Effect of Acute Fasting on Diaphragm Strength and Endurance 1- 3
c. SHINDOH,
A. DIMARCO, W. LUST, and G. SUPINSKI
Introduction
It is not uncommon for severely ill patients to be inadvertently fasted for several days after admission to an intensive care unit. Although conventional wisdom would hold that such temporary cessation of nutritional intake should be avoided, some experts have pointed out that there may be substantial morbidity associated with both parenteral and enteral nutritional support, but there are only limited data demonstrating that short periods of fasting have deleterious physiologic effects (1). One potential mechanism by which fasting could adversely affect critically ill patients is by affecting the functional capacities of the respiratory muscles. Studies of limb muscles to indicate that muscle endurance may be critically dependent upon muscle carbohydrate stores since depletion of these stores has been shown to result in muscle exhaustion, i.e., a reduction in muscle contractile function (2-4). If short-term starvation produces a depletion of respiratory muscle carbohydrate stores, it is possible that an associated impairment in respiratory muscle function could also occur. Such a change in respiratory muscle function could, in turn, predispose fasted patients to respiratory muscle fatigue and hypercapnic respiratory failure. The purpose of the present study was to examine the effects produced by short periods of fasting on diaphragm muscle glycogen stores and contractile function. To permit precise control over nutritional intake and direct measurement of diaphragm tension development, studies were performed in hamsters. Measurements of diaphragm tension and glycogen content were made on excised diaphragm muscle strips taken from several groups of animals killed after varying periods of acute fasting. Methods Studies were performed on 36 8-month-old male Syrian golden hamsters (Charles River Breeding Laboratories, Wilmington, MA). 488
SUMMARY The effects of short periods of fasting on diaphragm contractile function remain unclear. The purpose of the present study was (1) to examine the relationship between duration of acute fasting and diaphragm contractile performance, and (2) to assess the effects of fasting on diaphragm glycogen stores and the relationship between changes in diaphragm function and alterations in muscle glycogen stores. Studies were performed on four groups of Syrian hamsters (nine animals in each group). One group served as a control and was allowed to feed normally, whereas the other three groups were fasted for either 1, 2, or 3 days. Diaphragm strips from animals were studied in vitroby measuring tension during electrically induced contractions. Twostrips from each animal were studied; one strip was examined with a bath glucose equal to the prevailing blood glucose, and the second was preincubated in a high glucose solution (170 mg/dl) for 20 min. Fasting resulted in reductions in body weight, blood glucose concentrations, diaphragm strength, and diaphragm endurance in strips tested at the prevailing blood glucose levels. These effects were pronounced in animals fasted for 3 days, with little or no change in diaphragm contractility observed in animals fasted for shorter periods. Diaphragm weight, thickness, and glycogen content were unchanged in the fasted animals, as was the weight of the soleus muscle. Preincubation of strips from 3-day-fasted animals in a high glucose medium resulted in a significant increase in diaphragm strip strength and endurance. These data suggest that (1) fasting of sufficient duration to produce hypoglycemia may elicit reductions in diaphragm contractility, (2) these reductions in diaphragm function are not related to fasting-induced changes in diaphragm glycogen content, and (3) these decrements in contractility may be reversible by correcting the hypoglycemia. AM REV RESPIR DIS 1991; 144:488-493
Animals were divided into four equal groups, three of which were fasted for differing lengths of time (1,2, and 3 days), whereas the fourth group served as a control and were allowed to feed normally. Water was given ad libitum to all animals during the period of fasting, and animals were weighed at 24-h intervals. Animals were killed by decapitation and, within 1 min, the diaphragm was removed en bloc and placed in a dissecting dish containing oxygenated (95% Or5o/0 CO 2 ) , glucosefree Krebs-Hanselheit solution (pH, 7.40; Na, 154 mEq; K, 5.0 mEq; Ca, 5 mEq; Mg, 2.0 mEq; CI, 145 mEq; HC03 , 15 mEq; HP0 4, 1.9 mEq; S04' 2.0 mEq; insulin, 50 units/L) (5). Simultaneously, blood glucose concentration was assessed using a dextrostix, and an aliquot of blood was frozen in liquid nitrogen and stored (- 80° C) for later analysis. Four muscle strips were dissected from the diaphragm, two each from the right and the left hemidiaphragm. In brief: (1) one strip was placed in an organ bath containing oxygenated Krebs-Hanselheit solution (at 20° C) having a glucose concentration of 170 mg/dl (this strip was electrically stimulated to provide an assessment of muscle strength and fatigability); (2) a second strip was also placed in an organ bath containing Krebs-Hanselheit with a 170 mg/dl glucose concentration (this strip was not stimulated); (3) a third strip was placed in a bath containing Krebs- Hanselheit
solution with a glucose concentration adjusted to equal the blood glucose level as assessed by the dextrostix analysis (this strip was also stimulated to determine its strength and fatigability); (4) a fourth strip was placed in a Krebs-Hanselheit solution that had a glucose adjusted to equal the blood glucose (this strip was not stimulated). The two stimulated strips provided a means of comparing diaphragm contractility for a given number of days of fasting for strips incubated in a high glucose bath (170 mg/dl) and strips incubated in a bath with a lower glucose concentration equal to the blood glucose level. Assessment of glycogen stores in these two strips (to be described below) provided a measure of the glycogen remaining after a fatigue trial. Assessment of glyco-
(Received in original form March 6, 1990 and in revised form March 21, 1991) 1 From the Pulmonary Division, MetroHealth Medical Center, Case Western ReserveUniversity, Cleveland, Ohio. 2 Supported by Grant HL-38926 from the National Institutes of Health. 3 Correspondence and requests for reprints should be addressed to Gerald S. Supinski, M.D., MetroHealth Medical Center, Pulmonary Division, 3395 Scranton Road, Cleveland, OH 44109.
EFFECT OF FASTING ON THE DIAPHRAGM
gen stores in the two nonstimulated strips provided a measure of prefatigue glycogen levels. The two muscle strips (one in a 170 mg/dl organ bath and one in a blood-glucose bath) that were not subjected to electrical stimulation were simply kept in these organ baths for the remainder of the experiment. The two remaining muscle strips were mounted so that the origin of each strip was secured by steel needles, and the tendinous insertion of each strip was attached to a nondistensible stainless steel wire connected to a Grass FT-IO force transducer (Grass Instruments, Quincy, MA). These latter muscle strips werestimulated with supramaximal currents (1.2 to 1.3 times the current required to elicit a maximal tension) delivered via platinum field electrodes. Current (0.2-ms duration pulses) was supplied by a constant current stimulus isolation unit (Applied Neural Control Laboratory, Case Western Reserve University, Cleveland, OH) driven by a Grass S48 stimulator.
Experimental Protocol The two strips used to assessmuscle contractile characteristics (one placed in an organ bath containing a glucose concentration of 170 mg/L and the other placed in a bath with a glucose concentration that equaled the blood glucose) were allowed to equilibrate in the muscle bath solution for 20 min. Muscle length was then adjusted to that length at which twitch tension development was maximal. Muscle contractile characteristics were assessed from measurements of muscle twitch kinetics, of the diaphragm force-frequency relationship, and of diaphragm fatigability during a series of repetitive rhythmic contractions. Twitch kinetics were assessed by measuring the time to peak tension development (contraction time) and the time for peak tension to fall by 50010 (half relaxation time) during single muscle twitches. The diaphragm force-frequency relationship was studied by stimulating diaphragm strips at frequencies of 1, 10,20, 50, and 100 Hz (stimuli were applied until a clear plateau in tension was reached, e.g., 800 to 1,000ms). Muscle fatigability was assessed by calculating the rate of fall of tension over 5 min of rhythmic contractions (90/min). Contractions wereinduced by stimulating strips with trains of 20 Hz stimuli (train duration 200 ms, contraction duty cycle of 0.30). After completion of this protocol, these two muscle strips wereweighed and subsequently frozen in liquid nitrogen for later analysis of muscle glycogen concentrations. At this point in time, we also froze the two nonstimulated muscle strips (one kept in a 170 mg/L bath and the other in a bath adjusted to the blood glucose). For those strips used to assess contractile characteristics, strip cross-sectional area was calculated by dividing muscle mass by the product of fiber length, measured with a micrometer, and muscle density (1.06g/cm") (6). Muscle widths were also measured, and muscle thickness was calculated as muscle cross-sectional area divided by muscle width.
