Hypocaloric diets and ketogenesis in the management of obese gestational diabetic women.

Journal of the American College of Nutrition

ISSN: 0731-5724 (Print) 1541-1087 (Online) Journal homepage: http://www.tandfonline.com/loi/uacn20

Hypocaloric diets and ketogenesis in the management of obese gestational diabetic women. R H Knopp, M S Magee, V Raisys, T Benedetti & B Bonet To cite this article: R H Knopp, M S Magee, V Raisys, T Benedetti & B Bonet (1991) Hypocaloric diets and ketogenesis in the management of obese gestational diabetic women., Journal of the American College of Nutrition, 10:6, 649-667, DOI: 10.1080/07315724.1991.10718184 To link to this article: http://dx.doi.org/10.1080/07315724.1991.10718184

Published online: 02 Sep 2013.

Submit your article to this journal

Article views: 2

View related articles

Citing articles: 3 View citing articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=uacn20 Download by: [ECU Libraries]

Date: 23 September 2015, At: 04:00

Hypocaloric Diets and Ketogenesis in the Management of Obese Gestational Diabetic Women Robert H. Knopp, MD, FACN, M. Scott Magee, MD, Vidmantas Raisys, PhD, Thomas Benedetti, MD, and Bartolome Bonet, MD, PhD Northwest Lipid Research Clinic, and Departments of Medicine, Obstetrics and Gynecology, and Laboratory Medicine, University of Washington School of Medicine, Seattle

Downloaded by [ECU Libraries] at 04:00 23 September 2015

Key words: pregnancy, gestational diabetes, hypocaloric diet, carbohydrate, lipid metabolism The extent to which given levels of caloric restriction will improve glycémie status but increase plasma ketone bodies in gestational diabetic women has received little attention. After reviewing the underlying physiology, we present data on two feeding studies investigating the question. In the first, a weight-maintaining -2400-kcal/day diet was fed on a metabolic ward to 12 gestational diabetic women for 1 week. In the second week, subjects were randomized to a continuation of the 2400-kcal/day diet or to a 1200-kcal/day diet. Twenty-four-hour mean glucose levels remained unchanged in the control group but declined in the calorie-restricted group (6.7 mM or 121 mg/dl in week 1 vs 5.4 mM or 97.3 mg/dl in week 2) (p < 0.01). Nine-hour overnight fasting plasma insulin also declined but oral glucose tolerance did not improve with caloric restriction. Fasting plasma ß-hydroxybutyrate rose in the calorie-restricted group, along with an increase in ketonuria, but not in the control group. A second study compared the impact of a 33% calorie-restricted diet or insulin to a full-calorie diet in a similar 2-week experimental design and measured hepatic glucose output and insulin sensitivity with dideuterated glucose before and during an insulin clamp. Diet in three subjects improved fasting and 24-hr mean glucose by 22 and 10%, respectively, whereas prophylactic insulin in three subjects produced 0 and 4% reductions, respectively. On average, ketonuria after a 9-hr fast declined to an equivalent degree with both treatments. Hepatic glucose output and insulin sensitivity were not statistically significantly altered by gestational diabetes or the therapeutic interventions compared to nondiabetic normal weight or obese pregnant controls. In conclusion, 50% caloric restriction improves glycémie status in obese women with gestational diabetes but is associated with an increase in ketonuria, which is of uncertain significance. An intermediate 33% level of caloric restriction (to 1600-1800 kcal daily) may be more appropriate in dietary management of obese woman with gestation al diabetes mellitus and more effective than prophylactic insulin. Further studies are required to confirm these findings. Abbreviations: APE = atom percent excess, CRC = Clinical Research Center, DIEP = Diabetes in Early Pregnan cy project, FA = fatty acid, FRA = free fatty acid, GDM = gestational diabetes mellitus, HDL-C = high-densitylipoprotein cholesterol, HGO = hepatic glucose output, IRI = immunoreactive insulin, [6,6-2H2] = dideuterated, S| = insulin sensitivity index, VLDL - very-low-density lipoprotein

INTRODUCTION Gestational diabetes mellitus (GDM) is a condition in which fasting hyperglycémie or glucose intolerance ex ceeding specific norms are recognized for the first time in pregnancy [1-4] (see [5,6] for recent reviews). Al though insulin therapy even at low dose is of demonstrable benefit in reducing perinatal mortality and infant macrosomia [7] and has been recommended for more general use in the treatment of GDM [8], GDM is nonetheless an early manifestation of type II insulin-

resistant obesity-associated diabetes in which caloric restriction and weight loss are the cornerstones of initial therapy. Indeed, some authors have advocated caloric restriction in the treatment of GDM [9,10]. The issue of whether calorie-restricted diets are safe and effective in the management of GDM has not under gone sufficient clinical investigation. For instance, we have been unable to draw secure conclusions regarding the extent to which glycémie control is improved, plas ma ketone concentrations and ketogenesis are enhanced, and fetal outcome is improved by calorie restriction in

Presented in part at the 31 st Annual Meeting of the American College of Nutrition, October 13-15,1990, in Albuquerque, New Mexico. Address reprint requests to Dr. Robert H. Knopp, Northwest Lipid Research Clinic, 325 Ninth Avenue, Seattle, Washington 98104.

Journal of the American College of Nutrition, Vol. 10, No. 6, 649-667 (1991) © 1991 John Wiley & Sons, Inc.

CCC 0731-5724/91/0600649-19$04.00

Diet and Gestational Diabetes Table 1. Effects of Food Deprivation on Plasma Fuels and IRI in Rat Pregnancy1

Glucose (mM) IRI(pM) FFA(uEq/L) Ketones (uEq/L)

Fed

Fasted 48 hours

Days gestation

Days gestation

0

0

19

6.0 143 310 285

4.5* 242* 516* 145*

5.3 71 443 1781

19 2.8*

93 739* 6473*


"From Herrera et al [14]. *p < 0.01 vs nonpregnanL

Table 2. Adipose Tissue FFA and Glycerol Release in Vitro in Rat Pregnancy11 Fed

Fasted 48 hours

Days gestation

Days gestation

0 Glycerol (uM/mg) FFA (μΜ/mg)

0.094 0.124

19 0.151* 0.235**

0 0.148 0.223

19 0.286** 0.444**

From Knopp et al [16]. *p< 0.02, **p< 0.01.

GDM. The purpose of the present paper is to review the available information on the mechanism of ketogenesis, the effect of diets of differing degrees of caloric restric tion (including previously unpublished work), the ad visability of caloric restriction in GDM based on present knowledge, and directions for future research. The pre viously unpublished data concern the glycémie response to 33% caloric restriction, hepatic glucose output, and the sensitivity of hepatic glucose output to insulin in GDM.

EXPERIMENTAL STUDIES OF CALORIE RESTRICTION IN ANIMAL PREGNANCY The propensity for ketonemia and ketonuria to develop more rapidly in pregnancy with calorie restriction has been recognized since the late 1920s, when it was first noted by Bokelmann and Bock [11]. The effect was then confirmed in the fasted pregnant rat by MacKay and Barnes [12] in the 1930s, who found the increase in ketonuria to be as much as 175 times the nonpregnant level after 2 days of fasting in late rat pregnan-

650

cy [12]. The significance of the ketone rise for the fetus was addressed by Scow et al [13], who studied the relationship between maternal and fetal ketonemia in 1958. They found a linear proportionality between ketone body concentrations in the maternal and fetal circulations. The same relationship was found for blood glucose but not for total plasma lipids, which were elevated in the maternal but not in the fetal plasma. Their conclusion was that ketone bodies as well as glucose have free concentration-dependent access to the fetal circulation. An example of the increase in plasma ketone bodies occurring with 2 days fasting in the 19-day gestation pregnant rat and its mechanism are shown in Table 1, based on our own work [14]. In the fed state, mean plasma ketone body concentrations were 0.15 mM in the 19-day pregnant rat and 0.29 mM in the nonpregnant rat. After 48 hours of fasting, ketone concentrations increased to 1.8 mM in the nonpregnant rat and 6.5 mM in the pregnant rat. The ketone increment in fasted pregnant rats was approximately 50-fold above the fed pregnant rats. The fasting state was associated with a more marked drop in plasma glucose in the pregnant animal, about 1.65 mM (30 mg/dl), compared to 0.66 mM (12

VOL. 10, NO. 6

Diet and Gestational Diabetes Table 3. Effects of Food Deprivation on Hepatic Metabolites in Rat Pregnancy3

Glycogen (mg/μΜ DNA-P) Triglycéride (mg/μΜ DNA-P) Aceryl CoA (μΜ/μΜ DNA-P)

Fed

Fasted 48 hours

Days gestation

Days gestation

0

19

0

19

5.6 5.2 7.8

5.1 3.2* 7.7

0.3 3.0 6.0

0.4 10.6* 7.6*

"From Herrera et al [14].


