Hormonal and metabolic changes during exercise in cirrhotic patients

Hormonal Bernard Campillo,

and Metabolic

Chantal Chapelain,

Changes

Jean Claude Bonnet,

During Exercise

in Cirrhotic

Eric Frisdal, Michel Devanlay,

Evelyne Wirquin,

Patients

Philippe Bouissou, Paul Fouet,

and Guy Atlan

The metabolic response to exercise was compared in 10 cirrhotic patients (PI in a stable clinical condition and in 6 sedentary, age-matched, normal subjects (Cl performing 32 minutes of treadmill exercise with the same constant workload corresponding to three to four times their resting oxygen uptake. Taking indirect calorimetry as reference, respiratory exchanges indicated that cirrhotic patients consumed carbohydrates almost exclusively, unlike the normal controls, who consumed lipids and glucids in about the same proportions (RQ : 0.98 * 0.04 ~0.87 + 0.04, P < 6001). In the patients, this carbohydrate path of exercise metabolism lowered glycemia from the resting value of 5.23 + 0.16 mmol/L to 4.03 * 0.37 mmol/L (P < .OOOl) and raised the plasma lactate concentration from 2.08 * 0.24 mmol/L at rest to 3.48 r 0.32 mmol/L at the eighth minute of exercise (P < DOI ), thus suggesting defective liver glyconeogenesis. Fatty free acids and glycerol remained almost constant during exercise, whereas catecholamines increased. Insulin levels were high in patients at rest (67.1 f 14.5 U/mL v 15.1 * 3.5 U/mL); they declined sharply at the onset of exercise but nevertheless remained high compared to those observed in the controls (P < .OOOl1. Glucagon increased in exercising patients from 88.3 ? 21.3 pg/mL to 127.4 ? 30.6 pg/mL (NS). Esterified plasma carnitine declined in the patients from 13.0 + 2.2 wmol/L to 8.6 + 1.5 /.rmol/L (P < .05). Several possibilities might account for these results: (I 1insulin and exercise might act synergistically to increase glucose use; (2) insulin might not allow other hormones to express their effects. hence the lipolysis inhibition and reduced liver glycogen release; (3) the decrease in esterified carnitine might be due either to hyperinsulinemia and small lipid use or to some defect of synthesis connected with liver cell damage, a defect that becomes evident when carnitine requirements increase. 0 1990 by W.B. Saunders Company.

U

NDER MOST CONDITIONS, especially during exercise, the increased requirements of working muscles are met by precisely matched releases of glucose from the liver and of lipids from the adipose tissue. The exact mechanism of this coordination is not completely known, but many regulatory factors have been investigated. During exercise, the principal changes that may affect glucoregulation and lipid mobilization include a decrease in insulin secretion and an increase in catecholamine and glucagon secretion, to mention only the main hormonal systems involved. A few studies of patients with metabolic diseases have been made during exercise. They essentially concern diabetes mellitus’.’ and obesity,3-5 but no investigation has yet been made of cirrhotic patients during exercise, which is surprising because metabolic abnormalities are common in liver cirrhosis. They include (1) insulin resistance and glucose intolerance6-*; (2) fat mobilization and lipolysis, both of which are greater in fasting than in normal subjects’,“; (3) decrease in protein synthesis with distortion of amino acid metabolism”; and (4) abnormal carbohydrate storage and release.” Under any of these conditions it is not yet known how these patients respond to the requirements of exercise, what kind of fuels they consume, and how the hormonal regulatory factors interfere with each other. Therefore in this study a group of cirrhotic patients performed treadmill exercise for 32 minutes, and the results From the Unite de Recherches de Physiologie Respiratoire. Institut National de la Sante et de la Recherche Medicale, Centre Hospitalo-Vniversitaire Henri Condor, and the Service d’HepatoGastroenterologie. Hbpital Albert Chenevier. Creteil, France. Address reprint requests to Guy Atlan, MD, Inserm V 296, Fact&h de Medecine, Centre Hospitalo-Vniversitaire Henri Mondor, 8 Av du General Sarrail, 94010 Creteil, France. o 1990 by W.B. Saunders Company. 0026-0495/90/3901-0003$3.00/O 10

were compared

sedentary

to those for a group of normal, age-matched subjects after the same exercise. MATERIALS AND METHODS

