Elevated zinc protoporphyrin/heme ratios in umbilical cord blood after diabetic pregnancy

Transcription

1 ORIGINAL ARTICLE Elevated zinc protoporphyrin/heme ratios in umbilical cord blood after diabetic pregnancy KB Lesser 1, SB Schoel 2 and PJ Kling 3 (2006) 26, r 2006 Nature Publishing Group All rights reserved /06 $ Departments of Obstetrics and Gynecology and Pediatrics, The University of Arizona Health Sciences Center, Tucson, AZ, USA; 2 Coon Rapids Women s Health, Coon Rapids, MN, USA and 3 University of Wisconsin, Madison, WI, USA Objective: Offspring of diabetes patients may suffer from tissue iron deficiency. Erythrocyte zinc protoporphyrin/heme (ZnPP/H) ratios measure impaired iron status. The aim of the study was to examine whether cord ZnPP/H ratios were associated with pregnancy glycemic control. Methods: ZnPP/H was measured in cord blood from 31 pregnancies with insulin-treated diabetes (diabetes group) and compared to population normal values. Maternal glycemic control was assessed by daily glucose log, glycosylated hemoglobin and birth weight. Results: Median cord ZnPP/H was higher in the diabetes group than the population normal values (106 (65.2 to 146.8) mm/m vs 68.2 (37.6 to 98.8) mm/m, P<0.0001). Ratios were directly correlated to surrogates of control (glycosylated hemoglobin, P ¼ 0.05, and birth weight, P<0.04). Cord ZnPP/H ratios from pregnancies with pre-existing and gestational diabetes were similar. Conclusion: Because cord ZnPP/H was higher in large offspring of diabetic pregnancy, it might identify greater iron utilization for fetal erythropoiesis. (2006) 26, doi: /sj.jp ; published online 5 October 2006 Keywords: iron deficiency; diabetes; glycemic control; pregnancy; fetal growth Introduction Fetuses of women with poorly controlled diabetes mellitus experience hyperglycemia, hyperinsulinemia, macrosomia, chronic hypoxemia, polycythemia and decreased iron storage. 1 4 Both chronic fetal hyperglycemia and hyperinsulinemia increase cellular oxygen consumption. 1,3,4 The resulting hypoxemia stimulates fetal erythropoiesis and accelerates erythrocyte iron delivery. 1,3,4 Iron is shunted into erythrocyte mass and away from developing organs and tissues. 2 5 Correspondence: Dr PJ Kling, Department of Pediatrics and Center for Perinatal Care, University of Wisconsin & Meriter Hospital, 202 S, Park St, Madison, WI 53715, USA. pkling@wisc.edu Received 30 March 2006; revised 20 July 2006; accepted 18 August 2006; published online 5 October 2006 Iron is an essential nutrient for normal perinatal growth and development. Although infants born to women with poorly controlled diabetes may experience impaired childhood intellectual development, 6 only recently has the role of brain iron deficiency been explored. In otherwise normal infants, development of tissue iron deficiency during rapid first year growth is associated with disturbed brain development. 7,8 Recent work shows that long-term cognitive deficits in offspring of women with poorly controlled diabetes are similar to those seen in tissue iron deficiency The erythrocyte zinc protoporphyrin/heme (ZnPP/H) ratio measures inadequate erythrocyte iron incorporation. Heme is synthesized from iron and protoporphyrin IX. 13,14 When availability of iron is not sufficient, zinc replaces iron in the protoporphyrin ring. 13,14 The intrinsic fluorescence of zinc protoporphyrin within the erythrocyte is measured and expressed as a ratio relative to moles of heme within the erythrocyte. 15 Because ZnPP/H ratios measure incomplete iron incorporation, it detects iron deficiency at an earlier stage than classical measures of tissue iron stores. 13,16 In young children, ZnPP/H is more sensitive than plasma ferritin level (a measure of stores) or hemoglobin (a measure of anemia). 