Genetic and biochemical markers in relation to iron transport in obese and diabetics pregnant women. Marcadores genéticos y bioquímicos en relación al trasnporte de hierro en embarazadas obesas y diabéticas

Resumen

GENERAL SUMMARY 1. Background Optimal maternal nutrition is now widely recognised as being essential for optimal fetal growth and development and considerable interest is being shown in the way in which nutrition during pregnancy and after birth interacts to determine fetal and postnatal health. Once formed, the placenta is very efficient at transfer of nutrients to the conceptus because of the organization of the maternal and fetal vasculatures. Nonetheless, nutritional deprivation or other insults to placental function can compromise fetal development and cause adverse effects on the physiology of the offspring that can persist into adulthood. Iron deficiency anaemia (IDA) is a common problema in pregnancy, and has been associated with adverse pregnancy outcomes. The consequences are serious for both the mother and her infant. Now we know that it is a risk factor for pre-term delivery and subsequent low birth weight. During pregnancy, the absorbed iron is predominantly used to expand the woman's erythrocyte mass, fulfill the foetus's iron requirements and compensate for iron losses (i.e. blood losses) at delivery. Iron absorption is regulated by the size of body iron stores (Finch 1994). The increase in absorption is most pronounced after 20 weeks of gestation and peaks in late pregnancy. Growth of the fetus and of the placenta, and the larger amount of circulating blood in the pregnant woman, lead to increase in the demand of nutrients, one of them is iron. The daily requirements for iron for a woman in the last trimester of pregnancy are six times greater than for a non-pregnant woman (Christensen RD 2004). Markers of iron status are haemoglobine, haematocrite, ferritin, transferrin, serum transferrin receptor, plasma transferrin saturation and others such serum iron, mean cell haemoglobin (MCH), mean cell volume (MCV). Serum ferritin (sFnt) is considered to be a reliable blood test in the first trimester of pregnancy for judging whether iron supplementation is necessary. However, becomes less reliable after the 20th week due to the physiological dilution of the plasma, which reduces more or less the concentration of serum ferritin indenpendently of changes in the iron stores (Schwartz III, Thurnau 1995, Sandstad et al. 1996, Allen 2000). Serum ferritin usually falls markedly between 12 and 25 wk of gestation and the concentration reduces steadily to ~50% of normal at mid-gestation due to hemodilution and the mobilization of iron from stores to meet the increased needs of pregnancy and expansion of the maternal blood cell mass (Fenton, Cavill & Fisher 1977, Milman et al. 1999). Therefore, the status of iron stores in the second and third trimesters of pregnancy cannot be accurately determined by s-Fnt alone. The physiologic variations in ferritin during pregnancy may be compensated for by calculating body iron using the sTfR-to ferritin ratio (Cook, Flowers & Skikne 2003, Milman et al. 2006). Some investigators have suggested that the sTfR-to-ferritin ratio may be a better marker of body iron status than ferritin alone (Malope et al. 2001, Punnonen, Irjala & Rajamaki 1997). Iron is transported from the mother to the fetus across the placental membrane by an active process, which is mediated via binding of maternal transferrin bound iron to transferrin receptors in placenta and subsequent transfer of iron into the fetal circulation (Fletcher, Suter 1969, Brown, Molloy & Johnson 1982). The efficiency of this transport system implies that iron deficiency in the newborn is encountered only at extreme iron deficiency in the mother, so that iron deficiency in mature newborn babies is a rare event in the developed countries. In the study by Rusia et al. (Rusia et al. 1996), serum transferrin receptor concentrations were higher in infants born to anemic mothers. Furthermore, the ferritin-GD association was modified by level of obesity. Obese women with high ferritin levels had a 3.5-fold increased risk of developing GDM (95% CI: 1.35, 9.27; p=0.01), whereas results were not significant among nonobese women. Whereas placental transferrin receptor expression is increased in pregnancies complicated by diabetes mellitus, the affinity of the receptor to maternal transferrin is decreased, probably due to hyperglycosylation of the oligosaccharides present in the binding domain (Georgieff et al. 1997). Furthermore, placental vascular disease might be present in mothers with longstanding, poorly controlled diabetes mellitus, further limiting iron transport across the placenta. Tissue iron is depleted to support the iron needs of augmented erythropoiesis under these situations. Nearly 65% of infants of diabetic mothers (IDM) have perinatal iron deficiency, as suggested by cord serum ferritin concentration <60 mg/L. In approximately 25% of these infants cord serum ferritin is <35 mg/L, suggesting significant depletion of tissue iron, including brain iron (Georgieff et al. 1990, Petry et al. 1992). Until recently, few studies had considered body weight or body composition as factors related to