489
Tension was calculated as force per unit area (kg/em"),
Whole diaphragm weight was calculated by adding the weights of the four strips to the remaining portion of the diaphragm. One soleus muscle from each animal was dissected free and also weighed. Frozen blood samples were subsequently thawed and assessed for glucose concentrations, and diaphragm samples were thawed and assessed for glycogen levels as described in Passonneau and Lauderdale (7). Specifically, muscle samples were washed, placed in 0.03 N Hel, and homogenized at zero degrees. Homogenates were placed in a boiling water bath for 5 min. Glycogen was then enzymatically converted to glucose by adding a mixture containing amylo-alpha-1, 4-alpha-1, and 6 glucosidase (AG), and the resultant glucose produced was measured using a glucose-6-Pdehydrogenase/hexokinase reaction. Glucose assays were performed on samples to which no AG was added to provide a measure of tissue glucose and glucose-6-P levels, which werethen subtracted from the results obtained after AG preincubation. Tissue glycogen levels were expressed per gram of muscle wet weight.
Data Analysis Data are presented as mean ± 1 SE. Differences in sample means were statistically compared using t tests and ANOVA. Comparisons with a p value of less than 0.05 were taken as being statistically significant (8). Results
100
85
o
3
Fasting Duration (days) Fig. 1. Effect of fasting on body weight. Open circles = no fasting; open triangles = 1 day of fasting; closed circles = 2 days of fasting; closed triangles = 3 days of fasting. Bars indicate 1 SE.
ing, strips taken from control and from 1-, 2-, and 3-day-fasted animals were studied in organ baths with glucose concentrations of 90 ± 6, 75 ± 8, 60 ± 6, and 42 ± 8 mg/dl in this series of studies. Fasting had no effect on diaphragm twitch kinetics, with contraction times (CT) of approximately 105 ms for all four groups of animals and with half relaxation times (Y2 RT) of about 75 ms (table 2). Twitchtension fell significantly with fasting' however, with twitch tensions from
Effects of Fasting on Body Weight and Muscle Dimensions Prefasting hamster weight averaged 193 ± 7 g. With fasting, body weights declined, with progressively greater reductions in weight the longer fasting was continued. As shown in figure 1, weights fell by 4070 at the end of the first 24 h of 0.6 fasting, and by 6.5 and 10% by 48 and 72 h of fasting, respectively (p < 0.01 for comparison of 72 h with baseline weights). In contrast to this reduction in body 00 0.5 weight, diaphragm and soleus weights did .s: OJ) not change over the 3 days of fasting, as .0; shown in figure 2. In keeping with these ~ latter findings, diaphragm strip length, calculated cross-sectional area, and cal- ...92 u 0.4 culated thickness were also not changed ::I ~ by fasting, as shown in table 1.
rJ)
Effects of Fasting on Diaphragm Twitch Kinetics, on the Force-Frequency Relationship, and on Fatigability We will first present measurements of diaphragm contractile function for strips studied at blood glucose concentrations. Because blood glucose levelsfell progressively with increasing duration of fast-
2
~
Diaphragm
~ Soleus
0.3 3 Fasting Duration (days) Fig. 2. Effect of fasting on diaphragm (open symbols) and soleus (closed symbols) weights.
490
SHINDOH, DIMARCO, WST, AND SUPINSKI
TABLE 1 EFFECT OF FASTING ON DIAPHRAGM STRIP DIMENSIONS· Duration of Fasting
Length
(days)
(mm)
o
14.1 14.2 13.9 13.5
1 2 3
± ± ± ±
0.3 0.3 0.4 0.3
Crosssectional Area
Calculated Thickness
(cm~
(mm)
0.029 0.029 0.029 0.030
± ± ± ±
0.001 0.002 0.002 0.002
0.55 0.51 0.55 0.58
± ± ± ±
0.03 0.04 0.05 0.03
• Values are mean ± 1 SE.
TABLE 2 EFFECT OF FASTING ON DIAPHRAGM TWITCH KINETICS' Duration of Fasting
Twitch T, kg/cm Twitch CT, ms 112 RT, ms
2
o day
1 day
2 days
3 days
0.55 ± 0.03 105.9 ± 2.9 71.2 ± 2.8
0.58 ± 0.02 113.0 ± 4.2 85.3 ± 6.2
0.49 ± 0.02 105.6 ± 2.7 74.9 ± 3.3
0.36 ± 0.03 102.3 ± 1.9 71.9 ± 2.6
Definition of abbreviations: Twitch T • Values are mean ± 1 SE.
=
twitch tension; Twitch CT
nonfasted and 3-day-fasted animals of 0.55 ± 0.03 and 0.36 ± 0.03 kg/ern", respectively (p < 0.001 for this comparison). Fasting also affected the diaphragm force-frequency curve, as shown in figure 3. Although the force-frequency relationship was not appreciably altered by 2 days of fasting, marked reductions in diaphragm tension were noted in animals fasted for 3 days (p < 0.001 for comparison of force- frequency curves from nonfasted and from 3-day-fasted animals), with apparent reductions in both high and low frequency tension development.
=
twitch contraction time; 1/2 RT
=
half relaxation time.
As shown in figure 4, strips studied in organ baths adjusted to blood glucose concentrations also demonstrated greater fatigability with increasing duration of fasting. For example, tension fell by 69 ± 3070 over 5 min for strips taken from nonfasted animals and by 83 ± 2070 over 5 min in strips taken from animals fasted for 3 days (p < 0.01). The contractile characteristic of strips taken from control animals and from 1-day-fasted animals and studied in organ baths containing a "high" glucose concentration (i.e., 170 mg/dl) were not different from those assessed on strips studied in organ baths containing glu-
cose concentrations equal to that measured in the blood. For strips taken from animals fasted for longer periods of time, however, altering the bath glucose concentration had an appreciable effect on the diaphragm force-frequency relationship and on diaphragm fatigability. As shown in figure 5, strips taken from 3-day-fasted animals and studied in a high glucose bath had significantly higher tensions, in response to all frequencies of stimulation, than did strips studied in a bath having a "low" glucose concentration equal to the blood glucose measurement (p < 0.01 for comparison of the two force-frequency curves). Strips studied in a high glucose bath also had a lower rate of fatigue than did strips studied in a low glucose bath, as shown in figure 6 (p < 0.01 for the comparison of rate of fatigue for strips from 3-dayfasted animals studied in high and low glucose baths). Although incubation in a high glucose bath improved muscle function of diaphragms from fasted animals, this intervention did not entirely reverse the effects of fasting since strength and fatigability remained somewhat impaired when compared with values for nonfasted animals. For example, tension fell by 69 ± 3070 during fatigue trials for nonfasted animals and by 78 ± 3070 during fatigue trials for 3-day-fasted animals af-
2.0
N
1.6
E u --.. 0()
90
~ N
E
u
'E
00
Q)
~
lrl
0
';ji
80
c:
~
~
e:
coe: ._ 0 ~ .~
o
.iii
~
c:
"2
1.2
0.8
u.. ~ 0 .5
-Q)
~
~
0.4
~
70
0.4
n:l
~
u
Q)
0
~
60 110 20
1 10
20
50
50
100
100
Stimulation Frequency (Hz) Stimulation Frequency (Hz)
Fig. 3. Effect of fasting on the diaphragm force-frequency relationship. Open circles = no fasting; open triangles = 1 day of fasting; closed circles = 2 days of fasting; closed triangles = 3 days of fasting.
o
2
3
Fasting Duration (days) Fig. 4. Effect of fasting on diaphragm fatigability.
Fig. 5. Effect of altering bath glucose concentration on the force-frequency relationship of the diaphragm from fasted animals. Open and closed symbols indicate, respectively, data obtained from strips studied in high and low glucose concentration baths.
491
EFFECT OF FASTING ON THE DIAPHRAGM
30 100
90
85
Fig. 7. Effect of fasting on blood glucose and diaphragm glycogen content.
80
o
75
2
3
"
o !
3
Fasting Duration (days)
70
1_-----l.-_----l.....-__ Low Bath Glucose
High Bath Glucose
Fig. 6. Effect of altering bath glucose concentration on rate of fatigue of the diaphragm from fasted animals.
ter preincubation in a high glucose bath (p < 0.05 for this comparison).
