Preservation of glucose tolerance and insulin secretory response to repeated glucose loads by the feeding of minimal glucose during prolonged fasting

Preservation of Glucose Tolerance and Insulin Secretory to Repeated Glucose Loads by the Feeding of Minimal During Prolonged Fasting

Response Glucose

C. Abraira and A. M. Lawrence The facilitation occurring effects

of glucose disposal (Staub-Traugott

after closely spaced intravenous

of minimal

glucose tolerance

amounts

of glucose

tests, obese volunteers

glucose every 6 hr. and compared

effect) and potentiation

of serum insulin (IRI) concentration

glucose loads, are known to disappear

during

fasting

upon the insulin

response

dynamics

after

to totally fasted subjects without stimulation

modified fast, while baseline serum glucagon unchanged

glucose supplementation.

after the prolonged with accelerated potentiating

by repetitive

as expected

after the repeated

were

preferentially

However,

To study the

of repeated

intravenous

sparing

a delayed

G

LUCOSE DIRECTLY stimulates B cells to release insulin. In addition a priming exposure to glucose appears to potentiate subsequent insulin release to closely spaced glucose challenges. This potentiating effect of prior glucose administration has been shown in normal humans,‘.2.3 in acromegalic patients,4 in obese subjects’ and has also been demonstrated from rat pancreas studied in vitro.6 These observations are interesting, in part, because of the similarity of these circumstances to usual intermittent exposure of humans to successive nutrient loads throughout the day. The related phenomenon of improved disposition of administered glucose when given as closely spaced challenges is known as the Staub-Traugott effect.5 However, this normally enhanced response of insulin secretion to potentiating stimuli given in successive challenges has demonstrated defects in insulin release not clearly shown after a standard single stimulus. The Staub-Traugott effect is absent in decompensated diabetics,’ and Koncz and Soeldner have recently detected significant diminution in insulin release after a second intravenous glucose pulse in children of two diabetic parents when these to matched controls.* responses were compared

From the Endocrinology Diabetes Section, Medical Service, Veterans Administration Hospital. Hines. Illinois: Abraham Lincoln School of Medicine, University of Illinois, and Loyola University Stritch School of Medicine. Received for publication February 25. 1981. Address reprint requests to C. Abraira, M.D., Chief. Endocrinology Diabetes Section, Medical Service (I I I C). Veterans Administration Hospital. Hines, Illinois 60141. 0 1981 by Grune & Stratton, Inc. 002&0495/81/3012-00I0$0I.00/0

loss rate and the fall

preserved

lower, and the basal IRI/IRG

ratios were thus fasted

in the fed state, but failed to do so

glucose loads, in which a significant

releasable

effect and

by this glucose

similar in the fed and glucose modified

that minimal amounts of carbohydrate

including increased glycolysis, responsible

Weight

the Staub-Traugott

loading were

glucose injections

modified fast until the second and third repetitive

action of glucose,

glucose

levels (IRG) were significantly

glucose disposal. It is suggested

utilization mechanisms,

1200

intravenous

from the fed state. IRG and free fatty acid suppression

states. Lactic acid levels increased

and disposal

normally

starvation.

were fasted for a mean of 25 t 2 days, while receiving either 8 or 16 gm of oral

in basal IRI and glucose levels were similar to those of totally fasted subjects. insulin secretory

after prolonged

rise coincided

during fasting preserve the insulin

pool of insulin, while protecting

for the Staub-Traugott

the glucose

effect.

Although the quantitative insulin secretory response of these subjects to an initial glucose challenge is comparable to that reported from controls, impaired potentiation of insulin release was seen in a subsequent closely-timed glucose load. In addition, we have reported in totally fasted non-diabetic men that this facilitated insulin release also disappears.’ Because of these observations a defect in the putative inducible glucose sensor in the B cell membrane, brought about either by total fasting or even, apparently, by a genetic predisposition to diabetes, appears to be unmasked by review of glucose assimilation and insulin secretory dynamics to closely timed and repeated glucose challenges. In fasted rats, however, Kipnis et al.” reported that although insulin secretory dynamics to a single post fast glucose load was significantly diminished, the intraperitoneal administration of trivial amounts of calories, in the form of glucose, prevented the alteration in the insulin secretory profile from that seen in fed controls. Such observations have raised the intriguing question whether deficient insulin secretory responses in either the totally calorically deprived or in the predisposed diabetic are absolute or not. In the present study we have attempted to determine whether the allotment of “trivial” calories to starving non-diabetic men could preserve normal carbohydrate metabolism, perhaps by protecting the integrity of the putative glucose sensor in the beta cell membrane. We have quantitated the effects of small amounts of glucose on the insulin secretory response and glucose tolerance to repeated intravenous glucose loads in calorically deprived human subjects, known to show gross deterioration in facilitated glucose disposal following total starvation.’ Obese men were studied Merabofism,

Vol.