Patients

Ten male patients with histologically proven liver cirrhosis and six male control volunteers took part in this study. The controls were sedentary, with no regular practice of sport, and had a well-balanced diet. They were recruited among the staff of our healthy

laboratory. All patients and volunteers were informed of the nature and purpose of the study and the possible risks involved before they consented to participate. The aim of the study was approved by the Ethical Research Committee of the Centre Hospitalo-Universitaire Henri Mondor. In all patients the etiology of cirrhosis was alcoholism. The patients were in a stable clinical condition and were mobile when the study was performed. Alcohol intake had ceased for all of them at the time of their hospitalization, which lasted for at least one month. None of these patients had diabetes mellitus, hyperlipidemia, thyroid dysfunction, neoplasm or other chronic diseases; neither did they have evolutive cardiac or pulmonary illness. Furthermore, none of them was taking any drug known to impair glucidic or lipidic metabolism. The severity of disease was scored according to Pugh et al,* from 5 points for correct liver function to 15 for poor function. The mean score for the 10 patients was 6.8 _t 0.4 (range. 5 to 9). Nutritional status (Table I) was assessed by anthropometric measurements. Muscle mass was measured according to Heymsfield et alI0 from the arm muscle area and was calculated from the triceps skinfold thickness and midarm circumference. The body mass index (BMA; weight [kg]/H2 [cm]) was used to classify obesity.’ Only one patient had obesity (grade I). Mean hemoglobin level was slightly diminished in all patients (I I .9 + 0.3 g/dL). As regards the BMI of control subjects (Table I), two had grade I obesity, and one grade II obesity. The ages of patients and controls were similar (Table 1); the weight of control subjects was greater (P < .05); but body mass index was not significantly different. Muscle mass, in percentage of total body weight, was larger in the controls (P < .02). In patients, food intake was controlled by dietetic monitoring. Metabolism, Vol 39, No 1 (January), 1990: pp 18-24

19

CHANGES DURING EXERCISE IN CIRRHOTIC PATIENTS

Table 1. Means (+SEM)

of Physical, Pulmonary and Maximal Exercise Test Data in Patients and Normel Subjects

Weight

Age

BMI

Muscle Mass

vc

FEV,

(% body weight)

% Pred

% Pred

Patients

49.1

f.

7.8

63.6

* 7.45

22.9

+ 0.92

33.9

+ 1.33

Controls

45.3

+ 10.4

80.2

k 20.32

25.9

k 1.51

39.0

+ 0.87

83.7

+ 3.7

81.1

+ 5.3

vo,

max

ImL/kg)

HR “ax

19.6 + 0.6

155t

32.7

164 + 4

f 2.0

5

Signiticance threshold

PC

.05

NS

P < .02

Abbreviations: VC, vital capacity; FEV, , forced expiratory volume first second (1

P < ,001

* sm‘); % Prad. percent from

NS

the predicted values; HR max. heart rate

recorded at the maximal work level.

Daily intake was 2,500 t 260 kcal, with caloric proportions of proteins, lipids, and glucids of 14.8% + 0.2,3 I .2% k 1.9, and 54% f 2, respectively. Three hours prior to exercise, each patient had his or her usual breakfast, which supplied an average 590 kcal, with proportions of proteins, lipids, and glucids of 9%, 22%, and 69% respectively. Control subjects also had their usual breakfast three hours before the test, supplying an average 450 kcal with proportions of proteins, lipids, and glucids of IO%, 25%, and 65% respectively. One patient took 20 g of lactulose with his breakfast. Routine laboratory pulmonary function tests of all patients and controls were performed before exercise, and the mean results are given in Table 1; 7 of the 10 patients had had no severe pulmonary illness; 2 of them, who were exsmokers, had a history of mild bronchitis with, at the time of the study, normal arterial blood gases; I had had pleuropulmonary tuberculosis.

Exercise Testing All subjects came to the laboratory on two successive mornings. Before exercise, they all had at least half an hour’s rest in the sitting position. The exercise was performed on a treadmill. On the first day, maximal oxygen uptake (VO, max) was determined by serial 3-minute workbouts at increasing speeds and grades until subjective exhaustion. The speed and grade were the same for all patients. The work levels for control subjects were generally higher than those for patients. On the second day, the exercise period for both patients and controls lasted 32 minutes at a constant load. The work level for this period of long-duration exercise was not selected as a percentage of VO, max because the values were different for patients and controls. To ensure that both groups performed the same work, we fixed a work level at a constant speed and grade, resulting in 3 to 4 times the rest oxygen uptake determined during the 30-minute rest period preceding exercise. During this period respiratory data were collected for 12 minutes. The recovery time was intentionally limited to data collection lasting from 4 to 8 minutes to shorten the time spent by the patients in the laboratory. Electrocardiogram (ECG) was continuously monitored throughout all tests. During these tests the subjects breathed through a Hans Rudolph valve (dead space of 120 mL) equipped with a mouthpiece and noseclip. Ventilatory and respiratory gas exchange data were obtained by an open circuit in which a pneumotachograph (Fleisch no 3 plus Validyne PM 45 transducer, Paris, France) was positioned on the inspiratory tubing. The expiratory tubing was attached to a mixing box (8 L), and the mixed expired air was analyzed by a mass spectrometer (Centronics MGA 200, Paris, France). All the analog signals were converted at 20 CPS, and conventional calculations” and one-line printing were obtained with a PDP 1 l/23 Computer (Digital Equipment Company, Paris, France). Values were averaged over one minute.