Effect of Fasting on Blood Glucose and Diaphragm Glycogen Concentrations Blood glucose levels determined by the method of Passonneau and Lauderdale (7), like measurements made using dextrostix, fell progressively the longer fasting was continued. Specifically, Passonneau glucose levels for control and for 1-, 2-, and 3-day-fasted animals were 90 ± 5, 76 ± 5, 62 ± 3, and 45 ± 4 mg/dl. On the other hand, diaphragm glycogen content did not fall significantly with 3 days of fasting (figure 7). Specifically,diaphragm glycogen concentrations were 22 ± 5 and 18 ± 3 urnol/g tissue wet weight for nonstimulated samples of diaphragm taken from nonfasted and from 3-day-fasted animals, respectively. Glycogen concentrations of strips of stimulated muscles (i.e., at the conclusion of fatigue trials) were somewhat lower than for nonstimulated strips taken from the same animals, but they were also not significantly different for fasted and nonfasted animals. For example, glycogen concentrations averaged 11 ± 2, 11 ± 1, 9 ± 2, and 8 ± 2 umol/g wet weight for fatigued strips taken from control and from 1-, 2-, and 3-day-fasted animals, respectively. Discussion
This study demonstrated that short periods of fasting induce reductions in diaphragmatic strength in hamsters, as
manifested by a downward shift in the in vitro diaphragm force-frequency relationship, and also increased diaphragm fatigability. In addition, although fasting resulted in a reduction in diaphragm twitch tension, it had no effect on twitch contraction or relaxation times. These effects appeared to depend upon the duration of fasting, with appreciable changes in contractile function occurring only after 3 days of fasting. These alterations in contractile function could not be explained on the basis of muscle atrophy since diaphragm dimensions and weight were unchanged. These changes could also not be explained by a reduction in diaphragm glycogen content since only small, nonsignificant decrements in glycogen concentrations wereinduced by fasting. There were, however, significant reductions in blood glucose levels with fasting, and the effects of fasting on diaphragm strength and fatigability were largely reversed by incubating the diaphragm in a high glucose-containing solution prior to physiologic assessment.
Methodologic Considerations These studies wereperformed in animals to permit precise control of dietary intake and to allow for the direct assessment of diaphragm contractile characteristics using in vitro techniques. While permitting more exact measurements to be made, there are certain limitations of this animal model of acute fasting. First, the fact that measurements of diaphragm strength and fatigability were made in vitro after a 20-min equilibrium period may have affected our results byeliminating humoral and hormonal influences on diaphragm function. The hormonal milieu of the body is altered as the result of fasting, with changes in catecholamine, insulin, glucagon, and other substances that may affect in vivo muscle function in a manner that cannot be fully appreciated using in vitro methods of
measurements. There may also be significant species-related variation in the responses to a given duration of fasting. The basal metabolic rate of small animals is appreciably higher than that of humans, and it is possible that a longer period of fasting may be required in humans to produce alterations in diaphragmatic function that are comparable to those produced by 4 days of fasting in hamsters. It also seems possible that the effect of fasting on diaphragmatic function may be magnified in the presence of various disease states. A number of stressful conditions (e.g., trauma, infection) induce muscle catabolism and, as a result, the magnitude of the diaphragmatic dysfunction produced by fasting may be markedly accentuated in the presence of such stresses (9). It could also be argued that hormonal factors may have "protected" the diaphragm from the effects of prolonged systemic hypoglycemia in the intact animals' and that the reductions in strip strength and increased fatigability seen in fasted strips studied in vitro occurred because of incubation in a low glucose bath. Totest this possibility, wecompared the force-frequency curves and fatigability of strips from control, nonfasted animals incubated in a low glucose bath (40 mg/dl, three strips) with strips taken from the contralateral hemidiaphragm and incubated in a high glucose bath (170 mg/dl, three strips). As shown in table 3, tension and fatigue characteristics of these two groups of strips were identical, indicating that 20 min of incubation in a low glucose bath is not sufficient to depress diaphragm contractility. Studies were performed with an organ bath temperature of 20° C, a factor that may have influenced our results. For one thing, strip contraction time (CT) was 105.9 ms in nonfasted animals in the present study, which was substantially faster than the CT of 25.1 ms reported by Farkas and Roussos (10) for golden hamsters,
492
SHINDOH, DIMARCO, LUST, AND SUPINSKI
TABLE 3 COMPARISON OF STRENGTH AND ENDURANCE IN MUSCLES FROM NONFASTED ANIMALS INCUBATED IN HIGH AND LOW GLUCOSE BATHS*
1 Hz
10Hz
20 Hz
50 Hz
100 Hz
Rate of Fatigue (% decrease in tension over 5 min)
0.60 ± 0.05
0.96 ± 0.09
1.71 ± 0.12
2.14 ± 0.11
2.07 ± 0.10
87 ± 4
0.68 ± 0.03
1.15 ± 0.05
1.77 ± 0.09
2.02 ± 0.13
1.98±0.15
81 ± 4
Force-Frequency Curves
High glucose bath strips Low glucose bath strips
• Values are mean ± 1 SE.
and also greater than the 70.0 ms CT reported by Burbach and coworkers (11) for FIB strain hamsters. This value also exceeds that observed by Supinski and Kelsen (5) in a previous study of golden hamsters, which found a CT of 69.6 ms. These differences are probably caused by differences in the bath temperatures used in these various investigations, with Farkas and Roussos studying hamsters at 37° C, and with both Burbach and coworkers and Supinski and Kelsen using a bath temperature of 22° C. The fact that our bath temperature was lower than that used in these other studies probably accounts also for the fact that measurements of half relaxation time (Y2 RT) also differ for these various studies; Y2 RT was 71.2 ms in the present study, carried out at 20° C, whereas values of 95.0 ms were reported by Burbach and coworkers (11), 72.8 ms by Supinski and Kelsen (5), and 23.5 ms by Farkas and Roussos (10). There is somewhat less discrepancy between measurements of tetanic tension in the present study and those observed in previous golden hamster studies. Tetanic tension averaged 1.85 kg/ern- for all control nonfasted animals in the pres- . ent study, a value similar to the 1.85 kg/ern? value reported by Supinski and Kelsen and the 2.06 kg/em" value reported by Farkas and Roussos. Our tetanic tensions are far in excess of those reported by Burbach and coworkers who found this parameter to be 0.92 kg/ern'. It is difficult to explain this latter difference since tetanic tension is not thought to be highly temperature-dependent. The lower tetanic tensions reported by Burbach and coworkers maybe the result of the different strain of hamster (i.e., the FIB) studied by this latter group or other methodologic differences.
Effects of Fasting on Body Weight and Muscle Dimensions Our observation that diaphragm and soleus weights remained constant with fasting even though body weight de-
creased is consistent with previous reports by Li and Goldberg (12) demonstrating a maintenance of soleus weight in fasted rats despite a reduction in whole body weight. Although we did not examine the weights of other skeletal muscles, it has been reported that the effects of fasting may differ from muscle to muscle. It has been suggested that muscle activity may play a role in modulating the rate of muscle-wasting during prolonged caloric deprivation, and that more active muscles such as the diaphragm may be partially protected from the effects of fasting (9, 13).
Effects of Fasting on Diaphragm Strength and Endurance There are several recent reports in which the effects of fasting on respiratory muscle contractility have been examined. Lewis and Sieck (14) found that 4 days of fasting had no effect on the strength of the rat diaphragm and produced only a minor effect on diaphragm fatigability. As in the present study, they assessed strength and endurance in vitro using diaphragm muscle strips taken from animals fasted for different periods. Duriel and coworkers (15)also examined the effects of fasting on in vivo rat diaphragm function, assessing diaphragm contractility from measurements of the transdiaphragmatic pressure bilateral phrenic nerve stimulation. These investigators observed a decrease in maximal transdiaphragmatic pressure generation, increased diaphragm fatigability in response to repetitive contraction, and a reduction in diaphragm weight as the result of fasting. Only one study has examined the effect of acute caloric deprivation on respiratory muscle function in humans. This report found no effect of 7 days of decreased food intake on respiratory muscle strength or endurance (16). Unlike the studies performed in animals, in which food intake was completely stopped, this human study employed only a partial reduction in caloric intake.
