Acta Physiol Scand 1979, 107: 19-32

The effect of different diets and of insulin on the hormonal response to prolonged exercise H . GALBO, J. J. HOLST and N . J. CHRISTENSEN Department of Medical Physiology B, University of Copenhagen, Department of Clinical Chemistry, Bispebjerg Hospital, Medical Department F, Herlev Hospital, and 2nd Clinic of Internal Medicine, Aarhus Kommunehospital, Denmark

GALBO, H., HOLST, J. J. & CHRISTENSEN, N . J.: The effect of different diets and of insulin on the hormonal response to prolonged exercise. Acta Physiol Scand 1979, 107: 19-32. Received 12 Dec. 1978. ISSN 0001-6772. Department of Medical Physiology B, University of Copenhagen, Department of Clinical Chemistry, Bispebjerg Hospital and 2nd Clinic of Internal Medicine, Aarhus Kommunehospital, Denmark. The importance of carbohydrate availability during exercise for metabolism and plasma hormone levels was studied. Seven healthy men ran on a treadmill at 70% of individual maximal oxygen uptake having eaten a diet low (F) or high (CH) in carbohydrate through 4 days. At exhaustion the subjects were encouraged to continue to run while glucose infusion increased plasma glucose to preexercise levels. Forearm venous blood, biopsies from vastus muscle and expiratory gas were analyzed. Time to exhaustion was longer in CH(106f5 min (S.E.)) than in F-expts. (64+6). During exercise, overall carbohydrate combustion rate, muscular glycogen depletion and glucose ma lactate concentrations, carbohydrate metabolites in plasma, and estimated rate of hepatic glucose production were higher, fat metabolites lower, and the decrease in plasma glucose slower in CH- than in F-expts. Plasma norepinephrine increased and insulin decreased similarly in CH- and F-expts., whereas the increase in glucagon, epinephrine, growth hormone and cortisol was enhanced in F-expts. Glucose infusion eliminated hypoglycemic symptoms but did not substantially increase performance time. During the infusion epinephrine decreased markedly and glucagon even to preexercise levels. Infusion of insulin (to 436 % of preexercise concentration) in addition to glucose in F-expts. did not change the plasma levels of the other hormones more than infusion of glucose only but reduced fat metabolites in plasma. At exhaustion muscular glycogen depletion was slow, and the glucose gradient between plasma and sarcoplasma a s well as the muscular glucose 6-phosphate concentration had decreased. Conclusions: The preceding diet modifies the energy depots, the state of which (as regards size, receptors and enzymes) is of prime importance for metabolism during prolonged exercise. Plentiful carbohydrate stores favor both glucose oxidation and lactate production. During exercise norepinephrine increases and insulin decreases independent of plasma glucose changes whereas receptors sensitive to glucose privation but not to acute changes in insulin levels enhance the exercise-induced secretion of glucagon, epinephrine, growth hormone and cortisol. Abolition of cerebral hypoglycemia does not inevitably incrase performance time, because elimination of the hypoglycemia may not abolish muscular energy lack. Key words: Glucagon, growth hormone, cortisol, norepinephrine, epinephrine, glucose,

glycogen, lipid mobilization, metabolism, fatigue, hypoglycemia

Dming prolonged heavy exercise large increases in the uptake and combustion of glucose and free fatty acids occur in the working muscles. At the same time marked changes take place in the plasma concentrations of various hormones, changes which favor glucose production and lipid mobilization: The concentration of insulin decreases, whereas the concentrations of catecholamines, glucagon,

growth hormone and cortisol increase (Galbo et al. 1 9 7 7 ~ ) . We have previously investigated the mechanisms which give rise to these hormonal changes. Our studies led to the view that in exercising man plasma norepinephrine mainly reflects a work load dependent activity in t h e sympathetic nervous system and that a-adrenergic activity inhibits insulin secretion. Furthermore, the Acto Physiol Scund 107

hypothesis w a s put forward that in man during prolonged exercise, t h e secretion o f t h e o t h e r hermones is stimulated by cells in CNS and pancreas, which respond t o glucose privation (Galbo et

1977~).