Blood Samples On the morning of the long-duration exercise, an 18-gauge catheter with a three-way stopcock was introduced into an antecu-

bital vein at least half an hour before the exercise, when subjects were resting in the sitting position. The first sampling was done at the end of the rest period; subsequent samplings were done after every four minutes of exercise. Blood was sampled with fluoride-EDTA for measurement of plasma metabolic substrates (glucose, lactate, free fatty acids [FFA], glycerol, and fl-hydroxybutyrate [@-OHB]), with lithium heparin for measurement of insulin, catecholamines and carnitine plus aprotinin for glucagon. All samples were stored on ice; plasma was separated within one hour of collection. Plasma samples were analyzed within four hours of blcod collection for FFA and glycerol and within two days for other substrates. Plasma samples with Li-heparin were frozen at -40°C until assay.

Analytic Method Plasma glucose was assayed by a conventional enzymatic method. FFA were assayed by an enzymatic method with a Nefac.C kit purchased from Wako Chemical, GMBH (West Germany). Glycerol was assayed by a modified enzymatic method using a kit (triglyceride kit) purchased from Boehringer, Mannheim (West Germany). P-OHB was assayed by an enzymatic method.12 Lactate was not assayed in acid-deproteinized total blood but in plasma. The values obtained by this procedure are independent of erythrocyte lactate. We checked that the production of lactate by erythrocytes was completely inhibited by the fluoride-EDTA reagent recommended by the manufacturer (Boehringer-Mannheim, West Germany). Insulin and glucagon were determined by conventional radioimmunoassay (RIA) with kits purchased from Behring Marbur (West Germany) for insulin (Ria Gnost Insulin) and from Biodata Milan (Italy) for glucagon (glucagon kit). Epinephrine and norepinephrine were determined by a radioenzymatic technique according to Cuche et al.” Total, free, and esterified carnitine were measured by a radioisotopic assay according to McGarry and Forster.14

Statistical Analysis Data are expressed as means + SEM. Student’s paired and unpaired I tests were used. A repeated measures analysis of variance was used to determine the significance of changes during exercise for patients and controls. The levels of significance were set at P < .05. RESULTS

Respiratory Gas Exchange The maximal oxygen uptake (VO, max) was clearly higher in the controls than in the patients (P -=c.OOl for VO, max, Table 1) . Table 2 shows all the respiratory data recorded during rest, exercise, and recovery. Mean VO,, expressed as mL was not significantly different in the two kg-’ - mm’, groups. Ventilation was greater in cirrhotic patients than observed in controls at rest (P -C .Ol) but not during exercise; the ventilatory rate was not different in the two groups. Mean

20

CAMPILLO ET AL

Table 2. Values for Long-duration

VQ* (mL .

kg

’.

min ‘1

Exercise,

Expressed

as Means for All the Data for the Period of 32-minute

Rest

Exercise (Mean Values1

P

4.50

‘- 0.22

14.2 + 0.60

4.98

+ 0.25

3.60

+ 0.29

14.9 + 0.69

4.30

k 0.33

NS

P

0.99

+ 0.028

0.98

+ 0.016

1.11 f 0.028

C

0.84

f 0.016

0.87

+ 0.016

0.90

(1

. mm I)

P

13.4 * 0.70

34.2

f 2.20

15.4 & 0.60

C

10.2 * 0.73

31.2

+ 1.67

12.7 ? 0.45

P< VR (cvcle

. mm ‘)

NS

.Ol

P

16.2 * 0.76

25.3

k 1.20

20.6

+ 1.50

C

14.0 + 1.27

25.1

+ 1.35

27.9

f 2.99

NS HR (beats

.

min ‘)

+ 0.024

P < .OOOl

P < .OOOl V

Recovery Four Mm&s

C

P < .05

RQ

Exercise

NS

P

85.6

+ 3.8

119.3

f 4.8

106.5

& 2.6

C

69.8

+ 3.9

98.9

f 5.1

76.6

i 5.9

P < .OOl

PC .Ol NOTE. All values are means + SEM, and their significance thresholds are indicated.