The slightly different findings of these studies probably reflect the different species studied and variations in methodology employed in these different experiments. In particular, it seems possible that the degree of fasting may have varied between studies, with those studies demonstrating an effect on diaphragm contractility having employed a greater degree of caloric deprivation relative to animal metabolic rate. In the present study, alterations in diaphragm strength were noted only when fasting was sufficiently prolonged to induce marked systemic hypoglycemia. The other studies examining the effects of acute fasting on diaphragm contractility have either employed a degree of fasting insufficient to induce significant hypoglycemia (14) or have not documented the degree of hypoglycemia present at the time of physiologic assessment (15). It should also be noted that two studies have shown that prolonged reductions in caloric intake can induce significant decreases in diaphragm strength (17, 18). These previous experiments examined periods of caloric deprivation sufficiently long (several weeks) to produce atrophy of type II fibers within the diaphragm. The reduction in strength observed in these chronic studies appeared, therefore, to be the result of fiber atrophy and are probably unrelated to the mechanisms by which diaphragm contractility was reduced by acute fasting in the present study.
Relationship of Alterations in Diaphragm Function to Changes in Blood Glucose Levels Previous studies by Coyle and coworkers (19) and Hickson and colleagues (20) have suggested that limb muscle endurance may depend upon carbohydrate availability during certain forms of exercise, with limb muscle exhaustion occurring when cellular glycogen stores are depleted, Administration of exogenous lipid or carbohydrate increases muscle endurance by slowing the rate of depletion of endogenous muscle carbohydrate stores, forestalling glycogen depletion. One might expect, based on these previous results, that prolonged fasting might reduce respiratory muscle function simply by reducing muscle carbohydrate stores, thereby decreasing the time a given respiratory work load can be sustained before carbohydrate depletion occurs. Our findings indicate, however, that short duration fasting did not produce significant glycogen depletion in the diaphragm and that the phenomenon observed by
493
EFFECT OF FASTING ON THE DIAPHRAGM
Coyle and coworkers and Hickson and colleagues cannot account for the alteration in diaphragm fatigability produced by fasting in the present study. Instead, alterations in diaphragm function appeared to be related to alterations in glucose levels rather than to changes in glycogen stores. This latter possibility is supported by the observation that reductions in diaphragm function in fasted animals seemed to correlate with reductions in blood glucose levels and that increases in the concentration of glucose in the organ bath could partially reverse the effectsof fasting, improvingdiaphragm strength and reducing fatigability. The mechanism by which alterations in glucose concentrations may have affected diaphragm function is unclear. It is known that glycogenolysis can become inhibited during exercise, necessitating a switch to alternative substrates to meet muscle metabolic needs (21). It is conceivable that a similar phenomenon may have occurred during the development of diaphragm fatigue in the present study and that diaphragm strips with higher prefatigue glucose levels may have been better able to tolerate such a reduction in glycogenolysis. It is also possible, however, that alterations in glucose levels may have affected diaphragm function by some alternative mechanism. For one thing, it is conceivable that prolonged hypoglycemia could lead to such alterations in metabolism that either irreversible or poorly reversible damage would be done to cellular organelles. The fact that incubation in a high glucose bath for a relatively short time (20 min) could improve muscle contractility argues, however, against such a possibility. It would seem more likely that the effects of fasting and hypoglycemia were mediated by an effect on some easily reversiblecellular constituent that modulated contractile activity. It is known, for example, that action potention propagation is strongly dependent on transmembrane gradients of potassium and sodium. It is also
known that cellular phosphate concentration is an important determinant of cellular ATPase activity and that phosphate accumulation can impair contractile function (22). It is conceivable that low intra- and extracellular glucose levels may have indirectly affected diaphragm contractile function by altering cellular levels of these ions, and that restoration of normal glucose concentrations could have reversed this process, normalizing ionic gradients and/or reducing intracellular phosphate or hydrogen ion concentrations (22, 23).
Potential Implications Regardless of the mechanism by which acute fasting resulted in reductions in diaphragm strength and endurance, our observation that contractile function could largely be restored to prefasting levels by incubating strips from fasted animals in a high glucose medium may have important implications. Our data suggest that the presence of severehypoglycemia may be the key element in reducing muscle contractility under fasted conditions. Furthermore, maintenance or restoration of normal plasma glucose levels may prevent or reverse the diaphragm dysfunction resulting from acute fasting. As a corollary, durations of fasting insufficient to elicit significant reductions in plasma glucose levels are probably insufficient to produce appreciable decrements in diaphragm contractility. References 1. Koretz RL. Breathing and feeding: can you have one without the other? Chest 1984; 85:298. 2. Newsholme EA. The glucose/fatty acid cycle and physical exhaustion. In: Human muscle fatigue: physiological mechanisms. Ciba Foundation symposium 82. London: Pitman Medical, 1981; 89-101. 3. Karlsson 1, Saltin B. Diet, muscle glycogen and endurance performance. 1 Appl Physiol 1971; 31:203-6. 4. Gollnick PD, Piehl K, Saubert CW IV, Armstrong RB, Saltin B. Diet, exercise and glycogen changes in human muscle fibers. 1 Appl Physiol 1972; 33:421-5. 5. Supinski GS, Kelsen SG. Effect of elastase induced emphysema on the force generating ability
of the diaphragm. 1 Clin Invest 1982; 70:978-88. 6. Close RI. Dynamic properties of mammalian skeletal muscles. Physiol Rev 1972; 52:129-97. 7. Passonneau lV, Lauderdale VR. A comparison of three methods of glycogen measurement in tissues. Anal Biochem 1974; 60:405-12. 8. Snedecor GS, Cochran WG. Statistical methods. 6th ed. Ames: Iowa State University Press, 1967. 9. Odessey R. Amino acid and protein metabolism in the diaphragm. Am Rev Respir Dis 1979; 119:107-12. 10. Farkas GA, Roussos C. Adaptability of the hamster diaphragm to exerciseand/or emphysema. 1 Appl Physiol 1982; 53:1263-72. 11. Burbach lA, Schlenker EH, Johnson lL. Morphometry, histochemistry and contractility of dystrophic hamster diaphragm. Am 1 Physiol 1987; 253:275-84. 12. Li lB, Goldberg AL. Effects of food deprivation on protein synthesis and degradation in rat muscle. Am 1 Physiol 1976; 231:441-8. 13. Goldberg AL, Goodman HM. Relationship between cortisone and muscle work in determining muscle size. 1 Physiol (Lond) 1969;200:667-75. 14. LewisMI, Sieck Gc. The effect of acute nutritional deprivation on diaphragm performance (abstract). Am Rev Respir Dis 1989; 139:AI65. 15. Duriel B, Viires N, Weber B, et al. Effects of acute starvation on diaphragmatic function. Am Rev Respir Dis 1987; 135(Part 2:A329). 16. Bender PR, Martin Bl. Ventilatory and treadmill endurance during acute semistarvation. 1 Appl Physiol 1986; 60:1823-9. 17. Lewis MI, Sieck GS, Fournier M, et al. Effect of nutritional deprivation on diaphragm contractility and muscle fiber size. 1 Appl Physiol 1986; 60:596-603. 18. Kelsen SG, Ference M, Kapoor S. Effects of prolonged undernutrition on structure and function of the diaphragm. 1 Appl Physiol 1985; 58:1354-9. 19. Coyle EF, Hagberg 1M, Hurley BF, Martin WH, Eheavi M, Holloszy 10. Carbohydrate feeding during prolonged strenuous exercise can delay fatigue. 1 Appl Physiol 1983; 55:230-5. 20. Hickson RC, Reunie Ml, Conlee RK, Winder WW, Halloszy 10. Effects of increased plasma fatty acids on glycogen utilization and endurance. 1 Appl Physiol 1977; 43:829-33. 21. Constable SH, Favier Rl, Holloszy 10. Exercise and glycogen depletion: effects on ability to activate muscle phosphorylase. 1 Appl Physiol1986; 60:1518-23. 22. LewisSF, Hellar RG. The pathophysiology of McArdle's disease: clues to regulation in exercise and fatigue. 1 Appl Physiol 1986; 61:391-401. 23. Fabiato A, Fabiato T. Effects of pH on the myofilaments and the sarcoplasmic reticulum of skinned cells from cardiac and skeletal muscles. 1 Physiol (Land) 1978; 276:233-55.