In orde r t o examine the validity of this hypothesis we have now measured plasma hormone levels during prolonged exercise, which was performed after intake through 4 d ay s of a diet rich i n either fat O r carbohydrate. S u ch regimens h a v e been used sev-

eral times in Scandinavia in this century in studies o f exercise metabolism (Baje 1935, Christensen & Hansen 1939, Bergstrom e t al. 1967, H u l tm a n 1967, Hultman & Nilsson 1971). T h e y result in small a n d large glycogen stores, respectively, a n d during exe r c ise t h e plasma glucose concentration declines more rapidly after a fat diet th a n after a carbohydrate diet. Accordingly, if th e hormonal response t o exercise is en h an ced by receptors sensitive to glucose privation, also the hormonal c h a n g e s during exercise would be expected to t a k e place m o r e rapidly after a f at diet t h a n after a c a r b o h y d r a t e diet. I n or d er t o f u r t h er clarify t h e relationship between t h e hormonal response a n d changes in glucose concentrations, plasma glucose w a s a t exhaustion restored to pre-exercise levels by glucose infusion during continued exercise. A decrease in t h e tissue concentration o f insulin ma y aggravate intracellular glucose privation in glucose-sensitive cells regulating hormonal secretion (Muller, Faloona & Unger 1971, Raskin, Fujita & U nge r 1975, Szabo & S z a b o 1972). Accordingly, th e decrease of insulin concentration i n plasma, which ta kes place during prolonged exercise, may enha nc e hormonal secretion during exercise. If so, simultaneous infusion o f insulin a n d glucose m a y be more efficient in depressing t h e hormonal response t o exercise t h an infusion of glucose only. C o n sequently, in o r d e r t o elucidate t h e importance of insulin for t h e hormonal response t o exercise we infused insulin together with glucose at exhaustion in a gr oup of expts. with fat diet. Finally, t o substantiate t h e influence o f th e diets on t h e glycogen de pot s a n d t o study t h e effect of different diets o n th e metabolic state o f working muscles biopsies w e re t a ken from t h e vastus muscle.

METHODS Subjects and procedures. Seven healthy male students (26 (24-29) years (mean and range)) familiar with the laborato-

ry gave their informed consent to participate in the study. The students' mean maximal 0, uptake ( V O ? , , ddetermined during treadmill running was 4.33 (3.76-5.12) Mean weight was and 56 (5243) m l . kg-l. I. 77 (67-85.5) kg - and mean height 184 (18@188) cm. Each subject underwent three diet periods. These periods lasted 4 days each and were separated by 17 days during which no restrictions concerning physical activity and choice of food were ordered. At the start of each diet period the subjects arrived in the laboratory after an overnight fast. After 30 min rest in a chair they were weighed and had in the standing position blood drawn without stasis from a forearm vein. Then in order to produce glycogendepletion the subjects exercised 45 min at a work load corresponding to 80% of individual V,,,,,, 30 min on a treadmill and 15 min on a bicycle ergometer. Subsequently a liquid diet ( I 000 kcal . I-') consisting in the first (F,-exprs., in figures designated: - - -) and third (F,-expts., designated: ---) diet period of 76% fat and 13.5 5% protein, and in the second diet period (CH-expts., designated: -) of 77% carbohydrate and 13.5% protein, was delivered. The subjects were asked to consume an amount of diet corresponding to their estimated energy consumption in order to avoid a negative balance of energy in the diet periods. The consumed amount was determined by weighing. Besides the diet only intake of water and vitamin pills were allowed. In the diet periods physical training and smoking were not allowed. In the intervals between the diet periods a supplement of iron (Fe", 40 mgxday-')was given. Each diet period was finished by an overnight fast (10 h) after which the students arrived in the laboratory at 8 a.m. They were weighed and had a cannula inserted intravenously in the left forearm, whereupon they rested in a chair for 30 min. A sample of urine was examined for ketone bodies, glucose and albumin by stix (Ames). The electrocardiogram was registered with precordial electrodes. While the subjects were standing their oxygen uptake at rest was measured with the Douglas bag method and venous blood samples were drawn without stasis. The glucose concentration in capillary blood from a fingertip was determined at once with a reflectance meter (Ames) (Schersten et al. 1974). Exercise was performed on a motordriven treadmill with 3 % inclination. The speed of the treadmill was chosen so that the work load required 70% of individual maximal oxygen uptake. The subjects ran for repeated 30 min bouts separated by 10-min rest intervals. This sequence was continued until exhaustion. At exhaustion (Exhaustion 1) a 10 min rest was allowed while a cannula was inserted intravenously in the right forearm and connected to a motor-driven infusion pump. Then the subjects were encouraged to resume the running while a 25 % (9. (100 ml)-') glucose solution, adjusted to pH 7 with 1 M NaHCO,, was infused from a calibrated glass syringe. Every second min the glucose concentration in capillary blood from a fingertip was determined with the reflectance meter and the speed of the infusion pump adjusted to achieve the preexercise blood glucose level and to maintain this level until the subjects were unable to run longer (Exhaustion 2). In the third exercise test (F,-expts.) human insulin (Actrapid, Novo) was infused as well (1.15 m l . min-' of a solution of 121.8 nmol- 1-' of insulin in isotonic sodium chloride containing in addition 1 % (g .

