The Journal of Clinical Investigation Vol. 46, Nc.

10, 1967

Brain Metabolism during Fasting *
0. E. OWEN, A. P. MORGAN, H. G. KEMP, J. M. SULLIVAN, M. G. HERRERA, AND

G. F. CAHILL, JR.4
(From the Elliott P. Joslin Research Laboratory, the Departments of Medicine and Surgery, Harvard Medical School, the Cardiovascular Unit, the Peter Bent Brigham Hospital, and the Diabetes Foundation, Inc., Boston, Massachusetts)

Abstract. Catheterization of cerebral vessels in three obese patients undergoing 5-6 wk of starvation demonstrated that /8-hydroxybutyrate and acetoacetate replaced glucose as the predominant fuel for brain metabolism. A strikingly low respiratory quotient was also observed, suggesting a carboxylation mechanism as a means of disposing of some of the carbon of the consumed substrates.

Introduction Numerous studies performed after brief postabsorptive periods have shown that the only significant energy-yielding substrate consumed by brain is glucose, at rates of 110-145 g/24 hr (2-4). In adult man glucose and glycogen stores amount to 150-300 g, and the protein mass is approximately 6-8 kg (5). Although the fat depot may be massive and glycerol from neutral fat can be converted to glucose, net gluconeogenesis from fatty acid carbon has not been demonstrated in mammalian systems. If the central nervous system does maintain an absolute requirement for glucose during starvation, this substrate must be derived from the limited carbohydrate stores, from de novo synthesis from amino acids from protein, and from other precursors such as glycerol from adipose tissue, or lactate recycled from glycolytic tissues.
* Received for publication 17 April 1967 and in revised form 23 June 1967. This study was supported in part by grants from the U. S. Public Health Service AM-09584-02, AM-09748-02, 8 MO1-FR-31-06, Tl-AM-5077-11, 5-RO1-AM-02657-07, 1-Tl-HE-05679-02, and HE-08591-02; the Adler Foundation, Inc., Rye, N. Y.; the U. S. Army Medical Research and Development Command (DA-49-193-MO2337) ; the John A. Hartford Foundation, Inc., New York, N. Y.; and the Atomic Energy Commission. A preliminary abstract of part of this work has been

The use of prolonged starvation for the treatment of obesity has posed a fascinating problem; namely, that man is capable of fasting for periods of time beyond which he would have utilized all of his carbohydrate resources and all of his proteins for gluconeogenesis in order to provide adequate calories as glucose for the central nervous system. This study was designed to clarify the apparent paradox, and it was found that f?-hydroxybutyrate and acetoacetate replace glucose as the brain's primary fuel during starvation. In addition, a surprisingly low respiratory quotient was observed due to failure to recover substrate carbon as CO2, suggesting a carboxylation by a pathway not yet defined. Methods Subjects. Three obese subjects were admitted for study to The Clinical Center of the Peter Bent Brigham Hospital (Table I). Subject M. B. had intermittent precordial discomfort, initially thought to be angina pectoris; F. N. had hip and backaches attributed to osteoarthritis. He also had mild hypertension, latent diabetes mellitus as detected by a mildly diabetic intravenous glucose tolerance test and electrocardiographic changes compatible with early left ventricular hypertrophy. Subject M.L. had menstrual irregularities unresponsive to uterine curettage and cyclic hormonal therapy.
Before admission, all three were informed of the procedures to be followed and the potential risks. The investigators, and the patients as well, felt that the procedures were warranted in order to elucidate possible metabolic aberrations associated with obesity. On admission, all had normal hemograms, urinalyses, chest and abdominal roentgenograms, electrocardiograms (ex-

published (1). j Investigator, Howard Hughes Medical Institute. Address requests for reprints to Dr. G. F. Cahill, Jr., Joslin Research Laboratory, 170 Pilgrim Road, Boston, Mass. 02215.

