Endogenous urinary nitrogen (EUN) excretion is an important component of basal protein metabolism. Moreover, it is used in the calculation of the biological value of dietary proteins and in the factorial determination of crude protein requirements of adult animals.

Estimates of EUN excretion have been obtained by measuring the excretion of nitrogen in the urine of animals given a protein-free diet (Calloway and Margen 1971, Kendall et al. 1982, Moughan et al. 1987). This method, however, has been criticized because it reflects the nitrogen metabolism of animals adapted to protein restriction instead of nitrogen metabolism under normal physiological conditions (Berdanier et al. 1967, Holmes 1965). Alternatively, EUN excretion in animals can be determined under physiologically more normal conditions by regression to zero nitrogen intake of the urinary nitrogen excretion data of animals fed graded levels of dietary protein (Glem-Hansen and Jørgensen 1973, Greaves and Scott 1960, Moughan et al. 1987).

To the authors' knowledge there are no published estimates of EUN excretion for adult domestic cats determined using the protein-free method and only one estimate (Greaves and Scott 1960) in adult cats using the regression technique.

The objective of the present study was to determine and compare endogenous urinary total N, urea N and ammonia N excretion in adult domestic cats using both the protein-free and regression methods.

MATERIALS AND METHODS

Table 1. Ingredient compositions of the experimental diets1 [View Table]

Table 2. Nutrient composition of the fish offal used in the experimental diets [View Table]

Daily urinary N metabolite excretion and food intake data for the cats in Study 1 were subjected to a repeated measures ANOVA with gender as the variable and time (days after start of the study) as the repeated factor (Cody and Smith 1987). No effect of gender was found on the daily excretion of any of the urinary N metabolites or the food intake data of the cats in Study 1, and the data were thus pooled across genders. Daily urinary N metabolite excretion data for the cats in Study 2 were subjected to ANOVA using the General Linear Models procedure in SAS with diet and gender as variables. There was no effect of gender on the daily excretion of any of the urinary N metabolites of the cats in Study 2. The combined data for the daily urinary excretion of total N, urea N and ammonia N for the male and female cats were then subjected to a least-squares linear regression analysis with daily N intake as the independent variate and the daily urinary total N, urea N or ammonia N excretion as the dependent variate. Student's t test was used to determine the significance of differences between the daily endogenous urinary creatinine N excretions of the cats over the 12-d period in Study 1 and the cats over the 3-d period in Study 2. All statistical analyses were performed using the SAS statistical package (SAS Version 6.04, SAS Institute, Cary, NC) and effects were considered significant at P < 0.05. Values in the text are means ± SEM.

RESULTS

Daily weight loss for the cats in Study 1 over the 10-d period in which the protein-free diet was fed was 29 ± 2.4 g. Daily weight loss for the cats in Study 2 was 30 ± 3.5, 25 ± 3.9, 22 ± 3.7 and 23 ± 1.1 g for the cats receiving the 4, 7, 10 and 13 g/100 g crude protein diets, respectively. The average daily minimum and maximum air temperatures over the two studies were 12 and 20°C, respectively.

The daily urinary excretions of total N, urea N and ammonia N for the cats in Study 1 were high when the casein-based diet was fed and decreased abruptly during the first day when the protein-free diet was introduced, attaining a relatively constant level until the end of the study. The daily excretion of creatinine N in the urine of the cats remained relatively constant throughout Study 1. The mean fraction of undetermined N (total N  urea N  creatinine N  ammonia N) in the urine of the cats in Study 1 was slightly higher when the casein-based diet was fed [75.3 ± 4.81 mg/(kg body weight·d)] than when the protein-free diet was offered [56.9 ± 2.97 mg/(kg body weight·d)]. There was no significant effect of gender on the daily urinary excretion of total N, urea N, ammonia N or creatinine N. There was a significant effect of time on the daily urinary excretion of total N, urea N and ammonia N, but no effect of time on the daily urinary excretion of creatinine N or the fraction of undetermined N in the urine of the cats in Study 1. The overall mean daily urinary excretions of total N, urea N and ammonia N for the cats in Study 1 are presented in Figure 2. Daily urinary creatinine N excretion for the cats in Study 1 was 11.5 ± 0.34 mg/(kg body weight). Daily urinary excretions of total N, urea N, ammonia N and the daily amount of undetermined N for the six adult cats over d 5 to 12 in Study 1 were 277 ± 9.0, 187 ± 6.9, 21.3 ± 0.86 and 57.6 ± 3.50 mg/(kg body weight) and 360 ± 11.3, 243 ± 8.8, 27.6 ± 1.06 and 74.3 ± 4.50 mg/(kg body weight^0.75), respectively.