Hormonal regulation during prolonged exercise (100 ml)-') human serum albumin) after a priming-dose (0.0575 ml x kg bw-'). Expired air was collected through 1-1.5 min after 20 min of each exercise period. In each exercise period blood samples were drawn from the catheter after 15 min for analysis of glucose, insulin and glucagon and during the last 2 min for analysis of glucagon, insulin, growth hormone (GH), cortisol, catecholamines, glucose, glycerol, FFA (free fatty acids), @-hydroxybutyrate, alanine, lactate, pyruvate, and hematocrit. Capillary blood from fingertips was sampled at the end of each exercise period for glucose determination. Finally, blood samples and expired air were collected 30 min after exhaustion, the subjects having recovered for 25 min in the supine and for 5 min in the standing position. Mineral water was offered ad libitum. In the first (F,) as well as in the third (F,) experimental series on an average 170 ml of blood were drawn and in the second (CH) 230 ml were drawn. The blood was immediately replaced by isotonic sodium chloride. In the first and second exercise test needle biopsies (Bergstrom 1975) weighing about 30 mg were obtained from the lateral head of the quadriceps femoris muscle immediately before exercise and immediately after the first and second run and at exhaustion. The biopsies were obtained through 0.5 cm incisions in the skin 12-16 cm above the knee. The incisions were made in one leg in the first exercise test and in the other leg in the second exercise test. Analyses. The volume of the collected expired air was measured with a 150 I gasometer (Collins). The expired air was analysed with an infrared C0,-analyser (Beckman LJ3-I,OA 184) and a paramagnetic 0, analyser (Servomex). The accuracy of the analyses was verified with the Scholander microtechnique (deviation less than 0.06 % CO, and 0,, respectively, at 12 comparisons performed with intervals of 14 days). Blood for hormone determinations was drawn in iced test tubes and centrifuged in the cold. Plasma or serum was stored at -20°C until analysis. For catecholamine analysis 10 ml blood were sampled in test tubes, which contained 20 mg EDTA and 20 rng ascorbic acid. Plasma norepinephrine and plasma epinephrine were determined by a double-isotope derivative assay (Engelman & Portnoy 1970) with certain modifications (Christensen 1973). For glucagon and insulin determinations 5 ml blood were drawn in heparinized test tubes containing in addition 0.25 ml Trasylol (Bayer), 10000 kallikreininhibiting units . ml-'. Glucagon was measured with radio-immunoassay after ethanol extraction of plasma (Heding 1971). The antiserum employed, 4317 (Holst & Aasted 1974), does not cross-react with enteroglucagon. Purified porcine glucagon (NOVO) was used as standard and bound and free hormone were separated by ethanol. Detection limit was below 5.7 pmol . I-' and intra- and inter-assay coefficient of variation were about 5 % and 15 %, respectively. Insulin was determined by a radioimmunoassay relying on charcoal separation (Albano et al. 1972). Human insulin (NOVO) was used as standard. The detection limit for the assay system was 0.3 pmol . 1-' and the intra-assay coefficient of variation (n=20) was 3 % at 46 pmol . I-' (7 pmol . I-' of insulin equal 1 pU . ml-I). Measurements of glucagon and insulin concentration in plasma with added exogenous hormone, in dilutions of plasma, and in mixtures of plasma with different concen-