1589

serum enzymes (alkaline phosphatase. female 49.0 177.0 123. was excreted on three or four occasions by F. 7 catheters were passed into the internal jugular bulb and hepatic vein.54 2. lactic dehydrogenase. diluted 1: 5. female 41 38 40 173.B. B. Estes. ammonia (14). which were done in triplicate by both the AutoAnalyzer and Somogyi-Nelson techniques.35 0. MORGAN. using a supernatant fluid obtained after deproteinization with an equal volume of serum and 0. Blood which was removed for analysis was replaced isovolumetrically by 5% human albumin in isotonic saline.72 0.8 -23.2 -24. glutamic-oxalacetic transaminase). Blood samples were then drawn simultaneously during 10-sectimed intervals from the various catheters every 15 min for four to six study periods of time and immediately prepared for analysis of oxygen and carbon dioxide content and determination of metabolic substrates.8 109. 17 mEq of NaCl. Position of the catheters was repeatedly ascertained visually with image intensification.) the volume was measured and samples taken for total nitrogen determination (Kjeldahl).00 0. F.16 * Mean of 3 days measured during last week of fasting before catheterization.. mean wt. urea nitrogen. .5 176. the supernatant fluid was filtered and stored at -20'C.). thyroid indexes.26 1. N.00 0.N. All analyses were done in triplicate except those for glucose. Oxygen and carbon dioxide contents were measured manometrically by the method of van Slyke and Neill. The remainder of analyses were performed by immediately injecting 10 ml of blood into 10 ml of iced 1 M perchloric acid and mixing.26 2.63 0. L.10 0. daily walks.09 0. the aortic catheter was repositioned in the carotid artery.41 0. Mellanby. and informative dietary sessions. male 26.6 147. and heated for 5 min in a boiling water bath.37 +61 +65 +109 +28 +37 +76 From Metropolitan Life Insurance tables. KEMP. passed nothing per rectum for the starvation period. composed primarily of mucus.g-hydroxybutyrate and acetoacetate by a modification (10) of the method of Williamson. After initial colonic evacuation. Kalamazoo. Mean* 1985 1707 1773 1822 1888 1607 1691 1729 0 0 0 0 97 100 82 93 3. M. Scant fecal material.B. and Krebs (11) with 8-hydroxybutyric acid dehydrogenase. and Friedberg (12) and a-amino nitrogen by the colorimetric method of Moore and Stein (13). Free fatty acids in serum were measured by the method of Trout.06 g 0.32 0.27 3. For cerebral blood flow. The subjects were premedicated with 100 mg of sodium pentobarbital and 50 mg of meperidine sulfate intramuscularly 1-2 hr before the procedure.11 0. intra-arterial pressure and heart activity were continuously monitored. Blood analyses. Mich.05 M acetic acid.96 2.96 2. M.L.L. N. a Sones No.96 1. and 1500 ml of water each day.33 2. F.00 3. Goodale-Lubin No. Pyruvic and lactic acids were measured in the potassium bicarbonate-neutralized supernatant fluid by standard enzymatic techniques (9) and . electrolytes. Intake during starvation consisted of one multivitamin capsule (Unicap. Urine analyses. M.7 -22. blood was added to tubes containing oxalate-fluoride and analyzed in triplicate within several hours by both the SomogyiNelson (6. uric acid (AutoAnalyzer). The subjects were encouraged to participate in occupational therapy.49 0. lipids. 1959. Through the right and left antecubital veins. SULLIVAN. cept F. Catheterization. and normal concentrations of serum protein. Urine was collected in refrigerated plastic containers. AND CAHILL TABLE I Clinical data Weight Final Body surface area A Initial M2 Patient Age and sex Starvation days Height cm Initial Final Per cent deviation from pop.13 0.1 2.86 0.90 4.8 124.m. The Upjohn Co.46 0. * 42. Throughout the procedure. For glucose determination.14 2.N. creatinine.5 132.* Initial Final kg M. and uric acid before starvation.45 2.1590 OWEN. and electrolytes. at the end of 24 hr (at 7 :00 a. HERRERA. After the patients had fasted 38-41 days. 7 catheter was placed percutaneously through the right brachial artery and passed into the aortic root. After centrifugation in the cold.70 1. and the values for glucose are the average of all six determinations for each study period. 7) and Technicon AutoAnalyzer ferricyanide procedures (8).). and was not analyzed.3 98. and M. TABLE II Caloric expenditure and nitrogen excretion/24 hr Calories CarboUrinary nitrogen Total Urea Unidentified Patient Total Fat hydrate Protein Ammonia Uric acid Creatinihe M.

