Synergistic interactions among growing stressors increase risk to an Arctic ecosystem.

Oceans provide critical ecosystem services, but are subject to a growing number of external pressures, including overfishing, pollution, habitat destruction, and climate change. Current models typically treat stressors on species and ecosystems independently, though in reality, stressors often interact in ways that are not well understood. Here, we use a network interaction model (OSIRIS) to explicitly study stressor interactions in the Chukchi Sea (Arctic Ocean) due to its extensive climate-driven loss of sea ice and accelerated growth of other stressors, including shipping and oil exploration. The model includes numerous trophic levels ranging from phytoplankton to polar bears. We find that climate-related stressors have a larger impact on animal populations than do acute stressors like increased shipping and subsistence harvesting. In particular, organisms with a strong temperature-growth rate relationship show the greatest changes in biomass as interaction strength increased, but also exhibit the greatest variability. Neglecting interactions between stressors vastly underestimates the risk of population crashes. Our results indicate that models must account for stressor interactions to enable responsible management and decision-making.

Increased N-glycosylation and reduced transferrin-binding capacity of transferrin receptor isolated from placentae of diabetic women.

Infants of diabetic mothers are frequently born iron deficient because their fetal iron demand exceeds placental iron transport capacity. Although transferrin receptor (TfR) expression is increased, binding to diferric transferrin is decreased proportionately to the severity of maternal disease. It is hypothesized that TfR isolated from diabetic placentae has altered N-glycosylation since proper glycosylation of N-linked oligosaccharides is important for normal TfR binding kinetics to diferric transferrin. TfR was obtained from syncytiotrophoblastic membranes of six diabetic and six non-diabetic human placentae. Competitive binding to 125I-transferrin demonstrated a higher Kd in the diabetic TfR (P = 0.04), directly correlated to cord serum C-peptide concentration (r = 0.81, P < 0.001). The molecular weight of the monomeric form of TfR prior to treatment with glycopeptidase F (PNG-F) was greater in the diabetic group (P < 0.001) was directly related to the Kd (r = 0.77, P = 0.002). Treatment with PNG-F eliminated the molecular weight difference between the two groups. Increased glycosylation of the N-linked oligosaccharides of TfR isolated from diabetic placentae may alter the three-dimensional structure or charge of the receptor, thus reducing its binding affinity for transferrin.

Health, growth, and use of community services in NICU graduates at early school age.

PURPOSE: To examine outcomes related to health, growth, and use of community health and education services in children ages 6 to 8 years who received newborn intensive care because of prematurity or perinatal complications. METHOD: Parents of 81 children who had received neonatal intensive care at a Midwest US tertiary care center completed a mailed questionnaire. Three birth weight groups (very low birth weight [VLBW] < 1500 g, n = 35; low birth weight [LBW] 1501-2500 g, n = 24, and normal birth weight [NBW] > 2500 g, n = 22) were compared regarding growth, health, and use of community-based services using descriptive statistics and one-way analysis of variance. FINDINGS: VLBW and NBW groups had more ongoing health concerns. Growth patterns were similar in all groups. VLBW and NBW groups demonstrated greater use of community-based services, and service use increased at school age. CONCLUSIONS: Comprehensive systems are needed for follow-up of high-risk infants to detect and refer problems early. Neonatal histories must be tracked throughout childhood. Seriously ill term NBW infants are at risk for later morbidity and require follow-up similar to that provided for VLBW children.

Impact of Genetic potential and prematurity on growth outcomes.

PURPOSE: To explore the relationship between genetic potential and catch-up growth in school-age children who were born prematurely. STUDY DESIGN AND METHOD: This descriptive correlational study compared three groups of children who were born prematurely, sorted by birthweight groups into low, very low, and extremely low birthweight on measures of catch-up growth and body composition at school age (n = 45). Height and weight were compared to established norms for children of normal birthweight. Growth at school age and parental heights were also correlated. RESULTS: Children in all birthweight groups achieved growth within normal ranges (two standard deviations from the mean) by school age. The growth of the extremely low birthweight group was in the lower range of normal. Maternal height was the best predictor of children's heights at 8 to 10 years of age. CLINICAL IMPLICATIONS: Parents and providers can be reassured that many children overcome the adverse effects of prematurity on childhood growth. Throughout childhood, growth should be closely monitored using appropriate grids, and correcting for prematurity.

