Grafe And Graham, 1911

The lengthy observation of Grafe and Graham1 on excess feeding of a 20-kilogram dog has provoked an unusual amount of discussion among physiologists, in spite of the fact that relatively little adverse criticism has been printed. Although primarily considering the question of overfeeding, the experiments have such a bearing on the possible adjustment of basal metabolism to food intake and have received so much attention from physiologists that we feel justified in discussing them here. Personal acquaintance with Dr. Grafe has led to a thorough investigation of this remarkable research, and it is a source of much regret that the present war conditions do not make it possible, before publication, to communicate with Dr. Grafe regarding our critical examination of his study. Although we are forced to dissent from the main conclusions, we are fully aware of the important place that the research has taken in physiological circles and the stimulus it has been to thought and to research.

The dog was first starved for 21 days. It was then given presumably excess food (280 per cent of the basal requirement) until the normal weight was regained (a period of 7 days). This was followed by 29 days of gross excess feeding (300 per cent of the basal needs), and three subsequent periods of 11, 19, and 10 days, respectively, with a smaller amount of food, but still presumably above the normal requirement (200,130, and ca. 100 per cent). During the last three periods the dog remained with an essentially constant body-weight. Grafe argues that this indicates an adjustment of the body to the excess nourishment. While the main argument is based upon the fact that the body-weight did not change, Grafe presents in addition a number of respiration experiments made during the period of supposed excess feeding, which he compares to a basal value and thus satisfies himself that during the period of excessive feeding there was a very greatly increased nilchtern metabolism.

1 Grafe and Graham, Zeitachr. f. physiol. Chem., 1911, 73, p. 1.

At first sight the experiment seems to be carried out with unusual care and accuracy; furthermore, Grafe gives a very good exposition of the main points under discussion. Unfortunately we find it necessary to differ with him on several fundamental points regarding his main conclusion that the animal was greatly overfed. During the period of 21 days of starvation the animal lost 5.15 kg. in body-weight and approximately 4 grams daily of nitrogen. During the second period, when the body-weight was being regained, Grafe's figures show that the animal received 2,243.9 net calories per day, and the nitrogen storage almost exactly compensated the nitrogen loss during the hunger period. The measurements of the metabolism during the hunger period, however, show that the total caloric loss during the 21 days was 17,403.5 calories.1 Assuming that the caloric loss was the same, whether or not the animal was inside the respiration chamber, the daily loss would be not far from 829 calories.

During the second period of 7 days, when presumably excess food was given, it is reasonable to assume that the caloric output could not have been much less than 1,000 calories per day. Unfortunately no metabolism experiments after the ingestion of food were made during this period, but in the later experiments reported by Grafe, in which 2,600 or more calories were given per day, the caloric output was 1,200 calories or over; 1,000 calories is therefore a minimum rather than a maximum estimate. Accordingly, during these 7 days at least 7,000 calories would be required. Since the dog was given approximately 16,000 total net calories the amount available for replenishment of the lost tissue and fat in the body was about 9,000 calories, or about one-half that actually lost during the 21 days. Hence it is clear that it is illogical to reason that the animal body was in its original condition at the end of the 7-day period of feeding.

In discussing the metabolism in the 29-day period of greatest excess feeding, Grafe has compared it with the minimum metabolism of the dog on the last days of fasting, and has computed a value which shows he considers that if the animal weighed 20 kg., it would have a minimum caloric production of 822.8 calories. We believe that Grafe (like Benedict and Carpenter1 of this Laboratory) has fallen into the error of using fasting values as distinguished from post-absorptive values for his basis of comparison. Benedict and Carpenter have discussed in considerable detail the error in using true fasting days for this purpose, since unquestionably the fast has, per se, a specific influence upon metabolism.2 On this ground the basal value employed by Grafe is altogether too low; according to our computation, the total heat production of the dog would much more closely represent 1,200 calories per day during the period of heavy feeding. Since during this time Grafe states that 2,580 net calories per day were available, the excess would correspond to about 40,000 calories, i. e., 2,580 - 1,200 X 29 = 40,020.

1 Computed from table II on p. 18 of Grafe and Graham's paper (loc. cit.) as follows:

In the period of 11 days, when feeding with 1,660 calories per day, the maintenance requirement would, in all probability, be not less than 1,112 calories, as found by Grafe. Since this is a niichtern value, during the feeding period it would unquestionably be somewhat higher, leaving the net surplus smaller. Disregarding this point, the net surplus is 548 calories, or a total for 11 days of 6,028 calories. In the next period of 19 days, the food was hardly above maintenance. Using the actual values obtained on those days when respiration experiments were made and the animal was without food in the stomach, the basal value can be assumed to be 1,030 calories. Since the daily energy intake during this period was 1,120 net calories, the surplus energy above maintenance would be only 90 calories daily, with a total excess for the 19 days of 1,710 calories. At the end of this period, therefore, we have a deficiency of 17,404 calories in the first 21 days to be made up, and available for this replenishment 56,758 calories. On this basis one can assume that 39,354 calories were stored in the body. On the basis of 9.5 calories per gram of fat this would correspond to approximately 4,143 grams or 4 kg. of fat.