28 4.24 0.09 ±0.06 ±0. again an insignificant quantity compared to daily energy expenditure.N1 M. Hourly.004 ±0.34 ±0. Flow rates were calculated using the formulas of both Lassen et al. The samples were passed through an analysis train composed of an American Meter Company wet test flowmeter.08 ±0.96 -2.17 ±t0. . ±0.26 ±t0. N.003 0.08 ±0.06 ± 0.23 3.23 ±0. M.05 i 0. rather than the 138 g needed in the fed state. using the sampling routine validated by Kinney (15).029 -0. was Results In Table II are listed the mean daily caloric expenditures.09 0. Blood flow. Table III gives the arterial concentrations of the metabolic fuels. a 5 min sample of expired air was collected in an aluminized plastic bag.023 * Values in mmoles/liter whole blood (plasma for free fatty acids).38 3.07 ± 0.67 1 0.02 i 0.28 40.04 0. ketonuria depresses the nonprotein respiratory quotient.86 1.09 -0. if one corrects Loewy's original calculations (16) for subject M. this amounted to approximately 70 kcal/day.01 i 0. and mean daily nitrogen excretions.04 i 0. Urinary nitrogen averaged between 3.010 0. and the energy expenditure calculated from total oxidation of fat is overestimated.65 ±1=0.B.37 1.27 2. but this can be derived from the data presented in Table I.65 0.14 0.05 0..61 3. Energy expenditure was measured by indirect calorimetry.06 ±0. B. as well as the identified and unidentified nitrogenous components in the urine for each subject. and Table IV presents the cerebral arteriovenous differences with the mean and standard errors of the means of the multiplepaired.71 0. The conventional calculations of indirect calorimetry were applied. to 7:00 p. The data are not corrected for body surface area.96 0.92 6. (17) and Zierler (18).m.04 4± 0.05 ±0.3 and 3.01 0.081 0.57 ±0.12 2. 100 g protein oxidation requires only 112 g of 02. In addition.63 ±0.34 ±0. from 7:00 a.03 0. In F.27 -1.01 i 0. F.19 ±t 0.B.L.04 0.01 i 0. Two standard gas mixtures were analyzed before each sample was measured.arteriovenous analyses for each of the four TABLE IV Brain arteriovenous differences* Patient Oxygen Carbon dioxide Respiratory quotient Glucose Lactate Pyruvate P-Hydroxybutyrate Acetoacetate a-Amino nitrogen Free fatty acids M.05 i 0.38 0.05 ±0.003 -0. with the patient reclining in bed.06 1.15 ±t0.08 ±0.01 7.03 ±t0.44 3.032 -0. by allowing the decreased fecal nitrogen and altered urinary nitrogen partition.023 0.03 0.20 -0.26 -1.09 ± 0.037 ±0.000 Mean ±0. Cerebral blood flow was determined in subjects F. but the resulting error is far smaller than the sampling error.004 8.53 ± ± ± :1 0.03 0. t Six observation periods.36 ±0. Calorimetry.001 0.055 i 0.047 i 0.57 0.003 -0.49 ±.019 0.003 0.BRAIN METABOLISM DURING FASTING TABLE III 1591 Arterial levels of metabolic fuels Patient Glucose Lactate Pyruvate P-Hydroxybutyrate Acetoacetate a-Amino nitrogen Free fatty acids F. For example.01 0.03 0. L.04 ± 0.N.06 0.003 -0.27 4.41 ±0.43t 0. The usual assumptions are not strictly correct in the fasting patient. and M.09 i 0.* M.05 -0. Each value is the mean with SE of the mean for 4* or 6t observation periods.19 2. The patients were allowed to become accustomed to the face mask during the prestarvation period.08 ±0.05 0.90 0.72 ±0.01 -0.21 ±0.02 ±0.32 2. N.14 :i 0.02 2.02 0.07 4.033 ±0.89 2.57 ± 0. 'Kr dissolved in sterile saline and injected rapidly into the right common carotid artery.054 i 0.29 0.52 40.* Mean 3.061 :1 0.. simultaneous integrated sampling from the right jugular bulb was also done.26 ±0.05t ±0. B.82 2.01 i 0. using Krypton ('Kr).029 ±0.03 ±L0.0.07 0.060 ±0.91 5.17 ± 0. Each value is the mean with SE of the mean for four observation periods.43 Values in mmoles/liter of whole blood (plasma for free fatty acids).37 ±0.87 -1.15 +0. their calculated sources.26 0..9 g/24 hr.10 0.m. of which the main component was ammonia. and a Beckmann E-2 oxygen analyzer.L.06 i 0. a modified Beckmann LB-2 carbon dioxide analyzer. In M.41 0.04 ±t0. The clearance rate of 'Kr was monitored by a 2 inch external gamma scintillation detector with a 1 inch collimation tube placed over the right temporal region.08 0.06 0.