Second-line drug treatment in rheumatoid arthritis associated with depressed autologous mixed lymphocyte reaction.

The AMLR (autologous mixed lymphocyte reaction) was performed in 24 control subjects, 41 patients with rheumatoid arthritis (RA), 19 patients with ankylosing spondylitis (AS), and 7 patients with psoriatic arthritis (PSA). It was found to be depressed in 21 of the RA subjects and this was linked to medication with sodium aurothiomalate or D-penicillamine. Addition of ultrapure interleukin 1 (IL-1), indomethacin, or catalase did not cause any improvement in the AMLR but some improvement was noted in those RA patients with subnormal AMLR when pure, recombinant interleukin 2 (IL-2) was added to the cultures. Since autoreactive T-cells are generated during the AMLR which are capable of providing help for immunoglobulin production, it is proposed that the defective AMLR seen after second-line drug treatment may be the mechanism whereby rheumatoid factor (RF) production is down-regulated during such treatment.

Determinants of arm muscle and fat accretion during the first postnatal month in preterm newborn infants.

We measured arm muscle and fat areas in 22 preterm appropriate for gestational age infants at birth (mean +/- 1 SD birth weight: 1,640 +/- 484 g; gestational age: 31 +/- 2 weeks). Birth arm muscle and fat areas correlated significantly with gestational age (arm muscle: r = 0.86; p less than 0.001; arm fat: r = 0.75; p less than 0.001) and with birth weight. Deviations of birth weights from gestational age means (birth weight z-scores) were related more to variations in arm muscle area (r = 0.69; p less than 0.001) rather than arm fat area (r = 0.44; p = 0.04). Sixteen infants were followed over 4 weeks. They were most physiologically unstable (mean Physiologic Stability Index score = 5.3 +/- 3.5) during the first postnatal week when they also all lost weight. Their mean arm muscle area decreased significantly during the first week by greater than 10%, whereas the mean arm fat area remained unchanged. First week arm muscle losses were directly correlated with the lack of protein intake (r = 0.52; p less than 0.05). The regression equation predicted a protein intake of 4.06 g/kg/day (95% confidence interval: 2.3-6.4) to prevent first week muscle loss. Enteral intake and weight gain were established after week 1, accompanied by a significant reduction in physiologic instability (PSI score = 1.9 +/- 1.9; p less than 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)

Validation of two scoring systems which assess the degree of physiologic instability in critically ill newborn infants.

Modifications of the Physiologic Stability Index (PSI) and Therapeutic Intervention Scoring System (TISS) were used to evaluate the physiologic stability and need for therapeutic intervention in 55 infants hospitalized in the newborn ICU. After modifying the PSI to reflect neonatal physiology, we found that PSI scores correlated significantly with TISS values (r = .75, p less than .001) and Nursing Utilization Management Intervention System (NUMIS) classifications (r = .62, p less than .001). TISS values also correlated with NUMIS scores (r = .72, p less than .001). PSI and TISS scores increased significantly with each increase in NUMIS classification (p less than .001 for all comparisons). PSI and TISS scores decreased significantly between admission and either discharge (n = 41) or day 14 of hospitalization (n = 14, p less than .001). PSI and TISS scores were greater on days 1 and 5 in infants with hyaline membrane disease when compared with infants with transient tachypnea (p less than .001). Infants with PSI scores greater than or equal to 4 and TISS scores greater than or equal to 7 on day 1 took significantly longer to achieve adequate protein-calorie intakes than infants with lower first-day scores (p less than .002). The modified PSI and the TISS scoring systems are both useful objective measurements of the degree of physiologic instability in newborn infants requiring intensive care, and both scores identify those at increased risk for nutritional morbidity.

Growth failure and altered body composition are established by one month of age in infants with bronchopulmonary dysplasia.