*p < 0.05.

mg/dl) in the nonpregnant rat (Table 1). With fasting, plasma immunoreactive insulin fell more in pregnancy to a level almost equivalent to the nonpregnant rat. This fall was accompanied by a greater increase in free fatty acid (FFA) concentrations in the fasted pregnant rat to a level that was about 75% above the fasted nonpregnant level and about 50% above the fed pregnant level. Notice that, even in the fed state, plasma FFA concentrations were elevated in the late gestation rat while ketone bodies were below the level of the nonpregnant rat. Only in fasting was marked ketonemia observed, accompanied by a modest rise in FFA concentrations. Thus, although increased FFA concentrations in the plasma may help drive ketogenesis, it is apparent that ketonemia does not develop until plasma glucose and insulin concentrations fall below a critical level. A similar conclusion was reached by Felig and Lynch [15] in fasted first trimester humans. Further investigations were performed to assess the mechanism of the elevated plasma FFA in pregnancy [16]. As shown in Table 2, when adipose tissue samples from fed and fasted nonpregnant and 19-day gestation rats were incubated in vitro, the release of both glycerol and FFA into the medium was increased in the fed state and increased further in the fasted state. These data sug gest that the increase in plasma FFA concentrations is due to an increase in fatty acid (FA) mobilization from adipose tissue rather than an impairment of FA utiliza tion. Indeed, in unpublished work (Knopp RH, Hum phrey J, Montes A, Herrera EH), radiolabeled chylemicron triglycéride FAs administered to the pregnant rat recycle to the liver and reappear in the maternal circulation as triglycéride FA faster in pregnancy than in the nonpregnant rat, suggesting a heightened rate of hepatic FA uptake and reesterification. An index of the impact of fasting in pregnancy on hepatic metabolism is shown in Table 3, again based on previous work [14]. Liver glycogen was virtually absent

after 2 days of fasting, whereas liver triglycéride concentrations were increased threefold over the fed pregnant or nonpregnant rat levels, consistent with the heightened incorporation of FA into circulating triglycérides referred to above. Indeed, the process of FA mobilization and hepatic reesterification is so much accelerated in the rat that a marked hypertriglyceridemia develops in the fasted pregnant rat [17,18]. Evidence of altered intermediary metabolism in the liver is found in the heightened ketogenesis of fasted pregnancy with an increase in acetyl CoA concentrations in the fasted pregnant rat compared to the nonpregnant fasted control. However, acetyl CoA concentrations do not appear to be the driving mechanism, as the concentrations are not significantly higher in the fasted pregnant rat from the fed pregnant or nonpregnant rats (Table 3) [14]. A summary of the metabolic alterations provoking accelerated ketogenesis in pregnancy is shown in Figure 1 [19,20]. Utilization of glucose in both the fed and fasted state by the fetus is associated with a lower maternal plasma glucose concentration, more so in the fasted state (Table 1). This reduction in plasma glucose concentration then results in a reduction in maternal plasma insulin concentration, despite the heightened insulin secretion inherent in pregnancy (Table 1) [20]. The result is an enhanced mobilization of FFA from maternal adipose tissue as a consequence of the disinhibition of lipolysis by the reduction in plasma insulin concentrations. Augmenting this mobilization of FAs is the direct stimulation of lipolysis (Table 2), which is believed to be due to the increased concentrations of pregnancy hormones, specifically human chorionic somatomammotropin, cortisol, thyroid hormone, and conceivably the insulin resistance associated with prolactin and progesterone, which are also present in high concentrations in gestation [20]. Thus, lipolysis is stimulated at all times in pregnancy, both in the fed and fasted state as a result of these hormones. Finally, low plasma insulin levels stimulate

JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION

651

Diet and Gestational Diabetes PLACENTA MOTHER

FETUS

Glucose

■> Glucose use

i

I

Hypoglycémie

Hypoglycémie

glyi Hypoinsulinemia

sul Direct effect of hormones mobilization of pregnancy —► FFA mobi

1

+

i

Ketonemia

♦

Ketonemia


I Fig. 1. Mechanisms of maternal ketogenesis. Continued fetal utilization of glucose despite caloric deprivation in the mother is associated with a reduction in maternal plasma glucose concentrations and, as a result, a reduction in plasma insulin concentrations. This reduction in the insulin-induced inhibition of lipolysis results in an increased mobilization of fatty acids further driven by the primary stimulus to lipolysis in pregnancy associated with cir culating pregnancy hormones. The increased availability of free fatty acids to the liver in the face of a reduced plasma insulin concentration allows ketogenesis to proceed at an accelerated rate. These ketones then traverse the placenta freely and enter the umbilical circulation in concentrations proportional to the maternal level. (Reprinted from [20] with permission.)

hepatic ketogenesis. This sequence of events was labeled "accelerated starvation" by Freinkel [21,22]. There may also be a direct effect of estrogen and progesterone on glucose homeostasis and ketogenesis since some of the effects of starvation in pregnancy can be seen in starved nonpregnant rats given estrogen and progestin [23]. A clinically relevant point is that the pregnant woman with insulin-dependent diabetes mellitus is markedly susceptible to progress into ketoacidosis as a consequence of the sequence outlined above, which can proceed out of control proportional to the deficiency of insulin even without marked elevations in plasma glucose.

KETONEMIA IN NORMAL PREGNANCY AND THE EFFECT OF LENGTH OF FASTING At the time the above studies were being performed in rat pregnancy, other work indicated that maternal fat storage occurs in an asynchronous fashion and is greatest in the first half of gestation, both in the pregnant rat and human (see [20,24] for reviews). This process is re flected by an enhanced de novo synthesis of FA in preg

652

nant rat adipose tissue in midgestation and a reduction of this process in late gestation [25] in association with enhanced FA mobilization [24]. The propensity to en hanced fat accretion in early gestation and enhanced fat mobilization in late gestation is related to increased food intake [24] and hyperinsulinemia [14,18,22]. The at tenuation of fat storage in late gestation is a consequence of insulin resistance and enhanced FA mobilization due to the combined effect of increased pregnancy hormones (Fig. 2) [20]. We were then curious to learn what change, if any, occurred in plasma fuels and hormone levels throughout the course of gestation. To answer this question, 20 nor mal pregnant women were followed serially throughout gestation with measurements of plasma glucose, FFA, insulin, ketone bodies, and very-low-density-lipoprotein (VLDL) triglycéride concentrations (Fig. 3) In these women, studied at home after a 12-14-hr overnight fast, plasma FFA concentrations and plasma ketone body levels were only marginally higher in late gestation than in early gestation. These data indicated that heightened ketonemia is not necessarily a part of normal gestation given a 12-14-hr overnight fast. Metzger and associates [26] investigated this question further by observing the increase of plasma ketonemia at

VOL. 10, NO. 6

Diet and Gestational MaontSE

PLACENTA MOTHER

fc

β0

Diabetes

Delivery

^

GLUCOSE


0 | .| B-OH BUTYRATE

20

Fig. 2. Maternal metabolism in early and late gestation. Fat anabolism is primarily accomplished in the first half or twothirds of pregnancy when maternal weight gain is primarily due to increased fat storage stemming from increased maternal food intake. In late gestation, the metabolic situation is reversed, with an increase in plasma hormones of pregnancy which induce lipolysis. Concomitantly, insulin resistance develops as a result of the effect of the increasing levels of these same pregnancy hormones. Fat storage and glucose utilization are attenuated and glucose is now diverted to the fetus while maternal fatty acids are mobilized to serve as an alternative fuel for the mother. (Reprinted from [20] with permission.)

increasing time intervals after an overnight fast. In this investigation, they confirmed the observations made earlier in the rat by showing that plasma glucose and insulin concentrations fell to a greater extent after 16 and 18 hours of fasting than in nonpregnant women. In addition, plasma FFA concentrations increased sharply after 16 and 18 hours of fasting compared to 12 hours. The ketone increase was on the order of 0.35 mM from baseline in ß-hydroxybutyrate. Since in these late gestation normal women the 12-hr fasting ß-hydroxybutyrate level was 0.15 mM, the 18-hr level would represent a plasma concentration of approximately 0.5 mM. The 12hr concentration of ketone bodies is very similar to the third trimester level we found in our own work (Fig. 3) [20]. The extent to which these physiological increments in plasma ketone levels might lead to ketonuria received some attention at about the same time. Coetzee [27]

27 3 0 33 36 39 Weeks

6

20

Fig. 3. Serial changes in plasma fuels in pregnancy. Twenty pregnant women were studied serially throughout gestation and at 6 and 20 weeks postpartum. (Reprinted from [20] with permission.)

found that maternal ketonuria was observed at plasma ketone concentrations as low as 0.1 mM. More recently, Chez and associates [28] found in 1987 that eight of nine clinically normal pregnant women followed serially with urinary ketone testing had acetonuria present on 2-15 days each throughout gestation. Ketonuria was no more common in the third trimester than in the first. An earlier investigation also found similar results [29]. It is noteworthy, however, that these investigations underes timate the extent of urinary ketone excretion, as the uri nary ketone test reagent reacts only with acetoacetic acid and not ß-hydroxybutyric acid.