30, No. 12

K%xember).

198 1

1201

INSULIN RESPONSE DURING PROCONGED FAST

before and immediately following a prolonged period of severe caloric restriction where either 8 or 16 gms of glucose were consumed every six hours throughout the study period. These results were compared with previous studies carried out by our group on totally caloric deprived comparably matched obese controls.’ MATERIALS AND METHODS Nine severely obese male patients were studied (+ 888, range 55%142%, over ideal body weight). (Table I). The age range of the subjects in the present study (33-57 yr) essentially overlapped with that of those individuals studied earlier with total starvation (2449 yr), who also had severe obesity (mean + 94% range 53%180% over ideal body weight). All patients were ambulatory and in good health at the time of the study and none had a history of major illness other than severe obesity. None had a personal or family history of diabetes mellitus. Each subject had been on an unrestricted diet for at least 2 wk prior to admission, and received at least 250g of carbohydrate daily for three days before testing while in the hospital awaiting study. After an overnight fast and 1 hr of resting in bed between the hours of 8:00 a.m. and 900 a.m., an indwelling scalp vein needle was inserted into an antecubital vein of each arm, one for glucose administration, kept open with a slow normal saline drip, and the other for blood sample withdrawal. The latter was kept patent with a solution of heparin and saline (5 units/ml). After base-line samples were obtained, the first glucose challenge (0.5 g/kg ideal body weight administered as a 50% glucose solution) was infused over a period of 3 min. Post glucose samples were drawn at 5, 10, 15, 25, 35, 45, and 55 min after the beginning of the infusion. Five minutes following the last sample of the preceding intravenous glucose tolerance test, a second and a third glucose challenges were administered with identical timing of blood sampling as in the first test. These patients were then started on a prolonged period of fast, described below, during which they received a minimum of 2 liters of fluid daily, consisting of water, tea, and coffee with saccharine and dietetic lemonade (equivalent to a maximum of 1.6 g or 7 cal/day as carbohydrate), 300 mg of potassium bicarbonate three times daily and one U.S. Pharmacopeia multiple vitamin daily. Five subjects received 8 gms of glucose orally every 6 hr as a 50% solution, and the remaining four subjects received 16 gms of glucose every 6 hr. during the entire period of fasting. (Table 2). These patients also participated in daily scheduled and supervised activities (swimming, walking) and were instructed to remain

Table 1. Subjects Ideal wt Kg

Age

Initial wt

Final Fasting

wt

Days

Kg

A

51

75

121

21

Kg

108

8

49

75

116

30

99

C

48

76

157

18

146 114

D

41

73

130

22

E

34

71

115

19

102

F

40

76

184

34

165

G

50

77

166

29

149

Ii

33

70

154

30

138

I

57

62

96

21

84

Mean

44

73

139

25

123

SEM

(range 33-57)

1.6

2

9.7

9.1

ambulatory on the ward during the day. The mean duration of fast was 25 days, with a range of 18-34 days (Table 1). Of the eight totally starved subjects reported earlier,’ six fasted between 21 and 36 days, one 45 days and one 80 days. All patients tolerated the prolonged fasting without significant complaints. Hematocrit, hemoglobin, and serum electrolytes, checked weekly, remained normal throughout. Fasting was interrupted with a triple intravenous glucose tolerance test performed in the manner of the prefast protocol and at least IO hr after the last small dose of oral glucose. ANALYTICAL METHODS The first 1.5 ml of blood and saline was discarded and the sample collection performed with a clean syringe. Blood samples were aliquoted into tubes with either Fluoride-Versene additive for glucose determinations, Trasylol, 500 I.U. for glucagon determinations, perchloric acid for lactic acid determination, or tubes without additives. Plasma was separated and frozen for later glucose and free fatty acid determinations; serum was separated and frozen at - 2OC, pending analysis of immunoreactive insulin and glucagon. Glucose was measured by the autoanalyzer method of Hoffman.‘* A modified method of the radiochemical assay of Ho et al. was used for determination of free fatty acids.” A dextran-coated charcoal method was employed for determination, by radioimmunoassay, of serum insulin levels.” Serum glucagon was determined by a modification of double antibody method.” Blood lactate levels were assayed enzymatically using commercially available assay kits.” Glucose tolerance, expressed as the disappearance rate (K value, decrease of blood glucose in %/min), was obtained by plotting absolute glucose values semilogarithmically against time.” The

