Counterregulatory hormonal responses to hypoglycemia during pregnancy

Counterregulatory Hormonal Responses to Hypoglycemia During Pregnancy BARAK M. ROSENN, MD, MENACHEM AND TARIQ A. SIDDIQI, MD

MIODOVNIK,

Objective: To evaluate the counterregulatory responses to insulin-induced hypoglycemia in healthy women and in women with insulin-dependent diabetes during pregnancy and in the nonpregnant state. Methods: Hypoglycemia was induced using the hypoglycemic clamp technique in 17 women with insulin-dependent diabetes and in ten healthy controls, both in the nonpregnant state (study 11, at 24-28 weeks’ gestation (study 21, and at 32-34 weeks’ gestation (study 31. Plasma glucose concentrations were decreased to 60 mg/dL and maintained at this level for 1 hour. Blood samples were drawn every 15 minutes to measure epinephrine, glucagon, growth hormone, and cortisol concentrations. Statistical analyses compared counterregulatory responses between women with and without diabetes, and between the pregnant and nonpregnant state. Results: Women with diabetes had significantly diminished peak epinephrine responses to hypoglycemia compared with controls (mean f standard error of the mean [SEMI): 52 % 11 versus 191 f 42 pg/mL in study 1, 30 f 9 versus 102 f 47 pg/mL in study 2, and 38 f 10 versus 148 r: 38 pg/mL in study 3 (P < .051. Their responses during pregnancy were also diminished compared with their own nonpregnant epinephrine responses. Women with diabetes also had no recognizable cortisol or glucagon responses to hypoglycemia, and in healthy controls the glucagon responses were significantly diminished during pregnancy compared with their own nonpregnant responses. In both groups, growth hormone responses (mean 2 SEM) diminished progressively during pregnancy from study 1 (14.6 f 2.5 and 12.5 f 5.2 ng/mL) to study 2 (4.4 f 1.1 and 7.3 f 2.7 ng/mL) to study 3 (2.5 f 0.9 and 4.4 f 2.3 ng/mL) in women with diabetes and in controls, respectively. Conclusion: Counterregulatory epinephrine and growth hormone responses to hypoglycemia are diminished in women with insulin-dependent diabetes during pregnancy. This may be due, in part, to an independent effect of

From the Dioision of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Pcrinatal Research Institute, Uniaersity of Cincinnati College of Medicine, Cincinnati, Ohio. This work was supported by National lnstitufes ofHealth (NIH) grant HD 11725, Diabetes in Pregnancy (Program Project Grant), and by NIH grant RR 08048.

568

0029.7844/96/$15.00 SSDI 0029-7844(95)00495-5

MD, JANE C. KHOURY, MSc,

pregnancy, contributing to the increased incidence of hypoglycemia in these patients during pregnancy. (Obstet Gynecol 2996;87568-74)

Intensive insulin therapy has become a widely accepted mode of treatment for pregnant women with insulindependent (type I) diabetes mellitus. The goal of such therapy is to improve glycemic control, thereby improving pregnancy outcome by decreasing the rate of early pregnancy loss and major malformations, as well as reducing the incidence of obstetric and neonatal complications.‘f2 In subjects with type I diabetes, intensive insulin therapy increases the risks of hypoglycemia, a phenomenon associated with diminished hormonal counterregulatory responses and hypoglycemia unawareness.” The purpose of this study was to characterize the acute hormonal counterregulatory responses to insulin-induced hypoglycemia in healthy pregnant women and in pregnant women with type I diabetes, compared with their own responses during the nonpregnant state.

Materials and Methods The study was approved by the Institutional Review Board and all participating subjects signed an informed consent statement. Seventeen women with type I diabetes (study group) and ten healthy women (control group) underwent hypoglycemic clamp studies to evaluate counterregulatory hormonal responses to hypoglycemia. Women with diabetes were recruited from our diabetes in pregnancy program. Control subjects were healthy women recruited from our normal obstetrics and gynecology outpatient clinics, and all had negative screening tests for diabetes during pregnancy and normal pregnancies with normal outcomes. The purpose and details of the study were explained to the potential subjects, and those who agreed to participate in the