17 Our primary hypothesis was that fetal cord erythrocyte ZnPP/H from offspring of pregnancies with diabetes would be higher than from offspring of uncomplicated pregnancies. Our secondary hypothesis was that women with diabetes experiencing poorer glycemic control would deliver neonates with higher ZnPP/H ratios. In order to test these hypotheses, we collected and measured cord ZnPP/H ratios from offspring of women with diabetes (diabetes group), comparing ratios to retrospective data collected from the corresponding pregnancies and also comparing ratios to population normals at the University of Arizona. Materials and methods We enrolled offspring of diabetic pregnancies into a retrospective descriptive study that prospectively measured and compared cord blood ZnPP/H ratios and cord blood indices of erythropoiesis, to fetal growth and maternal chart information describing diabetes

2 672 control. As population normal values for ZnPP/H at the University of Arizona had not been described, we report and compare normal ratios to ratios in offspring of women with diabetes. Population normal values Umbilical cord blood was obtained from 120 consecutive deliveries at The University of Arizona University Medical Center in January 2001 and February Ninety-eight of the 120 deliveries were admitted to the newborn nursery. From the newborn nursery log, birth weight, gender and gestational age by best obstetrical estimate were used to determine birth weight z-scores between 1.75 and 1.75 (appropriate for gestation). Birth weight z-score calculation was performed utilizing the North American intrauterine growth curves of Arbuckle et al. 18 The z-score was obtained by subtracting expected birth weight for gestational age and sex from actual birth weight and dividing by the standard deviation at that gestational age ((actual weight expected weight)/sd pca ). No newborn in the diabetes group was greater than 40 weeks gestation. Because ZnPP/H falls with increasing gestation, and rises in infants small for dates, we modified the normal population to include newborns appropriately sized and less than 40 weeks gestation, resulting in 55 ZnPP/H ratios. Diabetes group We collected and examined cord blood samples from women treated with insulin during pregnancy, either with pre-existing diabetes or insulin-treated gestational diabetes. Gestational diabetes was diagnosed using the criteria of Carpenter and Coustan. 19 Women with gestational diabetes were treated with insulin when a majority of fasting blood glucose values were X95 mg/dl (5.2 mm/l) or 2-h postprandial values were X120 mg/dl (6.6 mm/l). Women were consecutive referrals to the Diabetes in Pregnancy Program, which captured most insulin-treated women delivering at University Medical Center from May 2001 to March We retrospectively examined medical records of mother and newborn. No major congenital anomalies were observed in any neonate. Birth weight, gender and gestational age by best obstetrical estimate determined birth weight z-scores and classification as appropriate or large for gestational age. 18 Women with diabetes received nutritional counseling and supplemental iron. Women performed fasting and 2-h postprandial blood glucose monitoring. The glycemic goal was defined as mean glucose value <105 mg/dl (5.8 mm/l), as these were the goals set forth by the Diabetes in Pregnancy interdisciplinary team. We examined only third trimester glucose values because glucose logs in gestational diabetes were only available for the last trimester. We examined two surrogate measures of diabetes control: glycosylated hemoglobin and fetal growth. We obtained available glycosylated hemoglobin values from the laboratory and