Long-term growth failure and altered body composition are common consequences of bronchopulmonary dysplasia (BPD). We hypothesized that these chronic findings are preceded by uncompensated, acute early growth failure. The purpose of this study was to evaluate the effects of developing bronchopulmonary dysplasia on body composition and growth of very-low-birth-weight (VLBW) infants during the first six postnatal weeks. Arm muscle and fat accretion and changes in weight, length and head circumference were evaluated in 16 very-low-birth-weight infants who developed bronchopulmonary dysplasia and compared with 16 birth-weight-matched control infants without bronchopulmonary dysplasia. During the 1st wk, both groups experienced similarly low nutritional intakes, wasting of arm muscle and fat stores, and reduced weight, length and head circumference growth velocities, compared with intrauterine growth standards. Between wk 2 and 4, infants with developing bronchopulmonary dysplasia consumed less protein and energy (P < 0.05), accreted less arm fat and muscle (P < 0.05), and grew more slowly than control infants in all measured variables (P < 0.05). When infants with bronchopulmonary dysplasia had achieved full enteral feedings and had similar protein-energy intakes to control infants, they demonstrated similar rates of growth and arm muscle and fat accretion, but did not demonstrate catch-up growth. These data support the speculation that early reductions in muscle and fat accretion and growth velocity contribute to the long-term growth failure in infants with bronchopulmonary dysplasia. Prevention may require greater attention to defining and delivering optimal nutritional therapy to physiologically unstable premature infants in the immediate postnatal period.

Reduced neonatal liver iron concentrations after uteroplacental insufficiency.

Neonatal liver (storage) but not heart (nonstorage) tissue iron concentrations were reduced by 60% at autopsy in 15 newborn infants who had gestations complicated by uteroplacental insufficiency because of maternal hypertension or Potter syndrome. The hepatic iron reductions in term and preterm infants, and with either antecedent condition, were similar.

Rate of change of blood pressure in premature and full term infants from birth to 4 months.

Systolic and diastolic blood pressures were evaluated in a cohort of 61 non-hypertensive premature [very low birth weight (VLBW), n = 16; low birth weight (LBW), n = 22] and full-term [normal birth weight (NBW), n = 23] newborn infants admitted to a neonatal intensive care unit (NICU) and followed to their 4-month age-adjusted outpatient examination. All were receiving routine postnatal care by 7 days of age. Blood pressure was measured at 7 days of age, at discharge from the NICU, and at the outpatient examination. Simple linear regression of blood pressure on weight was used to fit a straight line to the three measurements for each infant and the average regression line for each birth weight group was then obtained. There was a significant correlation between systolic blood pressure and both weight and length at each of the measurement points and also between the change in systolic blood pressure and change in weight from the discharge to the 4-month examination. Diastolic blood pressure tended to follow this same pattern. Gestational age was correlated significantly with the 7-day blood pressure, but postnatal age at the outpatient examination was not correlated with either systolic or diastolic blood pressure. The average slopes of systolic and diastolic blood pressure on weight (mmHg/kg body weight) were virtually identical for the LBW and NBW groups; in contrast, the average slope of the VLBW group was greater than the other two groups, and the difference was statistically significant for diastolic blood pressure. These results show significant group differences in mean blood pressure prior to 4 months of age between VLBW, LBW, and NBW groups and, for the VLBW infants, a steeper slope of the estimated regression line of blood pressure on weight between birth and 4 months.

The effect of prolonged intrauterine hyperinsulinemia on iron utilization in fetal sheep.

Newborn infants of poorly controlled insulin-dependent diabetic mothers demonstrate a redistribution of iron from serum and tissue stores into red blood cells. These changes may be due to increases in iron utilization during augmented Hb synthesis, which compensates for chronic intrauterine hypoxemia induced by prolonged fetal hyperinsulinemia. We tested this hypothesis by measuring plasma iron, total iron-binding capacity, percent iron-binding capacity saturation (total iron-binding capacity saturation), Hb concentration, total red cell Hb, and total red cell iron in the arterial blood of 11 chronically instrumented fetal sheep after 7-12 d of infusion with 15 U/day of insulin (n = 5) or placebo (n = 6). The insulin-infused fetal sheep had higher mean +/- SD plasma insulin concentrations (448 +/- 507 versus 11 +/- 8 mU/L; p less than 0.001) and lower arterial oxygen saturations (38 +/- 7 versus 54 +/- 9%; p less than 0.02). The insulin-infused group had a lower mean plasma iron concentration (20.8 +/- 10.9 versus 42.1 +/- 14.7 microM/L; p less than 0.02) and total iron-binding capacity saturation (36 +/- 20 versus 64 +/- 22%; p less than 0.02) and a higher total red cell Hb (45.4 +/- 8.7 versus 32.6 +/- 8.8 g; p less than 0.02) and total red cell iron content (154 +/- 29 versus 111 +/- 29 mg; p less than 0.02) when compared with the placebo group. Seven to 12 d of intrauterine hyperinsulinemia decreases serum iron and increases total red cell iron, most likely by stimulating increased Hb synthesis in response to low arterial oxygen saturation.(ABSTRACT TRUNCATED AT 250 WORDS)

Catch-up growth, muscle and fat accretion, and body proportionality of infants one year after newborn intensive care.