Obstetrics

f~ Gynecology

study gave written, informed consent. The major factor determining whether the patient was willing to participate in the study appeared to be her ability to accept the discomfort involved in the insertion of multiple intravenous (IV) lines in the course of the three studies. Originally, 20 women with diabetes were recruited, but three subjects dropped out without completing the study. The management of study patients with diabetes has been described elsewhere.4 In brief, women with type I diabetes were enrolled in the diabetes in pregnancy program before 9 weeks’ gestation and were managed with intensive insulin therapy that included two to four daily injections of short- and intermediate-acting insulin, four to six daily measurements of capillary glucose concentrations with memory reflectance glucometers, and weekly clinic visits throughout pregnancy. Glycohemoglobin concentrations were measured every 4 weeks throughout pregnancy. The goals of glycemic control were a preprandial blood glucose concentration less than 100 mg/dL and a 90-minute postprandial concentration less than 140 mg/dL. All women with diabetes were tested to exclude diabetic autonomic neuropathy. These tests included determination of R-R interval on electrocardiogram during quiet breathing, determination of blood pressure and heart rate before and after the assumption of the upright position, and measurement of catecholamine concentrations before and after the assumption of the upright position. Each subject from the study and control groups underwent three hypoglycemic studies: once in the nonpregnant state (study 1) and twice during pregnancy, at 24-28 weeks (study 2) and at 32-34 weeks’ gestation (study 3). Most subjects in both groups were recruited during pregnancy, and therefore had their nonpregnant study performed after pregnancy (up to 6 months postpartum). All studies were performed after an overnight fast of 8-12 hours. Women with diabetes were admitted to the Clinical Research Center on the evening before the study, and the evening insulin dose was omitted. Subjects remained supine throughout the night and blood glucose concentrations were determined every hour from a sample drawn from an indwelling IV catheter. A varying-rate infusion of regular insulin (150 U/500 mL 0.9% NaCl) was used to maintain the blood glucose concentration at 80-100 mg/dL. On the morning of the study, subjects were connected to the Miles Biostator (Miles Laboratories, Inc., Elkhart, IN), a computerized apparatus that continuously samples minute amounts of venous blood and measures the glucose concentration. In the hypoglycemia mode, the Biostator infuses minute-by-minute calculated volumes

VOL.

87, NO.

4, APRIL

1996

of insulin and 20% dextrose to lower the subject’s blood glucose concentration to a predetermined level in a controlled fashion. That level is then maintained by infusing a fixed rate of insulin and varying rates of 20% dextrose according to built-in operating algorithms. An 18-gauge double-lumen catheter was inserted into a vein in the left hand or forearm for continuous sampling by the Biostator. Additional blood samples were obtained from a different vein in the same arm, and solutions were infused by the Biostator into a vein in the contralateral arm. The left arm was placed in a heating device to arterialize venous blood. After an initial stabilizing period of 30 minutes, two baseline blood samples were drawn 15 minutes apart, for baseline levels of counterregulatory hormones. The Biostator was then activated in operating mode 3 (varying rates of glucose and insulin infusion) to lower plasma glucose concentrations to 60 mg/dL over approximately 1 hour, and this level of hypoglycemia was maintained for 60 minutes, using operating mode 9 (constant rate of insulin and varying rates of glucose infusion). Plasma glucose concentrations were measured every 5-15 minutes to verify Biostator glucose sensor accuracy, using the YSI Model 2300 STAT (Yellow Springs Instruments, Inc., Yellow Springs, OH). Blood samples were drawn every 15 minutes, and at 30 and 60 minutes after completion of the l-hour hypoglycemic phase. All samples were placed on ice, then separated and stored at -20C for later analysis. Fetal heart rate and uterine activity were monitored continuously during all studies in pregnant women. Plasma epinephrine was extracted in 0.1 mol/L perchloric acid and measured by high-pressure liquid chromatography (electrochemical detection), using lOO-PL samples.” The limit of detection is 10 pg/mL, and inter- and intra-assay coefficients of variation are less than 10%. Plasma glucagon was measured by radioimmunoassayh; the limit of detection is 12 pg/mL, and the inter- and intra-assay coefficients of variation are 12 and 4%, respectively. Plasma cortisol was measured using a specific enzyme immunoassay7; the interand intra-assay coefficients of variation are 13.8 and 7.2%, respectively. Growth hormone was measured by standard radioimmunoassay.R Baseline hormonal concentrations for each subject were represented by the average of the two baseline values. The peak hormonal response for each subject was determined from the highest hormonal concentration during the l-hour hypoglycemia period minus the basal hormonal concentration. From these values, the mean peak hormonal response was determined for each group, for each hormone, and for each study. Peak hormonal responses were then compared between the study and the control groups, and also between the