Abbreviations: P. cirrhotic patlents: C, controls. V, ventilation; VR, ventilatory rate; HR. heart rate

heart rate was higher in cirrhotic during exercise.

patients

both at rest and

Glycemia and Metabolite Substrates Figure I shows that at rest glycemia was similar in both groups (P = 5.38 -+ 0.25 mmol/L; C = 5.23 f 0.16 mmol/ L; (NS). In the controls it either did not change during exercise or decreased slightly. Conversely, in the patients, glycemia decreased regularly until the 24th minute of exercise and thereafter tended to form a plateau for every patient. At the 12th minute of exercise the mean glycemia level was significantly lower (P < 0.01) than that observed at rest, and at the 24th minute it was lower still (P < .OOl for rest v 24th within P). Analysis of variance with repeated measures showed a marked difference between patients and controls (P < .OOOl). As regards plasma lactate, no difference was observed in the controls between the concentrations at rest and during exercise (Fig 1). In the patients the resting value was high and different from that observed in the controls; thus during exercise plasma lactate rose significantly (P < .OOl) as from the fourth minute, peaked at the eighth minute (P < .OOOl), and regularly decreased thereafter. Analysis of variance showed a significant difference between patients and controls (P < .OOOl). In the controls, FFA concentrations during exercise decreased from the resting value to form a plateau (Fig 2), which began at the 16th minute, but no statistically significant difference was noted between the resting and 16th minute (paired t test). Four minutes after exercise had stopped, the average value increased to near the resting value. In the patients the evolution of FFA was similar but less marked: the decreases observed were slight and not significantly different from those recorded in the controls. Glycerol changed little in the patients but rose nonsignificantly in the controls. Analysis of variance for FFA and glycerol showed no statistically significant differences. Figure 2 shows that no changes occurred in P-OHB levels in

controls or patients during exercise and that resting values were almost the same in both groups. Hormonal Responses In the controls no significant differences were observed between resting and exercise insulin values, but a nonsignifiT

E

ES,3 .I H 0 4,3 -

. +- controk + Patients -,-

A

I

-4

.

I

’

.

1

”

16

.

I

.

I

26Time(miiF

26

RR Time (min)

Fig 1. (A) Changes in glycemia during rest, exercise, and recovery in patients and controls. A significant difference appeared at minute 12 in the patients and increased thereafter. (6) Plasma lactate concentrations in patients and controls. The patients’ values during exercise were significantly different from those observed at rest, with the exception of the two lest values. The highest significant threshold was observed at the eighth minute of exercise in the patients (P < .DDDl). Means and SEM are shown.

21

CHANGES DURING EXERCISE IN CIRRHOTIC PATIENTS

both groups. Esterified carnitine increased slightly in the controls during exercise; conversely in the patients it decreased significantly from 13 t 2.2 pmol/L to 8.6 2 1.5 Nmol/L (P < .05; paired t test). DISCUSSION

The main finding of this study is that during exercise of moderate duration, cirrhotic patients use carbohydrates almost exclusively as fuel for performing muscular work. This finding results from indirect calorimetry, which is a simple reliable method for evaluating the fuels consumed by the body, as proved by many investigators.6,‘5 Indeed, the RQ value was always near 1 throughout exercise, unlike what was observed in normal subjects. The choice of the exercise level was based on the following considerations: if we had fixed this level in relation to the maximal aerobic capacity, the exercise performed in both groups would have been very different, since the VO, max in the controls was greater than in the patients. To make meaningful comparisons between the two groups and to obtain a better estimation of the metabolic events resulting from mild exercise such as a short, fast walk, we preferred to select the same exercise level for both groups leading to the same metabolic expenditure. Consequently, resting 0, uptake was multiplied threefold or fourfold, which seems to

1

-4

6

.

I

16

.

1

26

.

t

36

Time (min) Fig 2. FFA (A). glycerol (6). and &hydroxybutyrate (Cl concentrations in patients and controls. No significant difference were observed between the two groups. SEW are omitted for clarity.