c. SHINDOH,
A. DIMARCO, W. LUST, and G. SUPINSKI
Introduction
It is not uncommon for severely ill patients to be inadvertently fasted for several days after admission to an intensive care unit. Although conventional wisdom would hold that such temporary cessation of nutritional intake should be avoided, some experts have pointed out that there may be substantial morbidity associated with both parenteral and enteral nutritional support, but there are only limited data demonstrating that short periods of fasting have deleterious physiologic effects (1). One potential mechanism by which fasting could adversely affect critically ill patients is by affecting the functional capacities of the respiratory muscles. Studies of limb muscles to indicate that muscle endurance may be critically dependent upon muscle carbohydrate stores since depletion of these stores has been shown to result in muscle exhaustion, i.e., a reduction in muscle contractile function (2-4). If short-term starvation produces a depletion of respiratory muscle carbohydrate stores, it is possible that an associated impairment in respiratory muscle function could also occur. Such a change in respiratory muscle function could, in turn, predispose fasted patients to respiratory muscle fatigue and hypercapnic respiratory failure. The purpose of the present study was to examine the effects produced by short periods of fasting on diaphragm muscle glycogen stores and contractile function. To permit precise control over nutritional intake and direct measurement of diaphragm tension development, studies were performed in hamsters. Measurements of diaphragm tension and glycogen content were made on excised diaphragm muscle strips taken from several groups of animals killed after varying periods of acute fasting. Methods Studies were performed on 36 8-month-old male Syrian golden hamsters (Charles River Breeding Laboratories, Wilmington, MA). 488
SUMMARY The effects of short periods of fasting on diaphragm contractile function remain unclear. The purpose of the present study was (1) to examine the relationship between duration of acute fasting and diaphragm contractile performance, and (2) to assess the effects of fasting on diaphragm glycogen stores and the relationship between changes in diaphragm function and alterations in muscle glycogen stores. Studies were performed on four groups of Syrian hamsters (nine animals in each group). One group served as a control and was allowed to feed normally, whereas the other three groups were fasted for either 1, 2, or 3 days. Diaphragm strips from animals were studied in vitroby measuring tension during electrically induced contractions. Twostrips from each animal were studied; one strip was examined with a bath glucose equal to the prevailing blood glucose, and the second was preincubated in a high glucose solution (170 mg/dl) for 20 min. Fasting resulted in reductions in body weight, blood glucose concentrations, diaphragm strength, and diaphragm endurance in strips tested at the prevailing blood glucose levels. These effects were pronounced in animals fasted for 3 days, with little or no change in diaphragm contractility observed in animals fasted for shorter periods. Diaphragm weight, thickness, and glycogen content were unchanged in the fasted animals, as was the weight of the soleus muscle. Preincubation of strips from 3-day-fasted animals in a high glucose medium resulted in a significant increase in diaphragm strip strength and endurance. These data suggest that (1) fasting of sufficient duration to produce hypoglycemia may elicit reductions in diaphragm contractility, (2) these reductions in diaphragm function are not related to fasting-induced changes in diaphragm glycogen content, and (3) these decrements in contractility may be reversible by correcting the hypoglycemia. AM REV RESPIR DIS 1991; 144:488-493
Animals were divided into four equal groups, three of which were fasted for differing lengths of time (1,2, and 3 days), whereas the fourth group served as a control and were allowed to feed normally. Water was given ad libitum to all animals during the period of fasting, and animals were weighed at 24-h intervals. Animals were killed by decapitation and, within 1 min, the diaphragm was removed en bloc and placed in a dissecting dish containing oxygenated (95% Or5o/0 CO 2 ) , glucosefree Krebs-Hanselheit solution (pH, 7.40; Na, 154 mEq; K, 5.0 mEq; Ca, 5 mEq; Mg, 2.0 mEq; CI, 145 mEq; HC03 , 15 mEq; HP0 4, 1.9 mEq; S04' 2.0 mEq; insulin, 50 units/L) (5). Simultaneously, blood glucose concentration was assessed using a dextrostix, and an aliquot of blood was frozen in liquid nitrogen and stored (- 80° C) for later analysis. Four muscle strips were dissected from the diaphragm, two each from the right and the left hemidiaphragm. In brief: (1) one strip was placed in an organ bath containing oxygenated Krebs-Hanselheit solution (at 20° C) having a glucose concentration of 170 mg/dl (this strip was electrically stimulated to provide an assessment of muscle strength and fatigability); (2) a second strip was also placed in an organ bath containing Krebs-Hanselheit with a 170 mg/dl glucose concentration (this strip was not stimulated); (3) a third strip was placed in a bath containing Krebs- Hanselheit
solution with a glucose concentration adjusted to equal the blood glucose level as assessed by the dextrostix analysis (this strip was also stimulated to determine its strength and fatigability); (4) a fourth strip was placed in a Krebs-Hanselheit solution that had a glucose adjusted to equal the blood glucose (this strip was not stimulated). The two stimulated strips provided a means of comparing diaphragm contractility for a given number of days of fasting for strips incubated in a high glucose bath (170 mg/dl) and strips incubated in a bath with a lower glucose concentration equal to the blood glucose level. Assessment of glycogen stores in these two strips (to be described below) provided a measure of the glycogen remaining after a fatigue trial. Assessment of glyco-
(Received in original form March 6, 1990 and in revised form March 21, 1991) 1 From the Pulmonary Division, MetroHealth Medical Center, Case Western ReserveUniversity, Cleveland, Ohio. 2 Supported by Grant HL-38926 from the National Institutes of Health. 3 Correspondence and requests for reprints should be addressed to Gerald S. Supinski, M.D., MetroHealth Medical Center, Pulmonary Division, 3395 Scranton Road, Cleveland, OH 44109.
EFFECT OF FASTING ON THE DIAPHRAGM
gen stores in the two nonstimulated strips provided a measure of prefatigue glycogen levels. The two muscle strips (one in a 170 mg/dl organ bath and one in a blood-glucose bath) that were not subjected to electrical stimulation were simply kept in these organ baths for the remainder of the experiment. The two remaining muscle strips were mounted so that the origin of each strip was secured by steel needles, and the tendinous insertion of each strip was attached to a nondistensible stainless steel wire connected to a Grass FT-IO force transducer (Grass Instruments, Quincy, MA). These latter muscle strips werestimulated with supramaximal currents (1.2 to 1.3 times the current required to elicit a maximal tension) delivered via platinum field electrodes. Current (0.2-ms duration pulses) was supplied by a constant current stimulus isolation unit (Applied Neural Control Laboratory, Case Western Reserve University, Cleveland, OH) driven by a Grass S48 stimulator.
Experimental Protocol The two strips used to assessmuscle contractile characteristics (one placed in an organ bath containing a glucose concentration of 170 mg/L and the other placed in a bath with a glucose concentration that equaled the blood glucose) were allowed to equilibrate in the muscle bath solution for 20 min. Muscle length was then adjusted to that length at which twitch tension development was maximal. Muscle contractile characteristics were assessed from measurements of muscle twitch kinetics, of the diaphragm force-frequency relationship, and of diaphragm fatigability during a series of repetitive rhythmic contractions. Twitch kinetics were assessed by measuring the time to peak tension development (contraction time) and the time for peak tension to fall by 50010 (half relaxation time) during single muscle twitches. The diaphragm force-frequency relationship was studied by stimulating diaphragm strips at frequencies of 1, 10,20, 50, and 100 Hz (stimuli were applied until a clear plateau in tension was reached, e.g., 800 to 1,000ms). Muscle fatigability was assessed by calculating the rate of fall of tension over 5 min of rhythmic contractions (90/min). Contractions wereinduced by stimulating strips with trains of 20 Hz stimuli (train duration 200 ms, contraction duty cycle of 0.30). After completion of this protocol, these two muscle strips wereweighed and subsequently frozen in liquid nitrogen for later analysis of muscle glycogen concentrations. At this point in time, we also froze the two nonstimulated muscle strips (one kept in a 170 mg/L bath and the other in a bath adjusted to the blood glucose). For those strips used to assess contractile characteristics, strip cross-sectional area was calculated by dividing muscle mass by the product of fiber length, measured with a micrometer, and muscle density (1.06g/cm") (6). Muscle widths were also measured, and muscle thickness was calculated as muscle cross-sectional area divided by muscle width.