iron deficiency. Many of them have shown that obesity might increase the risk of iron deficiency but, at the same time, obese subjects exhibit high serum ferritin levels. Obesity is associated with alterations in iron metabolism. The two major characteristics are a deficit in serum iron levels and an increase in ferritin. Iron deficiency in obesity appears to be multifactorial and includes (i) A decrease in iron food intake; (ii) An impairment of intestinal iron uptake and iron release from stores because of an overexpression of hepcidin and (iii) Inadequate iron bioavailability because of inflammation. In addition, abnormal ferritin concentrations can be explained by chronic inflammation rather than by iron overload. Moreover, it appears that hypoferremia could be explained by both a true iron deficiency and a functional iron deficiency (Zafon, Lecube & Simó 2010). Pinhas-Hamiel et al. (Pinhas-Hamiel et al. 2003, Nead et al. 2004) have reported that low iron levels were present in 38.8% of obese children, in 12.1% of the overweight children and in only 4.4% of children of normal weigh. Another study has demonstrated that the prevalence of iron deficiency increases as body mass index (BMI) increases from normal weigh to overweight in a sample of nearly 10 thousand children and adolescents (Nead, 2004). In the adult population, one analysis from the Third National Health and Nutrition Examination Survey (NHANES III) showed that BMI was associated with significantly lower mean serum iron concentrations in women but not in men (Micozzi, Albanes & Stevens 1989). Conversely, it has also been suggested that iron deficiency, especially perinatal iron deficiency, might lead to increased visceral adiposity (Komolova et al. 2008, McClung et al. 2008). Serum sTfR seems to be a sensitive marker of iron deficiency in pregnancy, and by combining serum sTfR and serum ferritin measurements, the entire spectrum of iron status in pregnancy can be assessed (Carriaga et al. 1991, Åkesson et al. 1998). Three recent articles have analysed sTfR, and all of them have reported that the levels of sTfR are elevated in obese patients (Lecube et al. 2006, Freixenet et al. 2009, Yanoff et al. 2007). In addition, the chronic inflammation and increased leptin production characteristic of obesity increase hepcidin secretion from the liver (Chung et al. 2007), which, along with hepcidin produced by adipose tissue (Bekri et al. 2006), could reduce dietary iron absorption (Laftah et al. 2004). Iron deficiency in obese individuals may result from different factors. Low iron intake, reduced iron absorption, and the sequestration of iron as a result of chronic inflammation in response to excess adiposity has been suggested among differents reasons. In regard to an iron-poor diet, low iron intake and increased iron needs have been reported among obese children and adolescents who are iron deficient (Pinhas-Hamiel et al. 2003, Nead et al. 2004, Hassapidou et al. 2006). Zimmerman et al. (Zimmermann et al. 2008) also reported that high BMI Z-scores were associated with decreased iron absorption in women independent of iron status and reduced improvement of iron status in iron-deficient children following intake of iron-fortified foods. They hypothesized that obesity may affect iron absorption through an inflammatory mediated mechanism. The infiltration by and activation of macrophages in adipose tissue has also been linked to obesity-induced IR (Apovian et al. 2008). In accordance with this conception, obesity-associated iron abnormality has been interpreted as a feature that mimics the so-called anaemia of chronic inflammation, which is characterized by hypoferremia and high to normal serum ferritin concentration (hypoferremia and anaemia despite adequate iron stores) (Ausk, Ioannou 2008). This entity is also caused by increased inflammatory cytokines, especially IL-6, inducing increased production of the iron-regulatory hormone hepcidin. Thus, hepcidin by means of its capacity to block iron release from macrophages, hepatocytes and enterocytes appears to be a major contributor to the hypoferremia associated with inflammation (Ganz 2006). Hepcidin acts as an inhibitory iron regulator. Increased plasma hepcidin inhibits intestinal iron uptake and acts sequestering iron at the macrophage (Knutson et al. 2005), which could lead to decreased iron stores and hypoferremia. In accordance with this homeostatic model, iron loading increases hepcidin gene expression. Also, its production is suppressed by anaemia and hypoxaemia. Furthermore, hepcidin synthesis is markedly induced by infection and inflammation and because chronic disease (Park et al. 2001, Nicolas et al. 2002, Weinstein et al. 2002), and regulation mediated by cytokines, predominantly IL-6 (Nemeth, Ganz 2006). The potential role of hepcidin in the development of iron deficiency in the obese is supported by the discovery of elevated hepcidin levels in tissue from patients with severe obesity, and the positive correlation between adipocyte hepcidin expression and BMI (Bekri et al. 2006). Besides, it has been reported that leptin up-regulates hepatic hepcidin expression, suggesting that increased leptinemia in obesity could be a contributor to aberrant iron metabolism (Chung et