Carbohydrate Tolerance Improves with Fasting in Obese Subjects with Noninsulin-Dependent (Type II) Diabetes

Carbohydrate Tolerance Improves with Fasting in Obese Subjects with Noninsulin-Dependent (Type II) Diabetes NELSON B. WATTS, MD,

MARIO DIGIROLAMO, MD

ABSTRACT: To determine the effects of shortterm fasting on carbohydrate tolerance, 10 obese women with noninsulin-dependent diabetes mellitus (NIDDM) were studied with meal tolerance tests before and after 3 days of fasting. After 3 days' fast, basal serum glucose declined from 15.2 ± 0.9 to 7.5 ± 0.7 mmol/L (273 ± 17 to 135 ± 13 mg/dL) (mean ± SEM, p < 0.001) and the glycemic response to the test meal (area under the glucose curve) improved by 31 %. There were no changes in basal or postprandial insulin levels but a slight increase in serum c-peptide. Resting metabolic rate and the thermic effect of food were unchanged. There was a slight but insignificant change in basal and postprandial free fatty acid levels and a significant elevation of basal beta-hydroxybutyrate levels. Blood lactate rose significantly (from 0.9 to 2.0 mM) during the initial meal tolerance test, but no rise in lactate was seen in the meal tolerance test after fasting. Two subgroups of patients were identified based on the degree of glycemic improvement after shortterm fasting. Those with lesser improvement in serum glucose showed overnight rises in serum glucose during the period of fasting (the dawn phenomenon), while those patients who norFrom the Division of Endocrinology and Metabolism, Department of Medicine, Emory University School of Medicine. Supported in part by a grant from the American Diabetes Association, Georgia Affiliate (NB W), a Clinical Research Center grant (RR 00039) and a grant (DK 39326) from the National Institutes of Health (MDG). The authors are indebted to the nurses and staff of the Clinical Research Center at Emory University Hospital where the study was conducted, and to the staff of the Radioimmunoassay Laboratory for assistance in the laboratory determinations. Furthermore, they gratefully acknowledge the h
250

malized serum glucose showed a steady fall in serum glucose. This finding may help to predict the glycemic response to long-term calorie restriction. Carbohydrate tolerance improves in obese diabetic (NIDDM) women after 3 days of fasting, in contrast to the impairment of glucose tolerance seen in lean or obese nondiabetic subjects after fasting. The improvement in carbohydrate tolerance is probably due to a combination of reduced hepatic glucose output and improved insulin action rather than enhanced sulin secretion. KEY INDEXING TERMS: Diabetes Mellitus; Fasting; Glucose Tolerance; Obesity; NIDDM; Insulin; C-peptide; Lactate. [Am J Med Sci 1990; 299(4):250-256.]

C

alorie deprivation has profound metabolic effects that vary with the severity and duration of the caloric deficit.1-9 Furthermore, the metabolic response to a short period of fasting (2-4 days) varies in different groups of subjects. 2 ,s In lean subjects with normal carbohydrate tolerance, short-term fasting results in marked impairment of glucose tolerance. s In obese nondiabetic subjects, short-term fasting also produces a deterioration in glucose tolerance, but less than is seen in lean nondiabetic individuals. 8 In contrast, obese subjects with noninsulin-dependent diabetes (NIDDM, type II diabetes mellitus) respond to long-term caloric deprivation and weight loss (fasting for more than 7 days or partial food restriction for weeks) with a lowering of blood glucose and markedly improved carbohydrate tolerance as assessed by oral glucose or a meal challenge. 3 ,4 Such findings support the use of hypocaloric diet as a major modality of treatment in obese patients with type II diabetes. However, little is known about the effect of short-term fasting (ie, 2-4 days) on carbohydrate tolerance in obese patients with type II diabetes. The present study was designed to provide this information.

