LaboratOl y Animals (1976) 10,291-296.
291
MEASUREMENT METABOLIZABLE
OF DIGESTIBLE AND
ENERGY
OF DIETS FOR RATS
by B. A. ROLLS,
S. N. HEGDE
and MARIE
E. COATES
National Institute for Research in Dairying, Shinfield, Reading, RG2 9AT SUMMARY
A balance method for the measurement of digestible and metabolizable energy in rats is described. Experimentally determined metabolizable energies of 2 diets for rats were closer to those values for metabolizable energy calculated from tabulated values for pigs rather than chickens. Some drawbacks in the use of tables of energy values to predict the energy of diets are discussed. Few measurements of metabolizable energy (ME) values have been reported for the rat and compounders of diets for laboratory rats are forced to use figures for ME taken from published data from pigs (on the grounds that the pig is a mammal) or from chickens (on the grounds that the fowl is closer in size to the rat). Metta & Mitchell (1954) strongly criticised the application of results obtained with one species of animal to dietary formulations for another, and showed that experimentally determined ME values for purified diets for rats differed markedly from the Atwater factors obtained in human nutrition. The question of the appropriate ME values to be applied to diets for the rat became important in the course of a collaborative experiment organised by the Medical Research Council Laboratory Animals Centre Laboratory Animal Diets Advisory Committee in which several diets were tested in different laboratories to determine the optimal protein to energy ratio for laboratory rats. In the work reported here the ME of 2 of the diets compounded for these trials were measured experimentally in rats. MATERIALS
AND METHODS
Diets The diets given were formulated to be of high and low metabolizable energy; their compositions are given in Table 1. They were supplied and used as mixed, unpelletted.
Downloaded from lan.sagepub.com at NORTH CAROLINA STATE UNIV on March 6, 2015
292
B. A. ROLLS, S. N. HEGDE Table 1. Composition
AND
M. E. COATES
of experimental
diets (g/kg).
Low energy Barley Wheat, soft Maize Oat feed Wheat feed Animal fat Soya-bean meal, extracted White fish meal Methionine Sodium chloride Limestone flour Dicalcium phosphate Vitamins and trace elements
High energy
198·2 100 106,5 100
91-9 200 250
250
8'3
20 346 75 1·1 6'7 1·7 6,6
1·0
1·0
196·9 25
7-1 7,0
Animals
Weanling rats, 6 male and 6 female, of closely similar bodyweight, were selected from the Institute colony of Norwegian Hooded rats. Experimental
The rats were housed individually in cylindrical glass jars 20 cm in diameter and 20 cm high, with stainless-steel mesh floors and accessories (Fig. 1). A Whatman's no. 3 filter paper in the bottom absorbed most of the urine, and faeces and spilt food could be collected manually. Each rat received daily diet to the amount of 10 % of its bodyweight, made into a slurry with water to reduce wastage. This quantity was selected with the aim that the animal would eat all the food offered, and that bodyweight and hence bodily energy requirements would not change markedly during the experiment. Half the rats ate each diet for the 1st period; for the 2nd period the diets were switched, so that during the course of the experiment each rat ate each diet. For larger numbers of diets a Latin square design would be appropriate. Each feeding period lasted 7 days: a 3-day adaptation period followed by a 4-day collection period. Sexes and diets were balanced with respect to cage position to remove obvious sources of bias. During the adaptation period the food was given as described and the filter paper was changed daily. At the start of the collection period each rat was transferred to a cage cleaned throughout with acid and with a fresh filter paper sprinkled with 0·1 N hydrochloric acid to reduce losses of ammonia from the urine. During the collection period the weighed food was given daily as before, faeces were collected daily, separated from any spilt food and stored at -20°C. Spilt food was also collected daily and, at the end of the
Downloaded from lan.sagepub.com at NORTH CAROLINA STATE UNIV on March 6, 2015
RAT DIET ENERGY MEASUREMENT
293
waler bulb
lid with air holes
\
~/=
==
D
o
o
o
000
o
o
'stainless steel mesh floor food in
glass dish
-----------.....-
tilter paper
-------
glass jar
----
Fig. 1. Metabolism jar.
