Many of the conceptions as to the influence of the ingestion of food upon the heat production have long been held and need no material modification. But as most of such evidence was obtained in experiments with animals rather than with men it seemed desirable for us to undertake a research upon the influence of the ingestion of food upon the metabolism of man. In making these experiments we have been greatly indebted to the earlier investigators, more especially to Magnus-Levy1 and to Johansson and his school,2 as their researches were in large part with men. It has been impracticable in our discussion to cite adequately the numerous observations made upon animals, particularly the classic experiments of Rubner and the more recent work of Lusk and his associates in New York. Believing that our problem was sufficiently extended if confined primarily to man, we have therefore intentionally omitted in this publication a review of practically all experiments made upon animals. In the decade or more that the results have been accumulating, numerous papers by other investigators have appeared, many of them reporting experiments with men. These we have considered carefully in our digest of the literature, as well as in the discussion of the several chapters.

The experimental evidence in this book as a result of our research presents little that is startlingly new. The mechanical work of chewing has been found to produce a definite increase in the metabolism. The drinking of liquids, especially in large amounts, likewise has been shown to increase the metabolism, although these increases are usually relatively small. The fact that the ingestion of all kinds of food in any amount results in an increment in the metabolism seems very clearly established. No conclusive evidence of a metabolism depressed below the basal value after food has been found in any case. As our work was with man, it was obviously impracticable for us to use pure nutrients save in the case of sugars, and our experiments are thus open to this criticism. Hence, if we attempt to establish mathematical relationships for the effects of carbohydrates, fat, and protein, we at once meet the criticism that while the carbohydrates selected were, for the most part, pure nutrients, the fat and protein food materials were mixed nutrients, as, for instance, beefsteak, in which the protein was combined with fat, which also supplied a certain amount of energy.

1Magnus-Levy, Arch. f. d. ges. Physiol., 1894, 55, p. 1.

2 Johansson, Skand. Arch. f. Physiol., 1897,7, p. 29; same journal, 1902,13, p. 251; same journal, 1904, 16, p. 263; same journal, 1908, 21, p. 1.

Notwithstanding this defect in our experimental plan, the evidence obtained with diets in large part protein agrees with that secured by other observers with a protein diet, as an effect was found which was more pronounced and extended than that of any other nutrient. It appeared to make no difference whether the protein used was an animal or a vegetable protein, for the experiments with glidine on the one hand and with beefsteak and plasmon on the other are usually comparable.

Unfortunately the evidence obtained regarding fat is not so convincing, for our experiments are admittedly too few in number to give conclusive results and in the diets used the fat was combined with other substances; still the available energy derived from fat in the food intake was so large in most instances that the increment in the metabolism must necessarily have been due to this factor. Although the effect obtained was by no means so great as that found with protein, it can not be considered as negligible.

The most sharply defined results were those secured in the series of experiments with carbohydrate diets. It was possible to make a careful analysis of these data, compare the results obtained with the individual carbohydrates, and determine not only the total effect upon the metabolism measured, but likewise the time relations and the rapidity of the action of the food material. These results show in a striking manner that all of the carbohydrates influence the total metabolism and differ but little in this respect, although levulose and sucrose appear to exert a somewhat more powerful influence than the other sugars.

The experiments with mixed diets, especially those with excessive amounts of food, showed that it was possible by the ingestion of a large meal to stimulate the metabolism to 40 per cent above the basal value for a number of hours, and to 20 per cent for at least 8 hours; indeed, there was every reason to believe that the stimulus to the metabolism would have been found to continue considerably longer than the experimental period of 8 hours if the observations had been prolonged. This fact has a special practical significance in its relation to the daily life of human individuals. While it is possible for a human being to live with greatly reduced activity when sound asleep, without food in the stomach, and without extraneous muscular activity, his efficiency as a member of human society in such a state would be negligible. It is therefore only as the cellular activity increases that we find him becoming more and more of service to humanity, and not until he is erect and ready to perform active external muscular work is he in a condition to live on a basal plane that is of practical value.