Rosenn

et al

Counterregulatory

Hormones

569

130 120

120 -

110

110 -

100

100 -

90

90

80

80

70

70 -

80

Figure (mean

80 -

1. Glucose 2 standard

mean) during hypoglycemic studies in pregnant women diabetes and controls. IDDM

50 40 -

sulin-dependent Minutes6

0 30 80 90120150180

non-pregnant . IDDM

0 30 80 90120150180

0 30 80 90120150180

24-28 wks

32-34 wks

Results There were no significant demographic differences between women with diabetes and controls: mean (5 standard deviation [SD]) age was 26.2 + 4.7 and 28.2 + 5.8 years, mean gravidity was 2.4 ? 1.3 and 2.5 i 1.4, mean parity was 0.7 ? 0.6 and 1.0 ? 0.8, and 94 and

oh--0

MinutesB

30 80 90120150180

180

570

Rosenn et al

Counterregulatory

mellitus.

100% were white, respectively. The mean (5 SD) duration of disease in women with diabetes was 15.7 + 7.1 years; the mean (+ SD) glycohemoglobin A, concentrations were 9.2 t 2.6% in the nonpregnant phase, 6.6 ? 0.87% at 24 weeks, and 6.7 2 0.9% at 34 weeks’ gestation; the normal range for nondiabetic subjects in our laboratory is 5.5-8.5%. Fetal heart rates were normal in all pregnant subjects (with or without diabetes), and no abnormalities of uterine activity were observed during any of the studies. Figures 1-5 depict mean (? standard error of the mean [SEMI) plasma glucose, epinephrine, glucagon, growth hormone, and cortisol concentrations in women with diabetes and controls in each of the three studies. Concentrations are depicted at baseline, during the 60 minutes before the hypoglycemic clamp, during the 60-minute hypoglycemic clamping period, and 60 min-

180

Figure

2. Counterregulatory

epi-

nephrine responses to hypoglycemia (mean t standard error of the mean) in women with diabetes and

80

controls. 1DDM = insulin-dependent diabetes mellitus. 0 B

non-pregnant . IDDM

diabetes

clamp with = in-

n Controls

pregnant and nonpregnant state for each group. In addition, mean hormonal concentrations were plotted for each study, by group, and at each time point. Statistical analyses used SAS software (SAS, Inc., Cary, NC). Chi square, Fisher exact test, and two-tailed t test were used as appropriate. Repeated measures analysis of variance was used for each group, with respect to each of the hormonal responses over the three studies, to determine possible differences in responses.

7 -A-

concentrations error of the

0 30 80 90120150180

24-28 wks

B

0 30 80 90120150180

32-34 wks

n Controls

Hormones

Obstetrics

& Gynecology

220

r

200 t 180 1

180

160

g

180

c 0

120

120

g

100

100

140

Figure 3. Counterregulatory glucagon responses to hypoglycemia (mean t standard error of the mean) in women with diabetes and controls. IDDM = insulin-dependent diabetes mellitus.

80

2

60

0 -

60 140

i

80

60 40

20

20f-

40 t, Minutes B 0 30 60 90120150180

non-pregnant 0 IDDM

utes after the clamp. The highest concentrations in these figures represent the peak of average concentrations, which are generally lower than the average of peak levels because individual subjects’ responses peak at different time points during the study. Women with diabetes had significantly higher baseline plasma glucose concentrations compared with controls in study 1 (101 ? 6 versus 89 ? 4 mg/dL respectively, mean i- SEM; P < .05). However, by 30 minutes, glucose concentrations were virtually identical in both groups (Figure 1). Plasma glucose concentrations were lowered and maintained at 60 -C2 mg/dL in a similar manner in both groups for 60 minutes during all studies. The peak epinephrine responses above baseline were diminished in women with diabetes compared with controls, both in the pregnant and nonpregnant phases

0 30 60 90120150160

24-28 n

2oJAd B

wks

0 30 60 90120150160

32-34 weeks

Controls

(mean + SEM): 52 + 11 versus 191 + 42 pg/mL in study 1 (P < .05), 30 i: 9 versus 102 i- 47 pg/mL in study 2, and 38 + 10 versus 148 % 38 pg/mL in study 3 (P < .051, respectively. In women with diabetes, the epinephrine responses were diminished during pregnancy compared with their own responses in the nonpregnant phase (P < .05). In control subjects, the responses also appeared to diminish during pregnancy, but these differences did not attain statistical significance. The peak glucagon responses above baseline were diminished in women with diabetes compared with controls in the nonpregnant phase (mean i SEMI: 15 2 5 versus 47 % 8 pg/mL in study 1 (P = .OOl). During pregnancy, the glucagon responses in control subjects were significantly diminished compared with their own nonpregnant phase (P < ,011. Indeed, during pregnancy both groups had similarly diminished and insignificant