We studied catch-up growth, muscle and fat accretion, and body proportionality at 4 and 12 months of age corrected for prematurity in 30 very low birth weight (VLBW) (less than 1500 gm), 30 low birth weight (LBW) (1500 to 2499 gm) and 30 normal birth weight (greater than or equal to 2500 gm) infants who required newborn intensive care. At 4 and 12 months, the VLBW infants had significantly lower mean weight and length (p less than 0.01), but not lower occipitofrontal circumference percentiles, than the LBW and normal birth weight groups, and showed no catch-up weight or length growth between 4 and 12 months. All three groups had significant increases in mean upper mid-arm circumferences, mid-arm muscle circumferences, and arm muscle areas between 4 and 12 months. Mean mid-arm muscle circumferences and arm muscle areas were similar among the three groups at 4 months but became significantly stratified by birth weight groups by 12 months, with VLBW infants having the lowest mean value. In contrast, analysis of fat stores by triceps skin-fold thickness and arm fat area demonstrated no significant increases in any group between 4 and 12 months, except for arm fat area in the LBW group. The VLBW infants had significantly less fat than normal birth weight infants at 4 and 12 months. All three groups had proportional growth at both visits, as assessed by mid-arm circumference/head circumference ratio and weight-length percentile for age. The VLBW infants were significantly lighter for their length than normal birth weight infants. We conclude that VLBW infants have no first-year catch-up growth, remaining smaller than higher birth weight infants, although appropriately proportional. Somatic growth during the first year is due more to muscle than to fat accretion, especially in VLBW infants.

Changes in nutritional management and outcome of very-low-birth-weight infants.

We compared the in-hospital and postdischarge growth of 47 preterm very-low-birth-weight infants born in 1982 with that of 29 born in 1986. Infants in the two groups were of comparable gestational age, size, and illness at birth. During hospitalization, the 1986 infants began parenteral and enteral nutrition earlier, had fewer days when they received less than 252 kJ/kg, were treated earlier for patent ductus arteriosus (6.1 +/- 4.5 days vs 14.5 +/- 7.7 days), and had a lower prevalence of severe medical complications. By hospital discharge, these infants had significantly higher mean growth percentiles and fewer of them had weights and occipitofrontal circumferences below the fifth percentile. Follow-up at 4 and 12 months corrected age showed that these infants continued to have significantly higher growth percentiles, and fewer of them had weights below the fifth percentile (49% vs 24%) or major neurologic abnormalities. Infants whose weights were below the fifth percentile had significantly poorer 12-month developmental outcomes. We speculate that more aggressive early neonatal nutritional management, changes in cardiopulmonary management, and lower incidences of chronic disease promote earlier onset of and more rapid rates of postnatal growth, which extend through the first year of follow-up.

Metabolic response of preterm infants to variable degrees of respiratory illness.

In older children and adults, physiologic instability associated with severe illness causes increased cellular oxygen consumption (VO2), increased serum lactate and cortisol levels, and more negative nitrogen balance. To determine the metabolic response of preterm infants to severity of respiratory illness, we analyzed VO2, nitrogen balance, urinary 3-methyl-histidine and norepinephrine concentrations, and serum levels of lactate and cortisol as a function of ventilatory index (VI). Twelve 2-day-old premature infants who were appropriate in size for gestational age (mean +/- SEM birth weight: 1460 +/- 251 gm) and who required mechanical ventilation for respiratory distress syndrome had VO2 and carbon dioxide production measured by indirect calorimetry and blood and urine samples obtained concurrently. All infants received amino acids, 1.0 gm/kg per day, and a mean energy intake of 27 +/- 3 kcal/kg per day, provided as a parenteral dextrose solution. The resting energy expenditure exceeded energy intake in all infants. The VO2 value ranged from 5.5 to 9.2 ml/kg per minute and was directly correlated with VI (r = 0.79; p = 0.002). Nitrogen balance ranged from -160 to 53 mg/kg per day (mean: -33 +/- 21 mg/kg per day) but was not dependent on VI (r = 0.04) or VO2 (r = 0.01). The serum lactate level correlated directly with VI (r = 0.82; p = 0.002) and VO2 (r = 0.60; p = 0.05), but cortisol and urinary norepinephrine levels did not. We conclude that preterm infants with respiratory distress syndrome have increased VO2 rates and serum lactate concentrations directly related to the degree of respiratory illness. They are generally in a state of mildly negative nitrogen balance, the degree of which is not related to severity of illness. Although these infants may require increased energy delivery during illness, they do not appear to require excessive amounts of amino acids.