April 1990 Volume 299 Number 4

Watts and DIGirolamo

Table 1. Characteristics of Study Subjects*

Age (Years)

Duration of Diabetes Mellitus (Years)

Initial Fasting Serum Glucose (mmol/L)

HbA lc (%)

1 2 3 4 5 6 7 8 9 10

60 51 41 33 59 54 41 54 58 61

7 6 12 3 10 6 1.5 18 9 3

14.2 19.0 14.2 11.7 20.0 12.3 18.9 14.0 13.8 13.6

Mean SEM

51 3

7 1.7

15.2 0.9

BMI (kg/m2)

Body Weight (kg)

Weight Loss During 3 Days' Fast (kg)

10.5 10.5 9.1 10.9 6.4 7.9 8.8 6.8 10.6 11.5

29.4 29.8 31.3 32.4 32.5 32.5 35.7 36.7 37.4 37.4

69.7 77.2 85.2 87.4 91.5 91.8 92.5 93.9 94.6 108.2

2.0 3.3 1.6 5.8 3.0 2.9 4.5 5.1 1.9 4.8

9.3 0.6

33.5 0.9

89.2 3.1

3.5 0.5

* All black women with NIDDM.

Materials and Methods Subjects. Ten obese black women with type II dia-

betes mellitus were studied in the Clinical Research Center of Emory University Hospital. None had received insulin or oral hypoglycemic agents in the previous 6 months, and none showed any significant change in body weight (>5 lbs) for at least 6 months prior to the study. Their body weight was between 150% and 200% of ideal based on the Hamwi formula lO and body mass index (kg/m 2 ) was between 29 and 38. Characteristics of these women are shown in Table 1. For 3 days prior to admission, the subjects were instructed to follow a weight-maintenance diet containing at least 300 g of carbohydrate. All subjects signed an informed consent. The protocol had been approved by the Human Investigations Committee of Emory University. Experimental Design. Subjects were admitted and began fasting after 8:00 PM on the day of admission. Beginning at 8:00 AM the following morning, resting metabolic rate was measured by indirect calorimetry. Then an indwelling venous catheter was placed and maintained patent with dilute heparin solution. After blood was obtained for basal parameters (glucose, insulin, c-peptide, lactate, free fatty acids, and beta-hydroxybutyrate), the subjects were given a test meal of 800 kcalories (50% carbohydrate, 15% protein, and 35% fat). Measurements for calorimetry and additional blood samples were obtained at 30, 60, 120, 240, and 360 minutes after ingestion of the meal. After the first meal tolerance test the subjects were fasted for 3 consecutive days. The meal tolerance test was repeated on the morning of day 4. During the fast, at least 2000 mL of noncaloric fluids per day were given orally. Blood was obtained for glucose measurement at 8:00 AM, 2:00 PM, and 10:00 PM daily; these glucose measurements were done on capillary blood

THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES

samples with Chemstrips® and AccuChek® meters (Bio-Dynamics, Indianapolis, IN). Metabolic Determinations. Blood samples were drawn from an indwelling venous catheter for glucose, insulin, c-peptide, lactate, free fatty acids, and betahydroxybutyrate. One mL of blood for lactate measurement was added to 2 mL of8% perchloric acid. Measurement of Energy Expenditure. After stirring and centrifugation at 4 C, 2000 X g for 15 min, the supernatant was aspirated and lactate measured with an enzymatic kit reagent (Boehringer Mannheim, FRG). The rest of the blood was allowed to clot at room temperature for 20-30 min and serum collected by centrifugation. Serum glucose was measured by an automated glucose oxidase method (Kodak Ektachem®, Rochester, NY); serum insulin and c-peptide were measured by radioimmunoassay (Immunophase, Corning Medical, Medfield, MA); free fatty acids were determined using a colorimetric method l l ; and betahydroxybutyrate was determined at National Medical Laboratories (Willow Grove, PA) using a gas chromatography method. Resting metabolic rate (RMR) was determined by indirect calorimetry,12 Two or three basal measurements were performed on each subject and the results were averaged. Once RMR was established, the subject was fed the test meal. The thermic effect of food was calculated from oxygen consumption measured every 30 minutes for 6 hours after the meal. Statistical Analysis of Data. All data in the text and figures are expressed as mean ± standard error. Statistical significance was determined by paired two-tailed student's test or by analysis of variance, as appropriate. 13 P < 0.05 was chosen to indicate a statistical difference between two groups of observations. 0

251

Carbohydrate Tolerance and Fasting In Obese NIDDM

Results

::J"

No significant adverse consequences of the 3 days' fasting were noted in any of the patients. A few patients complained of hunger, weakness, and slight dizziness. Weight Loss During Fasting. Variable but consistent weight loss was observed (Table 1), ranging from 1.65.8 kg (mean 3.5 kg). The amount of weight loss did not correlate with initial weight, body mass index, HbA 1c , initial or final serum glucose.