collection period, dried back so that with a knowledge of the current dry matters of the diets, total food consumption could be calculated. On the 2nd and 4th days of the collection periods, after collection of faeces and spilt food, hair and other adventitious material was discarded and the urine was washed from the filter paper and the cage and fittings with several changes of 0·1 N hydrochloric acid. The combined urine liquors were stored at -20°C. In this way total collections were made of urine and faeces. Strictly, shed
Downloaded from lan.sagepub.com at NORTH CAROLINA STATE UNIV on March 6, 2015
294
B. A. ROLLS,
S. N. HEGDE
AND
M. E. COATES
hair and skin are an element of energy expenditure, but this correction was so small that it was disregarded. The faeces were homogenized, freeze-dried and weighed. Samples of diet, dried faeces and urine liquors were taken for estimation of nitrogen by the micro-Kjeldahl method and of energy by bomb calorimetry in a Gallenkamp Automatic Adiabatic Bomb Calorimeter Type CB-lOO (A. Gallenkamp & Co. Ltd, Technico House, Christopher Street, London, EC2P 2ER). For urine, the energy sample was freeze-dried in a standard, weighed plastic bag and the whole was compressed into a pellet and ignited in the bomb calorimeter. So that corrections might be made, empty plastic bags were ignited and the washings from unused filter papers were analysed in the same way as were the urine liquors. Digestible and metabolizable energies were calculated as follows: digestible energy metabolizable
energy
=
gross energy (energy of diet)-faecal
= gross energy-faecal
energy-Curine
energy energy + protein correction)
It is necessary to correct the urine energy for any gain or loss of body protein during the trial period, since protein is oxidized less completely in the body than by combustion. Metta & Mitchell (1954) determined experimentally that the average ratio of energy to nitrogen in rat urine was 26·3 kJ/g. Thus the following correction factor was applied: protein correction
(kJ)
=
x body protein gain = 26·3 x 6·25
26·3
(food
N-faecal
N-urine
N)
The ME values from individual rats were analysed statistically using analysis of variance. It was found that the results from the 1st and 2nd feeding periods were not significantly different (P>O'I) so period was disregarded for subsequent treatments. RESULTS
The rats remained similar in weight during the trial. There was no refusal of food and no evidence of selection, even though in a separate -experiment rats fed ad libitum with the low energy diet tended to leave some fibrous material. The gross, digestible (DE) and metabolizable (ME) energies of the diets are Table 2. Gross, digestible and metabolizable energies of the low- and high-energy diets in MJ/kg dry diet, with values in keal/kg dry diet in parentheses. Means, with standard errors, of 12 rats. High energy
Low energy Gross energy Digestible energy Metabolizable energy
17-99 (4301) 12-60 ± 0'11 (3012 11'84 ± 0'11 (2829
± ±
26) 26)
18·62 (4451) 15·36 ± 0,09 (3671 14·22 ± 0,08 (3399
Downloaded from lan.sagepub.com at NORTH CAROLINA STATE UNIV on March 6, 2015
± ±
22) 19)
RAT DIET
ENERGY
MEASUREMENT
295
given in Table 2. In Table 3 the experimentally determined MEs are compared with values calculated from listed MEs for the dietary constituents (National Research Council, 1971). The moisture contents of diets kept under refrigeration in sealed bags remained constant at about 9 %, but diet kept unsealed in the open laboratory lost moisture. Table 3. Dry matters (DM) and metabolizable energies (ME) of the low- and highenergy diets calculated from published· ME and DM values of the dietary constituents for chickens and pigs, and determined experimentally with rats. ME values are given in MJ/kg diet as given and MJ/kg dry matter, with corresponding values in kcal/kg in parentheses. Source
Diet
DM%
ME (as given)
ME (dry matter)
Calculated: chicken data
low energy high energy
9'22 (2204) 11·72 (2802)
10'26 (2453) 13·13 (3138)
Calculated: pig data
low energy high energy
10,95 (2617) 12-91 (3085)