maternal charts and reported the mean value when multiple samples were obtained. Laboratory methods For both the groups, fetal cord blood was collected at delivery in EDTA anticoagulant. Blood was retrieved from the blood bank after storage at 41C for up to 7 days. Preliminary samples were analyzed daily for 10 days and it was found that ZnPP/H, red cell counts, hemoglobin and red cell indices were stable at 41C. In cord specimens from both the groups, ZnPP/H was measured by the clinically available, Front-Face Hematofluorimeter (Aviv Biomedical Co., Lakewood, NJ). Specimens were washed in saline or buffered saline to remove interfering pigments, and reconstituted to original hematocrit. 20 The wash step was deemed necessary as 98% of paired samples from the initial preliminary group of 120 fell after washing (mean fall was 39 mm/m). In cord samples from the diabetes group, we measured complete blood counts by Coulter-Counter MD-16 (Hialeah, FL) (n ¼ 19) or by the ADVIA 120 (Bayer, Tarrytown, NY) (n ¼ 12). Data analyses Sample size estimate to test our primary hypothesis estimated that 18 diabetes group samples would detect with 80% power, a-level of 0.05, that ZnPP/H is 30 mm/m higher in the setting of diabetes than in healthy deliveries. If those with excellent glycemic control exhibited normal ZnPP/H ratios, it was anticipated that we would need to enroll 36 women in our diabetes group to compare good vs poorer glycemic control. We were able to enroll 31 women in the time available to us. ZnPP/H and glycosylated hemoglobin exhibit nonparametric distributions and are log-transformed for comparison with parameters normally distributed. In the diabetes group, the ratio of birth weight/length was calculated. In the diabetes group, total body hemoglobin was calculated, based on an assumed fetoplacental blood volume of 120 ml/kg. Statistical analysis was performed using Statview 5.01, SAS Institute. Analysis included simple linear regression and logistic stepwise regression with P<0.1 in simple regression for placement into the model. We utilized Mann Whitney U-testing. Data are reported as median and interquartile range. The University of Arizona Human Subjects Committee approved the project. Written informed consent was obtained from the women. Results We modified the population normal group to include newborns <40 weeks gestation because diabetes group newborns were delivered early and ZnPP/H falls slightly with gestation. 21 In the diabetes group, 31 were enrolled (19/31 gestational). Median birth weight and gestation of population normals (39 weeks, 3285 g) and diabetes groups (38 weeks, 3225 g) were similar. Median ZnPP/H was higher in the diabetes group than the population normals (106 (65.2 to 146) mm/m vs 68.3 (37.6 to 98.8) mm/m, P<0.0001).

3 673 Table 1 Glycemic characteristics when diabetes logs were available, newborn demographic characteristics and newborn laboratory data of diabetes group Characteristics Number of mothers/fetuses Median (interquartile range) Mean glycosylated hemoglobin 20/ ( )% Third trimester fasting glucose 28/30 93 (87 99) mg/dl 5.16 ( ) Third trimester 2-h postprandial glucose 27/ ( ) mg/dl 6.72 ( ) mmol/l Third trimester mean glucose 29/ ( ) mg/dl 6.01 ( ) mmol/l Birth weight (g) 29/ ( ) Gestation (weeks) 29/31 38 ( ) Cord ZnPP/H (mm/m) 29/ ( ) Hemoglobin (g/l) 29/ ( ) Total body hemoglobin (g) 29/ ( ) Figure 1 (a) By simple linear regression, cord ZnPP/H in the diabetes group was associated with RDW, y ¼ 1.43 þ 0.19x; R 2 ¼ 0.40, P< (b) By simple linear regression, ZnPP/H in the diabetes group was also negatively associated with MCHC, y ¼ x; R 2 ¼ 0.13, P ¼ We observed neither hematocrit values greater than 65% nor hemoglobin greater than 210 g/l in the diabetes