al. 2007). Therefore, leptin might be part of the axis that links obesity, inflammation, and hepcidin release with aberrant iron metabolism. For other hand, iron overload and the associated oxidative stress contribute to the pathogenesis and increase risk of type 2 diabetes and other disorders. As mentioned before, in iron overload, the accumulation interferes with the extraction, synthesis and secretion of insulin (Fernandez-Real, Lopez-Bermejo & Ricart 2002) and moderately elevated iron stores also increase the risk of type 2 diabetes (Jiang et al. 2004a). In pregnancies complicated by maternal diabetes, the foetus is hyperglycaemic, and hiperleptinic (Cetin et al. 2000, Tapanainen et al. 2001). Newborns small for gestational age (SGA) also show a marked reduction in body fat mass at birth, which mainly reflects the decrease in lipid accumulation in adipocytes (Levy- Marchal, Jaquet 2004). Thus, a change in the programming of the synthesis, secretion or actions of leptin may be decisive in the early origins of obesity after exposure, both above and below the needs of fetal or early neonatal life. Recent studies seem to indicate that obesity is associated with iron deficiency although the aetiology appears to be multifactorial and includes: i) A decrease in iron food intake; ii) An impairment of intestinal iron uptake and iron release from stores because of an overexpression of hepcidin; and, iii) Inadequate iron bioavailability because of inflammation. In addition, abnormal ferritin concentrations can be explained by chronic inflammation rather than by iron overload (Yanoff et al. 2007). Recent studies are emerging suggesting an association between perinatal iron deficiency and programmed obesity in the adulthood, although the mechanism explaining this relationship is unclear. The link between chronic diseases and anemia is well characterized. Leptin (LEP), the human homolog of the mouse obesity (ob) gene, is positioned in the chromosome 7q22-35 and is the most prominent candidate gene linked to body mass index (BMI). The leptin receptor, also identified as the diabetes gene product, is a single transmembrane protein that is established in many tissues and has several alternatively spliced isoforms. The results of linkage studies done on obese human beings using markers near the leptin (LEP) and leptin receptor (LEPR) gene regions are still controversial. LEP has been linked to extreme obesity in a French study (Clement et al. 1996) and a Pennsylvanian population but not in Pima Indian sibling pairs (Reed et al. 691-94). In humans, LEP and LEPR have been mapped to 7q31.3 (Green et al. 1995) and 1p31 (Winick, Stoffel & Friedman 1996), respectively. There is an increasing body of evidence that suggests a direct link between being overweight or obese and having poor iron status (Yanoff et al. 2007, Bekri et al. 2006). Adipose tissue is an active endocrine organ and releases a number of cytokines and adipokines (Lago et al. 2007), which may in turn influence iron metabolism. Leptin, an adipocyte-specific hormone that regulates body weight, was the first adipokine to be discovered, and is intriguing in this regard for 3 reasons: 1) it belongs to the family of long-chain helical cytokines (Zhang et al. 1997); 2) its circulating levels are proportional to fat mass (Considine et al. 1996); and 3) its membrane receptors exhibit structural similarity to class I cytokine receptors. The hypoferremia noted in obese subjects appeared to arise from a combination of 2 distinct mechanisms: 1) the development of iron deficiency (Lecube et al. 2006, Yanoff et al. 2007) and 2) the presence of chronic low-grade inflammation that resulted from the enhanced production and release of a cocktail of proinflammatory cytokines and adipokines from the adipose tissue (Lago et al. 2007). These inflammatory stimuli in turn lead in the liver to an increase in the expression of hepcidin (Weinstein et al. 2002), a 25 amino acid peptide hormone, which once released into the circulation is thought to bind to the iron efflux protein ferroportin (Nemeth et al. 2004) inhibiting the release of iron recycled from senescent red blood cells by reticuloendothelial macrophages (Knutson et al. 2005) and the absorption of iron by duodenal enterocytes (Laftah et al. 2004, Yamaji et al. 2004), resulting in hypoferremia (Rivera et al. 2005). The leptin can directly regulate hepatic hepcidin expression. Increased production of hepcidin in the presence of leptin was predicted to result in decreased duodenal iron absorption and impaired iron recycling from reticuloendothelial macrophages because of the inhibitory actions of hepcidin on ferroportin protein expression. Together with other stimuli, such as proinflammatory cytokines, leptin can now be added to the list of adipose-derived factors that may contribute to the hypoferremia observed in the overweight and obese population. Obesity is a polygenic disorder, which has several candidate genes that play a role in determining the final severity. There are some SNPs of LEP gene involved in obesity physiopathology, such as A19G, A2548G in LEP gene, and Q223R in LEPR gene. So, mutations in the leptin gene lead to defective leptin production and cause recessively inherited early onset obesity (Mammes