12'18 (2912) 14'46 (3455)
Experimental: rats
low energy high energy
89'9 89-3 89,9 89'3 90,8 91'5
10·75 (2569) 13-01 (3109)
1l·84 (2829) 14'22 (3399)
*National Research Council (1971). The values determined from female rats of DE (12,84 ± 0·12 and 15·49 ± 0·08 for the 2 diets) and of ME (11·99 ± 0,07 and 14·32 ± 0·07) were slightly higher than those values with male rats (DE 12·36 ± 0,13 and 15·23 ± 0'16; ME 11·59 ± 0·11 and 14·12 ± 0'15) and these differences were just significant (P>0·05). As would be expected from the figures in Table 2, the differences between the MEs of the diets were highly significant (P>O·OOI). There were no interactions between sex and feeding period or between sex and diets. DISCUSSION
The method described here was essentially similar to that of Metta & Mitchell (1954). The procedure was without serious disadvantages, distinguished readily between the 2 diets and gave results with low coefficients of variation. Since the rats changed little in weight in the course of the experiment, the correction factor for body protein change was small. In such a case little error would result from ignoring the correction, and the need for nitrogen determinations would be eliminated. In both samples tested the ME for the rat was nearer to the calculated value using data for the pig than that from chicken data. This was unexpected, as the fowl is nearer in size to the rat, which it resembles in having an insulating body covering. However, it is possible that the greater size of the pig and its
Downloaded from lan.sagepub.com at NORTH CAROLINA STATE UNIV on March 6, 2015
296
B. A. ROLLS, S. N. HEGDE
AND
M. E. COATES
consequent higher volume to surface ratio confers a metabolic advantage that compensates for its relative lack of body hair. Moreover, the digestive systems of rat and pig are more similar than that of the chick, with its relatively short gut. Nelson, May & Miles (1974) reported that MEs were similar for rats and chickens in the case of grains, although with high protein materials MEs were higher for the rat, and closer to pig values. No systematic study seems to have been made on the differences between males and females in the absorption and utilization of nutrients. The differences found in this work could have arisen by chance and clearly more extensive studies would be necessary to establish any real influence of sex. Apart from the problem of deciding which species most closely corresponds to the rat, the use of standard feed tables presents many difficulties. Values for many listed dietary components may differ markedly not only between different tables but also within the same set of tables, and there is often little indication of the reliability of a particular figure. Moreover, if tables of foreign origin are used, varieties of materials may differ and differences in treatments, particularly in the separation and classification of cereal products, may mean that no exact equivalent exists. We found that diet stored cold in closed plastic bags maintained a moisture content of around 9 %, although polythene is by no means impervious to water. This would not be true under all conditions of storage and it is desirable that moisture contents should be checked during any experiment for which a knowledge of the dietary energy concentration is important. ACKNOWLEDGEMENTS
We should like to thank Mr H. E. Clarke of RHM Agriculture Ltd, Wimborne, Dorset for formulating and supplying the diets, and Miss M. Tyler and her staff for their care of the rats. We thank Dr D. Hewitt for his help with the statistical interpretation. REFERENCES Metta, V. C. & Mitchell, H. H. (1954). Determination of the metabolizable energy of organic nutrients for the rat. Journal of Nutrition 52, 601-611. National Research Council (1971). Atlas of nutritional data on United States and Canadian feeds. Washington: National Academy of Sciences. Nelson, T. S., May, M. A. & Miles, R. D. (1974). Digestible and metabolizable energy of feed ingredients for rats. Journal of Animal Science 38, 554-558.
