The response of rats to long-term feeding with diets containing oxidized fat . 1 . Thermooxidative changes in fat , body weight gain , feed consumption and utilisation *

A 5:3:2 mixture of rape seed oil, poultry fat, and beef tallow was saturated with oxygen and heated at 70-80°C for 48 to 96 h to obtain batches of fat having peroxide values of 5.1, 50.0, 100.5, and 208.6 meq 02/kg and respective anisidine values of 2.6, 12.1, 22.2, and 95.9. Six casein diets containing 10% oxidized fat with peroxide values of 5, 40, 80, 120, 160 or 200 meq 02/kg were prepared and fed for 8 weeks each to 12 individually housed rats with an initial body weight of 58.5 g. Body weight gain, feed intake and utilisation, PER, protein and fat digestibilities, VFA content in the caecum of two groups of rats, and P-glucuronidase activity in the caecal digesta, and phosphatase activity in the mucosa of the small intestine were measured. The degree to which fat was oxidized did not affect feed intake, body weight gain or protein digestibility. Feed utilisation was, however, lower in the groups fed diets with peroxide values of 160 and 200 meq 02/kg. Fat digestibility was lower only with the highest degree of oxidation. Phosphatase activity decreased while that of glucuronidase rose as the degree of fat oxidation increased. The VFA concentration in the digesta of the caecum of rats fed the diet with the most oxidized fat was significantly lower than in the control group, which points to the unfavourable effect of oxidized fat on bacterial activity in the large intestine.


INTRODUCTION
The component of feeds and food that is most susceptible to oxidation is fats.Numerous studies have provided much evidence for the very harmful effects of prolonged heating of fats (for instance, during frying) leading to the formation of polymerized and cyclic products (Corcos Benedetti et al., 1987;Lopez-Varela et al., 1995;Sanchez-Muniz et al., 1998).There is less information, however, on the negative effects of oxidizing fat at lower temperatures, e.g. as when fat is stored improperly or heated for short time during the production of meat meal.Earlier experiments of many authors have provided conflicting results on the nutritional and toxic effects of oxidized fat on animals (Carpenter et al., 1966).According to more recent experiments of Hochgraf et al. (1997), a 10% content of fat oxidized to 1300 meq 02/kg at a temperature of 37°C caused the growth rate of rats to decline mainly as a consequence of lower feed intake, but at equal intakes, differences in body weight of rats fed on fresh and oxidized fat were not significant.Hayam et al. (1993) reported contradictory results and did not find differences in feed intake and oil digestibility, but did find significant differences in the body weight between rats fed diets containing soyabean oil fresh or oxidized to 420 meq 02/kg.
Similar results were obtained by Sanchez-Muniz et al. (1998) who found that at similar feed intake weight gains were smaller in rats fed highly oxidized than fresh soya oil.According to the authors the lower growth rate was caused by the smaller linolenic acid content in oxidized oil.The diversity of the results of reported experiments justifies the further studies.
The objective of this study was to investigate the effect of feeding thermooxidized fats on the intake and utilisation of feed, growth of animals, and selected indicators of gastrointestinal tract function of rats (fat and protein digestibilities, activity of selected enzymes, large intestine microflora).A mixture of fats applied in animal nutrition was used; oxidation was conducted under conditions of intestive oxygen access similar to those found in fat-supplemented feed mixtures.

Animals and diets
Seventy-two male Wistar rats with body weights of 58.6±3.1g,housed under standard conditions (21-22°C, relative humidity 50-70%) were used in this experiment.The rats were divided into 6 groups of 12 each and were fed for 8 weeks with casein diets containing 10% fat oxidized to various degrees (Table 1).The diets contained about 13% crude protein.The amount of energy from particular dietary components (protein, 12%; carbohydrates, 48%; fat 40%) was similar to the Western-type diet (Dabai et al., 1996).