group. We observed no relationship between cord ZnPP/H and hemoglobin concentration (P ¼ 0.9), but found a significant relationship between cord ZnPP/H and estimated fetoplacental hemoglobin (P<0.04). Cord ZnPP/H in diabetes was directly correlated with red cell distribution width (RDW) (P<0.0001), and indirectly with mean cell hemoglobin concentration (MCHC) (P ¼ 0.05; Figure 1a and b). From 30 patients, we examined a direct measure of glycemic control, the diabetic logs. A summary of the glycemic characteristics of diabetes subjects is presented in Table 1. Eight women were well controlled by meeting the Diabetes in Pregnancy Program s mean third trimester glucose goal of <105 mg/dl (5.8 mmol/l). Median ZnPP/H in fetuses who met the goal were not statistically different from those not meeting the goal (P ¼ 0.4, Table 2). We examined two indirect estimates of glycemic control: glycosylated hemoglobin and fetal growth. Glycosylated hemoglobin levels from 10 women with pre-existing diabetes were available. Cord ZnPP/H was correlated with mean glycosylated hemoglobin percent in pre-existing diabetes (P ¼ 0.05; Figure 2). Large for gestation newborns tended towards higher cord ZnPP/H than appropriate for gestation (P ¼ 0.06; Table 3). Because the estimate is based on weight, estimated body circulating erythrocytes were greater in those large for gestation, compared to appropriate (P<0.007). Birth weight z-scores were directly correlated to Abbreviation: ZnPP/H, zinc protoporphyrin/heme ratio. Table 2 Birth weight, gestation, ZnPP/H and hemoglobin comparing those who met goals and did not meet goals (median, interquartile) Diabetes group Met goals (n ¼ 8) Did not meet (n ¼ 22) P-value Birth weight (g) 3172 ( ) 3225 ( ) 0.9 Gestation (weeks) 38 (37 39) 38 ( ) 0.8 ZnPP/H (mm/m) 93.3 ( ) 119 ( ) 0.4 Hemoglobin (g/l) 156 ( ) 161 ( ) 0.9 Total body hemoglobin (g) 60.1 ( ) 64.7 ( ) 0.9 Abbreviation: ZnPP/H, zinc protoporphyrin/heme ratio. Figure 2 By simple linear regression, cord ZnPP/H in pre-existing diabetes was associated with mean glycosylated hemoglobin values, n ¼ 14, y ¼ 2.65 þ 1.0x; R 2 ¼ 0.39, P ¼ ZnPP/H (P<0.04; Figure 3a). To estimate lean body mass, ZnPP/H was compared to the ratio of birth weight/length and a significant association was found (P<0.01; Figure 3b). In contrast, in the modified normal population, neither birth weight z-scores, nor the ratio of weight/length correlated to ZnPP/H.

4 674 Table 3 Birth weight, gestation, ZnPP/H and hemoglobin comparing those born AGA and LGA (median, interquartile) Diabetes group AGA (n ¼ 21) LGA (n ¼ 10) P-value Birth weight (g) 3190 ( ) 3975 ( ) Gestation (weeks) 38.1 ( ) 37.5 (36 39) 0.05 ZnPP/H (mm/m) 93.0 (71 115) 161 ( ) 0.06 Hemoglobin (g/l) 165 ( ) 155 ( ) 0.07 Total body hemoglobin (g) 62.3 ( ) 74.9 ( ) Abbreviations: AGA, appropriate for gestational age; LGA, large for gestational age; ZnPP/ H, zinc protoporphyrin/heme ratio. Table 4 Birth weight, gestation, ZnPP/H and hemoglobin comparing those born with gestational and pre-existing diabetes (median, interquartile) Diabetes group Gestational (n ¼ 18) Pre-existing (n ¼ 13) P-value Birth weight (g) 3222 ( ) 3355 ( ) 0.9 Gestation (weeks) 37.9 ( ) 38.0 ( ) 0.9 ZnPP/H (mm/m) 99.5 ( ) ( ) 0.4 Hemoglobin (g/l) 154 ( ) 165 ( ) 0.03 Total body hemoglobin (g) 61.5 ( ) 71.4 ( ) 0.2 Abbreviation: ZnPP/H, zinc protoporphyrin/heme ratio. Figure 3 (a) By simple regression, cord ZnPP/H in the diabetes group was positively associated with birth weight z-score, y ¼ 4.6 þ 0.10x; R 2 ¼ 0.20, P<0.04. Vertical line indicates mean birth weight (z-score of 0). (b) By simple regression, cord ZnPP/H in the diabetes group was positively associated with the ratio of birth weight over length, y ¼ 3.5 þ 0.02x; R 