A

Figure

4. Counterregulatory

growth

hormone responses to hypoglycemia (mean i standard error of the mean) in women with diabetes and controls. IDDM = insulin-dependent mellitus.

diabetes

2

16

2

14

g

12

0

'0

16

16 r-

r

I

14

t

E 5

8

= c

6

$

4

5

2

2 0 MinutesB

0 30 60 90120150180

t Ok-----= 0

non-pregnant . IDDM

VOL.

87, NO.

4, APRIL

1996

30 60 90120150160

0 30 60 90120150160

24-28 wks

32-34

wks

n Controls

Rosenn

et al

Coutltcrregz~lntory

Horniones

571

320

r

280

280

280

-E

240

240

240

g

200

T

160

Figure 5. Counterregulatory cortisol responses to hypoglycemia (mean k standard error of the mean) in women with diabetes and controls. IDDM = insulin-dependent diabetes mellitus.

160 i

80 40 bMinutes q

0 30 60 90120150180

q

non-pregnant l IDDM

24-28 wks

Discussion Epinephrine and glucagon are the two primary counterregulatory hormones that prevent hypoglycemia by inducing glycogenolysis and gluconeogenesis; epinephrine also reduces glucose uptake by peripheral tissues.’ Glycemic thresholds for activation of the counterregulatory hormonal responses to hypoglycemia, as well as

Rosenn et al

Coirnterrq@to~y

32-34 wks

n Controls

glucagon responses to hypoglycemia: 74 -C 5 and 26 +- 5 pg/mL in study 2 and 25 -C 6 and 16 ? 3 pg/mL in study 3 in women with diabetes and controls, respectively. Baseline growth hormone concentrations (mean % SEM) were higher in women with diabetes compared with controls in study 1 (4.1 2 0.9 versus 1.2 5 0.5 ng/mL; P = .Ol) and in study 2 (5.6 t- 0.9 versus 3.7 -+ 0.6 ng/mL; P = .07). Peak growth hormone responses above baseline were similar in both groups. However, these responses diminished progressively from study 1 (14.6 t 2.5 and 12.5 2 5.2 ng/mL) to study 2 (4.4 + 1.1 and 7.3 ? 2.7 ng/mL) to study 3 (2.5 +- 0.9 and 4.4 ? 2.3 ng/mL), both in women with diabetes and in controls, respectively. As expected, basal cortisol concentrations (mean -t SEM) were much higher during pregnancy in both groups compared with the nonpregnant state: 93 + 12, 177 2 16, and 190 2 20 pg/L in women with diabetes, and 87 ? 7,193 + 22, and 235 -C 36 wg/L in controls in studies 1,2, and 3, respectively (P < .Ol pregnant versus nonpregnant in both groups). There was no evident response to hypoglycemia in the women with diabetes either during pregnancy or in the nonpregnant state. There was a small cortisol response to hypoglycemia in controls: 16.8 + 6.2, 15.7 f 14.3, and 23.9 2 8.3 pg/L in studies 1, 2, and 3, respectively.

572

0 30 60 90120150180

Hormones

thresholds for hypoglycemia awareness and altered mentation, are subject to individual variability. In general, secretion of epinephrine and glucagon in normal subjects begins at glucose concentrations of 64-68 mg/dL in arterialized venous plasma, autonomic symptoms begin at 56-60 mg/dL, and symptoms of neuroglycopenia begin at 48-54 mg/dL.‘“,” Thus, the human body has a hierarchy of responses to progressive hypoglycemia that maximizes the opportunity to prevent, as well as correct, its deleterious effects. Most patients with type I diabetes lose the ability to mount a glucagon counterregulatory response to hypoglycemia within a few years after the onset of diabetes. Many patients with longstanding disease also have diminished epinephrine responses to hypoglycemia.‘2 Moreover, counterregulatory responses are further diminished during intensive insulin therapy, the kind of therapy that is commonly instituted during pregnancy.‘” As expected, subjects with diabetes did not have a recognizable glucagon response to hypoglycemia. In the absence of an effective glucagon response, acute counterregulation depends on the ability to mount an effective epinephrine response. Our findings show that the diminished epinephrine response in women with diabetes is further diminished during pregnancy, increasing the risk of recurrent hypoglycemic episodes. These data support the findings of Diamond et al,‘” who demonstrated that hormonal counterregulatory responses to hypoglycemia are both diminished and delayed in intensively treated pregnant women with type I diabetes compared with normal, nonpregnant, historical controls. This may be due, at least in part, to the institution of intensive insulin therapy during pregnancy. Indeed, glycohemoglobin concentrations in our

