. American Institute of Nutrition (1980) Second report of the ad hoc committee on standards for nutritional studies. J. Nutr. 110: 1726. American Oil Chemists' Society (1973) Official and Recommended Practices of the American Oil Chemists' Society. Vol. 1, Methods: cd 8-53 and Ba 3-38. AOCS, Champaign, IL. Buford-Coulter, R. (1988) Extending shelf life by using traditional phenolic antioxidants. Cereal Foods World 33: 207-210. Chan, H.W.S. (1987) The mechanism of autoxidation. In: Autoxi dation of Unsaturated Lipids (Chan, H.W.S., Ã©d.), pp. 1-16. Aca demic Press, London, U.K. Cuppett, S. L., Gray, J. I., Booren, A. M., Price, J. F. & Stachin, M. A. (19891 Effect of processing variables on lipid stability in smoked Great Lakes whitefish. J. Food Sci. 54: 52-54. Dahle, L. K., Hill, E. G. & Holman, R. T. (1962) The thiobarbituric acid reaction and the autoxidation of polyunsaturated fatty acid methyl esters. Arch. Biochem. Biophys. 98: 253-255. Esterbauer, H., Benedetti, A., Lang, J., Fulceri, R., Fauler, G. & Comporti, M. (1986) Studies on the mechanism of formation of 4-hydroxyl-nonenal during microsomal lipid peroxidation. Biochim. Biophys. Acta 876: 154-166. Pritsche, K. L. & Johnston, P. V. (1988) Rapid autoxidation of fish oil in diets without added antioxidants. J. Nutr. 118: 425-426. Gonzalez, M. J., Schemmel, R. A., Gray, J. L, Dugan, L., Jr., Sheffield, L. G. & Welsch, C. W. (1991) Effect of dietary fat on growth of MCF-7 and MDA-MB 231 human breast carcinomas


products in the fish oil diet. The TEARS values in the antioxidant-supplemented fish oil diets were still higher than that observed for the corn oil diet. In regard to the oxidative susceptibility of the dietary oils used in this study, it is clear that corn oil [-60% linoleic acid, 18:2(n-6)] is considerably less prone to autoxidation than is fish oil. This difference is no doubt due to the high concentration of long-chain polyunsaturated fatty acids (i.e., eicosapentaenoic acid and docosahexaenoic acid) in fish oil; such fatty acids are absent in corn oil. Significant differences in lipid peroxidation were found among fish oil diets prepared from oils originating from separate batches from the same sup plier (data not shown). The mean PV in freshly ob tained fish oil in our studies was 7.25 (for comparison purposes, fresh stripped corn oil had a mean PV value of 0.97). This particular problem may prove to be very difficult to solve and may be due to an array of variables such as storage, processing, air and light exposure, temperature and seasonal variations (e.g., type of phytoplankton available for fish consump tion). The rapid addition of a synthetic antioxidant to the oil as soon as the oil container is opened should deter or decrease further deterioration of the oil. Organs and whole-body carcasses of mice fed the corn oil and fish oil diets (Â±antioxidants) were also evaluated for TEARS. The TEARS values obtained from certain organs and carcasses were directly related to the concentration of oxidation products of the ingested diet, although it is conceivable that this process can be further enhanced in situ. Even large doses of antioxidants (lOOx the AIN recommended level of a synthetic antioxidant and approximately 30x the amount of tocopherol acetate supplied by the vitamin mix) added to the fish oil diet failed to prevent the accumulation of secondary products of oxidation in certain organs (heart, skeletal muscle, mammary gland) and in whole bodies (carcass) of mice fed the supplemented fish oil diet. Leibovitz et al. (1990) reported similar results in tissue slices ob tained from rats fed a fish oil diet. They reported a decrease in secondary oxidation products in tissue slices of organs from rats receiving high levels of dietary vitamin E, but they were also unable to fully suppress the formation of these oxidation products. In contrast to what was observed in our study with heart, skeletal muscle and mammary glands, the livers and kidneys of mice fed fish oils containing antioxidants had significantly lower TEARS values than did the livers and kidneys of mice fed corn oil. Although the exact explanation is not certain, it is conceivable that at least for the liver, the capacity for storage of antioxidant lipid-soluble components and/ or the function of this organ in the production of lipoproteins may explain this phenomenon. In summary, based on the present results, the ad dition of the AIN recommended level of a synthetic antioxidant (0.02%, wt/wt oil) to diets with high

ET AL.

LIPID PEROXIDATION

PRODUCTS

in athymic nude mice: relationship between carcinoma growth and lipid peroxidation levels. Carcinogenesis 12: 1231-1235. Ke, P.}., Nash, D. M. & Ackman, R. G. (1977) Mackerel skin lipids as an unsaturated fat model system for the determination of antioxidative potency of TBHQ and other antioxidant com pounds. J. Am. Oil Chem. Soc. 54: 417-420. Kosugi, H., Kato, T. & Kikugawa, K. (1987| Formation of yellow, orange and red pigments in the reaction of alk-2-enals with 2-thiobarbituric acid. Anal. Biochem. 165: 456-464.

IN FISH OIL DIETS

2195

Leibovitz, B. E., Hu, M. L. & Tappel, A. L. (1990) Lipid peroxidation in rat tissue slices: effect of dietary vitamin E, corn oil-lard and menhaden oil. Lipids 25: 125-129. Rasheed, A. A., Oldfield, J. E., Kaufmes, J. &. Sinhuber, R. O. (1963) Nutritive value of marine oils: menhaden oil at varying oxi dation levels with and without antioxidants in rat diets. J. Nutr. 79: 323^32. Wilkinson, L. (1989) Systat: The System for Statistics. Systat, Inc., Evanston, IL.