21

trations yielded results, which deviated less than 10% from the expected values. Serum growth hormone (GH) was determined by a single antibody radioimmunoassay employing Wick chromatography (Qrskov, Thomsen & Yde 1968). A Wilhelmi preparation (HS 968 C) was used as standard. Serum cortisol was measured by a commercially available solid phase radioimmunoassay-kit (CEA-IRESORIN). The intra- and interassay coefficients of variation of the cortisol assay (n=8, triplicate determinations) were 13% (at 606 nmol . I-') and 15%, respectively. Serum from adrenalectomized rats yielded values which were not significantly different from zero and recovery experiments yielded results, which deviated less than 15 % from the expected values. Every subject had all analyses of a hormone carried out in a single assay run. The concentration of glucose in blood and plasma was determined by the glucose-oxidase method (Werner, Rey & Wielinger 1970). In respect to glucose concentrations arterialized capillary blood obtained from fingertips did not differ significantly from blood in the cannulated forearm vein. Free fatty acids (FFA) in serum were determined colorimetrically (Duncombe 1964) with palmitic acid as standard. Glycerol (Eggstein & Kreutz 1966) and alanine (Williamson 1974) in serum and lactate (Gutmann & Wahlefeld 1974) in blood were determined by enzymatic spectrophotometric methods and #3-hydroxybutyrate and pyruvate by enzymatic fluorimetric micromethods (Olsen 1971). Muscle biopsies obtained after exercise were frozen in liquid nitrogen within 10 sec from the end of an exercise bout and stored at -80°C until analysis. About 10 mg of frozen tissue was divided into five pieces and weighed at -20°C on a Perkin-Elmer microbalance to the nearest 0.01 mg. Metabolites were extracted during I5 min in 3 N HCIO, at 4°C. After neutralization with KHC03 (2 mol . I-') on dry ice the test tubes were centrifuged in the cold. Glucose 6-phosphate, glucose I-phosphate, glucose, lactate and pyruvate were determined on the supernatant by enzymatic fluorimetric micromethods (Lowry & Passonneau 1973). 1 ml HCI (1 mol . 1-') was added to the residue and the glycogen was hydrolyzed during 2 h at 100°C. Determination of glucose in the hydrolysates was carried out fluorimetrically (Lowry and Passonneau 1973) and the glycogen concentration is presented as mmol glucosyl units ' kg-' (wet weight). For determination of the protein concentration about 7 mg of muscle tissue were heated for 30 min at 100°C in 1 ml 1 N NaOH. The protein concentration was determined spectrophotometrically with the Folin phenol reagent (Lowry et al 1951) with bovine serum albumin as the standard. To estimate the precision of the biopsy technique duplicate analyses were performed on five pieces cut from a 31 mg biopsy of human muscle and on 10 pieces cut from a rat leg muscle weighing 80 mg. Mean values (mmol . kg wet weight) and coefficients of variation for the different analyses on human and rat muscle were: Glucose 6-phosphate: 0.6, 11.5%; 3.29, 29%; glucose 1-phosphate: undetectable amounts; glucose: 0.7, 32%; 0.7, 17%; lactate: 2.9, 14.5%; 13.3, 12.8%; pyruvate: 0.363, 33%; 0.368, 5 8 % ; glycogen: 122, 7.5%; 36.3, 7.3 %. The coefficients of variation of the chemical part of the procedures were below 3 % except for the coefficient of variation of the pyruvate analysis which was 29%. In Actri Physiol Scund 107

22

H . Galbo et al.

D u r ~ l m no f

~ M C I S mm ~

Fig. Z . The effect of different diets on the carbohydrate combustion and muscular glycogen concentration during exercise. Mean values (+ S.E.) of glycogen in the vastus lateralis muscle (expressed as mmol glucosyl units per kg wet weight) and amount of carbohydrate combusted in seven subjects running on a treadmill. In the postabsorptive state each subject exercised three times having eaten through 4 days a diet rich in either carbohydrate or fat. Glycogen concentrations were only measured in one of the two experimental series including a fat enriched diet. =+ denotes value significantly different (Pc0.05) from preexercise (rest) value. 71 denotes significant difference between experiments including different diets.

the study all chemical analyses were performed at least in duplicate. Enzymes and cofactors were from Boehringer, Mannheim. Statistical evaluation of the data was made by means of correlation analysis and by means of Wilcoxon's nonparametric ranking test and the r-test for paired comparisons (Snedecor & Cochran 1%5). Differences were considered to be significant, i f P values of less than 0.05 were obtained with both these tests. Concentrations obtained at exhaustion in F,-and CH-expts. (n=14) have been used for correlation analysis when something else is not explicitly stated. The cited correlation coefficients are significant on the 5 % confidence level.

RESULTS Effects of the different diets Except for hematocrit [46.2k0.9%, mean and S.E. (F,-expts.) significantly higher than 44.1 k0.8 (CHexpts.) and 43.5k0.9 (F,-expts.)] and FFA [significantly higher in F,- than in F,-expts. (Fig. 4)] values obtained immediately before each of the three diet periods were similar. The energy intake was similar (P>0.05) in periods with fat enriched diet f2 492+ 159 kcal . 24 h-', mean and S.E. (F,-expts.) and 27903% (F,-expts.)] and in periods with carbohydrate enriched diet (2 904f 195). However, during the diet periods a significant loss of body

.

INSULIN pmd I-'

Fig. 2. The effects of different diets, of exercise and of

infusion of glucose and insulin during exercise on glucose, pancreatic glucagon and insulin in plasma. In the postabsorptive state seven subjects ran three times three weeks apart on a treadmill having eaten in advance a diet rich in either carbohydrate or fat (two experimental series). At exhaustion (exhaustion 1) a 10 min rest was allowed. Then the subjects were encouraged to run again as long as possible (to exhaustion 2) while glucose or both glucose and insulin (in one of the experimental series with fat diet, designated: ---) concentrations were restored by infusion. Values are mean f S.E. * denotes signifcant difference (P

The effect of different diets and of insulin on the hormonal response to prolonged exercise.

Acta Physiol Scand 1979, 107: 19-32 The effect of different diets and of insulin on the hormonal response to prolonged exercise H . GALBO, J. J. HOLS...
1MB Sizes 0 Downloads 0 Views