$ox 5

65 -

I =

50-

35-

cant increase appeared at recovery (Fig 3). Conversely, the patients displayed marked changes in insulin levels: resting values were much higher than those observed in normal subjects (P < .OOl) and during exercise dropped sharply. All patients exhibited high insulin levels at rest; these levels decreased with exercise but nevertheless remained higher than control levels. No increase was observed at recovery in patients, contrary to the controls. No significant changes in glucagon were observed in the controls (Fig 3). In patients, however, a regular but nonsignificant increase occurred during exercise. Analysis of variance for insulin and glucagon showed a large difference between patients and controls (P < .OOOl and P < .0025, respectively) The norepinephrine values for patients at rest were higher than the control values (P < .Ol) and increased a little more with exercise in patients (Fig 4). Epinephrine values at rest were almost the same in both groups but increased a little more with exercise in patients. Analysis of variance did not show any differences for norepinephrine or epinephrine. Figure 5 shows the changes in total, free, and esterified carnitine. Total carnitine decreased in the patients with exercise, unlike what occurred in the controls, in which it rose slightly. According to analysis of variance, these differences are significant (P < .05). Free carnitine remained constant in

Exercise-~

a-

+

- +

Colltrob Patients

20 5

I

6

-4

-

I

16

.

I

.

1

28

A

Time (mi”n,”

60

%.,, -4

B

.

. ;

I’S

I

2;

Time (miz8

Fig 3. (Al Changes in insulin values during rest, exercise, and recovery in patients and controls. The differences between the two groups were statistically 8ignificant (P < .OOOl). In patients group a significant difference appeared between resting and exercise values (P < .Ol st eight minutes). (El Changes in glucagon values in patients and controls. Values observed in the two groups were significantly different (PC 0025). Means and SEM are shown.

22

CAMPILLO ET AL

o-_-

* +

Exercise -.

conlmh Patients

(ml Id+-+-+

3001 8

Fig 4.

-4

.

,

6

I

l6

.

I

.

I

26Tlme(m;

Changes in the levels

of epinephrine (Al and norepinephrine (6) during rest and exercise. Resting nonepinephrine values for patients and controls were statistically different (P < .05). Means and SEM are shown.

correspond to one of the usual situations encountered by patients outside the hospital. At rest, cirrhosis of the liver is frequently associated with disorders of the glycoregulatory system, and excessively high levels of blood glucose, insulin, and glucagon have often been reported. The causes of these defects are multifactorial, and furthermore the levels of the parameters concerned are different under fasting and fed conditions. The origin of hyperinsulinism and hyperglucagonemia is controversial, and many factors have been evoked, including enhancement of secretion by pancreatic p or cy cells,‘6 portal systemic shunting,” and severe liver-cell damage.” Whatever the origin of hyperinsulinemia, it is generally recognized to lead, by a down-regulatory effect,” to a state of resistance to endogenous insulin attributed to either binding or postbinding defects in insulin target-organ cells. The present results for carbohydrate metabolism before and during exercise call for two series of comments. First, before exercise our patients were given a breakfast comprising about 600 kcal as energy supply, 70% of it carbohydrates. This method was chosen to simulate as closely as possible the situation of these patients when they are in a stable clinical condition and live a nearly normal life outside the hospital. When the same energy supply was given to these subjects under fasting conditions, the effect corresponded to that observed for a glucose tolerance test. For instance, both Proietto et al3 and Taylor et ak4 who gave equivalent amounts of glucose, found insulin levels close to those observed in our patients. None of these data reflecting glucose intolerance were observed in any normal groups by

ourselves or others. Furthermore, Stewart et al’(‘showed that in cirrhotic patients, unlike normal subjects, the insulin rises induced by a normal meal persisted for several hours. As regards the exercise period, our patients began with high insulin levels that, although they sharply decreased during exercise, never dropped very low. Insulin levels are known to decline with acute exercise of mild or severe intensity in both normal, lean individuals” and in obese, diabetic patients,‘5.” and the same applied to the cirrhotic patients in this study. It may be assumed that in cirrhotic subjects, as in certain hyperinsulinic patients, exercise reduces insulin resistance. However, sustained hyperinsulinemia in cirrhotic patients during exercise probably interferes with the carbohydrate metabolism by means of two mechanisms, whose effects combine to lower glycemia: (I) enhancement of glucose uptake by exercising muscles above the normal uptake during exercise; and (2) inhibition of glucose release from the liver. Both in normal subjects” and noninsulin-dependent diabetic obese patients,‘4 many authors have shown by various technics that liver glucose output was inhibited during rest and only rose slightly during exercise when high levels of insulin or glycemia were present prior to exercise. In control subjects, glucagon and catecholamines help to

56

7

T

“6.