489
Tension was calculated as force per unit area (kg/em"),
Whole diaphragm weight was calculated by adding the weights of the four strips to the remaining portion of the diaphragm. One soleus muscle from each animal was dissected free and also weighed. Frozen blood samples were subsequently thawed and assessed for glucose concentrations, and diaphragm samples were thawed and assessed for glycogen levels as described in Passonneau and Lauderdale (7). Specifically, muscle samples were washed, placed in 0.03 N Hel, and homogenized at zero degrees. Homogenates were placed in a boiling water bath for 5 min. Glycogen was then enzymatically converted to glucose by adding a mixture containing amylo-alpha-1, 4-alpha-1, and 6 glucosidase (AG), and the resultant glucose produced was measured using a glucose-6-Pdehydrogenase/hexokinase reaction. Glucose assays were performed on samples to which no AG was added to provide a measure of tissue glucose and glucose-6-P levels, which werethen subtracted from the results obtained after AG preincubation. Tissue glycogen levels were expressed per gram of muscle wet weight.
Data Analysis Data are presented as mean ± 1 SE. Differences in sample means were statistically compared using t tests and ANOVA. Comparisons with a p value of less than 0.05 were taken as being statistically significant (8). Results
100
85
o
3
Fasting Duration (days) Fig. 1. Effect of fasting on body weight. Open circles = no fasting; open triangles = 1 day of fasting; closed circles = 2 days of fasting; closed triangles = 3 days of fasting. Bars indicate 1 SE.
ing, strips taken from control and from 1-, 2-, and 3-day-fasted animals were studied in organ baths with glucose concentrations of 90 ± 6, 75 ± 8, 60 ± 6, and 42 ± 8 mg/dl in this series of studies. Fasting had no effect on diaphragm twitch kinetics, with contraction times (CT) of approximately 105 ms for all four groups of animals and with half relaxation times (Y2 RT) of about 75 ms (table 2). Twitchtension fell significantly with fasting' however, with twitch tensions from
Effects of Fasting on Body Weight and Muscle Dimensions Prefasting hamster weight averaged 193 ± 7 g. With fasting, body weights declined, with progressively greater reductions in weight the longer fasting was continued. As shown in figure 1, weights fell by 4070 at the end of the first 24 h of 0.6 fasting, and by 6.5 and 10% by 48 and 72 h of fasting, respectively (p < 0.01 for comparison of 72 h with baseline weights). In contrast to this reduction in body 00 0.5 weight, diaphragm and soleus weights did .s: OJ) not change over the 3 days of fasting, as .0; shown in figure 2. In keeping with these ~ latter findings, diaphragm strip length, calculated cross-sectional area, and cal- ...92 u 0.4 culated thickness were also not changed ::I ~ by fasting, as shown in table 1.
rJ)
Effects of Fasting on Diaphragm Twitch Kinetics, on the Force-Frequency Relationship, and on Fatigability We will first present measurements of diaphragm contractile function for strips studied at blood glucose concentrations. Because blood glucose levelsfell progressively with increasing duration of fast-
2
~
Diaphragm
~ Soleus
0.3 3 Fasting Duration (days) Fig. 2. Effect of fasting on diaphragm (open symbols) and soleus (closed symbols) weights.
490
SHINDOH, DIMARCO, WST, AND SUPINSKI
TABLE 1 EFFECT OF FASTING ON DIAPHRAGM STRIP DIMENSIONS· Duration of Fasting
Length
(days)
(mm)
o
14.1 14.2 13.9 13.5
1 2 3
± ± ± ±
0.3 0.3 0.4 0.3
Crosssectional Area
Calculated Thickness
(cm~
(mm)
0.029 0.029 0.029 0.030
± ± ± ±
0.001 0.002 0.002 0.002
0.55 0.51 0.55 0.58
± ± ± ±
0.03 0.04 0.05 0.03
• Values are mean ± 1 SE.
TABLE 2 EFFECT OF FASTING ON DIAPHRAGM TWITCH KINETICS' Duration of Fasting
Twitch T, kg/cm Twitch CT, ms 112 RT, ms
2
o day
1 day
2 days
3 days
0.55 ± 0.03 105.9 ± 2.9 71.2 ± 2.8
0.58 ± 0.02 113.0 ± 4.2 85.3 ± 6.2
0.49 ± 0.02 105.6 ± 2.7 74.9 ± 3.3
0.36 ± 0.03 102.3 ± 1.9 71.9 ± 2.6
Definition of abbreviations: Twitch T • Values are mean ± 1 SE.
=
twitch tension; Twitch CT
nonfasted and 3-day-fasted animals of 0.55 ± 0.03 and 0.36 ± 0.03 kg/ern", respectively (p < 0.001 for this comparison). Fasting also affected the diaphragm force-frequency curve, as shown in figure 3. Although the force-frequency relationship was not appreciably altered by 2 days of fasting, marked reductions in diaphragm tension were noted in animals fasted for 3 days (p < 0.001 for comparison of force- frequency curves from nonfasted and from 3-day-fasted animals), with apparent reductions in both high and low frequency tension development.
=
twitch contraction time; 1/2 RT
=
half relaxation time.
As shown in figure 4, strips studied in organ baths adjusted to blood glucose concentrations also demonstrated greater fatigability with increasing duration of fasting. For example, tension fell by 69 ± 3070 over 5 min for strips taken from nonfasted animals and by 83 ± 2070 over 5 min in strips taken from animals fasted for 3 days (p < 0.01). The contractile characteristic of strips taken from control animals and from 1-day-fasted animals and studied in organ baths containing a "high" glucose concentration (i.e., 170 mg/dl) were not different from those assessed on strips studied in organ baths containing glu-
cose concentrations equal to that measured in the blood. For strips taken from animals fasted for longer periods of time, however, altering the bath glucose concentration had an appreciable effect on the diaphragm force-frequency relationship and on diaphragm fatigability. As shown in figure 5, strips taken from 3-day-fasted animals and studied in a high glucose bath had significantly higher tensions, in response to all frequencies of stimulation, than did strips studied in a bath having a "low" glucose concentration equal to the blood glucose measurement (p < 0.01 for comparison of the two force-frequency curves). Strips studied in a high glucose bath also had a lower rate of fatigue than did strips studied in a low glucose bath, as shown in figure 6 (p < 0.01 for the comparison of rate of fatigue for strips from 3-dayfasted animals studied in high and low glucose baths). Although incubation in a high glucose bath improved muscle function of diaphragms from fasted animals, this intervention did not entirely reverse the effects of fasting since strength and fatigability remained somewhat impaired when compared with values for nonfasted animals. For example, tension fell by 69 ± 3070 during fatigue trials for nonfasted animals and by 78 ± 3070 during fatigue trials for 3-day-fasted animals af-
2.0
N
1.6
E u --.. 0()
90
~ N
E
u
'E
00
Q)
~
lrl
0
';ji
80
c:
~
~
e:
coe: ._ 0 ~ .~
o
.iii
~
c:
"2
1.2
0.8
u.. ~ 0 .5
-Q)
~
~
0.4
~
70
0.4
n:l
~
u
Q)
0
~
60 110 20
1 10
20
50
50
100
100
Stimulation Frequency (Hz) Stimulation Frequency (Hz)
Fig. 3. Effect of fasting on the diaphragm force-frequency relationship. Open circles = no fasting; open triangles = 1 day of fasting; closed circles = 2 days of fasting; closed triangles = 3 days of fasting.
o
2
3
Fasting Duration (days) Fig. 4. Effect of fasting on diaphragm fatigability.
Fig. 5. Effect of altering bath glucose concentration on the force-frequency relationship of the diaphragm from fasted animals. Open and closed symbols indicate, respectively, data obtained from strips studied in high and low glucose concentration baths.
491
EFFECT OF FASTING ON THE DIAPHRAGM
30 100
90
85
Fig. 7. Effect of fasting on blood glucose and diaphragm glycogen content.
80
o
75
2
3
"
o !
3
Fasting Duration (days)
70
1_-----l.-_----l.....-__ Low Bath Glucose
High Bath Glucose
Fig. 6. Effect of altering bath glucose concentration on rate of fatigue of the diaphragm from fasted animals.
ter preincubation in a high glucose bath (p < 0.05 for this comparison).