03 i 0. as well as an insignificant mean arteriovenous difference for the average of all three subjects. as well as the amount of CO2 theoretically produced from each substrate.81 mmoles/liter (6.37 A-V. This is slightly lower than the mean value reported in the literature ( 19).34 d 0. N. the significant production of lactate and pyruvate.7 mg/100 ml).06 ± 0.53 0. The large error in the determination of a-amino nitrogen. to six periods of study. one can simply calculate a theoretical value for glucose synthesized from nitrogen loss by simply equating 1 g of nitrogen to 6. the fact that F. pyruvate. SULLIVAN. Of more importance is the large standard error in each subject for the uptake of free fatty acids. AND CAHILL TABLE V Brain oxygen substrate equivalents in mmoles/liter of blood Mean values of three subjects after 38-41 days of fasting. Although in good agreement with the glucose: nitrogen ratio of Stiles and Lusk (21).81 Total substrate equivalents Measured oxygen consumption or CO2 production -1.06 2. diminishes the significance of the uptake of this substrate. and this would yield about 14 g of glucose. Substrate A-V difference Calc. since only a single assay was performed in each subject in contrast to the repeated steadystate values for gases and substrates.26 mmole/liter (4. By these techniques. L.85 volumes/100 ml) agrees well with that determined directly. the calculated CO2 which should have been produced. The average nitrogen loss per 24 hr in these three subjects during the last week of fasting was 3. supporting our desire that the values represent true steady-state conditions and therefore not be subject to the problems incurred by changes in levels or pool sizes. it is possibly in excess since a considerable per- . 2. as well as the variations in the differences for each study period. In contrast.90 mmoles/liter (4.96 mmoles/liter (6. 02 equivalent C02 equivalent Glucose Lactate Pyruvate Glucose + (L + P)/2 f3-Hydroxybutyrate Acetoacetate a-Amino nitrogen 0.87 -1.25 g of protein.003 0.24 -0.06 mmoles/liter (6.09 d 0. Nevertheless. however.30 volumes/ 100 ml).'s flow was the same when calculated from both internal and external techniques gives validation to the observation.63 volumes/100 ml). B.1592 OWEN. Table V presents the calculated mean values for substrates removed and produced for each liter of blood perfusing the brain.03 0. 1. Appropriate corrections are made for the lactate and pyruvate produced from glucose.145 0. Of note in Table IV are the relatively reduced glucose uptake of 0.72 g. N.05 0. N. It should be emphasized. L.26 :1= -0. lactate. cerebral blood flow in both subjects was exactly 45 ml/ 100 g per min.42 3.20 at -0.96 :1: 0. P. and M. Accidental unavailability of 85Kr prevented similar measurements in M.01 0. KEMP.36 -0. that the flow determinations are only approximations. arteriovenous. The concentration of the metabolites showed no significant decrease or increase throughout the catheterization study.34 -2. and also by internal monitoring in F. and the significant uptake of /8-hydroxybutyrate and acetoacetate. Also presented are the calculated oxygen equivalents which would be needed for total combustion of the substrates to CO2.90 i 0. The total calculated oxygen consumption of 3. This striking discrepancy in recovery of expected CO2 is also reflected in the extremely low respiratory quotients listed in Table IV.02 0.87 1. Also presented are the means of the arteriovenous differences for the three subjects.24 0. 57 g of glucose may be derived from 100 g of protein (20). MORGAN.34 -0. Theoretically. Discussion Assuming a quantitatively minimal loss of nitrogen by routes other than urine. Cerebral blood flow was determined by the external techniques in subjects F. is far in excess of that directly determined.25 volumes/100 ml). HERRERA.06 0. but yet within the normal range. 2. the mean being insignificant for each individual subject.