+

Exercise w

6

26

36

Time(min)

I-r

-4

I

I

6

16

I

I

26

36

Time(min)

-4

6

16

28

Time(men9

Fig 5. Changes in total (AI. free (Bl, and esterified (Ester) carnitine ICI in patients and controls. Analysis of variance showed a significant difference between the two groups for total carnitine W’ < .05). A significant difference was observed for Ester carnitine between resting and end-of-exercise values in the patients W < .05). Means and SEM are shown.

CHANGES DURING EXERCISE IN CIRRHOTIC PATIENTS

maintain liver glucose output, ‘5z6 since their secretion levels are regulated by glycemia and also, in the case of catecholamines. by the general level of sympathetic nervous activity. This was verified in the present patients, since the fall in glycemia was accompanied by a rise in glucagon; parallel increases in catecholamines were observed in the patients and controls as a result of the increased sympathetic activity induced by exercise; but for norepinephrine, the initial resting level was higher in the patients. Many studies of cirrhotic patients’7.‘8 have shown that adrenergic activity is often enhanced at rest, as observed here. However, despite the increases in glucagon and catecholamines, glycemia fell when our cirrhotic patients performed exercise. One may therefore conclude that the action of insulin prevents the expression of all the other hormones, the more so as insulin levels in our patients remained high throughout exercise. The changes in plasma lactate concentrations observed in these patients were large and very different from those observed in normal subjects, in whom they were consistent with the changes reported by other investigators for similar rates of exercise.‘3 In the present cirrhotic patients, plasma lactate concentrations were high at rest, which again corresponds to the findings of others”; they increased with exercise and remained high throughout almost all the exercise period. This might have been the result of a double effect, namely, greater use of carbohydrates and decreased liver uptake of lactate. In normal” and in diabetic subject’s3’ it was shown that insulin infusion raised the lactate concentration: the authors concerned interpreted this result as indicating that liver lactate uptake was reduced by the inhibition of gluconeogenesis. Lipids in the form of FFA make a major contribution to energy needs during exercise, since their mobilization from adipose tissue is ensured by the adrenergic nervous system, whose stimulation increases lipolysis. The FFA values we found during exercise in the normal group followed a course close to that described by many investigators.“.30 The same applies to the evolution of glycerol and /3-OHB during exercise. The most striking observation made in the present cirrhotic patients is that they used almost no lipids for energy expenditure during exercise. FFA concentrations displayed an evolution close to that observed in the normal group, but it was less marked; no statistically significant difference was noted between rest and exercise values, although the number of patients was larger than the number of controls. The same applied to the evolution of glycerol and fi-OHB values, indicating that no ketogenesis occurred during exercise. Inhibition of lipolysis is another effect of insulin that at high concentrations expresses itself by suppressing the adrenergic influence on adipose tissue. Thus Martin et alz9 observed. that in normal subjects insulin infusion reduced the FFA concentration during rest but that the expected rise in these concentrations during exercise did not occur; at the same time, RQ was significantly higher during exercise with insulin infusion than without it. Inhibition of lipolysis leads to greater use of carbohydrates, which seems to correspond to the observations made here with exercising cirrhotic subjects. ‘Thus lipolysis blockade, which exists in normal subjects with high insulin levels, would seem to be considerable in cirrhotic patients.