Downloaded from lan.sagepub.com at NORTH CAROLINA STATE UNIV on March 6, 2015
291
MEASUREMENT METABOLIZABLE
OF DIGESTIBLE AND
ENERGY
OF DIETS FOR RATS
by B. A. ROLLS,
S. N. HEGDE
and MARIE
E. COATES
National Institute for Research in Dairying, Shinfield, Reading, RG2 9AT SUMMARY
A balance method for the measurement of digestible and metabolizable energy in rats is described. Experimentally determined metabolizable energies of 2 diets for rats were closer to those values for metabolizable energy calculated from tabulated values for pigs rather than chickens. Some drawbacks in the use of tables of energy values to predict the energy of diets are discussed. Few measurements of metabolizable energy (ME) values have been reported for the rat and compounders of diets for laboratory rats are forced to use figures for ME taken from published data from pigs (on the grounds that the pig is a mammal) or from chickens (on the grounds that the fowl is closer in size to the rat). Metta & Mitchell (1954) strongly criticised the application of results obtained with one species of animal to dietary formulations for another, and showed that experimentally determined ME values for purified diets for rats differed markedly from the Atwater factors obtained in human nutrition. The question of the appropriate ME values to be applied to diets for the rat became important in the course of a collaborative experiment organised by the Medical Research Council Laboratory Animals Centre Laboratory Animal Diets Advisory Committee in which several diets were tested in different laboratories to determine the optimal protein to energy ratio for laboratory rats. In the work reported here the ME of 2 of the diets compounded for these trials were measured experimentally in rats. MATERIALS
AND METHODS
Diets The diets given were formulated to be of high and low metabolizable energy; their compositions are given in Table 1. They were supplied and used as mixed, unpelletted.
Downloaded from lan.sagepub.com at NORTH CAROLINA STATE UNIV on March 6, 2015
292
B. A. ROLLS, S. N. HEGDE Table 1. Composition
AND
M. E. COATES
of experimental
diets (g/kg).
Low energy Barley Wheat, soft Maize Oat feed Wheat feed Animal fat Soya-bean meal, extracted White fish meal Methionine Sodium chloride Limestone flour Dicalcium phosphate Vitamins and trace elements
High energy
198·2 100 106,5 100
91-9 200 250
250
8'3
20 346 75 1·1 6'7 1·7 6,6
1·0
1·0
196·9 25
7-1 7,0
Animals
Weanling rats, 6 male and 6 female, of closely similar bodyweight, were selected from the Institute colony of Norwegian Hooded rats. Experimental
The rats were housed individually in cylindrical glass jars 20 cm in diameter and 20 cm high, with stainless-steel mesh floors and accessories (Fig. 1). A Whatman's no. 3 filter paper in the bottom absorbed most of the urine, and faeces and spilt food could be collected manually. Each rat received daily diet to the amount of 10 % of its bodyweight, made into a slurry with water to reduce wastage. This quantity was selected with the aim that the animal would eat all the food offered, and that bodyweight and hence bodily energy requirements would not change markedly during the experiment. Half the rats ate each diet for the 1st period; for the 2nd period the diets were switched, so that during the course of the experiment each rat ate each diet. For larger numbers of diets a Latin square design would be appropriate. Each feeding period lasted 7 days: a 3-day adaptation period followed by a 4-day collection period. Sexes and diets were balanced with respect to cage position to remove obvious sources of bias. During the adaptation period the food was given as described and the filter paper was changed daily. At the start of the collection period each rat was transferred to a cage cleaned throughout with acid and with a fresh filter paper sprinkled with 0·1 N hydrochloric acid to reduce losses of ammonia from the urine. During the collection period the weighed food was given daily as before, faeces were collected daily, separated from any spilt food and stored at -20°C. Spilt food was also collected daily and, at the end of the
Downloaded from lan.sagepub.com at NORTH CAROLINA STATE UNIV on March 6, 2015
RAT DIET ENERGY MEASUREMENT
293
waler bulb
lid with air holes
\
~/=
==
D
o
o
o
000
o
o
'stainless steel mesh floor food in
glass dish
-----------.....-
tilter paper
-------
glass jar
----
Fig. 1. Metabolism jar.
