P rotein Metabolism in Insulin-Tre a t e d Gestational Diabetes. Diabetes Care 22: , 1999

Transcription

1 P a t h o p h y s i o l o g y / C o m p l i c a t i o n s O R I G I N A L A R T I C L E P rotein Metabolism in Insulin-Tre a t e d Gestational Diabetes NANCY F. BUTTE, PHD HELEN W. HSU, MD MARY THOTATHUCHERY, MS WILLIAM W. WONG, PHD JANE KHOURY, MS PETER REEDS, PHD O B J E C T I V E To test the hypothesis that protein metabolism is not totally normalized in i n s u l i n - t reated gestational diabetes mellitus (GDM) patients compared with normal, pre g n a n t c o n t rol subjects. RESEARCH DESIGN AND METHODS P rotein metabolism in eight Hispanic women with insulin-treated GDM and eight healthy Hispanic control women was studied in late gestation and at 6 weeks postpartum. Nitrogen flux was assessed from the disposal rate of [ 1 5 N]-labeled urea over 12 h after a dose of [ 1 5 N]-labeled leucine. Plasma amino acid concentrations were determined in fasting and 2-h postprandial samples using an amino acid analyzer. R E S U LT S P rotein turnover was normalized in insulin-treated GDM; however, fasting and postprandial plasma amino acids were elevated antepartum and postpartum. Nitrogen flux was s i g n i ficantly lower during pregnancy (P = ) and did not differ between groups. Fasting and postprandial plasma amino acids were elevated in GDM antepartum and postpart u m, despite satisfactory glycemic control. Fasting levels of taurine, hydro x y p roline, glutamic acid, glutamine, cystine, tyrosine, phenylalanine, tryptophan, and histidine were higher in GDM antepartum and postpartum (P 0.05). Postprandial concentrations of taurine, hydro x y p roline, valine, cystine, isoleucine, leucine, tyrosine, phenylalanine, tryptophan, ornithine, lysine, histidine, and a rginine were higher in GDM antepartum and postpartum (P 0.05). With few exceptions, plasma amino acid concentrations were lower antepartum than postpartum (P ). C O N C L U S I O N S P rotein turnover was normalized in insulin-treated women with GDM; h o w e v e r, fasting and postprandial plasma concentrations of amino acids were elevated in the a n t e p a rtum and postpartum periods, despite satisfactory maternal glycemic control. Diabetes Care 22: , 1999 Gestational diabetes mellitus (GDM) is a metabolic disorder affecting all insulindependent fuels. Approximately 5% of all pregnancies are afflicted with GDM. Despite conventional diet and insulin therapy, macrosomia and attendant complications are still prevalent among infants of mothers with GDM (1). GDM is characterized by elevations in fasting, postprandial, and integrated 24-h plasma levels of glucose, amino acids, and lipids (2). Fetal pancreatic -cell development may be enhanced by minimal elevations in plasma glucose and other fuels, particularly insulinogenic amino acids. The ensuing fetal hyperinsulinemia, together with excess nutrient supply, may enhance fetal g rowth, even if glycemic control is satisfactory in late gestation. Although the importance of correcting metabolic derangements in GDM is believed to be essential, it has not been achieved in practice. Research and clinical management of GDM F rom the U.S. Department of Agriculture/Agricultural Research Services (USDA/ARS) Childre n s Nutrition R e s e a rch Center (N.F.B., M.T., W. W. W., P.R.), the Department of Pediatrics, Baylor College of Medicine and Texas Childre n s Hospital, Houston, Texas; and the Department of Obstetrics and Gynecology (H.W.H., J.K.), University of Cincinnati Medical Center, Cincinnati, Ohio. A d d ress correspondence and reprint requests to Nancy F. Butte, PhD, Childre n s Nutrition Research Cent e r, 1100 Bates St., Houston, TX nbutte@bcm.tmc.edu. Received for publication 30 September 1998 and accepted in revised form 30 December A b b re v i a t i o n s : BUN, blood urea nitrogen; FFM, fat-free mass; GDM, gestational diabetes mellitus; NPN, neutral protamine Hagedorn; TBW, total body water. A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances. have generally focused on glycemic contro l. The few studies on protein metabolism in women with GDM have yielded confli c t i n g results. Rates of protein turnover (3 5) have been re p o rted to be increased or unaltere d (6) in GDM. Plasma amino acid concentrations were elevated in one study (7), but w e re normal in another study