Effect of Fasting on Blood Glucose and Diaphragm Glycogen Concentrations Blood glucose levels determined by the method of Passonneau and Lauderdale (7), like measurements made using dextrostix, fell progressively the longer fasting was continued. Specifically, Passonneau glucose levels for control and for 1-, 2-, and 3-day-fasted animals were 90 ± 5, 76 ± 5, 62 ± 3, and 45 ± 4 mg/dl. On the other hand, diaphragm glycogen content did not fall significantly with 3 days of fasting (figure 7). Specifically,diaphragm glycogen concentrations were 22 ± 5 and 18 ± 3 urnol/g tissue wet weight for nonstimulated samples of diaphragm taken from nonfasted and from 3-day-fasted animals, respectively. Glycogen concentrations of strips of stimulated muscles (i.e., at the conclusion of fatigue trials) were somewhat lower than for nonstimulated strips taken from the same animals, but they were also not significantly different for fasted and nonfasted animals. For example, glycogen concentrations averaged 11 ± 2, 11 ± 1, 9 ± 2, and 8 ± 2 umol/g wet weight for fatigued strips taken from control and from 1-, 2-, and 3-day-fasted animals, respectively. Discussion
This study demonstrated that short periods of fasting induce reductions in diaphragmatic strength in hamsters, as
manifested by a downward shift in the in vitro diaphragm force-frequency relationship, and also increased diaphragm fatigability. In addition, although fasting resulted in a reduction in diaphragm twitch tension, it had no effect on twitch contraction or relaxation times. These effects appeared to depend upon the duration of fasting, with appreciable changes in contractile function occurring only after 3 days of fasting. These alterations in contractile function could not be explained on the basis of muscle atrophy since diaphragm dimensions and weight were unchanged. These changes could also not be explained by a reduction in diaphragm glycogen content since only small, nonsignificant decrements in glycogen concentrations wereinduced by fasting. There were, however, significant reductions in blood glucose levels with fasting, and the effects of fasting on diaphragm strength and fatigability were largely reversed by incubating the diaphragm in a high glucose-containing solution prior to physiologic assessment.
Methodologic Considerations These studies wereperformed in animals to permit precise control of dietary intake and to allow for the direct assessment of diaphragm contractile characteristics using in vitro techniques. While permitting more exact measurements to be made, there are certain limitations of this animal model of acute fasting. First, the fact that measurements of diaphragm strength and fatigability were made in vitro after a 20-min equilibrium period may have affected our results byeliminating humoral and hormonal influences on diaphragm function. The hormonal milieu of the body is altered as the result of fasting, with changes in catecholamine, insulin, glucagon, and other substances that may affect in vivo muscle function in a manner that cannot be fully appreciated using in vitro methods of
measurements. There may also be significant species-related variation in the responses to a given duration of fasting. The basal metabolic rate of small animals is appreciably higher than that of humans, and it is possible that a longer period of fasting may be required in humans to produce alterations in diaphragmatic function that are comparable to those produced by 4 days of fasting in hamsters. It also seems possible that the effect of fasting on diaphragmatic function may be magnified in the presence of various disease states. A number of stressful conditions (e.g., trauma, infection) induce muscle catabolism and, as a result, the magnitude of the diaphragmatic dysfunction produced by fasting may be markedly accentuated in the presence of such stresses (9). It could also be argued that hormonal factors may have "protected" the diaphragm from the effects of prolonged systemic hypoglycemia in the intact animals' and that the reductions in strip strength and increased fatigability seen in fasted strips studied in vitro occurred because of incubation in a low glucose bath. Totest this possibility, wecompared the force-frequency curves and fatigability of strips from control, nonfasted animals incubated in a low glucose bath (40 mg/dl, three strips) with strips taken from the contralateral hemidiaphragm and incubated in a high glucose bath (170 mg/dl, three strips). As shown in table 3, tension and fatigue characteristics of these two groups of strips were identical, indicating that 20 min of incubation in a low glucose bath is not sufficient to depress diaphragm contractility. Studies were performed with an organ bath temperature of 20° C, a factor that may have influenced our results. For one thing, strip contraction time (CT) was 105.9 ms in nonfasted animals in the present study, which was substantially faster than the CT of 25.1 ms reported by Farkas and Roussos (10) for golden hamsters,
492
SHINDOH, DIMARCO, LUST, AND SUPINSKI
TABLE 3 COMPARISON OF STRENGTH AND ENDURANCE IN MUSCLES FROM NONFASTED ANIMALS INCUBATED IN HIGH AND LOW GLUCOSE BATHS*
1 Hz
10Hz
20 Hz
50 Hz
100 Hz
Rate of Fatigue (% decrease in tension over 5 min)
0.60 ± 0.05
0.96 ± 0.09
1.71 ± 0.12
2.14 ± 0.11
2.07 ± 0.10
87 ± 4
0.68 ± 0.03
1.15 ± 0.05
1.77 ± 0.09
2.02 ± 0.13
1.98±0.15
81 ± 4
Force-Frequency Curves
High glucose bath strips Low glucose bath strips
• Values are mean ± 1 SE.
and also greater than the 70.0 ms CT reported by Burbach and coworkers (11) for FIB strain hamsters. This value also exceeds that observed by Supinski and Kelsen (5) in a previous study of golden hamsters, which found a CT of 69.6 ms. These differences are probably caused by differences in the bath temperatures used in these various investigations, with Farkas and Roussos studying hamsters at 37° C, and with both Burbach and coworkers and Supinski and Kelsen using a bath temperature of 22° C. The fact that our bath temperature was lower than that used in these other studies probably accounts also for the fact that measurements of half relaxation time (Y2 RT) also differ for these various studies; Y2 RT was 71.2 ms in the present study, carried out at 20° C, whereas values of 95.0 ms were reported by Burbach and coworkers (11), 72.8 ms by Supinski and Kelsen (5), and 23.5 ms by Farkas and Roussos (10). There is somewhat less discrepancy between measurements of tetanic tension in the present study and those observed in previous golden hamster studies. Tetanic tension averaged 1.85 kg/ern- for all control nonfasted animals in the pres- . ent study, a value similar to the 1.85 kg/ern? value reported by Supinski and Kelsen and the 2.06 kg/em" value reported by Farkas and Roussos. Our tetanic tensions are far in excess of those reported by Burbach and coworkers who found this parameter to be 0.92 kg/ern'. It is difficult to explain this latter difference since tetanic tension is not thought to be highly temperature-dependent. The lower tetanic tensions reported by Burbach and coworkers maybe the result of the different strain of hamster (i.e., the FIB) studied by this latter group or other methodologic differences.
Effects of Fasting on Body Weight and Muscle Dimensions Our observation that diaphragm and soleus weights remained constant with fasting even though body weight de-
creased is consistent with previous reports by Li and Goldberg (12) demonstrating a maintenance of soleus weight in fasted rats despite a reduction in whole body weight. Although we did not examine the weights of other skeletal muscles, it has been reported that the effects of fasting may differ from muscle to muscle. It has been suggested that muscle activity may play a role in modulating the rate of muscle-wasting during prolonged caloric deprivation, and that more active muscles such as the diaphragm may be partially protected from the effects of fasting (9, 13).
Effects of Fasting on Diaphragm Strength and Endurance There are several recent reports in which the effects of fasting on respiratory muscle contractility have been examined. Lewis and Sieck (14) found that 4 days of fasting had no effect on the strength of the rat diaphragm and produced only a minor effect on diaphragm fatigability. As in the present study, they assessed strength and endurance in vitro using diaphragm muscle strips taken from animals fasted for different periods. Duriel and coworkers (15)also examined the effects of fasting on in vivo rat diaphragm function, assessing diaphragm contractility from measurements of the transdiaphragmatic pressure bilateral phrenic nerve stimulation. These investigators observed a decrease in maximal transdiaphragmatic pressure generation, increased diaphragm fatigability in response to repetitive contraction, and a reduction in diaphragm weight as the result of fasting. Only one study has examined the effect of acute caloric deprivation on respiratory muscle function in humans. This report found no effect of 7 days of decreased food intake on respiratory muscle strength or endurance (16). Unlike the studies performed in animals, in which food intake was completely stopped, this human study employed only a partial reduction in caloric intake.