et al. 1998). Obese individuals homozygous for the G-allele showed significantly lower leptin concentration compared to obese patients either heterozygous or homozygous for the A-allele after correction for BMI (Jiang et al. 2004b). Le Stunff et al. (Le Stunff et al. 2000) have confirmed that the recessive effect of the LEP G-2548A variant could potentially alter leptin expression, and female subjects with the A/A homozygote had 25% lower mean leptin levels than girls with other genotypes. Wang et al. observed (Wang et al. 2006) that the BMI of the G/G genotype was significantly higher than that of G/A and A/A genotypes in extreme obesity, and found that the LEP G-2548A polymorphism was associated with extreme obesity in Taiwanese aborigines. Recently, it has been shown that LEP -2548GG genotype appear to be important in regulating leptin levels, whereas the LEPR 223R allele might predispose healthy subjects to develop metabolic disturbances (Constantin, et al. 2010). The association of the LEPR p.Q223R polymorphism with obesity was related to the co-dominant and dominant model, but not with the recessive model. There is hypothesis that the p.Q223R LEPR variant is associated with a BMI increase. It has been proposed the hypothesis that variation of LEPR is participate in the union with leptin and influence on leptin serum levels. Turn, leptin levels can influence on iron metabolism. 2. General aims - To analyse the effect of mother obesity and/or gestational diabetes during pregnancy on iron status in the mother and in the offspring, and its implication on fetal growth . - To study the role of obesity and/or gestational diabetes in pregnant women on the placental expression of TfR, and the mechanism involved in iron transplacental transport related to this biomarker. - To explore the potential effect of leptin polymorphisms on the iron metabolism during pregnancy in obese and gestational diabetic mothers. 3. General Material and Methods 3.1. Subjects and study design The subjects were participants in a longitudinal study of maternal nutrition and genetic on the foetal adiposity programming (Preobe study P06-CTS-02341), supported by Consejería de Innovación, Ciencia y Empresa de la Junta de Andalucía, Spain. This prospective, observational study was developed during 2007 until 2010. A total of 350 pregnant women aged between 18 and 45, with singleton pregnancies, were recruited at from 12 to 20 weeks of pregnancy at the Clinical University Hospital San Cecilio', Ambulatorio Zaidín-Vergeles, and Hospital Materno-Infantil in the city of Granada, Spain. A number of 4 of them were underweight with pre-pregnancy BMI <18.5 kg/m2 and did not fulfil inclusion criteria and were therefore excluded. 61 women did not complete the study, attending only 1 or 2 visits, and were therefore not included in the results of this study. Only 285 of 350 women recruited were therefore included in this study. 3.2. Pre-pregnancy Body mass index classification At week 20 of gestation, women were classified according WHO 2009 criteria (Anonymous1995b) related to their pre-pregnancy BMI into three groups: normalweight women (n=158) with BMI 18.5-24.9 kg/m2; overweight women (n=65) with BMI 25-29.9 kg/m2 and obese women (n=62) with BMI ¿30 kg/m2 (Table 1; Figure 1). After the first visit at 20 weeks of pregnancy, women were examined by the obstetrician at 24 (second trimester) and 34 weeks (third trimester) and clinical parameters were recorded. Data on weight before pregnancy and before delivery were used to calculate weight gain during pregnancy. Normal weight gain ranges were from 11.5 to 16.0 kg for normal-weight women (BMI 18.5-24.9 kg/m2); from 7.0 to 11.5 kg for overweight women (BMI 25-29.9 kg/m2) and from 5.0-9.0 kg for obese women (BMI ¿30 kg/m2, respectively, over pregnancy according to the Institute of Medicine (IOM) criteria (Anonymous2009). Total weight gains above these values, 16 kg for normal-weight women; 11.5 kg for overweight women and 9.0 kg for obese women, were considered excessive weight gains. 3.3. Gestational diabetes mellitus diagnosis Pregnant women who were diagnosed as having pre-gestational diabetes mellitus (type-1 diabetes mellitus) were excluded from the study. A total of 16 women (5.6%) of the total 285 recruited from 12-20th week of pregnancy were diagnosed gestational diabetes mellitus (GDM). Anothers 41 women were recruited between 24th-26th week of pregnancy, as the time the diabetes was being diagnosed to achieve a reasonable size of the group. The women were asked to participate in the study if they had gestational diabetes diagnosed at the time they were recruited. Pregnant women were initially screened by measuring the plasma glucose concentration 1 h after a 50 g oral glucose challenge test (GCT) at 24-28 weeks of gestation. A diagnostic oral glucose tolerance test (OGTT) was performed on the subset of women whose plasma glucose concentrations reached or exceded the glucose threshold value (¿140 mg/dl, 7.8 mmol/L). A fasting plasma glucose level >126 mg/dl (7.0 mmol/L) or a casual plasma glucose >200 mg/dl (11.1 mmol/L) meets the threshold for the diagnosis of diabetes, if confirmed on a subsequent day, and precludes the need for any glucose challenge. In the absence of this