Kemp. resulting in a quotient of 0. the rate of utilization or production of specific metabolites can be quantitatively estimated. Another source of glucogenic precursor is adipose tissue glycerol. however. this deficit in CO2 production can only be explained by a carboxylation reaction with the venous effluent transporting the CO2 in a form not liberated by the acidification used in the standard manometric technique for determination of CO2 and HCO3. A third source for glucose is lactate and pyruvate returning to the liver and kidney from glycolysis. Morgan.. Additional indirect evidence for a novel carboxylation reaction which would result in a low quotient has been Sacks' studies on glucose-14C oxidation in human brain whereby only 50%o of glucose-carbon that 10. Similar data were observed in our laboratory while the techniques and methods used to study these fasted subjects were validated. H. (35) noted decreased respiratory quotients in patients in diabetic ketoacidosis. these two substrates are quantitatively reincorporated into glucose. which relate to the respiratory quotient and carbon balance across the brain. G. In addition. should have ben produced. G.90 mmoles/liter. Brain. To our knowledge. a sum of approximately 33 g of glucose or less is available for terminal oxidation by the body. and therefore are not lost by terminal oxidation and do not contribute to net carbohydrate loss. is 0. M.BRAIN METABOLISM DURING FASTING 1593 centage of the nitrogen loss in the urine is not from glucogenic sources. respectively. including fixation of CO2 (31. for a total. by the 5th or 6th wk of fasting this amount probably does not contribute significantly to daily glucose production. as stated earlier. of approximately 33 g of glucose synthesized daily. which agrees well with the theoretical maximum of 33 g calculated from nitrogen execretion and glycerol from adipose tissue as described above. Turning to Table V. Table II shows that about 1729 cal are derived from fat daily.g. Data in preparation. M. since the total sum.' demonstrated that the liver almost totally ceases to synthesize glucose from amino acids and that the kidney assumes the role of the major source of this diminished amount of glucose daily produced and consumed during starvation. 3. 32). which obtained from hepatic and renal vein catheterization studies. the two numbers are in good agreement. 2. A.96 mmoles/liter of blood. Thus. producing 19 g of glycerol which can be incorporated into glucose. Cahill. Owen. with a theoretical respiratory quotient of 0. when added to that derived from nitrogen. However. 25). Sullivan. According to most observations. E.6 mg/100 ml). P. is close to unity (2. 3. The fourth source of potential glucose is the stored liver and muscle glycogen (the latter available by glycolysis to lactate and pyruvate and then by resynthesis to glucose in the liver). Jr. after subtracting the amount glycolyzed to lactate and pyruvate. and oxidation of keto acids has been demonstrated in vitro (30. Kety et al. and compare this sum to the oxygen directly determined. and G. which in fat approximates 10% of the weight of the triglyceride molecule. With measured cerebral blood flow of 45 ml/100 g of tissue per min.81 mmoles/liter of CO. The third confirmatory evidence for this marked reduction in glucose metabolism has been data. the respiratory quotient of brain. but direct utilization of keto acids has not been found in this condition (3) or in fat-fed animals (36). which glucose serves as sole energy source. 33. creatinine and uric acid. By combining the measurement of blood flow with the arteriovenous differences. is only about 150-300 g. Also. and amino acid assuming the a-amino nitrogen as alanine. .06 and 2. and assuming a brain size of 1400 g the 24 hr glucose oxidation would approximate 24 g. as stated before.62. contains enzymes for all the major metabolic pathways (29-31).145 mmole/liter (2. the average glucose uptake. F. Herrera. If we add together the calculated 02 equivalents needed for combustion of the glucose consumed (corrected for lactate and pyruvate production) and the equivalents needed for combustion of the f3-hydroxybutyrate. G6ttstein et al. e. the greatest proportion of which is produced by the red cell mass. a quantity approximately 30-40 g/ day (22). Of interest are the data presented in Tables IV and V. This is markedly less than the usual 9-10 mg/100 ml observed in our laboratory while the techniques which little or no production of lactate or pyruvate was observed (25-28). 34).92 instead of the observed 1. also observed decreased respiratory quotients in patients with cerebral arteriosclerosis (37). J. acetoacetate.