~ 22 .2

25

-,----------------:A"""T'" 450

~

g.,

E

19.4

400 350

en

16.7

::::,

w

o

g CI

13.9 11.1

~

8.3

...J

;:)

a: w

en

Fasting Serum Glucose and Carbohydrate Tolerance. Figure 1 shows the improvement in basal serum

glucose after fasting 3 days, from 15.2 ± 0.9 to 7.5 ± 0.7 mmol/L (273 ± 17 to 135 ± 13 mg/dL) (mean ± SEM), (p < 0.001), an average fall of 52%. Also, the glucose response to the test meal after 3 days' fasting showed significant improvement at each point tested (Figure 1, top panel). As shown in Table 2, when the areas under the glucose tolerance curves were compared, the area for glucose was 31 % lower after 3 days' fasting than that obtained after the overnight fast (p < 0.001). Serum Insulin and C-Peptide Levels. The mean serum insulin concentration was 160 pmol/L (22 uU/mL) after the overnight fast and was unchanged after the 3 days' fast. The magnitude of the insulin response to the test meal (area under the curve) was not changed after fasting in spite of the lower serum glucose (Figure 1, middle panel and Table 2). The peak insulin response was, however, slightly delayed in the meal tolerance test after fasting (Figure 1, middle panel). Mean serum c-peptide concentration was 2.2 ± 0.21 ng/mL after the overnight fast and 1.8 ± 0.16 ng/mL after 3 days' fast, not significantly different. However, during the test meal challenge, the c-peptide response (area under the curve) increased 24 % after 3 days' fast when compared to overnight fast and reached borderline statistical significance (0.1 < p > .05) (Figure 1, bottom panel and Table 2). Blood Lactate. Basal lactate concentration was not affected by the 3 days' fast. A marked difference, however, was observed when the blood lactate levels during the first test meal (before fasting) were compared to lactate levels during the second test meal (Fig 2, top panel). During the first test meal, lactate levels rose from 0.9 mM to 2.0 mM at 1 hour and then slowly declined to basal levels by 4 hours. In contrast, during the second meal, lactate levels remained unchanged throughout. The area under the second lactate curve was 36% lower than the first (p < 0.02) (Table 2) . Serum Free Fatty Acids and Beta-Hydroxybutyrate.

Free fatty acid levels rose slightly with the 3 days' fast, but no significant differences were observed in basal levels or during the meal tolerance tests (before and after fasting) (Figure 2, middle panels). Basal levels of beta-hydroxybutyrate increased three-fold after fasting and the area under the second test meal curve

252

::J"

~,b--- -lr--_ ------i-______ _

, ,,

300

~

250 200

U ::J

150

~

2

4

6 ::J"

z

290

40

20

145

01---.- - , - - - . - - . - - , - - - + 0

o ::J" ~

5 .5

w

4 .5

i=

4

W

U :!E ;:) a: w en

~ ::J

w

w

11.

z

a:

a:

11.

:::i

en

en

a

Z ::J

;:)

~

E

........ ::J

::t..

Co

en

::J

w

01 o

60

~ ::J

CI

en

430

z

...J

a:

5.6

::::, o E :::i

E w

4

2

6

C

5

3 .5 3 2 .5-

::J"

5.5 5 4.5 4 3.5

3

2 1.5 2

4

E

g., c: w

c i=

fu11. U

2.5 2

:!E

1.5

w

01---.--,---.--.--,----+0

o

en

;:)

a:

en

6

HOURS OF STUDY

•

•

o---.()

Day 1 Day 4

Figure 1. Glucose (A), insulin (B), and c-peptide (C) responses to the test meal on day 1 (following an overnight fast, e - e) and day 4 (after 3 days of fasting, 0 - 0). Values shown are means ± SEM of repeated observations in 10 patients.• p < .05; •• p < 0.001 for the difference between respective time points in the repeated test.

was 113% greater than the first (p < 0.005) (Figure 2, bottom panel). Metabolic Rate. Resting metabolic rates averaged 60.4 ± 28 kcal/hr before fasting and 59.4 ± 3.5 after the 3 days' fast, not significantly different. The areas for RMR under the meal test curves (Table 2) were also similar, indicating that the short-term fast did not affect either the resting metabolic rate or the thermic effect of food.