Table 2. Insulin Summations Fed State Oral Glucose DOS0 Patient

EvmyGhr

Glucose-Modified I.V. Glucose

I.V. Glucose

I.V. Glucose

1st Load

2nd Load

3rd Load

1 st Load

Fast State

2nd Load

3rd Load

A

8

244

297

366

165

146

220

8

8

605

780

864

320

405

483

C

8

384

307

277

126

168

251

D

8

333

385

334

229

103

130

E

8

434

392

553

188

191

575 818

F

16

558

794

BOO

488

507

G

16

264

472

522

286

299

379

Ii

16

429

463

482

458

469

658

I

16

252

429

612

310

322

449

387

480

534

286

290

440

49

74

Mean SEM

43.8

61.6

67

41.7

1202

ABRAIRA

plasma glucose regression value was calculated from the IS min sample onwards as the period of C-l 5 min is generally interpreted as one of equilibrium prior to actual glucose uptake into tissuesI Insulin response to each glucose load was expressed as a summation of insulin levels from 5, 15, 25, 35, and 45 min after each intravenous load as well as in terms of absolute levels. Initial (5 min sample) and maximum increments were measured for each glucose load over baseline levels. For the second and third pulses of intravenous glucose, the insulin concentrations at 55 min of the preceding curve served as a baseline. The maximum or “peak” insulin increment, after each glucose injection, was determined from that baseline. Insulin/glucagon (IRI/IRG) molar ratios were calculated as described by Unger.17 Mean IRI/IRG were compared by averaging the values obtained following each intravenous glucose challenge. Results were statistically evaluated using the Wilcoxon Rank Sum-test, or the Student’s t test where indicated.” RESULTS

Weight Loss The initial mean weight for this patient population was 139 kg (range 96-184). Final weight after 25 + 2 days of fasting (range 18-34 days) was 123 kg (range 84-165). This represents a mean daily weight loss of 0.62 + 0.02 kg/day, entirely comparable to the loss of 0.57 k 0.04 kg/day reported in our earlier study of total fasted subjects.’

Insulin Levels After completing mean fasting insulin

the period of modified fast the concentration dropped to 12.3 +

AND

LAWRENCE

3.0 pU/ml, half the value obtained in the fed state, 23.6 f 5 pU/ml. Each of the subjects in this study demonstrated, prior to introduction of their modified fast, peripheral insulin levels that increased after each successive intravenous glucose administration. As can be seen from Fig. 1, after a vigorous initial insulin release in response to the first glucose pulse, the second glucose challenge elicited significantly higher levels of insulin and these increased further after administration of the third glucose load, as is also noted in healthy subjects undergoing successive interest is the glucose challenges.‘m3 Of particular observation that the summation of insulin levels in these normally fed men increased significantly from the first to the third load (p < 0.02) and reached borderline statistical significance when the pattern between the third and second load was examined (p < 0.10). Immediately interrupting the period of fasting, the first 1.V. glucose pulse was accompanied by a brisk increment in blood insulin levels. The only real difference during the first test period in the patterns of insulin secretion in this fasted group from when they were studied in the fed state was that levels dropped more rapidly. In addition to the incremental insulin rise in these modified fasted individuals they also demonstrated sustained insulin secretory responses throughout assimilation of the second and further increase in the

140 120 I

1

I

TIME lllllllllllllll~

5

l

25

45

5 25 45 TIME (MN)

5

25

INTERVAL

IO MIN

f

45

Fig. 1. Mean( r SEMI insulin values during each of three successive I.V. glucose tolerance tests during fed and glucose modified fasted states in obese patients. The inset shows values obtained during similar tests before end after total fast.

1203

INSULIN RESPONSE DURING PRDLONGED FAST

nous glucose loads were grossly similar to that of the same obese subjects studied while in the fed state before fasting (Table 3). The only significant alterations in patterns of insulin secretory responses following the modified fast was that the maximum increment in plasma insulin concentrations following the second intravenous glucose challenge was less than obtained when studied in the fed state. Also, these obese subjects, recovering from their modified fast, had delayed the time of maximum plasma insulin increment during the third intravenous glucose period when compared with the timing of peak patterns of insulin secretion obtained during a similar study protocol in the fed state (p -C0.05) (Table 3). The pattern of insulin increments after modified fasting in the third load were essentially preserved, contrasting with those seen after total fasting (Table 3) in which the values were drastically and significantly lower when compared with the insulin increments obtained in the same patients during the fed state.