EFFECT OF HIGH-CARBOHYDRATE DIET IN PREGNANCY Pertinent to the amount of FA mobilization and ketogenesis in the body is the amount of carbohydrate fed in the diet. In light of the publication of Brunzell and associates [30] in the early 1970s that a high-carbohydrate diet was beneficial in individuals with type Π diabetes, we investigated the effect of a 75% car bohydrate/15% protein/10% fat diet in two normal and two GDM pregnant women (Table 4) [31]. When three


653

Diet and Gestational

Diabetes

Table 4. Effect of Isocaloric High-Carbohydrate Diet on Plasma Fuels and IRI in the Third Trimester*

Subjects0 c

K Bc G R

Glucose (mM)

FFA(uEq/L)

IRI(pM)

%CHO

%CHO

%CHO

40

75

40

5.72 4.84 3.63 4.02

5.17 4.95 3.52 3.96

629 623 990 1223

75

40

75

632 588 788 1053

192 64 107

192 86 86

a


From Waith and Knopp [31]. "Values based on means of 3 baseline and 5-7 treatment days. c Subject K was gestationaliy diabetic on insulin, and subject B was borderline diabetic.

Table 5. Composition of the Diet" % of calories Carbohydrate Fat Protein Fiber (g)

50 30 20 11

a

From Magee et al [33].

Table 6. Characteristics of Subjects Studies* Prepregnancy prior

Controls Restricted

Age (years)

Weight (kg)

mw

n

(%)

Pregnancy number

Gestation (weeks)

5 7

36 30

88 96

147 171

7 5

30 31

'From Magee et al [33].

baseline and 5-7-day treatment means were compared, fasting plasma glucose concentrations were slightly im proved (about 10% lower) in the one overt insulin-re quiring diabetic but not in the other three study subjects. Plasma FFA concentrations were unchanged in the diabetic woman and lower in the other three. Fasting plasma insulin concentrations did not change consistent ly. Ketones were not measured in this investigation. The fact that plasma FFAs were slightly reduced in three of the four cases and fasting glucose improved in the diabetic suggested that high-carbohydrate feeding might be an avenue through which metabolic control might be

654

improved in the gestational diabetic as well as in the nonpregnant diabetic patient.

EFFECT OF 50% CALORIE RESTRICTION ON PLASMA FUELS IN GDM Another approach to the dietary management of GDM is caloric restriction. Previous studies by Borberg et al [32], Maresh and associates [9], and Algert et al [10] indicate that moderate caloric restriction can be

VOL. 10, NO. 6


Diabetes

Control Group

Randomize

Calorie Restricted Group

10


Day of Study 1 Weekl

11

12

13

14

Week 2

Fig. 4. Design of the study to investigate the effect of caloric restriction in pregnancy. Subjects were given an equivalent full calorie diet approximating 2400 calories daily for 1 week in the metabolic ward. A 24-hr glucose profile was performed on day 6 in both groups and a glucose tolerance test on day 7. In the second week, the control group was continued on its 2400-kcal diet while the calorie-restricted group received approximately 1200 kcal daily for 1 week with repeat 24-hr glucose profile and glucose tolerance tests performed on days 13 and 14,respectively.(© American Diabetes Association, 1990, reprinted from Diabetes [33] with permission.)

employed without harm to the fetus and with some improvement in glycémie status. To define more precisely the metabolic response to specific levels of caloric restriction in GDM, we performed two short-term (1week) studies of restriction on a metabolic ward. In the first study, calories were restricted by 50% in an otherwise high-carbohydrate diabetic diet. The diet composition is shown in Table 5 and consisted of 50% carbohydrate, 30% fat, 20% protein, and 11 g of fiber [33]. The calories were distributed as 25% at each of the major meals and 12.5% of calories each at 3:00 pm and 10:00 pm snacks. All blood samples were obtained the following morning, 9 hours after the last snack. Characteristics of the subjects are shown in Table 6. Twelve women were studied and randomized to either fullcalorie or calorie-restricted regimens. Control subjects tended to be older and somewhat lighter, but the two groups were equivalent in number of prior pregnancies and week of gestation in which they were studied. All subjects were obese, having body weights in excess of 120% of nonpregnant ideal body weight. The experimental design consisted of admitting subjects to the Clinical Research Center (CRC) at the University of Washington Hospital after informed consent had been obtained. Subjects were maintained on a

calorie-sufficient diet based on the Harris-Benedict formula (see [33]) for the first week of observation. Plasma fuels and hormones were measured on the sixth and seventh days of the hospital week. In addition, a 24-hr plasma glucose profile was obtained on day 6, and a glucose tolerance test was performed on day 7. In the second week, the control group continued its original diet and observations were repeated at days 13 and 14. The calorie-restricted group was reduced in caloric content to 50% of control, with the 24-hr glucose profile and glucose tolerance tests repeated on days 13 and 14, respectively (Fig. 4). The actual mean caloric intake of the subjects in the two studies is shown in Table 7. Control subjects consumed 2307 kcal during each week, whereas the restricted subjects received 2476 kcal during the first week and 1238 kcal in the second week. Because subjects were fed a 10:00 pm snack, the length of the overnight fast was 9 hours. Effects on plasma fuels and insulin are presented in Table 8. In the control group, fasting plasma glucose concentrations declined 4% between weeks 1 and 2, a nonsignificant difference. In addition, the means of 24 observations of plasma glucose over 24 hours were identical (Table 8), as were the curves describing the postprandial glycémie changes during the 2 weeks (Fig. 5).


655

Diet and Gestational Diabetes Table 7. Caloric Intake of Subjects in the Two Studies"

Controls (kcal/day)b Restricted (kcal/day)

Weekl

Week 2

2307 ± 174 2476 ± 205

2307 ± 174 1238 ±103

"From Magee et al [33]. Mean±SD.

b

Table 8. Effects of 50% Caloric Restriction on Plasma Fuels and Insulin in Obese Gestational Diabetics"


Control

Glucose (mM) Fasting 24-hr mean IRI(pM) FFA (mg/L) β-ΟΗΒ(μΜ) Urine ketones (+) TG(mM)

Restricted

Week 1

Week 2

%Δ

Weekl

Week 2

%Δ

5.4 6.7 165 330 220 0.0 3.2

5.2 6.7 165 390 210 0.0 3.7

-A 0 0 18 -5 0 16

5.9 6.8 265 220 290 1.0 3.1

4.9 5.4 146 360 780 2.1 2.9

-17 -21 -45 64 269 110 -7

"From Magee et al [33].