degree of hyperglycemia, evaluation for GDM in women with average or high-risk characteristics was developed. High risk women, including maternal age ¿ 35 years, BMI ¿ 30 Kg/m2, relevant past obstetric and family history, impaired glucose metabolism, etc., were screening and diagnosed in the first trimester of the pregnancy. The diagnosis of GDM was made for the clinicians at the hospital based on an oral glucose tolerance test (OGTT) and the results were interpreted according to the National Diabetes Data Group (NDDG) criteria (Anonymous1979) and the Third International Workshop- Conference on Gestational Diabetes Mellitus (Metzger 1991) (Table 2). A 100 g oral glucose tolerance test (OGTT) was arranged during the second trimester for all the women, except for those with risk factors that were made earlier during the first trimester. After the test of Glucose tolerance, at 34 weeks of pregnancy the diagnosis of gestational diabetes was established, determining the following groups (see Figure 1): 1. Control group, normal-weight women (n=153) with pre-pregnancy BMI 18.5-24.9 kg/m2. 2. Overweight women (n=61) with pre-pregnancy BMI 25-29.9 kg/m2. 3. Obese women (n=55) with pre-pregnancy BMI ¿30 kg/m2. 4. Pregnant women with gestational diabetes independently of prepregnancy BMI (n=57). 4. Results 4.1. Descriptive analysis: The conditions during pregnancy are listed in Table 5. This included age, smoking, hypertension, mode of delivery, duration of pregnancy, maternal bodyweight gain, family history for diabetes among others. Mother's sociological characteristics are listed in Table 6, including ethnicity, studies, employment, among others; and neonatal clinics characteristics accordingly to mother's classification after GD diagnosis are shown in Table 7. 4.1.1. Perinatal and maternal clinical characteristics (Table 5) In Table 5, the perinatal and clinical characteristics of the control, overweight and obese groups at 20 weeks of pregnancy are shown. The three groups were comparable in terms of age, height, conditions during pregnancy (weight gain recommendations during pregnancy by IOM 2009, smoking habits), parity and previous macrosomic offprings and abortions). Duration of pregnancy was as mean higher than 38 weeks with no difference among the groups (Table 7). Related to anthropometrics measurement, as expected, there were significant differences among groups, so that obese women showed higher weight and BMI before and during pregnancy. Also body fat assessed during pregnancy was higher in obese group. Almost half of the normal weight and obese groups (46.3%, 50% respectively) reached a weight gain over pregnancy less than recommended by IOM 2009, while almost the half of the overweight group (45.2%) reached a weight gain accordingly to IOM 2009 recommendations. Related to smoking habits, 34.8% of the total women used to do it before pregnancy and 14.6% were smoking during pregnancy. This habit was more usual among obese and overweight women before and during pregnancy respectively, but no significant differences among groups were found. Related to diabetes diagnosis during pregnancy, 4.3% of the total women were diagnosed of having impaired glucose tolerance (IGT), and 23.7% of having gestational diabetes (GD). The occurrence for IGT was significantly higher in the obese group compared to overweight and control groups, while GD diagnosis was significantly higher in the obese and overweight compared to control, thus, 5 women of 158 from control group developed gestational diabetes (3.16%), in front of 4 women of 65 from overweight (6.15%) and 7 of 62 women from obese (11.29%). Accordingly with iron status, the risk to develop GD was lower in women with iron deficiency (ferritin < 12¿g/L) during pregnancy than in women with ferritin levels higher than 12¿g/L. That is, women with ferritin levels lower than 12¿g/L at 24 weeks of pregnancy had a 1.2-fold greater probability of not develop GD than those with ferritin >12¿g/L) (95% CI: 0.63-2.36). At 34th week the probability increased (1.9-fold greater probability of not develop GD, P=0.061, Pearson X2=3.50, df=1). At delivery, women with ferritin levels > 12¿g/L was associated with a significantly higher rate of GD compared with women with iron deficiency (ferritin < 12¿g/L) (P=0.020, Pearson X2=5.39, df=1). Also, women with ferritin levels lower than 12¿g/L at delivery had a 7.92-fold greater probability of not develop GD than those with ferritin >12¿g/L) (95% CI: 1.03- 60.90). Also, women with ferritin levels lower than 41.1¿g/L at 24 weeks of pregnancy had a 2.45-fold greater probability of not develop GD than those with high ferritin levels (95% CI: 0.729-8.23). At 24th week of pregnancy, control group women with high ferritin levels (>41¿g/L) had a 5.62-fold greater probability of not develop GD than those obese and overweight women with high ferritin levels (95% CI: 0.50-63.28). Also, 14.8% of the women with high ferritin values at 24th of pregnancy developed GD, and 75.0% of the GD women were overweight or obese and only a 25.0% were normal. At 34th week, the probalility of not develop diabetes with ferritin levels lower than 41.1¿g/L was increased to 4-fold compare to those with high ferritin levels (95% CI: 1.46- 11.11). A percentage