third study periods (Fig. 1). By contrast,..starved obese subjects had failed to further increase their circuIating immunoreactive insulin levels when subjected to repeated glucose loads in the earlier study (Fig. 1, inset).5 The difference between the summation of insulin concentrations during the third and first glucose tolerance test, immediately following this modified fast, was significant at p -C0.02. Closer examination of the pattern of serum insulin levels following each of the three successive glucose challenges reveals not only the preservation of potentiation in insulin concentration in these modified fasted subjects, but also shows that the larger caloric supplement (16 gm every 6 hr) preserved greater enhanced insulin secretion over that observed in subjects receiving the smaller caloric feeding (Table 2). The insulin summations during the second load fell by a mean of 53% (range 45%73%) after the modified fast as compared with the fed state in the subjects receiving only 8 gms of glucose every 6 hr. The mean fall was only 25% (range +O.Ol% to -36%) on the subjects receiving 16 gms. The difference in percentage falls was significant (p < 0.01). In other words it would appear that, despite the rigorous caloric restriction and comparable weight loss, the higher caloric supplement, namely, 260 calories daily as glucose, was associated with earlier preservation of insulin secretory dynamics than that seen in comparable individuals receiving only 130 calories daily (Table 2).

Intravenous Glucose Disposal Rates As a group these obese men showed impaired to borderline low glucose disposal rates to a standard intravenous glucose challenge at the outset of these studies while in the fed state. These observations do not necessarily define this group as diabetic, however. Although obesity is associated with resistance to the biologic effects of insulin and frequently leads to impaired glucose tolerance, it has been shown that severely obese men can maintain normal oral glucose tolerance by at least three different diagnostic criteria while often demonstrating impaired intravenous glucose tolerance.19 In three of our subjects who also underwent oral glucose tolerance tests, this observation was confirmed.

Insulin Increments In essence the results of these studies in obese individuals subjected to a modified fast, characterized by the administration of trivial amounts of calories as glucose showed that the initial (5 min) and maximal or “peak” increments after the initial post-fast glucose challenge and following the second and third intrave-

Table 3. Insulin Increments 1 st Load

5’

Max

2nd Load

3rd Load

Time

Time

Max

MZi

fmin)

5’

Max

(min)

Time Max 5’

Max

fmin)

Glucose Modified Fast Fed State. Mean

77

04

S.E.M.

21

19

6.0

13

Modiied Fast. Mean

75

79

8

13.3

S.E.M. P.

7.1 NS

7.7 NS

18

20.3

2.4 NS

3.9 NS

58 8.0 31 5.4 <0.05

17 4.2 26

24

39

12

12

22

47

14 2.3 32

5.1

8.8

11

6.4

NS

NS

NS

<0.05

82

197

18

29

72

23

52

Total Starvation (Ref. 51 Fed State. Mean

186

196

S.E.M.

39

40

Total Fasted. Mean

130

157

S.E.M. P.

239

33

NS

NS

8

69

139

2.5

19

46

8

34

82

1.3 NS

19 NS

17 5.7 12

6.1 29

32

2.6

11

15

9.2


NS

<0.05

<0.05

NS

1204

ABRAIRA AND LAWRENCE

suppression in the first load, levels remained unchanged throughout and not significantly different from those obtained when patients were tested fed. This contrasts with the significant higher suppression observed after total fasting during the third repeated intravenous glucose tolerance in our earlier report.5 The baseline IRI/IRG molar ratios were virtually identical in the fed (3.49 t 0.6) and the modified fasted state (3.4 t 0.7). This contrasts with the fall of insulin glucagon molar ratios that characterizes total starvation.17 Mean IRI/IRG ratios during the intravenous glucose loads were very similar in both the fed and the modified fasted state, showing significantly

MEAN GLUCAGON CONCENTRATION 220 --

180 h

FED MOD. FAST

Immunoreactive Glucagon (IRG) and IRI/IRG Molar Ratios Unlike what was observed in totally starved subjects(s), fasting glucagon levels were significantly lower after the period of modified fasting (161 2 28 pg/ml after fasting, and 199 + 27 in the fed state (p < 0.02). A progressive suppression after repetitive glucose loads was observed in the fed state (Fig. 3). In the glucose-modified fasting state, after maximum

MEAN GLUCOSE DISAPPEARANCE

15

35

55

15 35

55

15 35

55

TIME (MINI

RATE (IO MEAN INSULIN IGLUCAGON MOLAR RATIOS

0 Iii

Ii2

lb

Fig. 2. Mean glucose disposal rates (Kj for obese patients receiving three successive intravenous glucose tests, obtained in the fed state and after prolonged fast modified with small amounts of glucose. The inset shows values obtained during similar tests before and after total fast.6

I ST

2ND fbUJ;OSE

3RD fbLl&OSE

k%osE Fig. 3. Upper Panel: Mean (+SEM) serum glucagon values during each of three successive I.V. glucose tolerance tests during fed and glucose-modified fasted states. Lower Panel: Mean (*SEW insulin glucagon molar ratios during each glucose tolerante test.