of women with GDM (8). In part, these discre p a n c i e s can be attributed to diff e rences in study populations, designs, stages of pre g n a n c y, techniques, and treatment modalities. In all these studies, protein metabolism was studied with routine insulin therapy withheld f rom the GDM women. In this study, we hypothesized that p rotein metabolism is not totally norm a l- ized in insulin-treated GDM women comp a red with healthy pregnant contro l women. We compared protein turn o v e r using [ 1 5 N]-labeled leucine and fasting and postprandial plasma amino acids of i n s u l i n - t reated GDM women re c e i v i n g their routine insulin therapy with those of n o rmal, pregnant control women. We also tested for diff e rences in protein metabolism at 6 weeks postpartum because insulin resistance in GDM women may not re s o l v e immediately at part u r i t i o n. RESEARCH DESIGN AND M E T H O D S Design P rotein metabolism in eight women with i n s u l i n - t reated GDM and eight healthy c o n t rol women was studied between 32 and 36 weeks of gestation and at 6 weeks p o s t p a rtum. Eligibility criteria included age of years, parity of 1 3, no known medical illnesses, no obstetrical complications, and no substance abuse. At each study interval, subjects were admitted for 5 days to the Metabolic Research Unit of the Childre n s Nutrition Research Center in Houston, Texas, for stable isotope studies in conjunction with room re s p i r a t i o n c a l o r i m e t ry, anthropometric, and body composition measurements. A companion study has been published (9). The study was approved by the Baylor Aff i l i a t e s Review Board for Human Subject Researc h. Written, informed consent was obtained f rom all women. 806 DIABETES CARE, VOLUME 22, NUMBER 5, MAY 1999

2 Butte and Associates Subjects All subjects were of Mexican or Central American descent. Mean age was 28 ± 5 and 23 ± 2 years for the GDM and contro l g roups, re s p e c t i v e l y. Median (range) gravidity and parity were 3.5 (1 6) and 1.5 (0 3) for the GDM group and 2.0 (1 4) and 1.0 (0 3) for the control group. All women gave birth to healthy, term infants, with the exception of one stillborn infant (weight 2,900 g) in the GDM group. For the GDM and control groups, gestational age of the infants (39.5 ± 0.8 vs ± 1.8 week, respectively), and birth weight (3.6 ± 0.4 vs. 3.7 ± 0.4 kg, respectively) did not differ between groups. Except for one GDM subject and her infant who left the c o u n t ry after delivery, all subjects re t u rn e d for the postpartum studies. All mothers fed their infants formula; in addition, one GDM subject and one control subject also minimally breast-fed their infants. Clinical management The clinical management of the GDM subjects was under the control of the diabetic clinic of Ben Taub General Hospital. After s c reening with a 1-h 50-g oral glucose test, GDM was confirmed by a 3-h, 100-g oral glucose tolerance test according to the O Sullivan criteria (10). Two abnorm a l glucose values were re q u i red to diagnose GDM, and insulin therapy was initiated. An initial dosage of 20 U neutral pro t a- mine Hagedorn (NPN) and 10 U re g u l a r insulin was given in the morning, and 5 U NPN and 5 U regular insulin were given in the afternoon. Home glucose monitoring was perf o rmed in the fasted state and 2 h after breakfast, lunch, and dinner. Insulin dosage was then modified according to each woman s blood glucose diary. A 2,100 kcal/day diet consisting of 20% protein, 45% carbohydrate, and 35% fat was p rescribed, and dietary instruction was given by the staff dietitian. At the time of our pregnancy study, the insulin pre s c r i p- tion consisted of 20.2 ± 1.7 U NPH and 10.0 ± 1.1 U regular insulin in the morning, and 5.5 ± 1.4 U NPH and 5.8 ± 2.1 U regular insulin in the early evening. All GDM subjects were stabilized on their respective dietary and insulin re g i m e n s 2 3 weeks before the first study. No insulin was prescribed postpartum. The clinical chemistries are presented in detail e l s e w h e re (9). A fasting blood sample was analyzed for glucose (Olympus 5000, Olympus Instrument Division, Lake Success, NY) and HbA 1 c by high-perf o rm a n c e liquid chro m a t o g r a p h y, with a norm a l range of % (Bio-Rad Laboratories, H e rcules, CA). At 6 weeks postpartum, a 3-h glucose tolerance test was perf o rm e d (Beckman CX5, Beckman Instru m e n t s, F u l l e rton, CA). Plasma amino acids Plasma amino acid concentrations were d e t e rmined in fasting and 2-h postprandial samples taken at 0700 and 1900, re s p e c- t i v e l y. An amino acid analyzer with a 12-cm h i g h - p e rf o rmance sodium column with sodium buffers (Beckman 6300 Amino Acid Analyzer, Beckman Instruments, Palo Alto, CA) was used. Body composition Total body water (TBW) was measured by dilution of an orally administered dose of deuterium (40 mg 2 H 2 O/kg). 