Similarly, plasma immunoreactive insulin concentra tions, FFA concentrations, and ß-hydroxybutyric acid concentrations were unchanged in the 2 weeks, and no ketonuria was observed in these women after 9 hours of overnight fasting. In contrast, calorie-restricted subjects had a 17% reduction in mean fasting glucose concentration and a 21% reduction in the 24-hr mean glucose level compared to the first week of full-calorie feeding. An even more marked decrease in fasting plasma immunoreactive insulin concentration was observed in the calorie-restricted group, a fall of 45% from baseline. Plasma FFA concentrations increased 64% and ßhydroxybutyric acid concentrations increased from 0.29 to 0.78 mM, an increase of 269%. Urine ketones measured with nitroprusside reagent tablets increased twofold (Table 8). As the restricted subjects were more hyperglycémie at baseline and also tended to be more obese (Table 6), they had a greater tendency for ketonuria, even during the first week on a calorie-sufficient diet (Table 7). In light of the marked hypertriglyceridemia associated with fasting in the rodent model mentioned above [17,18], it was interesting to find that plasma triglycéride concentrations tended to decline in the calorie-restricted subjects compared both

656

to full-calorie feeding in the first week and control subjects in the second week (Table 8). To determine if the extent of calorie restriction observed was associated with an improvement in glucose tolerance, the glucose tolerance tests obtained at the end of the first and second weeks in the control and calorierestricted groups were compared (Fig. 6). As can be seen, there was no indication of a difference in either group. This result suggests that the improvement in glycémie status is not related to fundamental improvement in glucose disposal. Although the overnight fasting plasma insulin concentrations were nearly halved in the calorie-restricted group (Table 8), the fasting plasma glucose concentration was also lower, so it cannot be determined from these data if insulin sensitivity during the glucose tolerance test was improved or not. Inspecting the 24-hr glucose profiles (Fig. 5) discloses that at least some of the improved glucose tolerance was seen after an overnight fast (and putatively greater insulin sensitivity) and the remainder due to a reduced caloric and carbohydrate intake. In conclusion, these observations indicate that a substantial improvement in glycémie status can be achieved with caloric restriction of as much as 50%. This im-

VOL. 10, NO. 6



4

6

8

10 12 14 16 18 20 22 24

60 120 Time (min)

Time (hrs)

Diabetes

180

Fig. 5. Twenty-four-hour glucose profile performed in control subjects (top) and calorie-restricted subjects (bottom). A 2400-kcal diet for both experimental weeks 1 and 2 was given to controls. Calorierestricted subjects received -1200 kcal in the second week. (© American Diabetes Association, 1990, reprinted from Diabetes [33] with permission.)

Fig. 6. Glucose tolerance tests performed at the end of the first and second weeks. No difference in glucose tolerance was seen between the first and second weeks in the full-calorie control group (top panel) or in the calorie-restricted group (bottom panel). (© American Diabetes Association, 1990, reprinted from Diabetes [33] with permission.)

provement in glycemia was achieved at the expense of a 60% increase in plasma FFA concentrations and a 2.5fold increase in plasma ß-hydroxybutyrate concentra tions, approximately 50% higher than those observed by Metzger and Freinkel [26] after an 18-hr fast. In other words, this amount of ketonemia is not unlike that which might be encountered after a fast somewhat exceeding 18 hours during normal pregnancy. Because the impact of this amount of caloric restriction maintained over a long period of time on fetal growth and development is unknown, a specific dietary recommendation based on these data could not be made.

the effect of a 33% reduction in caloric intake, reported previously in abstract form [34]. The design of this otherwise unpublished investigation followed that of the previous study, being carried out at the University of Washington CRC and comparing observations in week 1 on a full-caloric diet vs week 2 on a 33% restricted diet. The length of the overnight fast was again 9 hours. In this investigation, we attempted further to compare the effect of caloric restriction vs insulin therapy beginning with a prophylactic fixed dosage of 20 units of NPH and 10 units of regular insulin daily [8] to answer this ques tion: Does calorie restriction offer a better alternative to "prophylactic" insulin in the management of the obese GDM woman? The characteristics of the subjects studied are shown in Table 9. Nine normal, seven obese, and six obese GDM subjects were studied during the first week of a full-calorie diet. The obese GDM subjects were studied for a second week while undergoing treatment with a

EFFECT OF 33% CALORIE RESTRICTION IN GDM To examine the effect of a less extreme calorie restric tion in GDM subjects, we investigated in a second study


657

Diet and Gestational Diabetes Table 9. Subject Characteristics: 33% Caloric Restriction in Pregnancy Pregnancy group


Normals Obese Obese with GDM p value0 Obese with GDM, insulin treated Obese with GDM, calorie-restricted

Number

Age

Gravida

Pregnancy % of ideal body weight

1-hr 50-g GCT* result

Study diet (kcal/day)

Gestational week studied

9 7 6 3

24.1 ± 4.9b 25.7 ± 3.5 28.8 ± 5.6 0.025 26.3 ±2.1

2.0 ± 0.9 2.0 ±0.6 3.2 ±1.8 NS 3.0 ±2.6

100 ±7 128 ±6 132 ±15 < 0.005 124 ±9

5.8 ± 12 6.6 ±1.4 9.1 ± 1.0 < 0.005 8.9 ±1.0

2352 ±166 2374 ± 104 2415 ±119 NS 2504 ±69

13.4 ± 1.1 32.0 ± 1.4 31.8 ± 1.7 NS 33.3 ± 0.6

3

31.3 ±7.4

3.3 ± 1.2

139 ± 17

9.4 ±1.2

1655 ±51

30.3 ± 0.6

"GCT = glucose challenge test. b Mean±SD. c p value calculated according to the Jonkheere test, a one-way analysis of variance reflecting ordering of results in the direction of normals £ nonobese

-0.033

Wkl Wk2

HDL3-C Δ

0.45 0.50 0.05 0.25 0.30 0.05 0.45 0.45 0.05 0.033

Wkl Wk2

Δ

0.85 0.90 0.05 0.90 0.85 -0.05 1.00 0.95 -0.05 -0.017

0.50 0.40 -0.10 0.80 0.90 0.10 0.45 0.60 0.20

0.80 0.75 -0.05 0.85 0.70 -0.15 0.90 0.80 -0.10

0.67

-0.10

GDM subjects showed little or no improvement in plas ma glucose concentrations either after an overnight fast or over the 24-hr period. Plasma immunoreactive insulin was higher, as might be expected, and plasma FFA con centrations were unchanged. Plasma triglycéride con centrations were only slightly changed. Urinary ketones were reduced compared to the previous week's measure ment. With calorie restriction, however, fasting glucose concentrations tended to be lower, as were the 24-hr


659

Diet and Gestational Diabetes Table 13. Measurements of Insulin Sensitivity and Hepatic Glucose Production1

n Normal Obese Obese GDM p value Obese GDM-Insx Obese GDM-4Cal

5-8 6-7 4-5 2 3

Basal HGO (mg/kg/min)

ED50» (*)

MF (mg/kg/min/μυ/ΓηΙ)

s, d

2.99 ± 0.45 2.88 ± 0.45 2.96 ± 0.50 NS 3.21 ±0.10 2.80 ± 0.063

135 ± 33 115119 147141 NS 14016 157 122

12.713.8 10.613.4 9.411.5 NS 9.312.9 9.311.1

2.5 11.3 3.011.7 2.111.1 NS 2.110.9 1.5 10.2


"Magee MS, Knopp RH: Previously unpublished data; data given as mean ± SD. b Percent of basal insulin level at which hepatic glucose production is reduced 50%. c Amount of glucose infusion needed to maintain plasma glucose constant after a 2-hr insulin infusion designed to double the basal level. Results expressed as [glucose infusion rate (mg/kg/min)]/[mean insulin level (μυ/ml)] x 100. "Insulin sensitivity index.

glucose profile means. In fact, these means were slightly less than the normal-weight nondiabetic controls. Immunoreactive insulin concentrations also approached the normal nonobese level. Plasma FFA concentrations fell to normal levels or lower, possibly indicating an absence of enhanced FA mobilization under these circumstances. Plasma triglycéride concentrations also tended to be lower than in the insulin-treated group and were now comparable to normal subjects. In the calorie-restricted group, urinary ketones were similar to the insulin-treated group but tended to be lower than in the obese GDM group. When individual paired differences were examined, comparing end of week 1 and week 2 observations, the trends seen in the group means were confirmed (Table 11). Fasting plasma glucose and triglycéride reductions were both greater in the diet-treated group compared to the insulin-treated group. With respect to urinary ketone excretion by Ketostix (Ames Diagnostics) testing, it can be seen that only diet-treated patient 3 experienced a slight increase in ketonuria. A decrease in ketonuria with insulin therapy was consistently seen in all three subjects and was modest in two and marked in the one GDM subject with the more marked ketonuria on the baseline diet. Because plasma triglycéride lowering seemed greater with calorie restriction than with insulin therapy, we examined the effects of the two treatment approaches on lipoprotein cholesterol. These results are shown in Table 12. Total cholesterol fell consistently with insulin therapy and appeared to be related to a decline in lowdensity-lipoprotein cholesterol. A consistent decrease in high-density-lipoproteinj cholesterol (HDL3-C) was also