April 1990 Volum~ 299 Number 4

Watts and DiGirolamo

Table 2. Areas under the Curves* Day 1 Glucose (mmol.L- I /6 hr) Insulin (pmol·mL -1/6 hr) C-peptide (ng.mL- I /6 hr) Lactate (mEq.L- I /6 hr) FFA (uEq.L- I /6 hr) B-OHB (Jlg.mL- I /6 hr) Metabolic Rate (kcal.hr- I /6 hr)

112.7 ± 1930 ± 22.5 ± 8.5 ± 1682 ± 297 ± 275.2 ±

8.2t 145 2.0 1.0 339 62 15.9

Day4 78.3 ± 1940 ± 27.9 ± 5.4 ± 1990 ± 634 ± 251.2 ±

4.8 130 2.3 0.5 258 78 16.2

Percent Change:j:

P§

-31 0 +24 -36 +18 +113 -9

<0.001

NS 0.05-0.1 borderline <0.02

NS <0.005

NS

* Response to test meal on Day I (overnight fast) and Day 4 (after 3 days' fast). t Values shown are means ± SEM for repeated observations at day I and 4 in 10 patients. FFA = free fatty acids; B-OHB = Beta-hydroxybutyrate. :j: % Change = mean of day I - mean of day 4/mean of day I X 100. § P value indicates the difference between mean values on day I and day 4. NS = nonsignificant.

Differential Glycemic Response to Fasting in Two Subgroups of Patients. While all patients showed an im-

provement in serum glucose after 3 days of fasting, five showed normalization of fasting glucose (Group A), mean serum glucose 5.7 ± 0_3 mmol/L (103 ± 6 mgjdL), whereas in the remaining five patients the ending serum glucose was higher, 9.7 ± 0.5 mmol/L (174 ± 9 mgjdL) (Table 3). The mean relative fall in serum glucose at the end of 3 days' fast was twice as great in Group A patients, 65% versus 33% in Group B (p < 0.01). Initial and final serum insulin levels were the same in the two siIbgroups. There was no difference between the subgroups in any of the other variables measured, including initial serum glucose after overnight fast, body mass index, age, duration of diabetes, or HbA 1c ' Comparing these subgroups, an interesting pattern was observed when serum glucose, measured every 8 hours during fasting, was plotted versus time (Figure 3)_ Patients in Group A showed a steady decline in serum glucose over the 3 days of fasting. In Group B, despite a consistent overall downward trend, serum glucose concentrations increased 15-25% overnight during each of the 3 nights of observation. Discussion

Carbohydrate tolerance improved significantly after 3 days of fasting in our study subjects. The improvement in carbohydrate tolerance in these obese NIDDM patients following short-term fasting is somewhat surprising, since short-term fasting has been shown to impair carbohydrate tolerance in both lean and obese nondiabetic subjectss.9 ; in fact, realization that fasting or chronic undernourishment impairs carbohydrate tolerance in normal subjects has led to the recommendation for a high caloric and carbohydrate intake for at least 3 days prior to any carbohydrate tolerance test. 9 •14 The present study and published observations therefore suggest that the effect

THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES

of short-term fasting on carbohydrate tolerance is a spectrum, the results depending on the characteristics of the study subjects: (a) marked impairment of carbohydrate tolerance in lean nondiabetic individuals; (b) moderate impairment in obese nondiabetic subjects; but (c) definite improvement of carbohydrate tolerance in obese diabetics (NIDDM). Table 4 summarizes the information available on the changes resulting from short-term fasting in different types of subjects. The findings in our subjects after short-term fasting are remarkably similar to results obtained by Henry et a1 3 after long-term hypocaloric diet in obese type II diabetic patients who lost a significant amount of weight (average 19 kg) over 6-18 months. Also, subjects studies by Greenfield et a14 after 10 days fast showed improvement in basal glucose and glucose tolerance. Thus, it appears that in obese NIDDM patients, improvement in basal glycemia and carbohydrate tolerance can be obtained by either acute, total starvation for as short as 3 days, as shown in our subjects, or by partial, long-term caloric deprivation. 3 The improvement in carbohydrate tolerance in our study subjects occurred despite no change in basal or postprandial serum insulin levels. Similarly, the NIDDM subjects studied by Henry et a1 3 showed improved carbohydrate tolerance but no increase in serum insulin levels after weight loss and long-term calorie restriction. The lack of an increase in postprandial serum insulin in obese diabetic patients after fasting is in contrast to the increase in postprandial insulin levels after fasting in both lean and obese nondiabetic subjects. s As fasting has different effects on carbohydrate tolerance depending on characteristics of the study subjects, the effect of fasting on mealstimulated insulin secretion also differs: (a) marked increase in lean, nondiabetic individuals; (b) moderate increase in obese nondiabetic subjects; but (c) no change in postprandial serum insulin levels in obese