1205

INSULIN RESPONSE DURING PROLONGED FAST

MEAN BLOOD LACTIC ACID

progressive increments with each glucose load (Fig. 3). A similar pattern was shown earlier in the totally fasted subjects.’ Free fatty acid suppression was similar overall in both metabolic states studied, fed and modified fast (Fig. 4), in a pattern comparable to the total fasting studies reported earlier, ’ in spite of differing insulin secretory patterns.

IST

2ND

3RD

Blood Lactate Level: (Fig. 5) Blood lactate levels are known to increase during the course of assimilating a glucose load.*‘**’In the fed obese subjects this increment is especially noticeable 55 min following the first glucose challenge when compared with the initial 15 min measurement but is also present during the second and third successive loads (Fig. 5). Although lactate levels, basally and after the first intravenous glucose load, were lower following the period of modified fasting, blood lactate levels increased significantly during the end of the second glucose load (p c 0.05) and rose substantially and significantly throughout assimilation of the third glucose challenge where a blood lactate elevation was noted at 35 and at 55 min in all patients studied (Fig. 5, p -z 0.001 and p < 0.01). DISCUSSION

When glucose loads are administered at closely spaced intervals one can demonstrate that glucose tolerance or the rate of glucose disposal improves with successive glucose challenges, the so-called Staub-

MEAN FREE FATTY ACIDS IST 2ND GLUCOSE WJ;DSE 1100 LoAD

1vI

‘I

3RD

v

fizoSE

100,1 I5 3555

I5 3555 TIME (MINI

IS 3555

Fig. 4. Mean (iSEM) free fatty acid values during the triple tandem I.V. glucou tolerance teat on fed state and after prolonged fan modiied by small amounts of glucose.

I5

35 55 I5

35 55

I5

35

55

TIME (MINI Fig. 5.

Mean

I+ SEW

blood lactic acid levels during repeated

Traugott effect.‘-3 However, this phenomenon is significantly diminished or absent in individuals with insulin-dependent diabetes.7 In this regard it is of some particular note that one non-diabetic situation in which both the Staub-Traugott effect and the insulin secretion to repeated glucose loads are known to deteriorate is following a prolonged period of total caloric deprivation. As mentioned in the introduction, Grey and Kipnis’ reported that the intraperitoneal administration of glucose, preserved insulin secretion in fasted rats. In the present study we have attempted to determine whether a similar allocation of small amounts of glucose representing 3-9 percent of weight maintenance caloric needs would affect facilitated glucose disposal and insulin release in fasted men otherwise known for demonstrating not only absence of the Staub effect but actual reversal of that phenomenon.5 The present results indicate that this modified fast served to preserve significantly improved glucose assimilation during successive IV. glucose challenges. Basal insulin levels dropped significantly during the modified fast and were comparable to what has been observed in similar subjects totally fasted. Incremental insulin release, with the exception of the second test period, was, however, largely preserved. More impressive, was the insulin secretory profile generated during the third glucose load following interruption of the modified fast where an absolute and significant protentiation of insulin summation was achieved. It could be speculated that the hypothetical glucoseinducible receptor in the beta cell,** responsible for increased insulin release after repeated stimuli, is inhibited or somehow lost during the course of total

