In the historical development of the study on the influence of food upon the metabolism the first observations in which demonstrable increases could be determined were those with protein. These increases were so large that it was not at all strange that observers expected to find a considerable rise in the metabolism with both fat and carbohydrates. Accordingly, when a much smaller increment in the metabolism was found with carbohydrates than that obtained in experiments with protein, the influence of the former class of nutrients was without doubt given less consideration than it should have been.

The actual importance of the increases with carbohydrates may have been obscured by several causes. First, the effect of carbohydrate ingestion persists for a much shorter time than that following the ingestion of protein; hence, in the experimental period first used (24 hours) the increase in the metabolism in the hours immediately following the taking of the carbohydrate food may have disappeared when the values for the essentially basal metabolism in the later hours of the day were included; in other words, the "peak" effect of the carbohydrate ingestion was lost as a result of the lengthening of the experimental period.

Secondly, it has frequently happened that the basal value was determined in 24 hours, or even longer, of complete starvation. Experiments have shown1 that during period of this length without food there is a very considerable draft upon the carbohydrate storage in the body; consequently when carbohydrate is afterwards ingested, the body attempts first to replenish the store of this material. The effect on the metabolism due to the ingestion of food is thus considerably lessened by the fact that the carbohydrate is in large part not burned, but simply stored as glycogen.

In Rubner's experiment on man2 (and in this monograph we are dealing entirely with experiments on man) a series of experiments on 5 consecutive days was carried out. On the first day the subject fasted and did no work; on the second day he was given protein without work; on the third day protein with work; on the fourth day sugar without work; and on the fifth day sugar with work. Considering specifically the fourth day, when sugar was given without work, we find that the heat output per 24 hours was 2,023 calories as compared with a basal value of 1,976 calories, an increment of only 47 calories. A close examination of the experimental procedure shows that the 3 days prior to the sugar day, i. e., a day of hunger, a day with protein, and a day with protein and work, all contributed toward the depletion of the glycogen supply in the body, and it is not surprising that no larger increment in the metabolism was found.

1Johansson, Skand. Arch. f. Physiol., 1909, 21, p. 1. See, also, p. 70 of this monograph. 2Rubner, Sitzber. K. Preuss. Akad. Wiss., 1910, p. 316.

It is surprising, however, that according to the method of computing the energy used by Rubner at that time, it was assumed that the carbohydrates were first burned. As the 600 grams of cane sugar given on that day correspond to an energy consumption of approximately 2,400 calories, or more than was actually measured, it is presumable that the 2,023 calories given by Rubner as the value for the heat output is based upon the assumption that the combustion for the day was entirely of pure carbohydrate. This amount of energy would be produced by the combustion of approximately 510 grams of sugar. It is well known that 50 to 100 or more grams of glycogen may be withdrawn on the first day of fasting. It is quite likely, therefore, that an appreciable portion of the sugar ingested may have been used to form glycogen and not to contribute to the increased metabolism measured by Rubner.

Rubner's calculations are seriously hampered by the absence of data regarding the oxygen consumption, which would contribute more directly to the computation of the total energy transformations. The increment in the carbon-dioxide production has been more employed for such researches as these than any other factor. Even when used in special studies like those of Johansson and Gigon,1 an attempt to explain the processes on the basis of this increment immediately results in great confusion. A typical case will serve to illustrate this.

If a subject has been without food for 12 hours or more and is drawing upon body material to the extent of 15 per cent of the total energy in the form of protein and the other 85 per cent is apportioned between carbohydrate and fat, presumably in the proportion of 45 per cent of carbohydrate and 40 per cent of fat, the respiratory quotient will be approximately 0.85. When carbohydrate is ingested there is immediately a great rise in the respiratory quotient and an increase in the production of carbon dioxide. It may be argued, then, in common with the old conception of von Hoesslin,2 that a fat-carbohydrate-protein combustion is replaced by an exclusively protein-carbohydrate combustion, without altering in any way the total amount of energy transformed. This is one possibility.

Another possibility is that there may be a transformation of carbohydrate into fat. By this process, which has been definitely proved in several laboratories in a number of ways, there may be a formation of fat and a splitting off of carbohydrate when large amounts of carbohydrate are ingested, with a so-called "atypical" carbon-dioxide production, unaccompanied by an increase in the oxygen consumption. In other words, this process is entirely aside from the katabolic processes in the body and does not affect the total katabolism appreciably, though there is a slight energy output incidental to the transformation.

1Probably no research on the influence of the ingestion of pure carbohydrate has been more accurately and carefully carried out than that of Johansson at Stockholm (see pp. 34 and 35), which was supplemented by the subsequent experiments of Johansson's former assistant, Gigon (see p. 38). While Rubner's criticism (Die Gesetze des Energieverbrauchs bei der Ernahrung, 1902, p. 216) of Johansson's method of computing the carbon-dioxide production of a single individual in half-hour periods seems justified when we consider that the volume of air in the chamber was 100,000 liters, nevertheless personal visits to Stockholm have convinced us that the remarkable Sonden gas-analysis apparatus used by Johansson permits measurements of carbon dioxide with a sufficient degree of accuracy to justify recording values for half-hour periods with this chamber, if not, indeed, for 15-minute periods.

Von Hoesslin, Arch. f. path. Anat. u. Physiol., 1882, 89, p. 341.

Finally, there may be an actual increase in the total katabolism, which would of itself result in an increased carbon-dioxide production. This increase in the katabolism may be caused by an increased tonus and an increased activity in the digestive tract due to the stimulating effect of the absorbed food materials upon the body cells. It is thus clear that these three processes, which may take place simultaneously in varying degrees of intensity, greatly complicate the interpretation of results based only upon the carbon-dioxide increment.

The experiments in this research which were designed to study the influence of the ingestion of carbohydrates were planned to measure not only a prolonged effect but particularly to show the maximum carbon-dioxide production and oxygen consumption which might appear early in the observations. In the calorimeter experiments measurements were also made of the heat production. For the respiration experiments the heat production was calculated from the results obtained for the gaseous exchange. A number of carbohydrate food materials were used, including not only pure carbohydrates, such as cane sugar, dextrose, levulose, and milk sugar, but also those of a mixed nature, like bananas and popcorn. As in the series of experiments already discussed, the data were secured with the respiration calorimeter at Wesleyan University, Middletown, and with the chair calorimeter and two forms of respiration apparatus at the Nutrition Laboratory, Boston.