253

Carbohydrate Tolerance and Fasting In Obese NIDDM

2 1.8 :::r ..... 1.6 15 1.4 E 1.2 E w

I-

< I-

0

:3

A

-! -0.----6------ __y --------

1

00

0 .8 0 .6 0.4 0.2 0

0

2

0

:r .....

600

~

500

'0

en 0 U

6

4

B

400

<

>

1= < ~

W W

a:: ~

200 100 0

2 280<; ••

C

\ \

240

:r E .....

\

160

al

120

g

8 00 4

x

\ \

200

Cl

.5

6

4

b,

00

'l. - - _

'Lr

•• \5-_ ____ l: --~-------

-...

tIP )

.x:

0~--_r--~~--_r--~----~---4

o

2

4

6

lin secretion was sufficient to explain the dramatic improvement in carbohydrate tolerance. How fasting influences carbohydrate tolerance is not clear. In type II diabetic subjects, fasting hyperglycemia is proportional to hepatic glucose production, 17,18 and fasting seems to enhance insulin's ability to suppress endogenous glucose production. 19 It follows that a reduction in fasting serum glucose would result from a reduced rate of hepatic glucose production as a result of increased sensitivity of the liver to insulin, or decreased delivery of metabolic substrates, or both. Improvement in peripheral sensitivity to insulin with increased glucose use is also possible, but somewhat unlikely in view of the inhibitory effects on glucose use produced by fasting. Further studies will be needed to clarify these important points. Fasting did not significantly alter resting metabolic rate, the thermic effect of food, or levels of free fatty acids. As expected, beta-hydroxybutyrate was higher after fasting. In response to the initial test meal, serum lactate levels increased as expected20 but showed no change in response to the test meal following 3 days' fasting. This may reflect either reduced lactate production in peripheral tissues or, more likely, increased lactate use for restoration of hepatic glycogen 21 following resumption of eating after fasting, or both. This intriguing finding merits further investigation. We were able to separate our patients into two subgroups, differing in the degree of fall of serum glucose after fasting. There was an overnight rise in serum glucose (the dawn phenomenon 22-27 ) only in Group B patients, those with the least glycemic improvement.

HOURS OF STUDY

100 •

•

0-- 0

O
4

W

z

1

90

UJ-

Figure 2. Metabolic fuels in response to the test meal on day 1 (following an overnight fast,. - .) and day 4 (after 3 days of fasting,O - 0); (A) lactate, (B) free fatty acids, and (C) beta-hydroxybutyrate. See Figure 1 legend for details. **p < 0.001.

diabetics (NIDDM) (Table 4). The lack of a postprandial insulin increase in obese NIDDM subjects after 3 days' fast, as compared with obese nondiabetics, may be due to insufficient time for pancreatic B-cell recovery from glucotoxicity.15 Postprandial c-peptide levels in our subjects were significantly higher after the 3 days' fast. Hepatic extraction of insulin increases after fasting. 16 It is probable that there was some increase in meal-stimulated insulin secretion in our subjects after fasting that was not reflected in the serum insulin levels. Alternatively, clearance of c-peptide may be reduced by fasting. Regardless, it is unlikely that an increase in insu-

254

CIJ..J

O~

80

::;)111

70

/

..Ju.

,, ,,

Y

U~

"0 ::it::;)z

, , ,,, ,

'~

60

a:UJ 50 UJU ClJa: UJ 40 Il.

30

•

,

' 1,

.

,,' "i,

-, ,

.

"

~' I

••

,~

D--OGroup A ~---0 Group 8

a

20

40

60

80

HOURS OF STUDY Figure 3. Relative serum glucose change during 3 days of fasting in good responders (Group A) and poor responders (Group B); this was calculated by dividing the mean glucose value at each time point by the initial value (8 AM) then multiplying by 100. Note the steady decline in glucose in Group A and the overnight increases in glucose in Group B. *p < 0.05; **p < 0.001 for the difference between the two groups at the respective time points.