660

seen with variable changes in total HDL-C and HDLj-C. With caloric restriction there were no consistent changes. The reduction in total triglycéride (Table 10) would suggest a reduction in VLDL cholesterol, but this parameter was not measured directly. An attempt was made to measure hepatic glucose production and insulin sensitivity in these subjects using the insulin clamp technique and a stable nonradioactive isotope [δ,ο-2!^] (dideuterated) of glucose (see methods in the Appendix to this paper). As shown in Table 13, basal hepatic glucose output (HGO) was indistinguish able in the baseline week observations of normal obese and GDM obese subjects, confirming Kalhan et al [37]. HGO tended to be higher in the two insulin-treated obese subjects and was lowest in the calorie-restricted group, but none of these differences are significant. The calcu lated extent to which insulin must be increased above basal to reduce hepatic glucose production by 50% was not significantly changed with calorie restriction or in sulin therapy compared to the obese GDM subjects studied in the baseline week. On the other hand, the nondiabetic obese and lean subjects tended to have a greater sensitivity to the inhibitory effect of insulin. The percent suppression of hepatic glucose output for the increment in insulin observed during the last 30 minutes of the glucose clamp is illustrated in Figure 7. Individual values and changes for the treated subjects are shown in Table 14. The normal range is quite wide, and, if anything, the obese subjects suppress more readily than the lean nondiabetic subjects. On the right side of the illustration are the corresponding plots for GDM sub jects. They all fall within the range of the nondiabetic subjects. With diet therapy, two subjects moved in the

VOL. 10, NO. 6


Diabetes

% of basal IRI

100

100

90

80 70 60

%Of


basal HGO

*° -

20 10 -

0 -I

Fig. 7. Percent reduction in hepatic glucose output (HGO) from basal vs percent observed increase in immunoreactive insulin (IRI) from basal as measured in the last 30 minutes of the 2-hr insulin clamp. The insulin infusion was designed to raise IRI 114 pM/L (16 μΙΙ/ml) above basal and was maintained at a rate of 10 mU/m2/min or 71.3 umol/m2/min.

direction of greater insulin resistance and one toward more sensitivity. Similarly, the amount of glucose needed to maintain plasma glucose concentrations during a 2-hr insulin infusion designed to increase the basal insulin level by 114 pM/L (16 μυ/ml) (Table 13) was lower in GDM subjects compared to normal and obese subjects, with the greatest glucose infusion rates observed in normal subjects. The important point from these measurements is that the calorie restriction did not appear to enhance the amount of glucose metabolized compared to no treatment in the first week or compared to insulin therapy in the second week. Estimates of glucose utilization during the insulin clamp procedure (MI) showed no significant differences but some interesting trends. The greater drop compared to normal controls was seen in the obese nondiabetic group. The GDM group was only slightly lower than the obese. There was no apparent effect of either insulin or calorie restriction. Likewise, estimates of insulin sen sitivity (S,) from the insulin clamp (Table 13) were slightly lower in obese GDM subjects than in obese or normal nondiabetic subjects. The lowest insulin sen sitivity was observed in the obese GDM group, with

decreased caloric intake consistent with the changes in Figure 7. The absence of significantly diminished insulin sensitivity in GDM subjects is in keeping with the ab sence of any difference in insulin sensitivity in GDM seen by Buchanan et al [38] using the Bergman minimal model. The data do not confirm the markedly impaired insulin sensitivity reported by Ryan et al [39] in GDM subjects, although our insulin infusion rate of 10 mU/m2/min (71.3 μπκ>1/πι2Ληυΐ) was l/4th and l/24th of the two doses used by Ryan et al [39].

COMPARISON OF THE EFFECTS OF 50% AND 33% CALORIE RESTRICTION Because of the limited number of subjects studied, definite conclusions about the mechanism of the effects of the 33% calorie restriction vs insulin treatment on glucose homeostasis in GDM cannot be drawn. How ever, these observations are consistent with the absence of a change in glucose tolerance tests performed in the previous study at 50% caloric restriction, as no major change in hepatic glucose output or insulin sensitivity


661


Diabetes

Table 14. Effect of the Insulin Clamp Procedure on Plasma IRI and Hepatic Glucose Output (HGO) in Gestational Diabetes IRI (% of basal)

Calorie restriction, 30% Patient 1

2 3

Weekl

Week 2

Δ

Weekl

Week 2

Δ

124 218 145

215 177 189

91 -41

42 46 44

19 50 41

23 -4

44 31

Mean Insulin: 20 NPH, 10 Reg Patient 2

147

3


HGO (% of basal)

168 127

3 7

38 56

51

-20

5

'Following a 120-min period of basal observations, crystalline insulin was infused at a rate calculated to increase the insulin concentration by 16 μυ/ml (l 14 pmol/L). The increase in immunoreactive insulin (IRI) and decrease in hepatic glucose control are assessed as percent change in the last 30 minutes of a 2-hr primed insulin infusion (insulin clamp) compared to the final 30 minutes of the basal observation period.

Table 15. Comparative Effects of Two Levels of Caloric Restriction or Insulin on Metabolism in Gestational Diabetes* Percent change Calorie-restricted diet

Fasting glucose 24 hr x glucose Fasting IRI FFA Ketonuria Triglycéride

50%

33%

Insulin

-17 -21 -Λ5 64 210 -6

-22 -10 -31 -41 -43 -35

0 -4 28 12 -57 13

'Based on Magee et al [33] and previously unpublished data of Magee and Knopp.

was seen with 33% calorie restriction (Table 14 and Fig. 7). Again, the conclusion is suggested that the im proved glycémie pattern seen with caloric restriction is related to diminished carbohydrate intake. To compare the effect of the two levels of calorie restriction and insulin on plasma fuels in the two dif ferent investigations, the percentage responses are presented in Table 15. Here it can be seen that fasting plasma glucose concentrations were reduced by 17% with 50% calorie restriction and 22% with the 33% calorie restriction, while the insulin-treated group

662

achieved no improvement. Twenty-four-hour mean glucose concentrations were reduced by 21% with the 50% calorie restriction, 10% with the 33% restriction, and only 4% with insulin therapy. Thus, even at a level of 33% calorie restriction, these preliminary data suggest a greater benefit with calorie restriction than with prophylactic insulin therapy. In addition, even a 10% reduction in overall 24-hr glucose profile was sufficient to shift a mild GDM subject to the level of normal nonobese subjects (Table 10). Reductions in plasma im munoreactive insulin in the 33% calorie-restricted group

VOL. 10, NO. 6


Diet and Gestational Diabetes were about two-thirds of those in the 50%-restricted group, as expected. While not likely to be precise with these small numbers of observations, the extent of FFA mobilization and ketonuria gives no indication of being exaggerated with 33% calorie restriction. This finding is in marked contrast to the increases seen with 50% calorie restriction. Finally, a reduction in plasma triglycéride concentrations was seen in both calorie-restricted diets compared to insulin therapy or a full-calorie diet Though definite conclusions cannot be made, the impression is reinforced that caloric restriction improves plasma triglycéride levels, unlike the rise seen in animal models [17,18]. The significance of reducing plasma triglycéride concentrations in nondiabetic as well as in GDM subjects is associated more strongly with infant macrosomia than even fasting plasma glucose [36]. Since lipoprotein lipase is present in the placental trophoblast in high concentrations [40], it is possible that FFA derived from triglycérides can cross the placenta in increased amount and contribute to the macrosomia of the infant of a GDM mother. Overall, these observations regarding the advantages of 33% calorie restriction in obese GDM patients are similar to those of Algert et al [10], Borberg et al [32], and Maresh et al [9], which are that maternal glycémie status in GDM can be improved without adverse affect on ketone body metabolism with moderate caloric restriction to a range of 1600-1800 kcal daily and an overnight fast of ~ 9 hours.