2 H abundance in saliva was measured by gas-isotope ratio mass spectro m e t ry (Delta-E, Finnigan MAT, San Jose, CA). Body density was measured using an underwater weighing system (11). Residual lung volume was estimated by the simplified nitrogen washout method (12). Fat-free mass (FFM) was computed from body density and TBW using the Siri equation (13). Nitrogen flux A modification of the single-dose end pro d- uct method was used to estimate nitro g e n flux, accounting for the amount of [ 1 5 N ] - labeled urea retained in the body urea pool at the end of the experiment (14). Nitro g e n flux was assessed from the disposal rate of [ 1 5 N ] u rea over a 12-h period after a single dose (5 mg/kg) of [ 1 5 N]leucine (98 atom p e rcent excess; Cambridge Isotope Laboratories, Wo b u rn, MA) given orally at Urine was collected at baseline before dosing and continuously thereafter for 12 h. Baseline and 12-h blood samples were taken for determination of urea nitro g e n and 1 5 N abundance. For 1 5 N isotopic measurements, ure a was extracted from plasma and urine using a basic anion exchange resin (Dowex-1, mesh, 8% cross-linked, chloride f o rm; Sigma-Aldrich, St. Louis, MO). Before the extraction, plasma was subjected to protein precipitation using 10% sulfosalicylic acid and adjusted to a ph level of 11 with 10 N NaOH. The plasma supernatant or urine was loaded onto the column and eluded with deionized water. The eluate was neutralized and dried in the combustion thimble for gas-isotope ratio mass spect ro m e t ry (Europa Scientific Tr a c e rm a s s ; E u ropa Scientific, Franklin, OH). The sample was combusted to CO 2, H 2 O, and N x O y with O 2 ; any N x O y was reduced to N 2 t h rough a reduction furnace at 500 C. The 1 5 N content was measured against a laborat o ry working standard that had been calibrated against atmospheric N 2. U rea nitrogen was measured by the B e rthelot reaction after treatment with u rease on an automated COBAS FARA II System (Roche Analytical Instru m e n t s, N u t l e y, NJ). Nitrogen flux calculations With this single-dose method, the cumulative excretion of 1 5 N in the metabolic end p roduct (urea) will tend toward a maximum or asymptote, when all the isotope that is going to be excreted has been cleare d f rom the metabolic pools. The pro p o rt i o n of 1 5 N excreted is equal to the pro p o rtion of the disposal rate excre t e d : Q = D E U ( c o rre c t e d ) / e U, w h e re Q is the rate of nitrogen flux; E u ( c o rre c t e d ) is the excretion rate of urea in 12 h, c o rrected for changes in the size of the ure a pool in the body during the 12 h; D is the dose of 1 5 N; and e u is the amount of 1 5 N e x c reted in the urine as urea, plus the amount retained in the urea body pool at the end of 12 h. During the experiment the subjects w e re in the fed state, having received a stand a rdized diet consisting of 20% pro t e i n, 45% carbohydrate, and 35% fat, distributed into three meals and an evening snack. All food and beverages consumed were weighed and analyzed using the Minnesota Nutrition Data System (Nutrition Coord i- nating Center, University of Minnesota, Minneapolis, MN). Protein intake averaged 1.2 ± 0.3 g k g 1 d a y 1 a n t e p a rtum and 1.3 ± 0.4 g k g 1 d a y 1 p o s t p a rtum and did not differ between groups. Subjects w e re sedentary throughout the day. Wo m e n with GDM administered their insulin, as p rescribed only during pre g n a n c y. Statistical analysis Data were analyzed using SAS (SAS, Cary, NC). The nitrogen flux and fasting and 2-h postprandial amino acid concentrations w e re analyzed using a repeated measure analysis of variance design to incorporate the effects of group (GDM vs. control), time ( a n t e p a rtum vs. postpartum), and interaction between group and time. An a priori DIABETES CARE, VOLUME 22, NUMBER 5, MAY