of 1.1% and 35.4% of the total women participating in the study asserted to have a family history of obesity and diabetes, respectively. Familiar history of diabetes was more frequent in the obese and overweight group compared to the control group, and the difference was significant. Familiar history of obesity was more frequent in the obese group compared to the others and the differences were also significant. Most of the women were nuliparous or primiparous. Multiparous women (having 2 or more birth) were less frequent in all groups. Of the total women taking part in the study, previous macrosomic offspring were found in 1.7%; spontaneous previous abortions happened in 25.2% and cesarea delivery happened in 19.7%. Cesarean delivery was significantly more frequent in the obese and overweight group (25.6% and 31.6% respectively) compared with control women. Blood pressure assessed at 12 and 24 weeks of pregnancy was higher in the obese and overweight groups compared to control and the difference was significant. At 34 weeks of pregnancy blood pressure was significantly higher in obese group compared to control and no differences were found between overweight's SBP and others groups, while overweight's DBP showed differences with obese group but not with control. However, after GD diagnosis significant differences were found in mother's age (P=0.000), pre-pregnancy BMI (P=0.000). Women with GD were older (34.08±4.26 years) than obese (29.10±4.68 years, P=0.000), overweight (31.74±4.58 years, P=0.044) and control (30.83±4.29 years, P=0.000), and prepregnancy BMI was significantly different (27.89±5.95) compared to obese (33.41±3.60, P=0.000) and control (22.00±1.69, P=0.000), and no significant differences were found compared to overweight pre-pregnancy BMI (27.26±1.29, P=0.956). 4.1.2. Neonatal clinical characteristics (Table 7) The four groups of neonates were similar at birth in terms of gestational age, APGAR score and anthropometric parameters, except for birth weight and waist circumference. The neonates born from obese mother had significantly higher birth weight than those babies born from the healthy pregnant women (P=0.067 ANOVA). The neonates born from GD women had significantly higher waist circumference than those born from the control group of women (P=0.025 ANOVA). Upper arm circumference was higher in neonates born from GD mothers (11.63±2.46) compared to healthy mothers (10.88±0.96), but differences were not significant. Ponderal index was higher in neonates born from overweight women and smaller in neonates born from healthy mothers, but differences between groups were not significant. Preterm neonates were reported in a 3.1% of the total of the subjects. Obese women had a 5.6% of preterm deliveries and GD group had a 1.8%, but the differences in the occurrence of preterm birth were not significant among groups. Post term neonates were reported in 1.2% of the total neonates and were not found in those born from obese and overweight mothers. In the total of neonates, 1,2% were SGA, 84,3% were NGA and 14.5% were LGA. No significant differences were found among the 4 groups of neonates, but neonates born from obese women showed more incidence of LGA (25.0%) compared to neonates from others women groups (Control: 8.8%; Overweight: 17.9%, GD: 18.2%). No SGA was shown for obese group. 90.0% of the neonates born from healthy women were NGA. Low birth weight was present in a 4.2% of the total neonates. Low birth weight was not found in neonates born from obese women. Differences between groups were not significant but the highest incidence of low birth weight was found in GD and overweight group (5.3% and 5.1% respectively) compared to healthy and obese women (4.7% and 0.0% respectively). Macrosomia was present in a 7.3% of the newborn (19 newborn from a total of 261). The prevalence was greater among newborn from obese women (16.7% of newborn from obese were macrosomic; 10.3% overweight; 4.7% control and 5.3% GD). However, the difference didn't reach significance. Newborn from non ID mothers at 24th week of pregnancy (defined by sTfR < 28.1nmol/L) had 4.79- fold greater probability to be born normal weight and no macrosomic, than whose born from ID mother (defined by sTfR >28.1nmol/L). The probability at 34th week was lower 1.1-fold and increased up to 3.3-fold at delivery. But didn't reach significance. Placental weight was higher in the obese group compare to healthy one, and the differences were nearly to be significant (P=0.055). Placental long and small diameter seems to be higher in obese group compared to the others groups and also placental/fetal weight ratio, but the differences were also not significant. A 19.7% of the total neonates were born by cesarean delivery. The rate of cesarean delivery was significantly higher in obese, overweight and GD compared with control group. 