the nitrogen loss of the other nonobese "professional" fasters studied at the turn of the century (40).) 2. M. S. 221. Zierler. M. N-2b). HERRERA. AND CAHILL is oxidized is recovered in effluent CO2 and HCO:. Regional cerebral blood flow in man determined by krypton'M. MORGAN. A. A possible objection to this study is the use of obese subjects and the utilization. at least in part.1594 OWEN. but it is extremely doubtful if these subjects do differ from normal. 9. Acad. London. Richter. and B. 211: 907.. Total cerebral blood flow and metabolism. J. J. Krebs. 6. 8-Hydroxybutyrate and acetoacetate utilization by the brain shows that fat products may even satisfy the central nervous system's substrate requirement. and W. Oppenheimer. Conway. Morgan. Nelson. 1960. Fat has the greatest caloric potential per unit weight and is readily expendable. S0rensen. 82: 90. A photometric adaptation of the the Somogyi method for the determination of glucose. Widdowson. C. Herrera. Ltd. (1956). 10. 4: 279. N. Sokoloff. L. K. In The Metabolism of the Nervous System. Biol.. Finally. D. Y. M. D. Levy. KEMP. editor. and Maryellen Valzania for their technical assistance. M. and should be expendable without adverse effects.. S. J. Cahill. A modified ninhydrin reagent for the photometric determination of amino acids and related compounds. J. Gustav Fischer. 1957. S. J. Owen. SULLIVAN. J. Academic Press Inc. Friedberg. In Human Body Composition. J. M. 16. and G. and H. (38). the Metabolism of the Nervous System.. Circulation Res. Clin. Scheinberg. G. G.. E. Equations for measuring blood flow by external monitoring of radioisotopes. 98. and D. 3: 31. B. K. Crosby Lockwood & Sons. F. particularly in a primitive setting. should be capable of meeting substrate requirements for all tissues. 3. our subjects failed to show any deficit on psychometric testing. Chemical analysis of the body. J. Neurol. 2nd International Neurochemical Symposium. Hoffman. 0. London. 13: 719. 13. 1911. Jena. Metabolism of the central nervous system in vivo. 0. This means that the depot should have the highest calorie: weight ratio. Chem. Somogyi. Acknowledgments We wish to express our appreciation to Mrs. Determination of blood sugar. 5. L. N. Section I. Velta Ramolins. Bourne. (Abstr. J. J. 14. 1: 199. P. K. Jr. of which the major portion is muscle. 1966. Stein.. 36: 617. Neurophysiology. Reinmuth. Oxford. Steinke. Jr. 1966.. Sullivan. and D. and in addition our own prior study of nitrogen loss during a shorter period of fasting in normals (10). E. Ingvar. Chem. 16: 309. Chem. Kinney. A. A rapid photoelectric method for the determination of glucose in blood and urine. D. P. 1945. E. 12. Survival during prolonged fasting. L. J. Enzymatic determination of D (-) -B-Hydroxybutyric acid and acetoacetic acid in blood. New York. Williamson. 1963. 1944.. U. and which are superimposable with our observations in the three obese subjects. Man's capability to survive marked extremes in caloric intake depends. F. H. Brain metabolism during starvation. Biochem. S. Kipnis. Jr. Inivest. Chem. 1937. Soeldner. Bergmeyer. Neurology. Waverly Press Inc. Jr. M.of these data as representative of normal man. W. In Handbuch der Biochemie des Menschen und der Tiere. Cahill. Microdiffusion Analyses and Volumetric Error. Bodforss. Biol. and therefore circumvent the need for gluconeogenesis and concomitant nitrogen depletion. Baltimore. a few general comments seem in order. Gaswechsel und Stoffwechsel. which suggest the same progressive decrease in gluconeogenesis as brain adapts to keto acid utilization. 17. obviously necessitates maximal sparing of nitrogen depots. 1963. E. S.. Estes. and S.. S. S. Titration of free fatty acids of plasma: a study of current methods and a new modification. 153: 375. Brozek. and electroencephalographic tracings remained unchanged. Kety. Lipid Res. The general metabolism of the brain in vivo. References 1. Trout. 15. editor. Biol. A. 7. and to the nurses and staff of The Clinical Center and Cardiovascular Research Unit at the Peter Bent Brigham Hospital. Indirect evidence of this fact is the nitrogen excretion of Benedict's nonobese subject (39). 1965. 4. L. Arch.. 160: 69. J. 1960. H. S. Cronquist. . In Handbook of Physiology. P. 1962. Methods of Enzymatic Analysis. 1959. 1843. 12: 49. Moore. Clin. H. Pergamon Press.. Biol. nonobese individuals fasted over the same duration.. Pergamon Press. 4th edition. Mellanby. J. 1965. 11. Reichard. H. N. on his ability to store fuel in an economical form. E. A consideration of energy exchange in human trauma. 1965. Skenhoj. 1954. Finally. Med. 18. Res. Vilma Lauris. A. 120: 51 (as modified by Technicon AutoAnalyzer Methodology. 45: 1751. Hoedt-Rasmussen. J. 1957. 14: 351. which suggests that the keto acids did fulfill the predominant energy requirement in a satisfactory fashion. M. Lassen. J. G. H. Hormonefuel interrelationships during fasting. Bull.