Preparation of fat
A mixture composed of rape seed oil (50%), poultry fat (30%), and beef tallow (20%) was saturated with oxygen and intensively aerated while heating at 70-80°C for 48 to 96 h.The degree of oxidation was monitored twice daily by determining the peroxidation value expressed in miliequivalents of oxygen per kg of fat (meq 02/kg).Batches of fat in which the peroxidation value exceeded 50, 100 and 200 meq 02/kg were used for the experiment.The fat was stored at 4°C.The diets were prepared weekly.Fat with the required peroxidation value (5,40,80,120,160 i 200 meq 02/kg) was obtained by mixing fats of various degrees of oxidation in the appropriate proportions.

Feeding experiments
Feed intake, body weight gains, and feed utilisation ratio were computed for the first (weeks 1-4) and second (weeks 5-8) periods of the experiment.The protein utilisation ratio (PER) was calculated for the first period.In the fourth week of PERFORMANCE OF RAT FED OXIDIZED FAT the experiment faeces were collected to determine protein and fat digestibilities.The true digestibility of protein was computed adjusting for metabolic nitrogen according to Rakowska et al. (1978).After the experiment was completed the rats were anaesthetized with urethane (Sigma) and the first segment of the small intestine and the caecum with digesta were sampled.

Analytical methods
The degree of fat oxidation was assessed on the basis of peroxide, acid, and anisidine values.The peroxide was determined according to AOAC (1990), the acid value (AN) according to Pr MN-ISO 660 (1996), and the anisidine value according to ISO 6885 (1988).Alkaline phosphatase (AP) activity in the mucosa of the small intestine was determined in a homogenate prepared according to Franson et al. (1985) using a diagnostic kit from Alpha Diagnostics, ^-glucuronidase acitivty in the large intestine digesta was determined by a modified version of the method by Andrieux et al. (1992) with an enzyme standard and 4-nitrophenyl-(3-D-glucuronide as the substrate.The VFA content in the caecum of rats (from the control and VI group) was determined by gas chromatography (Ziotecki and Kwiatkowska, 1973), fatty acids were separated and quantified by GLC on a Pye Unicam 104 model gas chromatograph with a flame ionisation detector.
The results were subjected to statistical analysis using one-way analysis of variance and Duncan's multiple range test.