April 1990 Volume 299 Number 4

Watts and DiGirolamo

Table 3. Individual Serum Glucose Values (mg/dL) 0800 h on Day 1 (Following an Overnight Fast) and Day 4 (After 3 Days of Fasting) in Good Responders (Group A) and Poor Responders (Group B) Group A

Group B

Patient Number

Day 1

Day 4

Percent Change

Patient Number

Day 1

Day 4

Percent Change

(1) (2) (3) (4) (5)

14.2 19.0 14.2 11.7 20.0

4.9 5.3 5.2 4.9 7.0

58 72 63 65 65

(6) (7) (8) (9) (10)

12.3 18.9 14.0 13.8 13.6

8.2 11.2 8.8 10.4 9.7

33 41 37 24 29

Mean SEM

15.8 1.4

5.7 0.3

65

14.5 1.0

9.7 0.5

33

Percent change was calculated by subtracting glucose on day 4 from glucose on day I, then dividing by glucose on day I and multiplying by 100. Patient number in parentheses refers to Table I .

Final serum insulin levels were not significantly different between the subgroups. The explanation for the striking nocturnal rise in glucose in Group B subjects may relate to differences in insulin resistance between these subgroups or to differences in the levels of counterregulatory hormones (GH, cortisol, and so forth). More work is needed to determine the mechanism of this process and its significance. We have previously shown that many obese, noninsulin-dependent diabetic patients fail to show an improvement in blood glucose levels despite significant weight loss.28 It seems likely that patients who show the greatest fall in blood glucose levels after a shortterm fast will have a better glycemic response to mild or moderate weight loss with long-term diet therapy. The presence or absence of an overnight rise in serum glucose (the dawn phenomenon, demonstrated during an outpatient period of fasting, could provide a simple way of identifying patients as belonging to group A (good responders) or group B (poor responders), and might permit prediction of the glycemic response to therapy with hypocaloric diet. These preliminary observations need to be confirmed and expanded before such a conclusion is fully accepted. In summary, we have shown that a short-term fast in obese subjects with NIDDM improves not only basal glucose but also carbohydrate tolerance. This is in contrast to the impaired carbohydrate tolerance Table 4. Changes in Glucose Tolerance and Insulin Secretion after Short Term Fasting in Three Groups of Subjects

Nondiabetic Lean Obese NIDDM Obese

Glucosf' Tolerance

Insulin Secretion

Impaired Slight impairment

Increased Slight increase

Improved

No change

THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES

seen after fasting in nondiabetics. The improvement in glucose tolerance occurs with no increase in fasting or postprandial serum insulin levels. The observation that a short-term fast abolishes the meal-related increase in blood lactate level is of interest and merits further work. The finding that subjects with the least glycemic improvement after fasting show overnight rises in serum glucose level is intriguing, deserves confirmation and amplification, and could form the basis for rapid characterization of the anticipated glycemic response to hypocaloric diet. References 1. Fink G, Gutman RA, Cresto JC, Selawry H, Levine R, Recant L: Glucose·induced insulin release patterns: Effect of starva· tion. Diabetologia 10:421-425, 1974. 2. Cahill GF Jr, Herrera MG, Morgan AP, Soeldner JS, Steinke J, Levy PL, Reichard GA Jr, Kipnis DM: Hormone·fuel inter· relationships during fasting. J Clin Invest 45:1751-1769,1966. 3. Henry RR, Wallace P, Olefsky JM: Effects of weight loss on mechanisms of hyperglycemia in obese non-insulin-dependent diabetes mellitus. Diabetes 35:990-998, 1986. 4. Greenfield M, Kolterman 0, Olefsky JM, Reaven GM: The effect of ten days fasting on various aspects of carbohydrate metabolism in obese diabetic subjects with significant fasting hyperglycemia. Metabolism 27 (suppI2):1839-1852, 1978. 5. Jackson 1M, McKiddie MT, Buchanan KD: Effect of fasting on glucose and insulin metabolism of obese patients. Lancet 1: 285-287, 1969. 6. So Iter M, Sekso M: Effect of fasting on posthyperglycemic glucose homeostasis in obesity-experimental model for reactive hypoglycemia. Exp Clin EndocrinoI81:33-40, 1983. 7. Genuth SM: Effects of prolonged fasting on insulin secretion. Diabetes 15:798-806, 1966. 8. Goschke H: Mechanism of glucose intolerance during fasting: Differences between lean and obese subjects. Metabolism 26: 1147-1153,1977. 9. Himsworth HP: The dietetic factor determining the glucose tolerance and sensitivity to insulin of healthy men. Clin Sci 2: 67-94, 1935. 10. Hamwi GJ: Therapy: Changing dietary concepts. in Diabete.~ Mellitus, Diagnosis and Treatment. New York, American Diabetes Association, Inc, 1964, p 73-7R. 11. Itaya K: A more sensitive and stable calorimetric determination of free fatty acids in blood. J Lipid Res 18:663-665,1977.