Clin.. F. J. G. and G. Allison. 1967. Glucagon und Glucose-Insulin.. Stiles. Kleinerman. New York.. L. K. Scheinberg. Psychiat. Schmidt. Clin. 9th edition. and G. J. Pyruvic acid exchange of the brain. Pyruvate carboxylase activity of mammalian brain cortex. 33. Gottstein. L.. Banks. Biochem. Reinmuth. W. S. editors. Tschirgi. G. A. Neurophysiology. D. Takagaki. Invest. Effect of high spinal anaesthesia on cerebral circulation and metabolism in man. 40. Saunders Co. Baltimore. G. and H. Cahill. The metabolic fate of amino acids. Polis. Neuronal Metabolism. 24. Chem. and W. 274: 82. 1965. Nadler. 237: 2570. 32. Human cerebral lactate and pyruvate extraction. Carbon dioxide fixation in the brain.. Physiol. 168: 728. 25. J. 1940. 90: 464. A. Whelan. J. R. A. W. 9: 133. In Mammalian Protein Metabolism. Neurol. W. and W. A. 1942. and D. On the formation of dextrose in metabolism from the end-products of a pancreatic digest of meat. 22. 1928. J. Life Sci. and A. Physiol. A study in man of cerebral bood flow and cerebral glucose. McCann. Physiol. A. The blood flow and oxygen consumption of the human brain in diabetic acidosis and coma. A. 6: 133. S. J. Am. Cerebral metabolism in fat-fed dogs. Leng.. D. J... and R. 1966. C. and C. Invest. G. 27: 500.. A. L. In Carbohydrate Metabolism and its Disorders. M. Invest.. 1967. P. J. Ain. and E. F. 1959.BRAIN METABOLISM DURING FASTING 1595 19. J. K. Himwich. Neurol. 3: 1815. Gabrielli. Klin. Mulder. Academic Press Inc. Cerebral blood flow. F. Lusk. W. and L. Sancetta. Substrate and enzyme distribu- 37. Hackel. J. Psychiat. The metabolism of D (-) -B-hydroxybutyrate in sheep. 10: 37. 1952. C. Abood. J. F. Cerebral metabolism of isotopic glucose in normal human subjects. Am. F. B. J. Owen. J. Wortis. 39. 33: 1. Kety. and H. 137: 436. R. The Science of Nutrition. Felicioli. Rossi. N. In press. Insulin hypoglycemia. 1960. Sebening.. B. Neurophysiol. Goldfarb. H. J... 31. M. Clarke. G. G. Bowman. Dickens. Waelsch. R. 1965. P. F. Held. Gerard. Rowe. New York. A study of prolonged fasting. G. J. Muro and J. H.. 1957. D. . Randle. U. E. 1946. 43: 965. P.. Bourne. 1935. E. 38. 1962. Annison.. 23. S. G. Jr. J. 12: 246.. B. Maxwell. New Engl. H. Walpurger. Benedict. B. W.. Dameshek. W. M. 1957. J. Biol. C. 35. Carnegie Institute of Washington. S. The oxidation of ketone bodies by mitochondria from liver and peripheral tissues. 1958. Med. and 0. J. 38: 2154 29. Physiol. Philadelphia. 552. Am. tion in cells and cell fractions of the nervous system. 1964. 28. Sacks. Der Glucoseverbrauch des menschlicken Gehirns unter dem Einfluss intraven6ser Infusionen von Glucose. publication 203.. and 0. 36. Castillo. Crumpton. Himwich. Krebs. Jr. 27. lactate. Abood. 1: 125. M. Section I. S. Some observations on carbohydrate metabolism in man. 1915. Academic Press Inc. McHenry. Freeman. 34. D. W. Clin. Waverly Press Inc. Arch. Lusk. B. 26. Human brain metabolism. 21. G. editors.. Crandall. Myerson. 20. 1948.. 97: 30. G. Berl. Jr. and pyruvate metabolism before and after eating. and C. 1903. and C. A. J. Chem. G. Wochshr. Biol. In Handbook of Physiology. Appl. 226: 15. 9: 380. 37: 285. D. P. Arch.