4.1.3. Maternal iron status in the different groups of mothers. At 24 week of pregnancy 72.1% of the women were iron sufficient, 24.6% were iron deficient without anemia and only 3.4% were iron deficient with anemia. At 34 weeks of pregnancy the proportion of iron deficiency with anemia was increasing reaching their highest proportion, as the iron sufficiency was decreasing (53.2% iron sufficient; 27.4% iron deficient without anemia; 19.4% iron deficient with anemia). At delivery this situation changed with iron sufficiency proportion rising again (82.4%; 10.1%; 7.6% respectively). No significant differences between maternal groups were found. ID accordingly with the definition of ferritin < 12 ¿g/L, were found in 28.7% of the total women at 24 weeks of pregnancy; 43.6% at 34 weeks and 14.5% at delivery with the significantly lowest prevalence among GD (2.7%) and the highest prevalence among obese (31.6%). Iron depletion (ferritin < 20 ¿g/L) was found in 53.6%, 68.6%, 45.9% at 24th, 34th and at delivery respectively in the total women with the significantly lowest prevalence among GD (32.4%) and the highest prevalence among obese (78.9%). At 24th weeks of pregnancy 12.9% of the women had elevated ferritin levels (> 41 ¿g/L); at 34th week of pregnancy the women with high ferritin levels were reduced to 3.4% and at delivery the amount was 11.9%, with GD women showing the highest frequency being significantly different from the others groups at delivery (25% at 24th; 8% at 34th and 24.3% at delivery). ID accordingly with the definition of sTfR >28.1, were found in 4.4%, 22.7%, 10.7% at 24th, 34th and at delivery respectively in the total women. ID accordingly with sTfR/sFtn index was 18.3%, 46.5%, 21.3% at 24th, 34th and at delivery respectively in the total women. ID accordingly with the definition of TBI (negatives values means iron deficiency) were 64.9%, 80.6% and 71.0% at 24th, 34th and at delivery respectively in the total women with the GD women showing the lowest values of TBI < 0, reflecting a better iron status concern to this parameter. A multiple linear regression model with sTfR as dependent variable and serum iron, hemoglobin, ferritin, iron supplementation, dietary iron intake, birth weight, and maternal BMI as independent variable was developed in all the women with the method introduce, to predict iron deficiency. The model included as predictors serum iron, serum ferritin and neonatal birth weight (R2=0.29, P=0.000). No one of the others variables predicted sTfR. Serum iron and ferritin were inversely correlated to sTfR whereas birth weight was positively correlated. Ferritin alone contributed to 21.5% of the total variance predicted by this model, whereas serum iron alone accounted for 14.3% and birth weight alone 0.05%. No correlation between BMI and sTfR was found. Another multiple linear regression model with serum ferritin as dependent variable and maternal age, parity, smoking, iron supplementation, dietary iron intake, maternal BMI, gestational diabetes (as equivalent to C-R protein) and sTfR as independent variable was developed with the method introduce in all the women, to predict iron deficiency and risk of gestational diabetes. The model counted only as predictor variables to sTfR, parity and gestational diabetes. sTfR and parity were correlated to ferritin levels inversely whereas gestational diabetes was correlated directly (R2=0.28, P=0.000). sTfR alone explained 21.5% of the variance of ferritin levels, whereas gesattional diabetes explained 3.6% and parity explained only 1.2%. However, maternal BMI has been correlated inversely in control women, but not in the others groups of mothers. 4.1.4. Longitudinal and inter-groups analysis of the haematological and biochemical parameters during pregnancy The General Model for repeated measures showed not significant differences in GD group during pregnancy in all parameters measured except for transferrin. As was expected through pregnancy, transferrin increased significantly from 24th week to 34th week in all the groups (P=0.000 control and overweight; P= 0.001 obese; P=0.007 GD) and decreased slightly from 34th to delivery. Thus, transferrin at delivery was higher than at 24 weeks of pregnancy in all the groups (P=0.068 control and P=0.005 overweight). Only control group showed a significant decrease from 34th week to delivery (P=0.024). No significant changes in serum iron were found in any group of women during pregnancy, however a progressive increase was observed from the 24th week to delivery in control and overweight. In contrast, in obese and GD women, a progressive decrease during pregnancy was shown. Obese women showed low serum iron levels compared to others groups, but this difference was not significant. Ferritin and RBC decreased in all the groups from 24 to 34 weeks, but differences were significant only in overweight women for RBC (P=0.008). At delivery, maternal ferritin and RBC levels increased in all groups from 34th week of gestation, but this increase was significant only in control group (P= 0.000 for ferritin and RBC). At delivery, ferritin levels were significantly low in obese compare to control (P= 0.016), while RBC levels were significantly low in overweight and obese women compare to control group. GD diabetes group had high ferritin levels during all pregnancy compare to the others, however only at delivery significant differences were showed, that was between control and obese group who had the lowest levels. Hb, MCV and MCH decrease significantly in control, overweight and obese women from 24 to 34 weeks (Hb: P=0.000 in control and overweight; P=0.046 in obese; MCV: P=0.000 in