To study the influence upon metabolism of such a factor as the ingestion of food, the energy requirements of the quiescent body prior to the ingestion of the food must be known, for otherwise the measurement of metabolism after food can have no comparative significance. Thus the whole problem of demonstrating the influence of the ingestion of food upon metabolism depends upon two vitally important processes: (1) the establishment of a suitable base-line, and (2) the accurate measurement of metabolism following the ingestion of food.

While at first sight it might be assumed that the establishment of a base-line is relatively simple, close analysis shows that this is far from being the case. In the first place, there is no normal value for either male or female adults that may be taken, a priori, as a base-line for any subsequent measurements. Various attempts have been made to establish more or less crude "standard" values and results have been obtained which give rough indications of the major changes in metabolism due to disease, food, or muscular work. These so-called standard values can not, however, be used for any quantitative study of the influence of a specific factor upon metabolism. Each series of measurements accordingly demands its own basal determination.

In determining basal values, the conditions should preferably be as much as possible like those obtaining during the comparison experiments. Thus, in any research on the effects of bicycle riding, it may be fairly argued that the base-line should be determined not when the subject is lying in deep sleep, but when he is sitting in the ordinary position occupied by a bicycle rider. Again, when the work of horizontal walking is studied, the base-line would not logically be that obtained during deep sleep, but would be a value secured with the subject standing in readiness for walking.

The degree of care necessary in the selection of a base-line is dependent upon the size of the increment in the metabolism due to the superimposed factor. By active muscular work it is perfectly possible for a well-trained athlete to increase his basal metabolism tenfold or more, the professional bicycle rider studied by us1 and also the one studied by Benedict and Cathcart2 showing no difficulty in producing such increases. With values so large as these, it is clear that small differences in the base-line play a comparatively unimportant role. Indeed, it has been the custom in the researches on muscular work, published not only from this laboratory but also by investigators elsewhere, to use a basal value determined with the subject lying down but not asleep. While such a practice is theoretically unsound, the increments due to muscular work are so large that in comparisons of metabolism during muscular work and during rest the relatively slight differences between metabolism during sleep and that with the subject standing or sitting quietly or lying down awake may be neglected.

In studies on the influence of food upon metabolism, the increments are much smaller than in studies with muscular work. A glance at the literature (see pages 10 to 46) shows that the maximum effect due to this factor may be to increase metabolism for a short time, possibly 30 or even 40 per cent. When we consider the potential increment of 1,000 or more per cent with muscular work, even this maximum increase in metabolism after food seems comparatively insignificant. Accordingly, in a study on the effect of the ingestion of food, great care should be taken to secure a uniform base-line and a critical examination should be made of those factors liable to influence the determination of the basal metabolism.

The quiescent metabolism of the body may be affected by a number of factors, primarily by muscular activity. We have already seen that severe muscular work increases the metabolism largely, but we find that moderate activity or even the relatively few muscular movements that distinguish between complete rest and ordinary rest also have a definite influence. Furthermore, when the increment in metabolism to be measured is probably small, one has to consider not only minor muscular activity, but even the degree of muscular relaxation. Thus we find Johansson3 training himself and his co-workers to establish an arbitrarily complete muscular repose. Finally, experimental evidence4 obtained in the Nutrition Laboratory has shown positively that the quiescent metabolism of a subject asleep differs considerably from that of the same subject awake. In experiments with the subject in a profound sleep there was a noticeable decrease in pulse rate, which was almost invariably accompanied by a decrease in total metabolism. We may expect, therefore, that with the subject in deep sleep there will be a decrease in pulse rate, respiration rate, and muscle tonus, with consequently lower metabolism as compared with values obtained with the same subject awake and in complete muscular repose. Even though the body be muscularly quiet while lying on a couch or bed and the voluntary muscles be perfectly controlled, the involuntary muscles, such as those of circulation, digestion, and respiration, are active. These involuntary muscles continue their work in deep sleep at a somewhat lower level.

1Benedict and Carpenter, U. S. Dept. Agr., Office Exp. Stas. Bull. 208, 1909. 2Benedict and Cathcart, Carnegie Inst. Wash. Pub. No. 187, 1913. 3Johansson, Skand. Arch. f. Physiol., 1898, 8, p. 119. 4Benedict, Carnegie Inst. Wash. Pub. No. 203, 1915, p. 343.

A second factor which definitely affects the base-line is previously ingested food. It has been clearly demonstrated by practically all of the earlier workers that an increased metabolism follows the taking of food, particularly when protein and certain carbohydrates form a part of the diet. If possible, therefore, we must find a point in the digestive cycle at which the metabolism will not be influenced by the previous diet, but which will be prior to the severe drafts upon the body glycogen that have been found in researches at both Wesleyan University and the Nutrition Laboratory during several days of strict fasting. It seems to be the consensus of opinion of nearly all experimenters in this line of research that with normal man, unless the last meal has been excessively rich in protein, active digestion ceases 12 hours after the ingestion of food and the metabolism has then reached essentially the normal level, i. e., the level prior to the taking of food. This has been demonstrated in a number of researches, particularly those of Magnus-Levy.1 Hence it is now the custom of most experimenters to study the basal metabolism by making experiments 12 or 14 hours after the last meal or, as Benedict and Cathcart have expressed it, with the subject in the "post-absorptive condition,"2 and to assume that the influence of previously ingested food will in this way be eliminated.3 The metabolism at this time, however, does not always represent the minimum metabolism, as will be seen in a later discussion.