2 ¼ 0.2, P<0.01. Cord ZnPP/H in pre-existing diabetes was similar to that in insulin-treated gestational diabetes (P ¼ 0.2; Table 4). Stepwise logistic regression involving continuous variables that exhibited P<0.1 in simple regression was performed. Although gestational age (inversely), z-score for weight (positively), total body hemoglobin (positively), z-score was the only variable retaining significance in the stepwise regression (P<0.04). Discussion Studies have shown that newborn iron status correlates with glycemic control in diabetic pregnancy, 2,5 but this is the first study to show that cord ZnPP/H correlated with surrogates of glycemic control. Erythrocyte ZnPP/H measures iron-deficient erythropoiesis in older premature and term infants. 17,22 Limited data show that ZnPP/H from cord blood or the first week of life were inversely correlated to gestational age 21,23 and rise with exaggerated erythropoiesis observed with placental insufficiency and diabetes. 21,23 In a larger sample, our study confirms that mean cord ZnPP/H ratios after diabetic pregnancy were higher than population normal values. Our cord population normal values did not differ from others reported previously. 21,23 Previous work found brain iron deficiency in offspring of women with diabetes. 2,5 Children of diabetic pregnancies may also experience disturbed cognitive development and recognition memory, consistent with a hippocampal insult. 6,9 12,24 The fetal hippocampus is targeted by hypoxia ischemia, impaired iron delivery and postnatal hypoglycemia, all of which accompany poorly controlled diabetes during pregnancy. 11,12,25 In diabetes, specific defects in placental function are unable to compensate for increased iron needs accompanying fetal overgrowth and increased erythrocyte mass. 26 Although we found no relationship between cord ZnPP/H and hemoglobin concentration, postnatal polycythemia is influenced by mode of delivery, delay in cord clamping and postnatal fluid shifts, 27,28 such that hemoglobin from cord blood may not adequately estimate red cell mass or polycythemia. Another possibility is that hemoglobin concentration does not measure total hemoglobin produced. When we estimated circulating fetoplacental blood hemoglobin, we observed a small but significant correlation. It is also possible that glucose monitoring and interventions to improve glycemic control prevented polycythemia in these newborns. However, the finding that RDW was directly and MCHC indirectly associated with ZnPP/H supports the notion of insufficient erythrocyte iron delivery. Abnormalities in RDW (measure of erythrocyte size variation) occur during iron-deficient erythropoiesis, especially when hemoglobin content (MCHC) falls. 29,30

5 675 ZnPP/H ratios were similar in those with excellent and poor control. Because of limited number of women with excellent control, the study may have had insufficient power to show lower ZnPP/H ratios with better glycemic control from the glucose logs. However, ZnPP/H was correlated with surrogate measures of glycemic control (glycosylated hemoglobin, birth weight and birth weight/length). Although we were unable to obtain glycosylated hemoglobin values for all subjects, this study is the first to show that ZnPP/H was correlated with glycosylated hemoglobin and fetal growth. Because tight glycemic control is associated with improved neurodevelopmental outcomes, 24 it would be important to investigate whether cord ZnPP/H is a marker of later neurodevelopmental outcome. Cord ZnPP/H ratios in fetuses of mothers with gestational diabetes were similar to fetuses of mothers with pre-existing diabetes. This was not anticipated, as most studies support that offspring of women with pre-existing diabetes experience worse outcomes than offspring of women with gestational diabetes. 