3 Protein metabolism in GDM Table 1 Nitrogen flux measured by [ 15 N]leucine in GDM and control subjects antepartum and postpartum GDM Control Antepartum Postpartum Antepartum Postpartum Fasting plasma BUN (mg/dl)* 8.1 ± ± ± ± 3.0 Postprandial plasma BUN (mg/dl)* 10.1 ± ± ± ± h urinary urea nitrogen excretion Grams per 12 h 6.93 ± ± ± ± 1.98 Milligrams per kilogram per 12 h 79 ± ± ± ± 46 Milligrams per kilogram FFM per 12 h 125 ± ± ± ± 68 Nitrogen flux Grams per 12 h* 35.0 ± ± ± ± 13.3 Milligrams per kilogram per 12 h* 0.39 ± ± ± ± 0.23 Milligrams per kilogram FFM per 12 h* 0.64 ± ± ± ± 0.35 Data are means ± SD. *Significant time effect from repeated measures analysis of variance (P = ). P value of 0.05 was set as statistically significant. Data are presented as mean ± SD. R E S U LT S Mean weight and BMI w e re 89.7 ± 10.9 kg and 36.0 ± 4.5 kg/m 2 a n t e p a rtum and 83.5 ±13.2 kg and 33.3 ± 5.1 kg/m 2 p o s t p a rtum for the GDM subjects and 70.6 ± 10.8 kg and 29.4 ± 3.8 k g / m 2 a n t e p a rtum and 64.8 ± 10.1 kg and 26.6 ± 3.4 kg/m 2 p o s t p a rtum for the cont rol subjects, re s p e c t i v e l y. FFM was higher in the GDM group than in the contro l g roup (P = 0.03), but percent body fat (35% antepartum; 38% postpartum) was not diff e rent between groups. During the a n t e p a rtum study, glycemia was re l a t i v e l y well controlled in the GDM group, as indicated by their fasting concentrations of glucose (83.6 ± 11.8 vs ± 3.9 mg/dl in control group) and HbA 1 c (5.2 ± 0.7 vs. 5.1 ± 0.5% in control group). In addition, we re c o rded fasting blood glucose and 2-h postprandial values after each meal for 5 days while the subjects were admitted to the Metabolic Research Unit; all values w e re normal. At 6 weeks postpartum, the glucose tolerance test revealed signific a n t l y higher plasma glucose values in the GDM g roup compared with those values in the c o n t rol group (P = 0.001). Fasting and 2-h postprandial plasma concentrations of blood urea nitro g e n (BUN) were significantly (P = 0.001) lower during pregnancy compared with the postp a rtum period, but no significant diff e r- ences were seen between GDM and contro l g roups (Table 1). Urinary excretion of ure a n i t rogen tended to be elevated in the GDM g roup in the antepartum period, but this d i ff e rence did not attain statistical significance. Adjusted for weight or FFM, ure a n i t rogen excretion did not differ by gro u p or time. The estimated rates of nitrogen flu x a re given in Table 1. Nitrogen flux (g/12 h), adjusted or unadjusted for weight or FFM, was significantly lower in both groups during pregnancy (P = ). Nitro g e n flux did not differ between the GDM and c o n t rol gro u p s. All plasma amino acid concentrations for the GDM group were numerically higher than those concentrations of the c o n t rol group, re g a rdless of pregnancy or fasting status. With few exceptions, plasma amino acid concentrations were significantly lower antepartum than postpart u m (P 0.05). Fasting levels of taurine, h y d ro x y p roline, glutamic acid, glutamine, cystine, tyrosine, phenylalanine, try p t o- phan, and histidine were signific a n t l y higher in GDM subjects than control subjects antepartum and postpartum (P 0.05) (Table 2). Fasting concentrations of p roline, valine, isoleucine, leucine, lysine, and arginine were higher in the GDM g roup than in the control group in the p o s t p a rtum period only (P 0.05). Postprandial concentrations of taurine, h y d ro x y p roline, valine, cystine, isoleucine, leucine, tyrosine, phenylalanine, try p t o- phan, ornithine, lysine, histidine, and arg i- nine were statistically higher in GDM subjects than control subjects antepart u m and postpartum (P 0.05) (Table 3). Fasting plasma concentrations of h y d ro x y p roline, serine, glutamine, pro l i n e, valine, isoleucine, phenylalanine, and try p- tophan were positively correlated with birt h weight (r = ), but only serine attained statistical significance (P = 0.04), c o n t rolling for gestational age and sex. C O N C L U S I O N S P rotein turn o v e r was normalized in insulin-treated women with GDM; however, fasting and postprandial concentrations of amino acids were elevated in the antepartum and postpart u m periods. The postpartum elevation in