Obstetrics

& Gytzecolog~y

study population were markedly lower during pregnancy compared with the nonpregnant state. However, an independent effect of pregnancy-related factors on the counterregulatory epinephrine response cannot be ruled out. It is unclear why the glucagon response is diminished during pregnancy in nondiabetic women. A diminished glucagon response to a bolus injection of insulin in pregnant women has been reported previously by Luyckx et al.” Rossi et al,16 in their hypoglycemic clamp studies in pregnant rats, noted a markedly diminished glucagon response as well as a diminished epinephrine response to hypoglycemia compared with nonpregnant control rats. One possible explanation is that the diminished responses are related to a suppressive effect of placental hormones. Supporting this contention is the fact that arginine-induced secretion of glucagon is decreased in women using oral contraceptives,17,‘s suggesting that higher levels of estrogen and/or progesterone, such as occur during pregnancy, are associated with diminished glucagon and epinephrine responses to hypoglycemia. The counterregulatory roles of growth hormone and cortisol appear to be primarily in prolonging the effects of epinephrine and glucagon, and therefore they do not represent a primary defense mechanism against hypoglycemia.” The growth hormone response is usually preserved in subjects with diabetes, and may even be exaggerated in uncontrolled diabetes.‘” Our data demonstrate that this response diminishes progressively during pregnancy, both in women with diabetes and in healthy controls. This phenomenon may be due to the effect of progressively rising concentrations of placental hormones with somatostatin-like activity. It is possible that the threshold glucose concentrations for activation of the counterregulatory hormonal responses to hypoglycemia are lower during pregnancy, and that the level of hypoglycemia achieved in this study was insufficiently low to evoke these responses. However, Diamond et al’” demonstrated deficient counterregulatory responses in pregnant women with diabetes even when the plasma glucose concentration was lowered to 45 mg/dL. The short- and long-term benefits of improved glycemic control during pregnancy, both for the mother with diabetes and her offspring, have been clearly demonstrated in several studies and clinical trials, and should continue to be advocated. At the same time, it is also clear that intensive insulin therapy during pregnancy increases the risks of hypoglycemia, and that some women will experience severe and recurrent hypoglycemic episodes during this period of strict glycemic control.21 Our findings provide some insight into the mechanisms underlying the increased incidence of hy-

VOL.

87, NO.

4, APRIL

1996

poglycemia during pregnancy. However, several questions remain unanswered: 1) Does pregnancy, per se, independently affect the counterregulatory responses to hypoglycemia? 2) How does pregnancy affect hypoglycemia unawareness (an entity that is associated but not synonymous with deficient counterregulation)? 3) Could a less stringent regimen of glycemic control be associated with some preservation of the counterregulatory responses and with a decrease in the incidence of severe hypoglycemia, without jeopardizing the outcome of pregnancy? These issues will need to be addressed in future studies.

References 1. Rosenn glycemic

B, Miodovnik M, Combs CA, Khoury control and antepartum obstetric

J, Siddiqi TA. complications

Poor in

women with insulin-dependent diabetes. lnt J Gynecol Obstet 1993;43:21-8. diabetes on the fetus 2. Rosenn B, Tsang RC. The effects of maternal and neonate. Ann Clin Lab Sci 1991;21:153-70. 3. Cryer PE. Iatrogenic hypoglycemia as a cause of hypoglycemiaassociated autonomic failure in IDDM: A vicious cycle. Diabetes 1992;41:255-60. 4. Rosenn B, Miodovnik

M,

Dignan

I’S, Siddiqi

TA,

Khoury

J,

Mimouni F. Minor congenital malformations in infants of insulindependent diabetic women: Association with poor glycemic control. Obstet Gynecol 1990;76:745-9. 5. Goldstein Validity chemical

DS, Feuerstein G, laze JL Jr, Kopin IJ, Keiser HR. and reliability of liquid chromatography with electrodetection for measuring plasma levels of norepinephrine

and epinephrine in man. Life Sci 1981;28:467-75. 6. Kuku SF, Jaspan JB, Emmanourl DS, Zeidler A, Katz AI, Rubenstein AH. Heterogeneity of plasma glucagon circulating components in normal subjects and patients Clin Invest 1976;58:742-50.

with

chronic

renal

failure.