RELATIONSHIP OF INCREASED MATERNAL KETOGENESIS TO FETAL GROWTH AND DEVELOPMENT Concerns about the potential pathophysiologic effect of ketonemia on fetal growth and development had their origin in the study of Churchill and Berendes [41], which indicated that ketonuria measured on the day of delivery was associated with reduced IQ. However, a subsequent review of these data [42] indicated that these observations were not consistent, and were in fact probably related to the fact that women who had ketonuria had amniotic infection at the time of birth. Likewise, an increase in maternal ketonemia in association with reduced IQ in infants of diabetic mothers was reported [43], but then not confirmed [44]. Even if the ketonuria in GDM were associated with lower IQ in offspring [43], it seems more likely that the overall poorer diabetic control resulting in ketonemia is more responsible for the lower IQ than an effect of ketone bodies per se. A recent report from the Freinkel group [45] supports this view. It is further inter-

esting to learn that psychomotor evaluation of Dutch military recruits who were exposed to different degrees of maternal food restriction during their gestation in the Dutch famine of 1944 and 1945 showed no impairment in IQ compared to those not bom during this time [46]. Indeed, ample precedent for caloric restriction in obese women in pregnancy is available from the experience of Algert et al [10], Jovanovic-Peterson [47], Pedersen over many years [48], and, most recently, the National Academy of Sciences [49]. In any case, there is no consistent epidemiologie link between maternal ketonuria per se and impaired fetal growth and development. Another way to gauge the significance of maternal ketonemia in full or restricted caloric intake in GDM is to compare the extent to which plasma ketone body concentrations are elevated in diabetic subjects. In one such paper, Stangenberg et al [50] reported plasma ketone levels between 0.5 and 0.75 mM at 8:00 am in the morning after overnight fast in insulin-dependent diabetics, and concentrations of 0.52 to 0.40 mM in non-insulindependent subjects with residual ß cell function. These concentrations are equal to the level of Metzger et al [26]. Finally, in the recently reported Diabetes in Early Pregnancy project (DIEP) [51], plasma ketone bodies were measured at 6 weeks and approximated at 0.35 mM compared to 0.1 or 0.15 mM after overnight fast in nondiabetic control pregnant subjects (DIEP authors, manuscript in preparation). Thus, even in the first trimester, ketonemia is more likely to be observed in the diabetic than in the nondiabetic subject. Although treatment goals should be judged by the nondiabetic excursions in plasma fuels in pregnancy rather than in the diabetic, and although malformations have been reported to be associated with culture of rat fetuses in ketonemic incubation media [52,53], the occurrence of congenital malformations in the DIEP project was not in fact associated with the level of ketone bodies (DIEP authors, manuscript in preparation). Thus, even here there is no detectable link between ketonemia and infant outcome in a carefully performed study.

CONCLUSIONS In conclusion, the data indicate that a promising degree of glycémie and plasma triglycéride lowering can be seen with moderate caloric restriction of approximately 33% of calories, or 1600-1800 kcal/day in the management of the obese GDM patient. There is evidence for little or no increased plasma FFA or ketone body mobilization after a 9-hr fast under controlled metabolic ward conditions. None of these changes were associated with enhanced insulin sensitivity or altered


663



Diabetes

hepatic glucose production, suggesting that reduced caloric intake is important in yielding lower fasting and 24-hr glucose profiles. Preliminary data presented in this manuscript suggest that calorie restriction may prove to be a more beneficial therapeutic approach than prophylactic insulin itself in treating mild GDM with fasting glucose < 90 mg/dl (5 mM) and 1-hr pc glucose < 120 mg/dl (6.67 mM). The argument that maternal ketonemia is associated with impaired fetal outcome is not supported by a critique of earlier studies. Finally, caloric restriction is only proposed for use in the obese GDM patient, a condition where it is pathophysiologically most relevant, and at a time that is well removed from any possible implication regarding generation of malformations. More research is required in controlled metabolic ward and clinical trial designs to further compare the efficacy of insulin therapy vs calorie restriction in the management of the GDM patient.

ACKNOWLEDGMENTS The authors express their gratitude to Dr. Zane Brown for assistance in identifying patients and to Al Greeves for typing the manuscript, and Virginia Fitzpatrick for statistical analysis. This work was supported by Grant DK 35816 (CNRU), Grant MO1RR-0037 (CRC), and DK 17047, a Diabetes Research in Endocrinology Center Grant. This study was also supported by a training grant in Endocrinology Metabolism (DK 07247) to Dr. Magee, a Research Fellowship from the American Diabetes Association Washington Affiliate, and a generous grant from the Washington State Eagles.

APPENDIX Methods Used in the 33% Calorie Restriction Studies Because most women with GDM are obese, and GDM occurring in nonobese women may represent a subgroup with a different pathophysiology [54], we recruited women who were moderately obese, at approximately 28 ± 5 weeks of gestation, and met the criteria for the diagnosis of GDM as modified by Coustan [8]. Moderate obesity was defined as 120-185% of prepregnancy ideal body weight, with ideal body weight being defined by frame size determination [55], height measurement, and use of the modified 1959 Metropolitan Life Insurance tables. Patients were recruited through the University of Washington obstetrics clinics

664

and solicited referrals from area physicians. After an initial interview, a full explanation of the study was provided and signed informed consent obtained. The first week of the study was identical for all patients and consisted of prescription and teaching of a full-calorie American Diabetes Association diet [35 kcal/day/kg of pregnancy IBW plus 300 kcal/day pregnancy allowance (approximately 2200-2600 kcal/day)]. Food group distribution was 50% carbohydrate, 30% fat, 20% protein, with avoidance of sucrose or fructose. The distributions of daily calories were approximately 3/18 at 0800 hours; 5/18 at 1200 hours; 2/18 at 1500 hours; 5/18 at 1800 hours; and 3/18 at 2200 hours. Diet education and arrangements for the patients to pick up 2-3 days worth of prepackaged meals at a time from the CRC kitchen (University of Washington Hospital) were provided for days 1-4 of the study week. Subjects were taught the use of a desktop reflectance meter (Accucheck, Boehringer Mannheim Diagnostics, Indianapolis, IN) for determination of fingerstick glucose values. These values were obtained seven times per day 3 days per week (before each meal, 90-120 minutes after each meal, and before bedtime snack). Records of these values were kept in a logbook by the patients. On day 5, the patients were admitted to the CRC at the University of Washington Hospital after their evening meal. On the morning of days 6 and 7 and after an overnight fast, blood was drawn for glucose, insulin, FEA, ß-hydroxybutyrate, glycerol, and full lipoprotein quantifications. Additionally, first morning double voided urine samples were collected on these mornings and tested for ketones using Ketostix. On the morning of day 6, a heparin lock or nondextrose-containing IV was begun at 0700. Hourly glucose and insulin samples were drawn until 2300 that evening and at 0300 and 0700 hours the next morning. Additionally, samples were drawn every 30 minutes x4 after each major meal on day 6. On the morning of day 7, again after an overnight fast, patients underwent a hyperinsulinemic euglycemic clamp coupled to a 4-hr infusion of the stable isotope ([6,6-2HJ glucose). This euglycemic insulin clamp procedure was performed as follows: at -20, -10, and 0 minutes, samples were drawn for determination of plasma glucose and background concentration of ([6,6-2H2] glucose. At "time 0" a primed continuous infusion of ([6,6-^] glucose was begun and was continued until the two 2-hr steps of the procedure were completed. During the last 30 minutes of each step, four samples were drawn for determination of glucose, [6,6-2H2] glucose, and insulin levels. Beginning at 120 minutes, a primed continuous infusion of insulin was begun and concurrently every 10-min sampling of plasma glucose and a variable in-

VOL. 10, NO. 6


Diet and Gestational Diabetes fusion of 20% dextrose in water was begun. The dextrose infusion was adjusted to keep the plasma glucose value in the euglycemic range (80-90 mg/dl) using measurements of glucose with a glucose oxidase analyzer (Yellow Springs Instrument, Yellow Springs, OH). This procedure has been described in detail else where [56,57]. After completion of this test, obese GDM subjects were randomized to either of two interventions to be studied: a 33% calorie-restricted diet (1600-1800 calories) or insulin treatment. For those patients who were euglycemic during the first week of study (as judged from the 24" glucose profile on day 6) the insulin dosage was standardized as 20 U NPH and 10 U Regular each morning [8]. The patients were discharged either at the end of day 7 or day 8 and continued their assigned intervention for the duration of the second week. Home glucose monitoring through the use of the reflectance meter and dispensing of prepackaged meals from the CRC kitchen was continued as during the first week of the protocol. The patients were admitted after their eve ning meal on day 12 of the study, and on days 13 and 14 repeated the tests previously performed on days 6 and 7 (fasting blood sampling of insulin and intermediary fuels, 24° glucose/insulin profile, and hyperinsulinemic euglycemic clamp coupled to a 4* infusion of ([6,6-2HJ glucose).