255

Carbohydrate Tolerance and Fasting in Obese NIDDM

12. Hill JO, DiGirolamo M, Heymsfield SB: Thermic effect of food after ingested versus tube-delivered meals. Am J Physiol 248: E370-E374,1985. 13. Winer BJ: Statistical Principles in Experimental Design. New York, McGraw-Hill, 1971. 14. National Diabetes Data Group: Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 28:1039-1057, 1979. 15. Unger RH, Grundy S: Hyperglycaemia as an inducer as well as a consequence of impaired islet cell function and insulin resistance: Implications for the management of diabetes. Diabetologia 28:119-121, 1985. 16. Henry RR, Brechtel G, Griver K: Secretion and hepatic extraction of insulin after weight loss in obese noninsulin-dependent diabetes mellitus. J Clin Endocrinol Metab 66:979-986, 1988. 17. Bowen HF, Moorhouse JA: Glucose turnover and disposal in maturity-onset diabetes. J Clin Invest 52:3033-3045,1973. 18. DeFronzo RA, Ferrannini E, Koivisto V: New concepts in the pathogenesis and treatment of noninsulin-dependent diabetes mellitus. Am J Med 74:52-81, 1983. 19. Jensen MD, Milse JM, Gerich JE, Cryer PE, Haymond MW: Fasting enhances insulin's ability to suppress endogenous glucose production. Clin Res 34:927a (Abstract), 1986. 20. Doar JWH, Cramp DG, Maw DSJ, Seed M, Wynn V: Blood pyruvate and lactate levels during oral and intravenous glucose tolerance tests in diabetes mellitus. Clin Sci 39:259-269, 1970.

256

21. Foster DW: From glycogen to ketones-and back. Diabetes 33: 1188-1199, 1984. 22. Deckert T, Lorup B: Regulation of brittle diabetes by a preplanned insulin infusion programme. Diabetologia 12:573-579, 1976. 23. Clarke WL, Haymond MW, Santiago .JV: Overnight basal insulin requirements in fasting insulin-dependent diabetics. Diabetes 29:78-80, 1980. 24. Schmidt MI, Hadji-Georgopoulos A, Rendell M, Margolis S, Kowarski A: The dawn phenomenon, an early morning glucose rise: Implications for diabetic intraday blood glucose variation. Diabetes Care 4:570-585, 1981. 25. Bolli GB, Gerich JE: The "dawn phenomenon" -a common occurrence in both non-insulin-dependent and insulin-dependent diabetes mellitus. N Engl J Med 310:746-750, 1984. 26. Schmidt MI, Lin QL, Gwynne JT, Jacobs S: Fasting early morning rise in peripheral insulin: Evidence of the dawn phenomenon in nondiabetes. Diabetes Care 7:32-35, 1984. 27. Campbell PJ, Bolli GB, Cryer PE, Gerich JE: Pathogenesis of the dawn phenomenon in patients with insulin-dependent diabetes mellitus: accelerated glucose production and impaired glucose utilization due to nocturnal surges in growth hormone secretion. N Engl J Med 312:1473-1479,1985. 28. Spanheimer RG, DiGirolamo M, Watts NB, Musey VC, Siddiq YK, Gebhart SSP, Goldstein S, Davidson JK, Phillips LS: Glycemic response to weight loss in obese diabetic patients. Clinical Res 34:198A (Abstract), 1986.

April 1990 Volume 299 Number 4