In general, the daily urinary metabolite excretions for the cats in Study 2 increased with increasing dietary N intake, although the average daily urinary excretions of total N and urea N for the cats receiving the 10 g/100 g crude protein diet were slightly higher than the corresponding values for the cats receiving the 13 g/100 g crude protein diet. The daily urinary excretion of creatinine N remained relatively constant with increasing intake of dietary N. The daily amount of undetermined N in the urine of the cats fed the four protein-containing diets showed no obvious increase or decrease with increasing intake of dietary N. There was no significant effect of gender on the daily urinary excretions of total N, urea N, ammonia N, creatinine N or the amount of undetermined N. There was a significant effect of dietary treatment, however, on the daily urinary excretions of total N, urea N and ammonia N, but no significant effect of diet on the daily excretion of creatinine N or the daily amount of undetermined N in the urine. Figures 3, 4 and 5 show the linear regression lines relating the daily urinary excretion of total N, urea N and ammonia N to daily dietary N intake, respectively. The slope of each line was significant at the 5% probability level, whereas the intercept of each regression line was significant at the 1% probability level. The daily urinary creatinine N excretion and the daily amount of undetermined N in the urine of the cats in Study 2 were 11.8 ± 0.70 and 38.0 ± 7.82 mg/(kg body weight), respectively. The estimates ± SEM of endogenous urinary total N, urea N and ammonia N excretion as determined by the extrapolation of the regression relationship to zero protein intake were 227 ± 41.2, 166 ± 34.8 and 24.1 ± 3.96 mg/(kg body weight·d) and 316 ± 53.9, 232 ± 43.4 and 33.7 ± 5.68 mg·kg body weight^0.75·d¹, respectively.

The mean daily creatinine N excretion of the cats in Study 1, measured over the 12-d period [14.4 ± 0.49 mg/(kg body weight^0.75)], was not significantly (P = 0.64) different than the mean daily creatinine N excretion for the cats in Study 2, measured over the 3-d period [15.9 ± 1.05 mg/(kg body weight^0.75)].

DISCUSSION

The rate of weight loss of the cats in Study 1 as measured over the 10-d period in which the protein-free diet was fed was 0.99 ± 0.024%/d which is similar to the rate of weight loss of 1.04%/d for adult cats fed a protein-free diet for 10 d observed by Hendriks et al. (1996). The rate of weight loss of the cats in Study 2 over the 9-d period decreased with increasing protein content of the diet, although the cats fed the 13 g/100 g protein diet lost weight at a rate similar to the cats fed the 7 g/100 g protein diet. Weight loss of the cats fed the 4, 7, 10, and 13 g/100 g protein diets were 0.86 ± 0.033, 0.73 ± 0.148, 0.68 ± 0.132 and 0.73 ± 0.08%/d, respectively.

The cats in the present study were previously trained to use the urine collection apparatus and did so habitually during the two studies. Urine voided outside the urine collection apparatus was easily detected on the white surface of the metabolism cage. Urine could have been sprayed through the partially covered barred door of the metabolism cage. However, during the daily inspections (six times, between 0700 and 1800 h), no urine was ever found on the floor in front of the metabolism cages.