23

As regards lipid metabolism, carnitine metabolism during exercise was also assessed. Only plasma concentrations were studied, since risks of hemorrhage precluded muscle biopsies in these patients. Carnitine is known to play an important part in lipid catabolism and energy production as a cofactor of a “shuttle” mechanism whereby long-chain fatty acids are converted into acylcarnitine derivatives and transported across the inner mitochondrial membrane for energy production via P-oxidation. Carnitine metabolism during exercise has been studied in normal subjects by two groups,3’,3’ who both found that esterified plasma carnitine increased. A divergency exists, however, between these authors: according to Lennon et a13’ a net loss in muscle carnitine was observed that might have been responsible for the rises in esterified plasma carnitine, while according to Carlin et a13’ these increases did not result from a loss of muscle carnitine that in any case was observed but might have been due to an exchange of plasma carnitine with the hepatic carnitine pool. Moreover, several authors have shown similar rises in plasmaesteritied carnitine during conditions including fasting33,34 and diabetes,‘” which were not associated with an increase in muscle activity. The results reported for cirrhosis of the liver are diverse: Rudman et al36 observed very low levels of carnitine at rest in patients with serious hepatocellular damage, while Fuller and Hoppe13’ reported elevated concentrations. The levels of total, free, and esterified carnitine observed in the present study were the same in both groups at rest. During exercise, no change occurred in the normal group for total, free, or esterified carnitine. It should, however, be noted that the duration of exercise in this study was short compared to that of other studies; in the investigation by Carlin et al,32 for instance, the rises in esterified carnitine only appeared after the 30th minute. Conversely, in the present cirrhotic patients, a significant decrease compared to resting values was observed for esterified carnitine. The significance of this decrease is not evident, but two explanations may be suggested. (I) It may essentially reflect the small participation of lipids in supplying the energy requirements of exercise, and the high levels of insulin might help to supply this energy by acting indirectly on enzymatic carnitine activity. (2) Alternatively, the hepatocellular damage is such that it is not possible to maintain a stable concentration of esterified carnitine. As the present results show, the cirrhotic patients who performed exercise in this study essentially used carbohydrates as fuel. This situation, assumed to be linked to hyperinsulinism, probably leads to fast depletion of carbohydrate stores, the more so as the latter are reduced by the liver disease. One may therefore wonder what relationship exists between exercise and the fast development of the catabolic state of starvation described by Owen et al.6 Furthermore, this probable, fast depletion of carbohydrate stores during exercise might explain the weak maximal aerobic capacity observed in our patients compared to normal subjects. In a larger group of cirrhotic patients, we observed a strong link between the maximum oxygen uptake measured and the liver impairment identified through the Pugh score (unpublished observations, 1989). This raises the important question of whether a muscular impairment is

CAMPILLO ET AL

24

present in cirrhotic patients and the part played in this connection by the hormonal and metabolic disturbances described here. To sum up, the cirrhotic patients in this study, who were observed at a time when their clinical condition was good, exercised at a constant work load for 32 minutes and had high levels of insulin. Under these conditions their almost exclusive use of carbohydrates as fuel for energy expenditure induced a fall in glycemia. The increases observed in cate-

cholamines and glucagon did not express their physiologic effects, as if hyperinsulinemia had suppressed them. ACKNOWLEDGMENT wish to thank J.L. Cuche for his assistance in the The excellent secretarial assistance of M. Roulet is gratefully acknowledged. We would also like to thank M. Dreyfus for revising the English of this manuscript and J. Meunier for his help in statistical treatment. The authors

catecholamine assay.

REFERENCES I Richter EA. Ruderman NB: Diabetes and exercise. Am J Med 70:201-209, 1981 2. Bray GA, Whipp BJ, Koyal SN, et al: Some respiratory and metabolic effects of exercise in moderately obese men. Metabolism 26:44 l-446, 1977 3. Proietto J, Alford FP, Dupley FJ: The mechanism of the carbohydrate intolerance of cirrhosis. J Clin Endocrinol Metab 55524-530, 1980 4. Taylor R, Heine RJ, Collins J, et al: Insulin action in cirrhosis. Hepatology 5:65-7 I, 1985 5. Riggio 0, Merli M, Cantafora A, et al: Total and individual free fatty acid concentrations in liver cirrhosis. Metabolism 33:646651. 1984 6. Owen OE, Trapp TE, Reichard Jr GA, et al: Nature and quantity of fuels consumed in patients with alcoholic cirrhosis. J Clin Invest 72:1821-1832, 1983 7. Owen OE, Reichle FA, Mozzoli MA, et al: Hepatic, gut, and renal substrate flux rates in patients with hepatic cirrhosis. J Clin Invest 68:240-252, 198 I 8. Pugh RNH, Murray-Lyon JM, Dawson JL, et al: Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 60:646-649, 1913 9. Jequier E: Energy, obesity and body weight standards. Am J Clin Nutr 45:1035-1036, 1987 10. Heymstield SB, McManus C, Smith J, et al: Anthropometric measurement of muscle mass: Revised equations for calculating bone-free arm muscle area. Am J Clin Nutr 36:680-690, 1982 11. Wasserman K. Hansen JE, Sue DY. et al: Principles of Exercise Testing and Interpretation. Appendix D. Philadelphia, PA, Lea & Febiger, 1987. pp 253-262 12. Li PK, Lee JT, McGillivray MH, et al: Direct, fixed-time kinetic assays for o-hydroxybutyrate and aceto Acetate with a centrifugal analyzer or a computer-backed spectrophotometer. Clin Chem26:1713-1717, 1980 13. Cuche JL, Prinseau J, Selz F, et al: Oral load of tyrosine or L Dopa and plasma levels of free and sulfoconjugated catecholamines in healthy men. Hypertension 7:81-89, 1985 14. McGarry JD, Forster DW: An improved and simplified radioisotopic assay for the determination of free and esteritied carnitine. J Lipid Res 17:277-28 I, 1976 15. Costill DL, Coyle E, Dalsky G, et al: Effects of elevated plasma FFA and insulin on muscle glycogen usage during exercise. J Appl Physiol Respir Environ Exert Physiol43:695-699, 1977 16. Sherwin RS, Fischer M, Bessoff J: Hyperglucagonemia in cirrhosis: Altered secretion and sensitivity to glucagon. Gastroenterology 74: 1224- 1228, 1978 17. Shurberg JL. Resnick RH, Koff RS, et al: Serum lipids, insulin and glucagon after portacaval shunt in cirrhosis. Gastroenterology 72:301-304, 1977 18. lwasaki Y, Ohkubo A, Kajinuma H, et al: Degradation and secretion of insulin in hepatic cirrhosis. J Clin Endocrinol Metab 471774-779. 1978 19. Chong Shing Teng, Ho PWM, Yeung RTT et al: Down-