Calculations The following formulas were used in the hyperinsulinemic-euglycemic clamp coupled with the ([6,6-2HJ glucose infusion. 1. Insulin dosing. A priming dose of 1.56 μπκ>1/ηι2Αηίη (220 mU/m2) was given 120 minutes into the procedure (end of the basal step). This dose was immediately followed by initiation of a continuous in fusion of 71.3 jAmol/m2/min (10 mU/m2/min) and yielded an insulin step above basal of approximately 114 pM/L (16 μυ/ml) (derived from [58]). The achieved insulin levels allowed plotting at the approximate midpoint of the line relating HGO to ambient insulin (derived from [59,60]). 2. Insulin sensitivity index S,. S, is a ratio of the change in insulin action to the change in insulin level that produced it and may be calculated from [58]

the mean glucose during the two clamp steps being ex amined (mg/dl). 3. Stable isotope priming doses and infusion rates. The doses necessary for attainment of a target concentra tion [commonly expressed as atom percent excess (APE) and in the range of 1-2% above background] are dis cussed in detail in [61]. Briefly, the calculations are p/i - (C · vyR,,

where P = bolus dose in mg of tracer, I = infusion rate in mg/min of tracer, C = basal concentration in the plasma of the substance pool being enriched (in this case, plas ma glucose expressed as mg/ml), V = rapid volume of distribution for glucose (40 ml/kg), and R, = rate of en dogenous glucose appearance, i.e., HGO. This estimate is based upon the fasting plasma glucose concentration [56] and is commonly in the 1.8-4.0 mg/kg/min range. This calculation gives P = (X)I, where X is an integer. I is then derived from the formula R, = [(IE infusate - 1)/(Œ plasma)] I,

where R, = as above, IE infusate = tracer enrichment of the infusate in percent, commonly 98-99% as supplied in the purity information accompanying each vial of [6,6-2HJ glucose from the manufacturer (Merck, St. Louis, MO), and IE plasma = targeted enrichment of the isotope above background: P = (X) I. 4. Calculation of HGO. HGO is calculated using Steele's equations [62,63] and is discussed in detail in [61]. For HGO: R, - F[(IE infusateMIE plasma x 0.01)]"1,

where R, = HGO (in this case = "endogenous R,," since there is no concurrent glucose infusion during the basal 2-hr study and "exogenous R," (glucose infusion rate in mg/kg/min) was subtracted from "total R," when calculating HGO for the second 2-hr period during which insulin and D20W were infused). Also, F = infusion rate of tracer in mg/kg/min, IE infusate = tracer enrichment of infusate (approximately 99%), and IE plasma = mean percent enrichment above background achieved during the last 30 min of the steady state infusion. Additional tracer was added to the DMW infusion per [61].

REFERENCES

S, (damp) = (ΔINF/Δ 1RIVG,

where ΔINF is the change in mean glucose infusion (mg/kg/min) required to maintain euglycemia between the two segments of the clamp being evaluated; ΔIRI is the change in mean insulin levels between the two seg ments of the clamp being evaluated (μυ/ml); and G is


1. Hoet JP, Lukens FDW: Carbohydrate metabolism during pregnancy. Diabetes 3:1-12,1954. 2. O'Sullivan JB, Mahan CM: Criteria for the oral glucose tolerance test in pregnancy. Diabetes 13:278-285,1964. 3. Knopp RH, Montes A, Warm MR: Carbohydrate and lipid metabolism in normal pregnancy. In "Laboratory Indices

665

Diet and Gestational Diabetes

4.

5.

6.

7.


8.

9.

10.

11.

12.

13. 14.

15.

16.

17.

18.

19. 20.

666

of Nutrional Status in Preganancy." Washington, DC: Food and Nutrition Board, National Academy of Sciences, pp 35-88, 1978. National Diabetes Data Group: Classification and diag nosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 28:1039-1057,1979. Sepe SJ, Connell FA, Geiss LS, Teutsch SM: Gestational diabetes: incidence, maternal characteristics and perinatal outcome. Diabetes 34(Suppl 2):13—16,1985. Cousins L: Pregnancy complications among diabetic women: review 1965-1985. Obstet Gynecol Surv 42:140149, 1987. O'Sullivan JB: Insulin treatment for gestational diabetes. In Camerini-Davalos RA, Cole HS (eds): "Early Diabetes in Early Life." New York: Academic, pp 474-454,1975. Coustan DR, Imarah J: Prophylactic insulin treatment of gestational diabetes reduces the incidence of macrosomia, operative delivery and birth trauma. Am J Obstet Gynecol 150:836-843,1984. Maresh M, Gillmer MDG, Beard RW, Alderson CS, Bloxham BA, Elkeles RS: The effect of diet and insulin on metabolic profiles of women with gestational diabetes mellitus. Diabetes 34(Suppl 2):89-93, 1985. Algert S, Shragg P, Hollingsworth DR: Moderate caloric restriction in obese women with gestational diabetes. Obstet Gynecol 65:487-491,1985. Bokelmann O, Bock A: Beitrag zum intermediären Fet tsoffwechsel in der Schwangerschaft: nahrungsfett und Acetokörper. ArchGynak 129:541-549,1927. MacKay EM, Barnes RH: Fasting ketosis in the pregnant rat as influenced by adrenalectomy. Proc Soc Exp Biol Med 34:682-683, 1983. Scow RO, Chernick SS, Smith B-B: Ketosis in the rat fetus. Proc Soc Exp Biol Med 98:833-835,1958. Herrera E, Knopp RH, Freinkel N: Carbohydrate metabo lism in pregnancy. VI. Plasma fuels, insulin, liver com position, gluconeogenesis, and nitrogen metabolism during late gestation in the fed and fasted pregnant rat. J Clin Invest 48:2260-2272, 1969. Felig P, Lynch V: Starvation in human pregnancy: hypoglycemia, hypoinsulinemia, and hyperketonemia. Science 170:990-992,1970. Knopp RH, Herrera E, Freinkel N: Carbohydrate metabo lism in pregnancy. VIII. Metabolism of adipose tissue iso lated from fed and fasted pregnant rats during late gesta tion. J Clin Invest 49:1438-1446, 1970. Scow RO, Chernick SS, Brinley MS: Hyperlipemia and ketosis in the pregnant rat. Am J Physiol 206:796-804, 1964. Knopp RH, Warth MR, Charles D, Childs M, Li JR, Mabuchi H, Van Allen M: Lipoprotein metabolism in pregnancy, fat transport to the fetus and the effects of diabetes. Biol Neonate (Paris) 50:297-317,1986. Knopp RH: Diabetes in pregnancy: the internist's view. Univ Washington Med 7(2):20-28,1980. Knopp RH, Montes A, Childs M, Li JR, Mabuchi H: Me tabolic adjustments in normal and diabetic pregnancy. Clin Obstet Gynecol 24:21-49,1981.

21. Freinkel N: Effects of the concepts on maternal metabo lism during pregnancy. In Leibel BS, Wrenshall GA (eds): "On the Nature and Treatment of Diabetes." Amsterdam: Excerpta Medica Foundation, pp 679-691, 1965. 22. Freinkel N: Of pregnancy and progeny. Diabetes 29:10231035, 1980. 23. Morrow PG, Marshall WP, Kim H-J, Kalkhoff R: Metabo lic response to starvation. I. Relative effects of pregnancy and sex steroid administration in the rat. Metabolism 30:268-278,1981. 24. Knopp RH, Childs MT, Warth MR: Dietary management of the pregnant diabetic. In Winick M (ed): "Current Con cepts in Nutrition." New York: Wiley, Vol 6, pp 119-139, 1979. 25. Knopp RH, Saudek CD, Arky RA, O'Sullivan JB: Two phases of adipose tissue metabolism in pregnancy: mater nal adaptations for fetal growth. Endocrinology 92:984988, 1973. 26. Metzger BE, Ravnikar V, Vileisis RA, Freinkel N: "Ac celerated starvation" and the skipped breakfast in late nor mal pregnancy. Lancet 1:588-593,1982. 27. Coetzee EJ, Jackson WPU, Berman PA: Ketonuria in pregnancy with special reference to calorie-restricted food intake in obese diabetics. Diabetes 29:177,1980. 28. Chez RA, Curcio FD III: Ketonuria in normal pregnancy. Obstet Gynecol 69:272-274,1987. 29. Paterson P, Sheath J, Taft P, Wood C: Maternal and fetal ketone concentrations in plasma and urine. Lancet 1:862, 1967. 30. Brunzell JD, Lerner RL, Hazzard WR, Porte D Jr, Bierman EL: Improved glucose tolerance with high car bohydrate feeding in mild diabetes. N Engl J Med 284: 521-524, 1972. 31. Warth MR, Knopp RH: Lipid metabolism in pregnancy. V. Interactions of diabetes, body weight, age and high car bohydrate diet. Diabetes 26:1056-1062, 1977. 32. Borberg C, Gillmer MDG, Brunner EJ, Gunn PJ, Oakley NW, Beard RW: Obesity in pregnancy. The effect of dietary advice. Diabetes Care 3:476-481, 1980. 33. Magee MS, Knopp RH, Benedetti TJ: Metabolic effects of a 1200 calorie diet in obese pregnant women with gesta tional diabetes. Diabetes 39:234-240, 1990. 34. Magee MS, Knopp R, Raisys V, Prince E, Brown Z, Benedetti T: Pathophysiology and response to insulin or dietary treatment in gestational diabetes. Diabetes 37(Suppl I):1HA, 1988 (abstract). 35. Hollander M, Wolfe DA: "Nonparametric Statistical Methods." Wiley: New York, pp 120-123,1973. 36. Knopp RH, Magee MS, Larson MP, Benedetti T: Alterna tive screening tests and birth weight associations in preg nant women with abnormal glucose screening. Diabetes 37(Suppl I):110A, 1988 (abstract). 37. Kalhan SC, D'Angelo LJ, Savin SM, Adam PAJ: Glucose production in pregnant women at term gestation. Sources of glucose for human fetus. J Clin Invest 63:388-394, 1979. 38. Buchanan TA, Metzger BE, Freinkel N, Bergman RN: In sulin sensitivity and B-cell responsiveness to glucose

VOL. 10, NO. 6

Diet and Gestational Diabetes

39.