amino acids was consistent with re s i d u a l insulin resistance in the GDM group, as indicated by a glucose tolerance test admini s t e red at 6 weeks postpart u m. Our study also confirms previous findings of nitrogen conservation during p regnancy (8). We observed signific a n t l y lower fasting and postprandial BUN and n i t rogen flux during pregnancy in both g roups. During pre g n a n c y, plasma ure a concentration and u rea synthesis rates are attenuated, conserving the maternal supply of nitrogen. Kalhan et al. (8) re p o rt e d d e c reased plasma urea concentration and rates of urea synthesis during late pre g- nancy; however, plasma urea concentrations and urea synthesis were similar in seven type 1 diabetic subjects, three GDM subjects, and ten nonpregnant control subjects. Tests were perf o rmed after a 12-h fast and at least 12 h after the last insulin dose. The mechanism whereby urea synthesis is decreased is unknown, but it may involve pregnancy hormones or substrate d e l i v e ry because all plasma amino acids a re decreased during pregnancy due to i n c reased extraction of amino acids by the fetus and hemodilution. Whole-body protein turnover has been studied in normal pre g n a n c y. Pro t e i n t u rnover and synthesis estimated from oral [ 1 5 N]glycine in pregnant women in the t h i rd trimester of pregnancy were not s i g n i ficantly diff e rent from nonpre g n a n t women (15,5). Fitch and King (5) comp a red protein turnover of eight pre g n a n t women and two women with diet-cont rolled GDM with nine nonpregnant contro l women. The pregnant and nonpre g n a n t women had similar rates of protein turn o v e r 808 DIABETES CARE, VOLUME 22, NUMBER 5, MAY 1999

4 Butte and Associates Table 2 Fasting plasma amino acid concentrations in nanomoles per milliliter measured antepartum and postpartum GDM Control P value Antepartum Postpartum Antepartum Postpartum Group effect Time eff e c t G roup time interact i o n Taurine 32.9 ± ± ± ± Hydroxyproline 12.5 ± ± ± ± Threonine ± ± ± ± Serine 80.1 ± ± ± ± Glutamic acid 68.1 ± ± ± ± Glutamine ± ± ± ± Proline ± ± ± ± * Glycine ± ± ± ± Alanine ± ± ± ± Valine ± ± ± ± * Cystine 37.5 ± ± ± ± Methionine 10.2 ± ± ± ± Isoleucine 46.1 ± ± ± ± * Leucine 84.8 ± ± ± ± * Tyrosine 46.0 ± ± ± ± Phenylalanine 50.3 ± ± ± ± Tryptophan 89.9 ± ± ± ± Ornithine 32.8 ± ± ± ± Lysine ± ± ± ± * Histidine 84.8 ± ± ± ± Arginine 46.4 ± ± ± ± * Data are means ± SD. *Significant differences between GDM and control groups in postpartum period only (P = ). and synthesis, while rates were 16% higher in the women with GDM. Thompson and Halliday (16) stre s s e d the importance of standardizing pro t e i n t u rnover rates by FFM, especially in pre g- n a n c y. Using a continuous infusion of [ 1 3 C]leucine, Thompson and Halliday (16) re p o rted higher rates of protein synthesis and catabolism (per kilogram FFM) in p regnant women at 24 and 35 weeks of gestation, but not at 13 weeks, than nonp regnant control women during fasting. Although protein turnover was elevated in late pre g n a n c y, catabolism was also i n c reased, resulting in similar net pro t e i n losses. In contrast, total rates of leucine flux were similar in pregnant women at 30 weeks of gestation and nonpregnant c o n t rol women; however, expressed in t e rms of body weight or FFM, pro t e o l y s i s was significantly lower in pregnant women after a 17-h fast (17). Leucine oxidation did not differ between groups. Urinary ure a n i t rogen excretion was lower in pre g n a n t women, indicating conservation of nitrogen during a brief fast. Findings re g a rding protein turnover in GDM have been equivocal. Robinson and P re n d e rgast (4) studied protein turnover in 10 pregnant women with a history of GDM and in 10 pregnant and in 10 nonpre g n a n t c o n t rol women using a continuous infusion of [ 1 3 C]leucine. Total leucine flux was similar in all three groups, but leucine turn o v e r (per kilogram weight) was reduced in the p regnant women compared with contro l women, and it was increased in GDM women versus normal pregnant women. Reduced leucine incorporation into pro t e i n and increased leucine oxidation was seen in the GDM group. At the time of the study, the women in the GDM group had norm a l glucose tolerance, but developed GDM later in their pre g n a n c y. Zimmer et