J

7. Thomas MA, Rebar RW, LaBarbera AR, Pennington EJ, Liu JH. Dose-response effects of exogenous pulsatile human corticotropinreleasing hormone on adrenocorticotropin, cortisol, and gonadotropin concentrations 1991;72:1249-54.

in agonadal

women.

J Clin

8. Frohman LA, Horton ES, Lebovitz HE. Growth action of a pseudomonas endotoxin (piromen). 16~57-67. 9. Gerich JE, Campbell PJ. Overview abnormalities in diabetes mellitus Metab Rev 1988;4:93-111. 10. Mitrakou Hierarchy

Endocrinol

Metab

hormone releasing Metabolism 1967;

of counterregulation and other conditions.

and its Diabetes

A, Ryan C, Veneman T, Mokan M, Jenssen T, Kiss I, et al. of glycemic thresholds for counterregulatory hormone

secretion, symptoms, 1991;260:E67-74.

and

cerebral

dysfunction.

Am

J Physiol

11. Schwartz NS, Clutter WE, Shah SD, Cryer PE. Glycemic thresholds for activation of glucose counterregulatory systems are higher than the thresholds for symptoms. J Clin Invest 1987;79:777-81. 12. Bolli 6, De Feo P, Compagnucci Santcusanio F, et al. Abnormal

I’, Cartechini MG, Angeletti G, glucose counterregulation in insu-

lin-dependent diabetes mellitus. Interaction bodies and impaired glucagon and epinephrine 1983;.32:134-41.

of anti-insulin antisecretion. Diabetes

13. Simonson DC, Tamborlane WV, DeFronzo RA, Sherwin RS. Intensive insulin therapy reduces the counterregulatory hormone re-

Rosenn

et al

Cou~&wegul~tory

Hormones

573

sponses to hypoglycemia in patients with type I diabetes. Ann Int Med 1985;103:184-90. 14. Diamond ME’, Reece EA, Caprio S, Jones TW, Amiel S, DeCemaro N, et al. Impairment of counterregulatory hormone responses to hypoglycemia in pregnant women with insulin-dependent diabetes mellitus. Am J Obstet Gynecol 1992;166:70-7. 15. Luyckx AS, Gaspard U, Lefebvre free fatty acids on the glucagon normal and in pregnant women.

PJ. Influence of elevated plasma response to hypoglycemia in Metabolism 1978;27:103~40.

16. Rossi G, Lapaczewski P, Diamond Ml’, Jacob RJ, Shulman GI, Sherwin RS. Inhibitory effect of pregnancy on counterregulatory hormone responses to hypoglycemia in awake rat. Diabetes 1993; 42:1440-j. 17. Beck I’, Eaton

RI’, Arnett

DM.

Effect

of contraceptive

arginine-stimulated glucagon and insulin secretion Lipid physiology. Metabolism 1975;24:1055-6. 18. Beck P, Arnett DM, Alsever RN. Effect of contraceptive arginine-stimulated glucagon and Carbohydrate and lipid physiology Metabolism 1976;25:23%31.

steroids

on

in women. steroids

I:

574

Rosenn

et al

Counterregulatory

hormone

Hormones

and

cortisol.

1991;40:1251-8. 21. Rosenn BM, Miodovnik M, Holcberg Hypoglycemia: The price of intensive with insulin-dependent 417-22.

Address Barak

reprint

requests

M. Rosenn, MD University of Cincinnati P.O. Box 670526 Cirzcinnafi, OH 45267

diabetes

G, Khoury JC, Siddiqi TA. insulin therapy for women

mellitus.

Obstet

Gynecol

1995;85:

to:

Medical

Center

on

insulin secretion in women. II: in insulin-dependent diabetics.

19. Gerich JE, Rizza R, Haymond M, Cryer I’. Hormone mechanisms in acute glucose counterregulation: The relative roles of glucagon, epinephrine, norepinephrine, growth tabolism 1980;29(suppl. 1):1164-75.

20. Cohen RM, Abplanalp WA. Resistance of growth hormone secretion to somatostatin in men with type 1 diabetes mellitus. Diabetes

Me-

Received Received Accepted

Azzgust 4, 2995. in veuisrd form Novetzzber December 14, 1995.

Copyright 0 Gynecologists.

1996 by The American Published by Elsevier

30, 1995

College of Obstetricians Science Inc.

Obstetrics

& Gynecology

and