31,32 This finding could be explained by our entry criteria of insulin use, because those with gestational diabetes using insulin were more severely affected than those with diet management. However, if elevated cord ZnPP/H indicates impaired erythropoiesis and thus impaired tissue iron delivery, this finding could contribute to ongoing discussions regarding the importance of universal screening for gestational diabetes. 31,32 Limitations of our study include retrospective data collection, sample size, self-reporting diabetic logs and lack of glycosylated hemoglobin percentages on all subjects. Too few enrollees experienced chronic hypertension, smoking or premature birth to be analyzed as a subgroup. There is potential for a Type II error in the three subgroups analyzed. Of note, the subgroup meeting goals was smaller than anticipated. If more enrollees were studied or we defined the meeting/not meeting subgroups differently, our results may have attained statistical significance, as those meeting goals exhibited numerically but not significantly lower median ZnPP/H ratios than those not meeting goals. This observation may be clinically relevant, as the surrogates of glycemic control, the glycosylated hemoglobin percentages and growth parameters support the notion that ZnPP/H ratios are higher with poorer control. Despite these limitations, our study demonstrates that offspring of women with diabetes exhibit elevated cord ZnPP/H ratios indicative of tissue iron deficiency, with ratios correlating to surrogates of poorer glycemic control, that is, glycosylated hemoglobin and fetal growth. Future work should compare cord and postnatal ZnPP/H and the long-term clinical outcomes in offspring of diabetic pregnancies. If elevated ZnPP/H ratios reflect those with impaired memory, as has been reported in offspring of women with diabetes, 11,12 ZnPP/H could be a marker of insufficient iron delivery to other tissues in addition to impaired delivery to erythrocytes. It would then be helpful in designing intervention or therapeutic trials. Acknowledgments We acknowledge the assistance of John A Widness, MD, Michael K Georgieff, MD, David G Lott, MD and Carrie Daniel. This work was supported by grant from the American Heart Association, Southwest Affiliate (SW-GS-16-98) (PJK) and a gift from the Sparks family to the University of Arizona Children s Research Center (PJK). References 1 Widness JA, Susa JB, Garcia JF, Singer DB, Sehgal P, Oh W et al. Increased erythropoiesis and elevated erythropoietin in infants born to diabetic mothers and in hyperinsulinemic rhesus fetuses. J Clin Invest 1981; 67: Georgieff MK, Landon MB, Mills MM, Hedlund BE, Faassen AE, Schmidt RL et al. Abnormal iron distribution in infants of diabetic mothers: spectrum and maternal antecedents. J Pediatr 1990; 117: Widness JA, Teramo KA, Clemons GK, Voutilainen P, Stenman U-H, McKinlay S et al. Direct relationship of antepartum glucose control and fetal erythropoietin in human Type 1 (insulin-dependent) diabetic pregnancy. Diabetologia 1990; 33: Schwartz R, Teramo KA. Effects of diabetic pregnancy on the fetus and newborn. Semin Perinatol 2000; 24: Petry CD, Eaton MA, Wobkin JD, Mills MM, Johnson DE, Georgieff MK. Iron deficiency of liver, heart, and brain in newborn infants of diabetic mothers. J Pediatr 1992; 121: Stehbens JA, Baker GL, Kitchell M. Outcomes at ages 1, 3, and 5 years of children born to diabetic mothers. Am J Obstet Gynecol 1977; 127: Lozoff B, Jimenez E, Wolf A. Long term developmental outcome of infants with iron deficiency. N Engl J Med 1991; 325: Lozoff B, Jimenez E, Hagen J, Mollen E, Wolf AW. Poorer behavioral and developmental outcome more than 10 years after treatment for iron deficiency in infancy. Pediatrics 2000; 105: e51. 