Fetal iron and cytochrome c status after intrauterine hypoxemia and erythropoietin administration.

Chronic fetal hypoxemia stimulates erythropoiesis and may result in a redistribution of fetal iron from plasma into erythrocytes. We studied the response of fetal plasma erythropoietin (Ep) to hypoxemia, the role of Ep in stimulating erythropoiesis in utero, and the effect of augmented erythropoiesis on fetal plasma Ep and iron and tissue cytochrome c concentrations in 19 chronically instrumented late-gestation fetal sheep. The fetuses were stimulated to produce 28 erythropoietic responses after exposure to 1) acute hypoxemia (1-5 days), 2) chronic hypoxemia (greater than 7 days), and/or 3) administration of 1,500 U recombinant human Ep concurrently during normoxemia. Plasma Ep peaked less than 12 h after the onset of hypoxemia or Ep bolus. Plasma iron decreased 24-48 h later and returned to baseline 48-96 h after normalization of Ep levels to baseline. The plasma iron response was directly related to the erythropoietin stimulus (r = 0.79, P less than 0.001) and inversely related to liver iron concentration at death (r = -0.84, P less than 0.001). Nine fetuses with depleted liver iron concentrations at autopsy had significantly lower heart and skeletal muscle iron concentrations compared with animals with 10% of control liver iron remaining. Skeletal muscle and heart iron and cytochrome c concentrations were significantly correlated. Ep has a potent biological effect on fetal erythropoiesis and iron metabolism. Augmented fetal erythropoiesis, mediated by Ep, results in decreased plasma iron, hepatic storage iron, and skeletal and cardiac muscle iron and cytochrome c. The model potentially explains the iron abnormalities found in newborn infants after fetal hypoxia.

Iron deficiency of liver, heart, and brain in newborn infants of diabetic mothers.

Infants of diabetic mothers frequently have polycythemia, elevated serum erythropoietin concentrations, and decreased serum iron and ferritin concentrations, likely representing a redistribution of fetal iron into erythrocytes to support augmented fetal hemoglobin synthesis. We hypothesized that fetal liver, heart, and brain iron concentrations are also reduced in these infants. After obtaining autopsy tissue from infants who had died before 7 days of age, we measured liver, heart, and brain iron concentrations using atomic absorption spectrophotometry. Seven infants of diabetic mothers and seven gestational age-matched control infants were studied. All infants of diabetic mothers had pancreatic islet cell hyperplasia, indicating fetal hyperglycemia and hyperinsulinemia. Liver iron concentrations in the infants of diabetic mothers were 6.6% of control values (489.0 +/- 154.4 vs 7379.7 +/- 1473.8 micrograms/gm dry tissue weight (mean +/- SEM); p less than 0.001), heart iron concentrations were 43.9% of control values (124.7 +/- 20.5 vs 284.1 +/- 34.8 micrograms/gm dry tissue weight; p less than 0.002), and brain iron concentrations were 60.6% of control values (106.1 +/- 13.7 vs 175.2 +/- 10.7 micrograms/gm dry tissue weight; p less than 0.003). Heart and brain iron concentrations were directly correlated with liver iron concentrations (r = 0.80 for both; p less than 0.001) and indicated that hepatic iron was greater than 75% depleted before heart and brain iron reduction. We conclude that severely affected infants of diabetic mothers have reduced liver, heart, and brain iron concentrations. The role of tissue iron deficiency in the genesis of the abnormal clinical findings in these infants deserves further consideration.

Skin and radioallergosorbent tests in patients with sensitivity to bee and wasp venom.