collection period, dried back so that with a knowledge of the current dry matters of the diets, total food consumption could be calculated. On the 2nd and 4th days of the collection periods, after collection of faeces and spilt food, hair and other adventitious material was discarded and the urine was washed from the filter paper and the cage and fittings with several changes of 0·1 N hydrochloric acid. The combined urine liquors were stored at -20°C. In this way total collections were made of urine and faeces. Strictly, shed
Downloaded from lan.sagepub.com at NORTH CAROLINA STATE UNIV on March 6, 2015
294
B. A. ROLLS,
S. N. HEGDE
AND
M. E. COATES
hair and skin are an element of energy expenditure, but this correction was so small that it was disregarded. The faeces were homogenized, freeze-dried and weighed. Samples of diet, dried faeces and urine liquors were taken for estimation of nitrogen by the micro-Kjeldahl method and of energy by bomb calorimetry in a Gallenkamp Automatic Adiabatic Bomb Calorimeter Type CB-lOO (A. Gallenkamp & Co. Ltd, Technico House, Christopher Street, London, EC2P 2ER). For urine, the energy sample was freeze-dried in a standard, weighed plastic bag and the whole was compressed into a pellet and ignited in the bomb calorimeter. So that corrections might be made, empty plastic bags were ignited and the washings from unused filter papers were analysed in the same way as were the urine liquors. Digestible and metabolizable energies were calculated as follows: digestible energy metabolizable
energy
=
gross energy (energy of diet)-faecal
= gross energy-faecal
energy-Curine
energy energy + protein correction)
It is necessary to correct the urine energy for any gain or loss of body protein during the trial period, since protein is oxidized less completely in the body than by combustion. Metta & Mitchell (1954) determined experimentally that the average ratio of energy to nitrogen in rat urine was 26·3 kJ/g. Thus the following correction factor was applied: protein correction
(kJ)
=
x body protein gain = 26·3 x 6·25
26·3
(food
N-faecal
N-urine
N)
The ME values from individual rats were analysed statistically using analysis of variance. It was found that the results from the 1st and 2nd feeding periods were not significantly different (P>O'I) so period was disregarded for subsequent treatments. RESULTS
The rats remained similar in weight during the trial. There was no refusal of food and no evidence of selection, even though in a separate -experiment rats fed ad libitum with the low energy diet tended to leave some fibrous material. The gross, digestible (DE) and metabolizable (ME) energies of the diets are Table 2. Gross, digestible and metabolizable energies of the low- and high-energy diets in MJ/kg dry diet, with values in keal/kg dry diet in parentheses. Means, with standard errors, of 12 rats. High energy
Low energy Gross energy Digestible energy Metabolizable energy
17-99 (4301) 12-60 ± 0'11 (3012 11'84 ± 0'11 (2829
± ±
26) 26)
18·62 (4451) 15·36 ± 0,09 (3671 14·22 ± 0,08 (3399
Downloaded from lan.sagepub.com at NORTH CAROLINA STATE UNIV on March 6, 2015
± ±
22) 19)
RAT DIET
ENERGY
MEASUREMENT
295
given in Table 2. In Table 3 the experimentally determined MEs are compared with values calculated from listed MEs for the dietary constituents (National Research Council, 1971). The moisture contents of diets kept under refrigeration in sealed bags remained constant at about 9 %, but diet kept unsealed in the open laboratory lost moisture. Table 3. Dry matters (DM) and metabolizable energies (ME) of the low- and highenergy diets calculated from published· ME and DM values of the dietary constituents for chickens and pigs, and determined experimentally with rats. ME values are given in MJ/kg diet as given and MJ/kg dry matter, with corresponding values in kcal/kg in parentheses. Source
Diet
DM%
ME (as given)
ME (dry matter)
Calculated: chicken data
low energy high energy
9'22 (2204) 11·72 (2802)
10'26 (2453) 13·13 (3138)
Calculated: pig data
low energy high energy
10,95 (2617) 12-91 (3085)
12'18 (2912) 14'46 (3455)