RESULTS AND DISCUSSION
The characteristics of fats oxidized to various degrees are presented in Table 2. Oxidation at 70-80°C caused a rapid increase in primary oxidation products (lipid peroxides) and a slower increase in secondary products (aldehydes, ketones, and others).This is indicated by the peroxide values (5 to 200 meq 02/kg) that is many times higher than the anisidine values (2.6 to 95.9).An inverse proportion of peroxide and anisidine values is found in fats subjected to intensive heating without additional aeration.In the experiment of Tynek and Drozdowski (1998) after 4.5 and 45 h of frying fat, the respective peroxide values were 8.9 and 5.4 meq 02/kg, while the anisidine values were 113.7 and 167.9, respectively.In these conditions, the formation of polymers is accompanied by distinct losses in unsaturated fats (Corcos Benedetti, 1987).In our experiments, we did not find significant changes in the composition of fatty acids as the result of oxidizing fat (Table 2).The tendency towards a reduction in the proportion of C183 and C182acids, and rise in C181 and C18.0 acids was similar to the results of other authors (Yoshida and Kajimoto, 1989;Hayam et al., 1993;Hochgraf et al., 1997), who oxidized fat under similar conditions.Batches of fat prepared for the experiment differed the most in terms of their peroxide contents, to a lesser degree in their contents of secondary oxidation products, and did not differ at all in their fatty acid composition (Table 2).The various degrees of oxidation of fat in the diets did not cause significant differences in protein digestibility (Table 3), which ranged from 92.5 to 94.1% and corresponded with the values given by Eggum (1973) for casein diets.Significant lowering (P<0.05) was found, however, of the apparent digestibility of fat with the highest peroxide value (200 meq 02/kg) in comparison with the control group: 96.9 vs 97.8%.A much larger decline in fat digestibility was found by Corcos Benedetti (1987) in rats fed a diet containing a high proportion (over 70%) total fat) of oxidized and polymerized soyabean oil fraction (98 vs 68%).The presence of peroxides from oxidized fat caused the alkaline phosphatase activity (AP) in the mucosa of the small intestine (Table 4) to decline from 60.5 U/g in the control group to 50.6-55.2U/g in the experimental groups, although the change was not significant statistically.The lower activity of this enzyme (which affects the absorptive functions of the intestinal mucosa) corresponded with the lower digestibility of fat in group VI, in which rats received fat with a peroxide value of 200 meq 02/kg.
The activity of P-glucuronidase in the caecal digesta increased from 0.48 U/g in the control group to 0.65 U/g in group VI (nonsignificant difference).This may indicate that the oxidized fat caused unfavourable changes in the composition of the microflora of the large intestine.The increased activity of P-glucuronidase that occurs, among others, when the fat content of the diet is increased (Manoj et al., 1993), can cause the cellulolytic activity of large intestine microflora to decline (Indira et al., 1980).This suggestion is corroborated by the difference of short-chain fatty acids content in the caecal digesta of rats fed diets differing most in degree of fat oxidation (groups I and VI).The use of highly oxidized fat in the diet reduced the sum of volatile fatty acids (C2-C5) in comparison with the group receiving non-oxidized fat (Table 5).The oxidized fat caused an approximately 2% reduction in the proportion of acetic and propionic acids in the sum of total short-chain fatty acids, and similar increase of the proportion of butyric and valeric acids.
Adding highly oxidized fat to the diet did not cause a reduction in feed intake (Table 6).Other investigators have obtained different results using polymerized soyabean oil (Corcos Benedetti et al., 1987) or oil oxidized to 1300 meq 02/kg (Hochgraf et al., 1997).In our experiment, when fat with a low peroxide value was replaced with one having higher number (160 and 200 meq 02/kg) we found a reduction in PER from 3.08 to 2.96 and 2.98, respectively.In the first 4 weeks of the experiment FER was also significantly reduced in groups V and VI, but in the second period of the experiment this ratio did not differ.Throughout the whole experiment, FER was higher in group V. Hayam et al. (1995) reported poorer feed utilisation in a 7-week experiment on rats fed fat oxidized to 420 meq 02/kg.Also in an experiment by Sanchez-Muniz et al. (1998) the addition of 15% highly oxidized sunflower oil, which had four times more of the polar fraction than non-oxidized oil, to the diets of rats did not affect feed consumption although a significant decline in body weight gain, poorer feed utilisation, and lowering of PER were noted.In the experiments of the authors cited above mainly secondary products of fat oxidation, including polymers of fatty acids, had a negative effect on feed consumption and utilisation.

CONCLUSIONS
The thermooxidative changes in fats occurring under conditions of intensive oxygen access were mainly a rapid rise in the primary products of oxidation (expressed as the peroxide value) and slower increase in secondary oxidation products (anisidine value).Oxidized fat did not reduce feed intake, body weight gain, or protein digestibility.A significant reduction in the digestibility of fat occurred only in the group receiving the most oxidized fat (200 meq 02/kg) and corresponded with a reduction in phosphatase activity in the small intestine of rats in this group.A rise in the activity of P-glucuronidase and lowering of the concentration of volatile fatty acids in the digesta of the caecum indicates that a high level of products of fat oxidation in the diets had an adverse effect on the activity of the microflora of the large intestine.The use of highly oxidized fats (160 and 200 meq 02/kg) worsened feed utilisation in the first 4 weeks of the experiment.

TABLE 5
Sum and composition of volatile fatty acids in the caecal content 1

TABLE 6
Diet intake, body weight gain, protein efficiency ratio (PER) and feed efficiency ratio (FER) 1