The ingestion of food with its attendant increase in metabolism appears at first thought like a highly inefficient process, this increase being comparable to the extra energy required by a donkey engine to stoke the boilers in a large factory; so far as the direct mechanical output of the factory is concerned, the energy thus used appears as waste, and yet it is necessary in order to secure a supply of fuel to the boilers. The increment in the metabolism or excess energy given off by the body as a result of the ingestion of food may be considered as the energy required for the preparation of material for use in the body tissues, and on this basis may be regarded as waste energy. Indeed, it is the belief of some writers that heat is invariably a waste product and that this factor has interest only in that it is developed in connection with muscular or glandular activity. Another phase of the situation appears, however, when we consider that the extra heat developed under these conditions may possibly be looked upon as a normal physiological stimulus to cellular activity. In this connection the practical experience of many investigators may be mentioned, especially those making observations with severe muscular work in studies with a protein diet and, in many cases, with a carbohydrate diet, such as sugar or sweet chocolate. If it be true that the increase in the metabolism resulting from the ingestion of such diets has a specific influence in stimulating the whole cellular system of the body to greater activity, then we may not properly regard this excess heat as a waste product.

Continuing the discussion in the terms of the efficiency engineer, it may be possible to consider the increase in heat production due to food as a measure of the "cost of digestion." For instance, the ingestion of 1,000 calories of food in the form of sugar requires the excess production of 60 calories of heat in order to have the sugar ready for an actual share in the muscular work. On this basis, one might compute that this excess heat was lost and that when 1,000 calories in the form of cane sugar are transformed into material ready for combustion in the body only 940 calories are available for such use. If, then, the increments in heat production obtained in our various experiments are computed and compared with the fuel value of the food ingested, the proportion of the energy in the ingested food which was given off as excess heat may be determined. One great difficulty in securing such data is the fact that in many instances the experiments did not continue long enough to include the entire heat increment. This is particularly true in the protein experiments, for frequently (see table 215, page 284) the basal value was not reached before the end of the experiment.

The data showing the relationship between the excess heat and the fuel value of the intake, which are given in tables 249 and 250 (see pages 336and 338), vary considerably with the length of the experiment, the total amount of the food intake, and the nature of the diet. While, for lack of a better terminology, the values are designated as the "cost of digestion", the use of such a term is distinctly misleading, as implying that this excess heat is waste heat. We are firmly convinced that the excess heat produced from the ingestion of protein or carbohydrates, like sugars, may not properly be considered as purely a waste process, but that it is far more logical to consider it as a general stimulation of all of the cells in preparation for the drafts of muscular activity.

Our results give no basis for recommending an exclusively protein diet or an exclusively sugar diet prior to severe muscular work. That the presence of glycogen in the body has an important bearing on the efficiency of the muscular system is, in general, we think, proved without doubt. That any food substance that will contribute toward the replenishment of a depleted glycogen store or will maintain this at a high level is important in the preparation for muscular work, we likewise may consider as being thoroughly established. The formation of glycogen from sugar is unquestionably proved; the formation of glycogen or sugars from protein is likewise demonstrated;but there is as yet no evidence that sugar is formed from fat. It follows, therefore, that diets preceding muscular work should contain liberal quantities of carbohydrates or protein, although our evidence does not allow us to determine which is the more important, the furnishing of glycogen or the normal stimulus to the body. There is no question but that protein is a more prolonged stimulus to the metabolism than carbohydrate. On the other hand, in the digestion of protein extra work is thrown upon the organs of excretion. Too much significance may be given to this, but nevertheless, since the ingestion of carbohydrates does not require such work, there appears to be a legitimate ground for questioning whether an excessive protein diet or an excessive carbohydrate diet would be the more desirable to provide a glycogen storage as preparation for muscular work. The value of large diets of either protein, carbohydrate, or mixed nutrients to replenish the glycogen depots and stimulate the whole body to cellular activity is plainly shown by our experiments. The practical application of this fact would seem to lie more particularly in the preparation for those bodily activities calling for considerable muscular work.