al. (6) found no diff e re n c e s in fasting hepatic glucose release and pro t e- olysis between untreated GDM women and p regnant control women using [ 1 3 C ] - labeled tracers. Similar rates of appearance of leucine, phenylalanine, and tyro s i n e ( re flecting proteolysis) after an overn i g h t fast were achieved by fasting plasma insulin concentrations that were three to five times higher in the GDM group, indicating re s i s- tance to the antiproteolytic effects of insulin. Rates of leucine appearance and leucine oxidation measured in the fasted state using [1-1 3 C]leucine were higher in 12 type 1 diabetic and 7 GDM patients (insulin withheld), but nonoxidative disposal of leucine and urea nitrogen excre t i o n w e re similar to those of pregnant contro l subjects (3). In contrast to our study, Kalhan et al. (3) concluded that vigoro u s management of insulin-treated GDM did not normalize whole-body protein bre a k- down. Diff e rent tracers and methods were employed, but more import a n t l y, our i n s u l i n - t reated GDM patients were on their routine insulin therapy and in the fed state at the time of the [ 1 5 N]leucine study. Our results of lower fasting plasma amino acid concentrations in late gestation a re corroborated by others (18,8). Gluconeogenic amino acids (alanine, glutamine, glutamate, serine, threonine) display a gre a t e r decline with fasting. Lower plasma amino acids during pregnancy have been attributed to combined hormonal effects, amino acid uptake by the placenta and fetus, and dilution by expanded maternal extracellular and other fluid compartments (2). Our higher plasma amino acid concentrations in GDM a re in agreement with Metzger et al. (7) who re p o rted higher branched-chained amino acid concentrations (leucine, isoleucine, valine) in GDM, compared with norm a l p regnant women after an overnight fast. The postprandial rise in plasma amino acids was blunted during pre g n a n c y, but to a lesser extent in the GDM gro u p. Our nitrogen flux data did not pro v i d e an explanation for the elevated plasma DIABETES CARE, VOLUME 22, NUMBER 5, MAY

5 Protein metabolism in GDM Table 3 The 2-h postprandial plasma amino acid concentrations in nanomoles per milliliter measured antepartum and postpartum GDM Control P value Antepartum Postpartum Antepartum Postpartum Group effect Time eff e c t G roup time interact i o n Taurine 41.1 ± ± ± ± Hydroxyproline 18.9 ± ± ± ± Threonine ± ± ± ± Serine ± ± ± ± Glutamic acid 73.4 ± ± ± ± Glutamine ± ± ± ± Proline ± ± ± ± Glycine ± ± ± ± Alanine ± ± ± ± Valine ± ± ± ± Cystine 27.5 ± ± ± ± Methionine 15.7 ± ± ± ± Isoleucine 68.4 ± ± ± ± Leucine ± ± ± ± Tyrosine 67.7 ± ± ± ± Phenylalanine 73.2 ± ± ± ± Tryptophan ± ± ± ± Ornithine 52.0 ± ± ± ± Lysine ± ± ± ± Histidine ± ± ± ± Arginine 96.8 ± ± ± ± Data are means ± SD. amino acids seen in the fasted or fed state. The single-dose end product model assumes that whole-body flux through a homogeneous precursor pool can be estimated from the disposal of 1 5 N urea. It is possible that the plasma amino acid concentrations were not re flective of the whole-body amino acid concentrations. A l t e rn a t i v e l y, our model may not have been sensitive enough to detect small diff e re n c e s in nitrogen flux with our relatively small sample size. The elevation of plasma amino acids, h o w e v e r, is consistent with insulin re s i s- tance during pregnancy and with its persistence postpartum in the women with GDM. Fasting insulin levels were significantly higher in the GDM group than in the c o n t rol group antepartum (16.4 ± 7.9 vs ± 4.2 µu/ml) and postpartum (17.3 ± 6.2 vs.10.6 ± 2.2 µu/ml) (P = 0.05). Plasma glucagon concentrations antepartum (194 ± 92 vs. 145 ± 27 pg/ml) and postpart u m (237 ± 141 vs. 130 ± 21 pg/ml) tended to be higher in the women with GDM than in the control women (P = 0.09) (9). Higher glucagon levels may depress amino acid incorporation into protein or enhance proteolysis, even though our nitrogen flux and u rea excretion data did not support d e c reased protein synthesis or incre a s e d p ro t e o l y s i s. We speculate that elevated plasma amino acids may promote fetal hyperinsulinemia and fetal growth, despite satisfact o ry maternal glycemic control in women with GDM. The transfer of amino acids f rom the mother to the fetus is a complex, e n e