regulation of insulin receptors in postnecrotic cirrhosis of liver. J Clin Endocrinol Metab 55:524-530, 1982 20. Stewart A, Johnston DG, Alberti KGMM, et al: Hormone and metabolic profiles in alcoholic liver disease. Eur J Clin Invest 13:397-403, 1983 21. Wahren J, Felig P. Ahlborg G, et al: Glucose metabolism during leg exercise in man. J Clin Invest 50:27 15-2725, 197 1 22. Marliss EB: The role of insulin in the metabolic response to exercise in diabetic man. Diabetes 28:76-81, 1979 (suppl I) 23. Felig P, Wahren J: Role of insulin and glucagon in the regulation of hepatic glucose production during exercise. Diabetes 28:71-75. 1979 (suppl 1) 24. Minuk HL, Vranic M. Marliss EB: Glucoregulatory and metabolic response to exercise in obese non insulin-dependent diabetes. Am J Physiol240:E458-E464, 1981 25. Issekutz B, Vranic M: Role of glucagon in regulation of glucose production in exercising dogs. Am J Physiol 280:E13-E20, 1980 26. Galbo H, Holst JJ, Christensen NJ: Glucagon and plasma catecholamine responses to graded and prolonged exercise in man. J Appl Physiol38:70-76. 1975 27. Ring-Larsen H, Hesse B. Henriksen JH, et al: Sympathetic nervous activity and renal and systemic hemodynamics in cirrhosis: Plasma norepinephrine concentration, hepatic extraction. and renal release. Hepatology 2:304-310, 1982 28. Ramond MJ, Comoy E, Lebrec D: Alterations in isoprenaline sensitivity in patients with cirrhosis: Evidence of abnormality of the sympathetic nervous activity. Br J Pharmacol2 1: I9 l- 196, 1986 29. Martin MJ, Horwitz DL, Natrass M, et al: Effects of mild hyperinsulinemia on the metabolic response to exercise. Metabolism 30:688-694. 198 I 30. Murray FT, Zinman B, McClean PA, et al: The metabolic response to moderate exercise in diabetic man receiving intravenous and subcutaneous insulin. J Clin Endocrinol Metab 44:708-720. 1977 31. Lennon DLF. Stratman FW, Shrago E, et al: Effects of acute moderate-intensity exercise on carnitine metabolism in men and women. J Appl Physiol Respir Environ Exert Physiol 55: I275- 1278. 1983 32. Carlin JI. Reddan WG, Sanjak M, et al: Carnitine metabolism during prolonged exercise. J Appl Physiol 54: 1275- 1278, 1986 33. Hoppel CL, Genuth SM: Carnitine metabolism in normalweight and obese human subjects during fasting. Am J Physiol 238:E409-E415. 1980 34. Hoppel CL. Genuth SM: Urinary excretion of acetyl carnitine during human diabetic and fasting ketosis. Am J Physiol 243:E168-El72,1982 35. Genuth SM, Hoppel CL: Plasma and urine carnitine in diabetic ketosis. Diabetes 28:1083-1087, 1979 36. Rudman D, Sewell CW, Ansley JD: Deficiency of carnitine in cachectic cirrhotic patients. J Clin Invest 60:7 16-723, 1977 37. Fuller RK, Hoppel CL: Elevated plasma carnitine in hepatic cirrhosis. Hepatology 3:554-558. 1983