40.

41.

42.


43.

44.

45.

46.

47.

48. 49.

50.

51.

during late pregnancy in lean and moderately obese women with normal glucose tolerance or mild gestational diabetes. Am J Obstet Gynecol 162:1008-1014,1990. Ryan EA, O'Sullivan MJ, Skyler JS: Insulin action during pregnancy. Studies with the euglycemic clamp technique. Diabetes 34:380-389, 1985. Bonet B, Gown AM, Brunzell JD, Knopp RH: Role of lipoprotein lipase in triglycéride metabolism by placental trophoblast and macrophages. Arteriosclerosis 7:642A, 1987 (abstract). Churchill JA, Berendes HW: Intelligence of children whose mothers have acetonuria in pregnancy. In "Perinatal Factors Affecting Human Development," Sci Pubi No 185. Washington DC: Pan American Health Orgization, 1969. Naeye RL, Chez RA: Effects of maternal acetonuria and low pregnancy weight gain on children's psychomotor development. Am J Obstet Gynecol 139:189-193,1981. Stehbens JA, Baker Gì, Kitchell M: Outcome at ages 1,3, and 5 years of children born to diabetic mothers. Am J Obstet Gynecol 127:408-413,1977. Naeye RL: The outcome of diabetic pregnancies. A prospective study. In "Pregnancy Metabolism, Diabetes, and the Fetus," Ciba Foundation Symposium 63. Amster dam: Excerpta Medica, 1979. Rizzo T, Freinkel N, Metzger BE, Hatcher R, Burns W, Barglow P: Fuel mediated teratogenesis: diabetic pregnan cies and Brazelton tests in the newborn. Diabetes 37:86a, 1988. Stein Z, Susser M, Saenger G, Marolla F: "Famine and Human Development." New York: Oxford University Press, 1975. Jovanovic-Peterson L, Peterson CM: Dietary manipulation as a primary strategy for pregnancies complicated by diabetes. J Am Coll Nutr 9:320-325,1990. Pedersen J: "The Pregnant Diabetic and Her Newborn," 2nd ed. Baltimore: Williams and Wilkins, pp 1-280,1977. National Academy of Sciences: "Nutrition in Pregnancy, Part I: Weight Gain." Washington, DC: National Academy Press, pp 10-13, 1990. Stangenberg M, Persson B, Fredholm BB, Lunell NO: Profiles of intermediary metabolites in insulin-dependent pregnant diabetic women with or without endogenous in sulin production. Diabetes Care 5:409-413,1982. Mills JL, Knopp RH, Simpson JL, Jovanovic-Peterson L, Metzger BE, Holmes LB, Aarons JH, Brown Z, Reed GF, Bieber FR, Van Allen MI, Holzman I, Ober C, Peterson

52.

53.

54.

55. 56.

57.

58.

59.

60.

61. 62.

63.

CM, Withiam MJ, Duckies A, Mueller-Heubach E, Polk BF: NICHD-Diabetes in Early Pregnancy project: in creased malformation rates in infants of diabetic mothers is unrelated to glycémie control during organogenesis. N Engl J Med 318:671-676, 1988. Horton WE, Sadler TW: Effects of maternal diabetes on early embryogenesis: alterations in morphogenesis pro duced by the ketone body, ß-hydroxybutyrate. Diabetes 32:610-616,1983. Weigensburg M, Sobel R, Garcia-Palmer F, Freinkel N: Temporal differences in vulnerability to fuel mediated teratogenesis. Diabetes 37(Suppl 1):85A, 1988 (abstract). Ward WK, Jonston CL, Beard JC, Benedetti TJ, Porte D Jn Abnormalities of islet B-cell function, insulin action and fat distribution in women with histories of gestational diabetes: relationship to obesity. J Clin Endocrinol Metab 61:1039-1045, 1985. Grant JP: "Handbook of Total Parenteral Nutrition." Philadelphia: WB Saunders, p 15,1980. Revers RR, Fink RI, Griffin J, Olefsky JM, Letteman OG: Influence of hyperglycémie on insulin's in vivo effects in type II diabetes. J Clin Invest 72:664-677,1984. Bogardus C, Ravussin E, Robbins DC, Wolfe RR, Horton ES, Sims EAH: Effects of physical training in diet therapy on carbohydrate metabolism in patients with glucose in tolerance and NIDDM. Diabetes 33:311-318,1984. Beard JC, Bergman RN, Ward WK, Porte D Jr: The in sulin sensitivity index in nondiabetic man. Diabetes 35:362-369,1986. Firth RG, Bell PM, Rizza RA: Effects of tolazamide and exogenous insulin on insulin action in patients with NIDDM. N Engl J Med 314:1280-1286,1986. Rizza RA, Mandarino LJ, Gerich JE: Dose-response char acteristics of insulin on production and utilization of glucose in man. Am J Physiol 240:E630-E639,1981. Wolfe RR: 'Tracers in Metabolic Research." New York: Wiley-Liss, 1984. Steele R: Influence of glucose loading and of injected in sulin on hepatic glucose output. Ann NY Acad Sci 82:420, 1959. Chaisson JL, Liljenquist JE, Lacy WW, Jenning AS, Cherrington AD: Gluconeogenesis: methodological approaches in vivo. Fed Proc 36:279,1977.

Received December1990;


revision accepted March 1991.

667

Metabolic effects of hypocaloric diets in management of gestational diabetes.

The metabolic response to hypocaloric protein diets in obese man.

Nutritional management of the obese gestational diabetic pregnant woman.

The influence of gestational weight gain on the development of gestational hypertension in obese women.

Brain Functional Connectivity Is Modified by a Hypocaloric Mediterranean Diet and Physical Activity in Obese Women.

Low Gestational Weight Gain in Obese Women and Pregnancy Outcomes.

Menopause care for obese and diabetic women.

Low Gestational Weight Gain and Risk of Adverse Perinatal Outcomes in Obese and Severely Obese Women.

Overweight PCOS Women under Hypocaloric Diet.

OS003. The influence of gestational weight gain on the development ofgestational hypertension in obese women.

Energy-restricted, high-protein diets more effectively impact cardiometabolic profile in overweight and obese women than lower-protein diets.

Is exercise safe or useful for gestational diabetic women?

Gestational weight gain and maternal and neonatal outcomes in term twin pregnancies in obese women.

Accelerated starvation in late pregnancy: a comparison between obese women with and without gestational diabetes mellitus.

Comparison of the effects of gestational weight gain on pregnancy outcomes between non-diabetic and diabetic women.

The effects of inhibition of gluconeogenesis on ketogenesis in starved and diabetic rats.

Excess Gestational Weight Gain in Low-Income Overweight and Obese Women: A Qualitative Study.

Elevated serum squalene and cholesterol synthesis markers in pregnant obese women with gestational diabetes mellitus.

The Effect of Low Carbohydrate Diets on Fertility Hormones and Outcomes in Overweight and Obese Women: A Systematic Review.

Effects of gestational weight gain in morbidly obese women: II: Fetal morbidity.

Pregnancy outcome of overweight and obese Chinese women with gestational diabetes.

Metabolic effects of 1200-kcal diet in obese pregnant women with gestational diabetes.

Effects of gestational weight gain in morbidly obese women: I. Maternal morbidity.

Exercise in pregnancy does not alter gestational weight gain, MCP-1 or leptin in obese women.