Experimental: rats
low energy high energy
89'9 89-3 89,9 89'3 90,8 91'5
10·75 (2569) 13-01 (3109)
1l·84 (2829) 14'22 (3399)
*National Research Council (1971). The values determined from female rats of DE (12,84 ± 0·12 and 15·49 ± 0·08 for the 2 diets) and of ME (11·99 ± 0,07 and 14·32 ± 0·07) were slightly higher than those values with male rats (DE 12·36 ± 0,13 and 15·23 ± 0'16; ME 11·59 ± 0·11 and 14·12 ± 0'15) and these differences were just significant (P>0·05). As would be expected from the figures in Table 2, the differences between the MEs of the diets were highly significant (P>O·OOI). There were no interactions between sex and feeding period or between sex and diets. DISCUSSION
The method described here was essentially similar to that of Metta & Mitchell (1954). The procedure was without serious disadvantages, distinguished readily between the 2 diets and gave results with low coefficients of variation. Since the rats changed little in weight in the course of the experiment, the correction factor for body protein change was small. In such a case little error would result from ignoring the correction, and the need for nitrogen determinations would be eliminated. In both samples tested the ME for the rat was nearer to the calculated value using data for the pig than that from chicken data. This was unexpected, as the fowl is nearer in size to the rat, which it resembles in having an insulating body covering. However, it is possible that the greater size of the pig and its
Downloaded from lan.sagepub.com at NORTH CAROLINA STATE UNIV on March 6, 2015
296
B. A. ROLLS, S. N. HEGDE
AND
M. E. COATES
consequent higher volume to surface ratio confers a metabolic advantage that compensates for its relative lack of body hair. Moreover, the digestive systems of rat and pig are more similar than that of the chick, with its relatively short gut. Nelson, May & Miles (1974) reported that MEs were similar for rats and chickens in the case of grains, although with high protein materials MEs were higher for the rat, and closer to pig values. No systematic study seems to have been made on the differences between males and females in the absorption and utilization of nutrients. The differences found in this work could have arisen by chance and clearly more extensive studies would be necessary to establish any real influence of sex. Apart from the problem of deciding which species most closely corresponds to the rat, the use of standard feed tables presents many difficulties. Values for many listed dietary components may differ markedly not only between different tables but also within the same set of tables, and there is often little indication of the reliability of a particular figure. Moreover, if tables of foreign origin are used, varieties of materials may differ and differences in treatments, particularly in the separation and classification of cereal products, may mean that no exact equivalent exists. We found that diet stored cold in closed plastic bags maintained a moisture content of around 9 %, although polythene is by no means impervious to water. This would not be true under all conditions of storage and it is desirable that moisture contents should be checked during any experiment for which a knowledge of the dietary energy concentration is important. ACKNOWLEDGEMENTS
We should like to thank Mr H. E. Clarke of RHM Agriculture Ltd, Wimborne, Dorset for formulating and supplying the diets, and Miss M. Tyler and her staff for their care of the rats. We thank Dr D. Hewitt for his help with the statistical interpretation. REFERENCES Metta, V. C. & Mitchell, H. H. (1954). Determination of the metabolizable energy of organic nutrients for the rat. Journal of Nutrition 52, 601-611. National Research Council (1971). Atlas of nutritional data on United States and Canadian feeds. Washington: National Academy of Sciences. Nelson, T. S., May, M. A. & Miles, R. D. (1974). Digestible and metabolizable energy of feed ingredients for rats. Journal of Animal Science 38, 554-558.
Downloaded from lan.sagepub.com at NORTH CAROLINA STATE UNIV on March 6, 2015