In mammals such as humans, marmosets, dogs and pigs, there is a rapid decline in urinary total N excretion during the first 3 d of consuming a protein-free diet, followed by a slower decline of ~3 d until a relatively steady excretion of urinary total N is achieved over d 6 to 10 (Deuel et al. 1928, Flurer et al. 1988, Kendall et al. 1982, Moughan et al. 1987). The daily urinary excretions of total N, urea N and ammonia N for the cats in Study 1 were high when the casein-based diet was fed and rapidly declined during the first day when the protein-free diet was offered. The daily urinary excretions of total N, urea N and ammonia N reached a relatively steady level during d 4 to 12 (d 2 to 10 after the protein-free diet was offered) of the study. This pattern of decline in the excretion of urinary N metabolites for the adult cats in the present study indicates that cats metabolize dietary protein rapidly and that the catabolism of body protein is set at a relatively constant level or only slowly adjusts to dietary changes. These data are in accordance with the view that in cats, adaptation of hepatic enzymes involved in the catabolism of amino acids is limited (Rogers et al. 1977). The mean excretions of total N, urea N and ammonia N over the period in which the food intake of the cats fed the protein-free diet in Study 1 was relatively constant (d 5 to 12 of Study 1), which was assumed to represent the endogenous urinary excretion of these metabolites by the adult cat, were 360 ± 11.3, 243 ± 8.8 and 27.6 ± 1.06 mg·kg body weight^0.75·d¹, respectively.

The value for the EUN excretion of adult cats seems to be much higher than the values found for other animals using the same technique. Studies using the protein-free technique have reported EUN excretions of 62, 110, 128, 210 and 163 mg·kg body weight^0.75·d¹ in humans (Calloway and Margen 1971), marmosets (Flurer et al. 1988), rats (Yokogoshi et al. 1977), dogs (Kendall et al. 1982) and domestic pigs (Moughan et al. 1987), respectively. The main cause of the higher daily EUN excretion of adult cats in comparison with these other animals was the higher excretion of urea N. Endogenous urinary urea N excretions using the protein-free technique of 60, 70 and 116 mg·kg body weight^0.75·d¹ have been reported for rats (Yokogoshi et al. 1977), pigs (Moughan et al. 1987) and dogs (Kendall et al. 1982), respectively. These values are much lower than the endogenous urinary urea N excretion of 243 mg·kg body weight^0.75·d¹ found for adult cats in the present study. The higher endogenous urinary urea N excretion of cats in comparison with other mammals is not unexpected; this obligatory carnivore has been shown to have a limited ability to conserve nitrogen because of nonadaptive hepatic enzymes involved in the catabolism of amino acids (Rogers et al. 1977). These enzymes, furthermore, are set to handle a high protein diet and the cat, therefore, loses a large amount of amino acid nitrogen even when fed a low protein or protein-free diet (Rogers and Morris 1980).

The protein-free technique for the determination of EUN excretion in animals has been criticized because it reflects nitrogen metabolism in animals deprived of protein rather than nitrogen metabolism in animals under normal physiological conditions (Berdanier et al. 1967, Holmes 1965). Another criticism of this method is that EUN excretion is not constant but declines with time and that a lower EUN excretion may be obtained by increasing the period of protein-free feeding (Dawson and Allen 1961, Flurer et al. 1988, Holmes 1965). Although this has been shown to occur in other mammals such as dogs, humans and marmosets (Deuel et al. 1928, Flurer et al. 1988, Kendall et al. 1982), the present study indicates that this may not occur to such an extent in adult cats. The regression technique, however, allows determination of EUN excretion under a physiologically more normal situation and involves the regression to zero protein intake of the urinary N metabolite excretion data of animals fed graded levels of dietary nitrogen. When the regression method is used for the determination of endogenous urinary total N, urea N and ammonia N excretion in adult animals, it is important to supply dietary protein below the maintenance protein requirement of the animal and to minimize the catabolism of amino acids to supply energy. In Study 2, the four protein-containing diets were formulated to contain levels of protein below the minimum protein requirement for adult cats (National Research Council 1986), and the daily energy intake of the cats in Study 2 was set at 251 kJ ME/(kg body weight·d), the dietary energy intake level required by adult cats housed in metabolism cages (Goggin et al. 1993, Kane et al. 1981, Miller and Allison 1958). Cats develop severe hyperammonemia when an arginine-free diet is fed because this essential amino acid is required for the conversion of ammonia, generated by the catabolism of amino acids, to urea in the liver (Morris and Rogers 1978). To obtain accurate estimates for endogenous ammonia N excretion in adult cats by the regression technique, it is important, therefore, to supply sufficient dietary arginine to allow maximal conversion of ammonia to urea. In cats fed a protein-free diet, there is no need to supply a dietary source of arginine because the obligatory body protein degradation of animals fed a protein-free diet will supply the necessary arginine to prevent hyperammonamia (Morris and Rogers 1978). Growing kittens require approximately 42 g of arginine/kg of dietary crude protein (National Research Council 1986). In the present study, the fish offal protein (Table 2) contained ~100 g of arginine/kg which should have been sufficient to allow maximal conversion of ammonia to urea in adult cats.