1206

starvation,’ but is protected by minimal carbohydrate intake during prolonged caloric restriction. The intake of minimal amounts of carbohydrate in this context could be viewed as involving preservation of a “slowly” releasable pool of insulin,23 which might be glycolytically24 or at least metabolically maintained.2s726 It has been suggested that periods of total caloric deprivation of more than 48 hr are necessary for islet cell metabolism to be significantly depressed.25 Were such the case this might explain why potentiated insulin release has been reported to be preserved following glucose exposure in shorter term fasted rats.6 In our current study, very small numbers of calories daily as glucose taken orally throughout the fast, appeared to preferentially maintain the delayed insulin release pool (third load), a period that is consistent with the time requirement for newly synthesized insulin to be released.*’ Although this and other studies dealing with the Staub-Traugott effect suggest an apparent correlation between insulin secretion and absolute insulin levels or increments with measured glucose disposal rates, such a correlation may not be the only explanation for the dramatic recovery of the Staub effect during the third glucose load in these partially fasted men. Dissociation between insulin levels and the Staub effect has been reported by us and by others3s8 and we have suggested that its presence may depend more upon glucose metabolism through glycolytic and glycogenic enzymatic pathways in turn dependent on optimal nutrition resulting in hormonal preparedness.3V435 Furthermore, although insulin’s pivotal role is acknowledged, it has been shown by us*’ and several other authors that the utilization of glucose loads may be independent of acute insulin secretory increments.29m32 For example, in this study, basal insulin levels, following the prolonged modified fast, fell as low as that seen in totally starved human subjects, suggesting, at least, that basal insulin levels have little predictive value with regard to facilitated assimilation of closely timed glucose challenges immediately following interruption of severe caloric deprivation. Furthermore, Lilavivathana and associates33 showed that basal insulin levels dropped to below 4 pU/ml in normal subjects during three days of total caloric restriction whether glucose was infused to maintain a normal circulating plasma glucose or whether the blood sugar was allowed to drop below normal in the course of this short fast. In a complementary study Schteingart et a1.34 noted that recovery of post-fast insulinopenia could not be achieved even if large amounts of dietary carbohydrates were given unless adequate protein was available in the post-fast diet. Thus, the insulinopenia associated with prolonged caloric restriction appears not to be related to the

ABRAIRA AND LAWRENCE

prevailing blood sugar, and it is not entirely surprising that even despite small carbohydrate allotments in our calorically restricted subjects, basal insulin levels fell significantly. This basal phenomenon was countered by marked improvement in absolute and incremental insulin release during the subsequent closely timed intravenous glucose challenges immediately following the modified fast. In keeping with the likelihood that trivial amounts of glucose calories, consumed during a prolonged fast, would contribute to more normal kinetic functions of the beta cell and thus earlier and greater availability of releasable insulin pools it was seen that subjects given 16 as compared to 8 gms of glucose every 6 hr demonstrated, as a group, better preserved insulin summation levels to the second intravenous glucose test. Furthermore, although the amounts of glucose administered during this modified fast represents only a very small fraction of weight maintenance calories in these men it is obviously not metabolically insignificant. For example, in other studies dealing with modified fasts, comparable amounts of calories as glucose have also been shown to be associated with intermittent increases in blood sugar and insulin levels, signihcantly diminished negative nitrogen balance, hyperuricemia, ketosis, and metabolic acidosis, despite comparable weight losses.35,N Trivial amounts of carbohydrate administered during prolonged fasting have also been reported to protect liver function.35 It is likely that the intermittent feeding of trivial amounts of calories during an otherwise prolonged fast would serve to protect glycolytic pathways which have generally been shown to be seriously impaired during total starvation.36-38 Furthermore, blood lactate levels ordinarily increase following the administration of an oral or intravenous glucose load in the fed state, and this rise is presumed to reflect increased glycolytic activity, either in muscle*‘.*’ or most likely, in liver.39 Because lactate levels fail to rise when glucose is administered following a prolonged period of total starvation. it has been interpreted that glycolysis is severely interfered with, following interruption of total starvation.*‘.*’ In the present study, however, where obese men were subjected to a modified fast there was an initial lack of lactate rise following intravenous administration of the first and second glucose load, an observation comparable to what we have previously reported in totally fasted men. However, in these same subjects, following this modified fast, we noted a significant rise in blood lactate levels nearing the end of the second and particularly during assimilation of the third glucose load. This observation, during the third test period, coincided with significantly accelerated glucose disposal over that seen following either the first or

1207

INSULIN RESPONSE DURING PROLONGED FAST

second intravenous glucose tolerance test. The rise in serum lactate noted late in the second and throughout the third study period would be expected if the accelerated glucose disposal during the third test period was actually associated with the presence of increased glycolytic activity, not present during the previous two hours of repeated glucose challenges. An alternate explanation would be a delayed depression of hepatic uptake of lactate. However, the splanchnic uptake of gluconeogenic precursors is reported to be already further suppressed by the daily addition of glucose during prolonged fasting.* No significant difference in either suppression of serum glucagon levels or in the insulin-glucagon ratios was seen during these repeated glucose tolerance tests in these particular obese subjects studied either in the fed or immediately following the period of modified fasting. By contrast glucagon suppression was significantly enhanced during the third intravenous glucose load in totally starved subjects.’ In addition, unlike the absence of changes in basal serum glucagon levels after total starvation, baseline glucagon levels decreased’ significantly in our caloricahy deprived patients receiving minimal glucose intake during prolonged fast. It would appear that these effects upon glucagon levels should be added to the other metabolic alterations obtainable by low doses of glucose modifying long-term fast. An important consequence is that the baseline insulin/glucagon molar ratio remained constant and similar to that accompanying a balanced diet” in spite of the severe weight loss of these men. We have earlier characterized the metabolic states with absent Staub-Traugott effect as having low basal insulin levels.’ The single exception would be the glucose-modified fast reported here, where low basal insulin levels still permit a preserved Staub effect.