The slightly different findings of these studies probably reflect the different species studied and variations in methodology employed in these different experiments. In particular, it seems possible that the degree of fasting may have varied between studies, with those studies demonstrating an effect on diaphragm contractility having employed a greater degree of caloric deprivation relative to animal metabolic rate. In the present study, alterations in diaphragm strength were noted only when fasting was sufficiently prolonged to induce marked systemic hypoglycemia. The other studies examining the effects of acute fasting on diaphragm contractility have either employed a degree of fasting insufficient to induce significant hypoglycemia (14) or have not documented the degree of hypoglycemia present at the time of physiologic assessment (15). It should also be noted that two studies have shown that prolonged reductions in caloric intake can induce significant decreases in diaphragm strength (17, 18). These previous experiments examined periods of caloric deprivation sufficiently long (several weeks) to produce atrophy of type II fibers within the diaphragm. The reduction in strength observed in these chronic studies appeared, therefore, to be the result of fiber atrophy and are probably unrelated to the mechanisms by which diaphragm contractility was reduced by acute fasting in the present study.
Relationship of Alterations in Diaphragm Function to Changes in Blood Glucose Levels Previous studies by Coyle and coworkers (19) and Hickson and colleagues (20) have suggested that limb muscle endurance may depend upon carbohydrate availability during certain forms of exercise, with limb muscle exhaustion occurring when cellular glycogen stores are depleted, Administration of exogenous lipid or carbohydrate increases muscle endurance by slowing the rate of depletion of endogenous muscle carbohydrate stores, forestalling glycogen depletion. One might expect, based on these previous results, that prolonged fasting might reduce respiratory muscle function simply by reducing muscle carbohydrate stores, thereby decreasing the time a given respiratory work load can be sustained before carbohydrate depletion occurs. Our findings indicate, however, that short duration fasting did not produce significant glycogen depletion in the diaphragm and that the phenomenon observed by
493
EFFECT OF FASTING ON THE DIAPHRAGM
Coyle and coworkers and Hickson and colleagues cannot account for the alteration in diaphragm fatigability produced by fasting in the present study. Instead, alterations in diaphragm function appeared to be related to alterations in glucose levels rather than to changes in glycogen stores. This latter possibility is supported by the observation that reductions in diaphragm function in fasted animals seemed to correlate with reductions in blood glucose levels and that increases in the concentration of glucose in the organ bath could partially reverse the effectsof fasting, improvingdiaphragm strength and reducing fatigability. The mechanism by which alterations in glucose concentrations may have affected diaphragm function is unclear. It is known that glycogenolysis can become inhibited during exercise, necessitating a switch to alternative substrates to meet muscle metabolic needs (21). It is conceivable that a similar phenomenon may have occurred during the development of diaphragm fatigue in the present study and that diaphragm strips with higher prefatigue glucose levels may have been better able to tolerate such a reduction in glycogenolysis. It is also possible, however, that alterations in glucose levels may have affected diaphragm function by some alternative mechanism. For one thing, it is conceivable that prolonged hypoglycemia could lead to such alterations in metabolism that either irreversible or poorly reversible damage would be done to cellular organelles. The fact that incubation in a high glucose bath for a relatively short time (20 min) could improve muscle contractility argues, however, against such a possibility. It would seem more likely that the effects of fasting and hypoglycemia were mediated by an effect on some easily reversiblecellular constituent that modulated contractile activity. It is known, for example, that action potention propagation is strongly dependent on transmembrane gradients of potassium and sodium. It is also
known that cellular phosphate concentration is an important determinant of cellular ATPase activity and that phosphate accumulation can impair contractile function (22). It is conceivable that low intra- and extracellular glucose levels may have indirectly affected diaphragm contractile function by altering cellular levels of these ions, and that restoration of normal glucose concentrations could have reversed this process, normalizing ionic gradients and/or reducing intracellular phosphate or hydrogen ion concentrations (22, 23).
Potential Implications Regardless of the mechanism by which acute fasting resulted in reductions in diaphragm strength and endurance, our observation that contractile function could largely be restored to prefasting levels by incubating strips from fasted animals in a high glucose medium may have important implications. Our data suggest that the presence of severehypoglycemia may be the key element in reducing muscle contractility under fasted conditions. Furthermore, maintenance or restoration of normal plasma glucose levels may prevent or reverse the diaphragm dysfunction resulting from acute fasting. As a corollary, durations of fasting insufficient to elicit significant reductions in plasma glucose levels are probably insufficient to produce appreciable decrements in diaphragm contractility. References 1. Koretz RL. Breathing and feeding: can you have one without the other? Chest 1984; 85:298. 2. Newsholme EA. The glucose/fatty acid cycle and physical exhaustion. In: Human muscle fatigue: physiological mechanisms. Ciba Foundation symposium 82. London: Pitman Medical, 1981; 89-101. 3. Karlsson 1, Saltin B. Diet, muscle glycogen and endurance performance. 1 Appl Physiol 1971; 31:203-6. 4. Gollnick PD, Piehl K, Saubert CW IV, Armstrong RB, Saltin B. Diet, exercise and glycogen changes in human muscle fibers. 1 Appl Physiol 1972; 33:421-5. 5. Supinski GS, Kelsen SG. Effect of elastase induced emphysema on the force generating ability
of the diaphragm. 1 Clin Invest 1982; 70:978-88. 6. Close RI. Dynamic properties of mammalian skeletal muscles. Physiol Rev 1972; 52:129-97. 7. Passonneau lV, Lauderdale VR. A comparison of three methods of glycogen measurement in tissues. Anal Biochem 1974; 60:405-12. 8. Snedecor GS, Cochran WG. Statistical methods. 6th ed. Ames: Iowa State University Press, 1967. 9. Odessey R. Amino acid and protein metabolism in the diaphragm. Am Rev Respir Dis 1979; 119:107-12. 10. Farkas GA, Roussos C. Adaptability of the hamster diaphragm to exerciseand/or emphysema. 1 Appl Physiol 1982; 53:1263-72. 11. Burbach lA, Schlenker EH, Johnson lL. Morphometry, histochemistry and contractility of dystrophic hamster diaphragm. Am 1 Physiol 1987; 253:275-84. 12. Li lB, Goldberg AL. Effects of food deprivation on protein synthesis and degradation in rat muscle. Am 1 Physiol 1976; 231:441-8. 13. Goldberg AL, Goodman HM. Relationship between cortisone and muscle work in determining muscle size. 1 Physiol (Lond) 1969;200:667-75. 14. LewisMI, Sieck Gc. The effect of acute nutritional deprivation on diaphragm performance (abstract). Am Rev Respir Dis 1989; 139:AI65. 15. Duriel B, Viires N, Weber B, et al. Effects of acute starvation on diaphragmatic function. Am Rev Respir Dis 1987; 135(Part 2:A329). 16. Bender PR, Martin Bl. Ventilatory and treadmill endurance during acute semistarvation. 1 Appl Physiol 1986; 60:1823-9. 17. Lewis MI, Sieck GS, Fournier M, et al. Effect of nutritional deprivation on diaphragm contractility and muscle fiber size. 1 Appl Physiol 1986; 60:596-603. 18. Kelsen SG, Ference M, Kapoor S. Effects of prolonged undernutrition on structure and function of the diaphragm. 1 Appl Physiol 1985; 58:1354-9. 19. Coyle EF, Hagberg 1M, Hurley BF, Martin WH, Eheavi M, Holloszy 10. Carbohydrate feeding during prolonged strenuous exercise can delay fatigue. 1 Appl Physiol 1983; 55:230-5. 20. Hickson RC, Reunie Ml, Conlee RK, Winder WW, Halloszy 10. Effects of increased plasma fatty acids on glycogen utilization and endurance. 1 Appl Physiol 1977; 43:829-33. 21. Constable SH, Favier Rl, Holloszy 10. Exercise and glycogen depletion: effects on ability to activate muscle phosphorylase. 1 Appl Physiol1986; 60:1518-23. 22. LewisSF, Hellar RG. The pathophysiology of McArdle's disease: clues to regulation in exercise and fatigue. 1 Appl Physiol 1986; 61:391-401. 23. Fabiato A, Fabiato T. Effects of pH on the myofilaments and the sarcoplasmic reticulum of skinned cells from cardiac and skeletal muscles. 1 Physiol (Land) 1978; 276:233-55.