The estimates for the endogenous urinary total N and urea N made using the regression technique [227 ± 41.2 and 166 ± 34.8 mg/(kg body weight·d), respectively] tended to be lower than the corresponding estimates made using the protein-free technique [277 ± 9.0 and 187 ± 6.9 mg/(kg body weight·d), respectively], whereas the endogenous urinary ammonia N excretion was slightly higher than the estimate obtained by the protein-free technique [24.1 ± 3.96 vs. 21.3 ± 0.86 mg/(kg body weight·d)]. The estimates of endogenous total N, urea N and ammonia N obtained by the regression technique, however, had large standard errors, and the corresponding estimates for the N metabolite excretions obtained by the protein-free technique all lay within the 95% confidence limit of the regression estimates, making the differences in the estimates between the two techniques statistically nonsignificant. The daily urinary excretion of creatinine N of the cats was not significantly different between the two studies and the mean daily excretion of creatinine N as measured over all of the cats was 11.7 ± 0.51 mg/(kg body weight). This value is similar to the urinary creatinine N excretion of 14.8 ± 0.61 mg/(kg body weight·d) for a male cat found by Hammett (1915).

The lower values for the endogenous urinary total N and urea N excretion obtained by regression compared with the protein-free method are in accordance with observations made in other animals. Berdanier et al. (1967) found that the regression technique gave lower estimates for EUN excretion in rats than the protein-free technique; Moughan et al. (1987) obtained estimates of endogenous urinary total N and urea N by the regression method in pigs which were 11 and 17% lower than the corresponding estimates obtained by the protein-free method, respectively. The reason for the lower EUN excretions found for cats in the present study by regression in comparison with the protein-free method is difficult to discern. Both methods determined EUN excretions of adult cats at zero protein intake and would, therefore, be expected to give similar results. The regression method, however, determined the EUN excretions under the condition that the dietary energy intake of the cats was met. This is unlike the protein-free diet approach in which the dietary energy intake of the cats was on average 53% of the energy requirements of cats housed in metabolism cages (Goggin et al. 1993, Kane et al. 1981, Miller and Allison 1958). The cats fed the protein-free diet, therefore, may have increased the catabolism of body amino acids to supply energy, resulting in an increased EUN excretion. The endogenous urinary urea N excretion of the cats fed the protein-free diet was higher, although not significantly, than the endogenous urinary urea N excretion determined by the regression approach [187 ± 6.9 vs. 166 ± 34.8 mg/(kg body weight·d)], supporting this hypothesis. However, the higher EUN excretions obtained by the protein-free diet in comparison with the regression approach, in the present study, were also the result of a higher excretion of undetermined N [57.6 ± 3.50 vs. 38.0 ± 7.82 mg/(kg body weight·d)]. The reason for the higher amount of endogenous N excreted in the latter fraction by the cats fed the protein-free diet is unknown. However, besides being a physiologically more normal approach, the value for the EUN excretion obtained by the regression technique can be expected to be more accurate than this value obtained by the traditional protein-free diet approach.

Greaves and Scott (1960) determined EUN excretion for adult cats by the regression technique, and re-analysis of the data of Greaves and Scott (1960) resulted in an estimate of 25 ± 48.6 mg/(kg body weight·d). This low value is likely to be the result of the dietary protein intake which was in excess of the requirement for body maintenance in the majority of the cats in this study, as evident from the positive nitrogen balance. Fitting a linear regression line through urinary N excretion data of adult animals ingesting protein in excess of requirement will result in a steeper regression line than when a linear regression line is fitted through urinary N excretion data of adult animals fed protein below the requirement for body maintenance.

The present study provides hitherto unavailable data on the endogenous excretion of nitrogen metabolites in the urine of adult domestic cats. These data can be used for the factorial determination of the protein and amino acid requirements of adult domestic cats.