rgy-dependent process involving selective amino acid transporters, each specific for a select group of amino acids (19). Placental transport of amino acids results in both a high intracellular concentration of amino acids within the placenta and a fetalm a t e rnal ratio 1.0 for all amino acids (20). It is also clear that amino acids are t r a n s p o rted to the fetus in excess of their re q u i rement for protein synthesis. For some amino acids, transport across the placenta appears direct with little or no intrat rophoblast metabolism; others amino acids are modified by the placenta to a seco n d a ry product; and still others are metabolized to other amino acids and then t r a n s p o rted to the fetus. The placenta t h e re f o re regulates both the quality and quantity of amino acids transferred to the f e t u s. T h e re are many factors which aff e c t the placental transport of amino acids in addition to maternal plasma concentrations, such as placental surface area, maternal and fetal blood flo w, placental p e rfusion, placental membrane perm e a b i l- i t y, concentration of carrier proteins, ion gradients, placental metabolism, energ y s u p p l y, circulating hormones, and dru g s (19,20). While the placenta clearly concentrates amino acids intracellularly, presumably for transfer to the fetus, umbilical uptake of amino acids surprisingly has only been studied in the fetal lamb (20). Until m o re species are studied, we will not know if the specific features of sheep placental t r a n s p o rt are common to other species. Insulinogenic amino acids may stimulate fetal pancre a t i c -cell development; h o w e v e r, experimental data are not available in humans and are limited in animals. Fetal rat pancreatic rudiments underw e n t -cell hyperplasia and increased insulin s e c retion in response to increasing amino acid concentrations in the medium (21). A rginine and leucine act as potent insulin s e c retagogues, when injected into the fetus, p roducing relative high plasma insulin levels, but whether they play a role in pancreatic -cell secretion under norm a l physiological or diseased states is unknown (19). Aberrations in maternal amino acid metabolism in GDM may have an effect on the fetal -cell and may contribute to fetal complications associated with fetal hyperinsulinism, but any conclusive statement re g a rding placental amino acid metabolism in diabetes is not possible at pre s e n t. 810 DIABETES CARE, VOLUME 22, NUMBER 5, MAY 1999

6 Butte and Associates S i g n i ficant associations between maternal plasma amino acid concentrations and fetal growth have been seen in a number of studies. While mean birth weight in our study was not significantly diff e re n t between groups, we did observe a level of c o rrelation (r = ) between individual amino acids and birth weight that was similar to other published values that attained statistical significance with larg e r sample sizes. Birth weight was shown to be c o rrelated with plasma concentrations of a s p a rtic acid, serine, alanine, tyro s i n e, o rnithine, and arginine in normal pre g n a n t women at 25 weeks of gestation (22). Birt h weight was significantly correlated with plasma concentrations of serine, pro l i n e, lysine, threonine, arginine, and citrulline in p regnant women with type 1 diabetes and lean control women (23). Birth weight correlated with fasting plasma amino acids ( t h reonine, serine, glycine, alanine, valine, isoleucine, leucine, tyrosine, phenylalanine) in women with diet-treated GDM in late gestation (2). Maternal plasma amino acids (proline, valine, isoleucine, tyro s i n e, phenylalanine) were also positively correlated with amniotic fluid insulin at weeks of gestation, suggesting that fetal -cells may be stimulated by maternal circulating amino acids. In conclusion, protein turnover was normalized in insulin-treated women with GDM; however, fasting and postprandial plasma concentrations of amino acids were elevated in the antepartum and postpart u m periods, despite satisfactory glycemic contro l. A c k n o w l e d g m e n t s This work is a publication of the U.S. Department of Agriculture (USDA)/Agricultural Research Service (ARS) C h i l d re n s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine and Texas Childre n s Hospital, Houston, TX. This project has been funded in