Perhaps the intact basal insulin/glucagon molar ratio characterizes better than insulin levels the hormonal milieu necessary for the preservation of the ability of the liver to utilize glucose.4’ In summary, apparently trivial amounts ofcarbohydrate calories seem to protect beta cell secretory dynamics during prolonged and severe caloric ,testriction, especially those beta cell activities associated with the ability to modify the insulin seeretory response to glucose stimuli sensed during closely spaced glucose challenges following an initial primary glucose load. This caloric allotment, although insignificant to prevent weight loss and significant lowering in basal insulin levels to what is obtained in subjects totally fasted, appears to be associated with the capability to induce glycolytic activity so that immediately following interruption of the modified fast essentially normal potentiation of glucose tolerance is achieved. The results of this study suggest that nutritional factors are essential to maintenance of the putative beta cell glucose sensor, and that this sensor or receptor can be modified quantitatively. Whether some comparable inducible manipulations might improve carbohydrate homeostasis in diabetics or in individuals genetically predisposed to diabetes where facilitated insulin secretion is relatively impaired remains to be studied. In any event a potential survival advantage during prolonged periods of severe caloric deprivation may accrue from access to minimal numbers of carbohydrate calories. ACKNOWLEDGMENT

r

The authors express their appreciation to Joan Ellinger for technical work, Michele DeBartolo, R.D. for dietary advice and Laverne Hodges for technical work and assistance in the preparation of the graphs and statistical analysis of the data.

REFERENCES 1. Szabo AI, Maier J, Szabo 0, Camerini-Davaios RA: Improved glucose disappearance following repeated glucose administration. Diabetes 18:232, 1969 2. Metz R, Friedenbert R: Effects of repetitive glucose loads on plasma concentrations of glucose, insulin and free fatty acids; paradoxical insulin responses in subjects with mild glucose intolerance. J Clin Endocrinol30:602,1970 3. Abraira C, Graham L, Lawrence AM: Absence of facilitated glucose disposal (Staub-Traugott effect) in subjects with hypopituitarism. Metabolism 241145, 1975 4. Abraira C, Lawrence AM: Improved tolerance to successive glucose loads in acromegaly. Metabolism 26:287-293, 1977 5. Abraira C, Lawrence AM: The Staub-Traugott phenomenon. Effects of starvation. Am J Clin Nutr 3 1:213-221, 1978 6. Efendic S. Cerasi E, Luft R, et al: Potentiation of glucoseinduced insulin release by glucose in the isolated pancreas of fed and fasted rats. Diabetes 25:949, 1976 7. Streeten DHP, Gerstein M, Woolfold D, et al: Measurement of glucose disposal rates in normal and diabetic human subjects

after repeated intravenous injections of glucose. J Clin Endocrinol 24:761, 1964 8. Koncz L, Soeldner JS, Otto H, et al: Deranged insulinsecretory dynamics in offspring of two diabetic parents after double stimulation with intravenous glucose. Diabetes 26:1184-l 191, 1977 9. Grey NJ, Goldring S, Kipnis DM: The effects of fasting, diet and actinomycin D on insulin secretion in the rat. J Clin Invest 4988 l-890,1970 10. Hoffman WS: Rapid photoelectric method for determination of glucose in blood and urine. J Biol Chem I20:5 1, 1937 11. Ho RJ, Meng HC: A simple and ultrasensitive method for determination of free fatty acids by radiochemical assay. Anal Biochem 31:426,1969 12. Herbert J, Lau KS, Gottlieb CW, et al: Coated charcoal radioimmunoassay of insulin. J Clin Endocrinol Metab 251375, 1965 13. Eisentraut AM, Whissen N, et al: Incubation damage in the radioimmunoassay for human plasma glucagon and its prevention with trasylol. Am J Med Sci 255:137, 1968