Intradermal (ID) and prick tests with bee or wasp venom (Pharmalgen) have been performed on 102 subjects with a history of adverse reactions to stings and forty-six control subjects giving no such history. Venom was diluted 100, 10 and 1 microgram/ml for prick testing and 10(-2), 10(-2), 10(-3) and 10(-4) micrograms/ml for ID injections. In forty-six control subjects all were tested with the highest concentration of prick testing solution (100 micrograms/ml), eight (17%) had positive reactions, a similar reaction rate to that reported in control subjects using 10(-1) micrograms/ml ID. In our 102 test subjects skin tests were therefore regarded as positive only if the reaction was elicited by 10 micrograms/ml or less by prick test of 10(-2) micrograms/ml or less ID. In general the results with skin prick tests and ID tests were comparable when the prick solution was 1000 times the concentration of that used for ID testing. ID tests were positive in thirteen with negative skin prick, seven of whom had detectable antibodies when tested by RAST. Conversely four with a positive skin prick test (two of whom were RAST positive) were considered negative on ID testing. As judged either by RAST or skin tests it appeared that sensitivity diminished with the time interval from the last sting (P less than 0.001).

The relationship between decreased iron stores, serum iron and neonatal hypoglycemia in large-for-date newborn infants.

We assessed the relationship between neonatal hypoglycemia and newborn iron status in 15 hypoglycemic, large-for-date newborn infants, 12 of whom were infants of diabetic mothers. These infants had significantly lower mean serum iron concentrations, ferritin concentrations, percent iron-binding saturation and calculated iron stores, and significantly higher mean transferrin concentrations, total iron-binding capacity concentrations and mid-arm circumference:head circumference ratios when compared with either 15 euglycemic large-for-date or 15 euglycemic appropriate-for-date control infants (p less than 0.001 for all comparisons). All hypoglycemic infants had ferritin concentrations below the 5th percentile as compared to 3% of controls (p less than 0.001), and 67% had transferrin concentrations above the 95th percentile (controls: 0%; p less than 0.001). Only the hypoglycemic infants demonstrated a significant negative linear correlation between ferritin and transferrin concentrations (r = -0.83; p less than 0.001). Decreased serum iron concentrations were associated with size at birth (r = -0.60; p = 0.01) and with increased red cell iron (r = -0.60; p = 0.01), implying a redistribution of iron dependent on the degree of fetal hyperglycemia and hyperinsulinemia. Infants with increased red cell iron had more profound neonatal hypoglycemia. These results show a significant association between decreased iron stores and neonatal hypoglycemia in macrosomic newborn infants associated with a significant shift of iron into red blood cells.

Abnormal iron distribution in infants of diabetic mothers: spectrum and maternal antecedents.

Because chronic hypoxemia causes a redistribution of iron from serum and storage pools into an expanding erythrocyte mass, and because infants of diabetic mothers are often hypoxemic in utero and have a high prevalence of polycythemia at birth, we studied iron distribution in 43 term infants of diabetic mothers. Twenty-four infants were at an appropriate size for gestational age; 19 were large for gestational age. At birth, 28 infants (65%) had abnormal serum iron profiles; eight had decreased ferritin concentrations only (stage 1), nine had decreased ferritin and increased total iron-binding capacity values (stage 2), and 11 had these serum findings plus elevated free erythrocyte protoporphyrin concentrations (stage 3). The hypoglycemic infants who were large for gestational age (n = 14) had a higher prevalence of abnormal iron profiles than euglycemic infants who were appropriate in size for gestational age (n = 20; 93% vs 50%; p = 0.009). Progressively abnormal iron profiles were associated with higher glycosylated fetal hemoglobin values, greater degrees of macrosomia, increased hemoglobin and erythropoietin concentrations, and increased erythrocyte/storage iron ratios. Erythropoietin concentrations were inversely linearly correlated with serum iron values (n = 32, r = -0.54; p = 0.003). The combined erythrocyte and storage iron pools were significantly lower in infants with abnormal iron values whose mothers were diabetic, particularly in infants of women with confirmed diabetic vasculopathy. We speculate that these findings are likely due to (1) increased fetal iron utilization during compensatory hemoglobin synthesis in response to chronic hypoxemia and (2) reduced iron transfer during late gestation complicated by diabetes.