part with federal funds from the USDA/ARS under Cooperative Agreement The contents of this publication do not necessarily re flect the views or policies of the USDA, nor does mention of trade names, commercial product, or o rganization imply endorsement by the U.S. G o v e rn m e n t. We thank the women who participated in this study and acknowledge the contributions of Patricia Wellington, Alice Ponce, Paula Clark, and Sandra Kattner for re c ruitment and clinical management of the patients; Zahira Colon, Nitesh Mehta, Maurice Puyau, and Firoz Vo h r a for technical assistance; Idelle Tapper and Adam Gillum for manuscript preparation and Leslie Loddeke for editorial re v i e w. R e f e re n c e s 1. Stenninger E, Schollin J, Aman J: Neonatal macrosomia and hypoglycaemia in c h i l d ren of mothers with insulin-tre a t e d gestational diabetes mellitus. Acta Paediatr S c a n d 80: , Metzger BE: Biphasic effects of matern a l metabolism on fetal growth: quintessential e x p ression of fuel-mediated teratogenesis. D i a b e t e s 40:99 105, Kalhan SC, Denne SC, Patel DM, Nuamah I F, Savin SM: Leucine kinetics during a brief fast in diabetes in pre g n a n c y. M e t a b - o l i s m 43: , Robinson S, Pre n d e rgast CH: Pro t e i n metabolism in pre g n a n c y. B a i l l i è re Clin Endocrinol Metab 10: , Fitch WL, King JC: Protein turnover and 3- methylhistidine excretion in non-pre g n a n t, p regnant and gestational diabetic w o m e n. Hum Nutr Clin Nutr 41C: , Zimmer DM, Golichowski AM, Karn CA, B rechtel G, Baron AD, Denne SC: Glucose and amino acid turnover in untreated gestational diabetes. Diabetes Care 19: , Metzger BE, Phelps RL, Freinkel N, Navickas IA: Effects of gestational diabetes on d i u rnal pro files of plasma glucose, lipids, and individual amino acids. Diabetes Care 3: , Kalhan SC, Ts e rng K, Gilfillan C, Dierker LJ: Metabolism of urea and glucose in normal and diabetic pre g n a n c y. M e t a b o l i s m 31: , Hsu HW, Butte NF, Wong WW, Moon JK, Ellis KJ, Klein PD, Moise KJ: Oxidative metabolism in insulin-treated gestational diabetes mellitus. Am J Physiol : E1099 E1107, Committee on Technical Bulletins: Diabetes and pre g n a n c y. ACOG Technical Bul - l e t i n 200:1 8, Akers R, Buskirk ER: An underw a t e r weighing system utilizing force cube transducers. J Appl Physiol 2 6 : , Wi l m o re JH: A simplified method for d e t e rmination of residual lung volumes. J Appl Physiol 27:96 100, Siri WE: Body composition from flu i d spaces and density: an analysis of methods. In Techniques for Measuring Body Com - p o s i t i o n. Broszek J, Henschel A, Eds. Washington, D.C., National Academy of Science-National Research Council, 1961, p F e rn EB, Garlick PJ, Waterlow JC: The concept of the single body pool of metabolic nitrogen in determining the rate of whole-body nitrogen turn o v e r. Hum Nutr Clin Nutr 39C:85 99, De Benoist B, Jackson AA, Hall JSE, Persaud C: Whole-body protein turnover in Jamaican women during normal pre g- n a n c y. Hum Nutr Clin Nutr 3 9 C : , Thompson GN, Halliday D: Pro t e i n t u rnover in pre g n a n c y. Eur J Clin Nutr 46: , Denne SC, Patel D, Kalhan SC: Leucine kinetics and fuel utilization during a brief fast in human pre g n a n c y. M e t a b o l i s m 40: , Metzger BE, Unger RH, Frienkel N: Carbohydrate metabolism in pre g n a n c y. Part X I V. Relationships between circ u l a t i n g glucagon, insulin, glucose and amino acids in response to a mixed meal in late pre g- n a n c y. M e t a b o l i s m 26: , Hay WW Jr: Placental transport of nutrients to the fetus. H o rm Res 42: , C a rter BS, Moores RR, Battaglia FC: Placental transport and fetal and placental metabolism of amino acids. J Nutr Biochem 2:4 13, De Gasparo M, Milner GR, Norris PD, Milner RDG: Effect of glucose and amino acids on fetal rat pancreatic growth and insulin secretion in vitro. J Endocrinology 77: , McClain PE, Metcoff J, Crosby WM, Costiloe JP: Relationship of maternal amino acid p ro files at 25 weeks of gestation to fetal g rowth. Am J Clin Nutr 31: , K a l k h o ff RK, Kandaraki E, Morrow PG, Mitchell TH, Kelber S, Borkowf HI: Relationship between neonatal birth weight and maternal plasma amino acid pro fil e s in lean and obese nondiabetic women and in type I diabetic pregnant women. M e t a b - o l i s m 37: , 1988 DIABETES CARE, VOLUME 22, NUMBER 5, MAY