9 Nelson CA, Wewerka SW, Thomas KM, Tribby-Waldbridge S, deregnier R-A, Geogieff MK. Neurocognitive sequelae of infants of diabetic mothers. Behav Neurosci 2000; 114: deungria M, Rao R, Wobken JD, Luciana M, Nelson CA, Georgieff MK. Perinatal iron deficiency decreases cytochrome c oxidase (CytOx) activity in selected regions of neonatal rat brain. J Pediatr 2000; 48: deregnier R-A, Nelson CA, Thomas KM, Wewerka S, Georgieff MK. Neurophysiologic evaluation of auditory recognition memory in healthy newborn infants and infants of diabetic mothers. J Pediatr 2000; 137: Siddappa AM, Georgieff MK, Wewerka SW, Worwa C, Nelson CA, deregnier R-A. Iron deficiency alters auditory recognition memory in newborn infants of diabetic mothers. Pediatr Res 2004; 55: Labbe RF, Finch CA, Smith NJ, Doan RN, Sood SK, Madan N. Erythrocyte protoporphyrin/heme ratio in the assessment of iron status. Clin Chem 1979; 25: Schifman RB, Rivers SL, Finley PR, Thies C. RBC zinc protoporphyrin to screen blood donors for iron deficiency anemia. JAMA 1982; 248:

6 676 Cord zinc protoporphyrin/heme ratios in diabetes 15 NCCLS. Erythrocyte Protoporphyrin Testing: Approved Guideline (NCCLS Document C42-A). National Committee on Clinical Laboratory Standards: Villanova, PA, Labbe RF, Dewanji A. Iron assessment tests: transferrin receptor vis-a-vis zinc protoporphyrin. Clin Biochem 2004; 37: Rettmer RL, Carlson TH, Origenes ML, Jack RM, Labbe RF. Zinc protoporphyrin/heme ratio for diagnosis of preanemic iron deficiency. Pediatrics 1999; 104: e Arbuckle TE, Wilkins R, Sherman GJ. Birth weight percentiles by gestational age in Canada. Obstet Gynecol 1993; 81: Carpenter MW, Coustan DR. Criteria for screening tests for gestational diabetes. Am J Obstet Gynecol 1982; 144: Hastka J, Lasserre JJ, Schwarzbeck A, Strauch M, Hehlmann R. Washing erythrocytes to remove interferents in measurements of zinc protoporphyrin by front-face hematofluorometry. Clin Chem 1992; 38: Lott DG, Labbe RF, Widness JA. Erythrocyte zinc protoporphyrin is elevated with prematurity and fetal hypoxemia. Pediatrics 2005; 116: Winzerling JJ, Kling PJ. Iron deficient erythropoiesis in premature infants measured by blood zinc protoporphyrin/heme. JPediatr2001; 139: Juul SE, Zerzan JC, Strandjord TP, Woodrum DE. Zinc protoporphyrin/ heme as an indicator of iron status in NICU patients. J Pediatr 2003; 142: Rizzo TA, Metzger BE, Dooley SL, Nam CH. Early malnutrition and child neurobehavioral development: insights from the study of children of diabetic mothers. Child Dev 1997; 68: Nylund L, Lunell NO, Lewander R, Persson B, Sarby B. Uteroplacental blood flow in diabetic pregnancy: measurements with indium 113m and computer-linked gamma camera. Am J Obstet Gynecol 1982; 144: Petry CD, Wobkin JD, McKay H, Eaton MA, Seybold VS, Johnson DE et al. Placental transferrin receptor in diabetic pregnancies with increased fetal iron demand. Am J Physiol 1994; 267: E507 E Shannon K, Davis JC, Kitzmiller JL, Fulcher SA, Koenig HM. Erythropoiesis in infants of diabetic mothers. Pediatr Res 1986; 20: Lubetzky R, Ben-shachar S, Mimouni FB, Dollberg S. Mode of delivery and neonatal hematocrit. Am J Perinatol 2000; 17: Monzon CM, Beaver BD, Dillon TD. Evaluation of erythrocyte disorders with mean corpuscular volume (MCV) and red cell distribution width (RDW). Clin Pediatr 1987; 26: Novak RW. Red blood cell distribution width in pediatric microcytic anemias. Pediatrics 1987; 80: Homko CJ, Reece EA. To screen or not to screen for gestational diabetes mellitus the clinical quagmire. Clin Perinatol 2001; 28: Coustan DR. Management of gestational diabetes: a self-fulfilling prophecy. JAMA 1996; 275: 1199.