Atwater and Benedict determined directly by means of the respiration calorimeter the heat production of the same man during five fasting experiments of one to two days each, and during a four-day experiment with food about sufficient for maintenance. The average total metabolism on the fasting days was about 9 per cent lower than on the days when food was taken.

In longer fasts there may be a somewhat greater decrease in heat production. Thus, Benedict found that a man who weighed at the start 59.5 kilograms (131 pounds) metabolized, on the successive days of a seven-day fast, 1765, 1768, 1797, 1775, 1649, 1553 and 1568 Calories respectively. Naturally in long fasts factors other than the simple sparing of the direct effect of food come into play.*

Tigerstedt studied by means of the carbon and nitrogen balance the metabolism of a man who fasted for five days and for the next two days took a very liberal diet. The following data were obtained:

* For a detailed account of the results obtained in a fasting experiment of 31 days' duration, see Benedict, A Study of Prolonged Fasting, Publication No. 203 of the Carnegie Institution of Washington.

These results show for man (as had previously been shown with dogs) that in fasting the total metabolism continues at a fairly constant rate in spite of the fact that the energy is obtained entirely at the expense of body material. In this case, the diet given at the end of the fasting period (4141 Calories) was approximately double what would have been required for maintenance, but the increase in energy metabolism was only 22.5 per cent over that of fasting.

The results of fasting experiments thus make it evident that the body has but little power in the direction of adjusting its energy metabolism to the energy value of its food supply.

Rubner found that each type of food exerted a more or less specific influence upon the energy metabolism, so that when the foodstuffs were fed separately, somewhat different energy values were required for the maintenance of body equilibrium. Thus, if the total metabolism of a dog fasting at 330 C. be represented by 100 Calories, he must be fed, in order to prevent loss of body substance, about 106.5 Calories of sugar, or 114.5 Calories of fat, or 140 Calories of protein. A man observed by Rubner metabolized in fasting 2042 Calories; when fed 2450 Calories in the form of sugar alone, he metabolized 2087 Calories; when fed 2450 Calories in the form of meat alone, he metabolized 2566 Calories.

* These figures are slightly too high because the loss of carbon on these days was due in part to combustion of glycogen, but is calculated as if due simply to protein and fat.

Recently Lusk and his coworkers have investigated the influence of the foodstuffs upon metabolism ("specific dynamic action") very extensively and have developed the subject to such an extent that for an adequate discussion of their results the original articles1 or Lusk's own summary2 should be consulted. It appears from this work that when the digestion products of carbohydrate or fat are carried by the blood to the tissues the energy metabolism (rate of oxidation) rises simply because of the increased concentration of oxidizable material; but that some of the products of the digestion and intermediary metabolism of protein increase metabolism not only to a greater extent, but also in a somewhat different manner, since they seem to act as stimuli rather than merely as fuel. On an ordinary mixed diet, however, this apparent loss of energy due to eating of protein is not a very large factor in the total metabolism, since the total specific dynamic action makes the metabolism of energy for the day only about one tenth higher on a full maintenance ration than when no food is eaten.

Benedict and Roth have studied the energy metabolism of vegetarians as compared with non-vegetarians of the same height and weight in order to determine whether or not the former maintain a lower plane of basal metabolism than do people who eat meat and who are sometimes held to be unduly stimulated by the protein of their food. The energy metabolism was computed from the carbon dioxide production and oxygen consumption determined when the subjects were at complete rest and in the "post-absorptive condition," i.e. at least 12 hours after the last meal, the immediate specific dynamic action of the food being thus practically excluded. Under these conditions the vegetarian men and women showed average basal metabolism of 1.06 and 1.025 Calories per kilogram of body weight per hour respectively, while the corresponding data for non-vegetarian men and women were 1.10 and 1.04 Calories respectively. Benedict holds that these differences are too small to establish any essential difference in the basal energy metabolism of vegetarians and non-vegetarians of like height and weight.

1 Lusk. Journal of Biological Chemistry, Vol. 20, pages vii-xvii and 555-617. Murlin and Lusk, Ibid., Vol. 22, pages 15-29.

2 Lusk. Science of Nutrition, Third Edition, Chapter VII (Conditions Governing Energy Metabolism And Total Food Requirement. Basal Metabolism Of The Adult).

It is sometimes thought that superior preparation or very thorough mastication of food results in such improvement in its utilization that a material saving may be effected in the amount of food required. But it will be remembered that under average conditions only about 5 per cent of the energy value of the food is lost in digestion or expended upon the digestive process. Any improvement in those conditions through superior preparation or mastication of the food can therefore at most effect a saving of less than five per cent of the energy value. Thus the influence upon total food requirement is scarcely appreciable. The advantages of good preparation and thorough chewing of the food are very important, but they lie in other directions than reduction in the amount of food required.

Recent scientific evidence supports the view that chronic undernutrition or even simple restriction of food consumption in health, if continued sufficiently, may bring the organism to a lower level of energy metabolism than would be indicated by the weight or surface.