1208

14. Calbiochem-Behring: Rapid Lactate Reagents. Pub #LO3053. La Jolla, Cal, May I, 1978 15. Lundbaek K: Intravenous glucose tolerance as a tool in definition and diagnosis of diabetes mellitus. Brit Med J 1:1507. 1962 16. Butterfield WJH, Abrams ME, Whichelow MJ: The 25 gram intravenous glucose tolerance test: A critical appraisal. Metabolism 201255, I97 1 17. Unger RH: Insulin/glucagon ratio. Israel J Med Sci 8:252. 1972 18. Arm&age P: Statistical methods in medical research. New York, Wiley, I97 I, p 398 19. Abraira C, Lawrence AM: Glucose metabolism in obese men: Abnormal intravenous and normal oral glucose tolerance. Proc Sot Exp Biol Med 157:407-410, 1978 20. Sussman KE: Effect of prolonged fasting on glucose and insulin metabolism in exogenous obesity. Arch Int Med 117:343, 1966 21. Jackson RA, Moloney M, Lowy C, et al: Differences between metabolic responses to fasting in obese diabetic and obese nondiabetic subjects. Diabetes 20:214, 1971 22. Kipnis DM, Permutt MA: Inductive effects of glucose on the insulin secretory and synthetic apparatus of the pancreatic beta cell. Israel J Med Sci 8:224, 1972 23. Koncz L. Soeldner JS, Otto M, et al: Insulin secretory dynamics after two consecutive intravenous stimulations with glucose and/or tolbutamide. Metabolism 28: 1 183- 1197, 1980 24. Malaisse WJ, Sener A, Levy J: The stimulus secretion coupling of glucose induced insulin release: Fasting induced adaptation of key glycolytic enzymes in isolated islets. J Biol Chem 251:1731-1737, 1976 25. Bouman PR, Wolters GHJ, Konijnendigk W: Insulin secretion and cyclic adenosine 3’,-S’-monophosphate levels in pancreatic islets of fed and fasted rats. Diabetes 28: 132-l 40. I979 26. Trus M, Warner H, Matschinsky F: Effects of glucose on insulin release and on intermediary metabolism of isolated perifused pancreatic islets from fed fasted rats. Diabetes 29: I - 14. 1980 27. Howell S, Taylor KW: The secretion of newly synthesized insulin in vitro. Biochem J 102:922-927, I967

ABRAIRA AND LAWRENCE

28. Abraira C, Goldstein MS: Acute glucose loads in insulinized pancreatectomized dogs. J Lab Clin Med 8 I :696-703, 1973 29. Bergman RN: Integrated control of hepatic glucose metabolism. Fed Proc 36:265-270. I977 30. Sacca H, Hendler R, Sherwin RS: Hyperglycemia inhibits glucose production in man independent of changes in glucoregulatory hormones. J Clin Endocrinol Metab 47: 1160-I 163. 1978 3 1. Liljenquist JE, Mueller GL, Cherington AD, et al: Hyperglycemia per se (insulin and glucagon withdrawn) can inhibit hepatic glucose production in man. J Clin Endocrinol Metab 48:171 -17s. 1979 32. Fineberg SE, Schneider SH. Howey DC: Insulin-independent glucose disposal during physiologic hyperglycemia. Diabetes 29(2):40A, 1980 (abstract) 33. Lilavivathana U. Campbell RG, Brodows RG: Control of insulin secretion during fasting in man. Metabolism 27:815-821. 1978 34. Schteingart DE, McKenzie AK, Victoria RS. et al: Suppression of insulin secretion by protein deprivation in obesity. Metabolism 28:943-949, 1979 35. Abraira C, Virupannavar C, Nemchausky B: Protective effect of small amounts of glucose on abnormal liver function tests during starvation. Metabolism 29:943-948. 1980 36. Perez N, Clark-Turri L. Rabajille E, et al: Regulation of rat liver enzymes by natural components of the diet. J Biol Chem 239:2420, 1964 37. Weber G, Singhal RL, Stamm NB. et al: Sinchronous behavior pattern of key glycolytic enzymes: glucokinase: phosphofructo-kinase and pyruvate kinase. Adv Enzyme Regul4:59. 1966 38. Pilkis SJ, Krahl ME: Glucokinase and glycogen synthesis under hormonal regulation. Fed Proc 28:888, I969 39. Jackson RA, Peters N. Advani V, et al: Forearm glucose uptake during the oral glucose tolerance test in normal subjects. Diabetes 22:4422458, 1973 40. Aoki T. Muller NA, Brennan MF. et al: Metabolic effects of glucose in brief and prolonged fasted man. Am J Clin Nutr 28:50775 I I, 1975 41. Parilla R. Goodman MN, Toews CJ: Etfect of glucagon: insulin ratios on hepatic metabolism. Diabetes 23:725- 73 I, 1974