control, overweight and obese; MCH: P=0.000 in control and overweight and P=0.001 in obese), and in GD women this decrease was not significant. At delivery, Hb levels increased significantly in control women from 34 weeks (P= 0.001), reaching the same levels that had at 24 weeks. In contrast, overweight women showed a significantly decrease in Hb levels from 24th to 34th week and delivery (P=0.000). At delivery, RBC, Hc and Hb levels in control women were significantly higher than those found in obese and overweight with lower levels. MCV continued to significantly decrease between the 24th week of gestation and delivery in all groups except for GD group (P=0.000). MCH decreased significantly from 34th week to delivery only in control group (P=0.000), and this decrease was not significant in the others group of pregnant. A significant drop in MCH and MCHC was found in control group from 24th to 34th week and from 34th to delivery (MCH: P=0.000; MCHC: P=0.027; P=0.005 respectively). In contrast, not significant changes in MCHC were found during pregnancy in the others groups. TSAT index showed a progressive drop during all time-points assesed during pregnancy in all groups except for control, in which no changes were observed during pregnancy. The decrease observed between 24th and 34th was significant only in overweight group (P=0.053). The decrease observed between 34th and delivery was not significant in any group. A progressive increase during pregnancy in RDW was observed in all groups. The increase observed between 24th and 34th week was significant only in control (P=0.001). From 34th to delivery RDW was increased in control (P=0.000), overweight (P=0.004) and obese (P=0.005) and between 24th and delivery (control: P=0.000; overweight: P=0.008); obese: P=0.015). Hc decreased from 24th to 34th week in all group in while sTfR was increased. This changes were shown significant only in control and overweight groups for both parameters (Control P=0.005, P=0.000 respectively; Overweight: P=0.000 for both). From 34th week to delivery Hc levels were increased significantly in control group (P=0.000) and not significantly in GD group. In overweight and obese women no significant changes in Hc from 34th week to delivery were observed. sTfR showed significant changes from 34th week to delivery only in overweight women. Finally, TBI and sTfR/sFnt ratio showed a significant drop from 24th to 34th weeks, rising from 34th to delivery. This change in TBI levels from 34th to delivery was only significant for control and overweight groups. 4.3. Influence of maternal iron status on neonatal iron status There were significant differences between the mothers and their neonates in all the groups studied. In general, haematological and biochemical parameters were higher in neonates compare to their mothers. Differences between the four neonates group's studied were found in serum ferritin and MCV levels. Neonates born from control mothers had the highest value of ferritin and MCV compared to other groups, whereas neonates born from obese women had the lowest value. The same circumstance happened when it was compared between their mother's levels, suggesting that maternal condition and maternal iron stores may have effects on neonatal iron stores early after birth (differences between neonates groups and between their mothers were not significant). Significant differences were found in MCV levels between neonates born from control and GD women compare to neonates from obese. No statistical differences were found in serum iron, Hb, sTfR levels and others haematological parameters between the neonates groups. For other hand, accordingly to iron maternal iron status classification in each time-point assessed during pregnancy, significant differences were found related to neonatal sTfR at 24th week. 4.7. Relation between birth weight and preterm delivery and maternal iron status and consequences for the neonatal health. Accordingly to iron maternal iron status classification in each time-point assessed during pregnancy, no significant differences were found related to gestational age, longitude, upper arm and waist circumferences (data not shown), head circumference, placental/birth weight ratio, and placental weight and small diameter, whereas significant differences were found in birth weight at 24th and 34th weeks of pregnancy. Related to birth weight, differences were between neonates born from sufficient women and iron deficient without anemia at 24th weeks. The highest birth weight was found in those whose mothers were iron deficient without anemia at 24th weeks. At 34th weeks, differences in birth weight were between iron deficient with anemia and the others, showing the highest birth weight compare to iron sufficient and iron deficient women without anemia. Related to longer diameter no significant differences were found between the groups of iron status, but seems like iron deficiency with anemia showed higher longer diameter compare to others two situations. To determine the effect of maternal iron status on birth weight, multiple regression analysis was performed adjusting for gestational age, pre-pregnancy BMI, offspring gender's, placental weight and sTfR/Fnt index. Maternal iron status assessed by sTfR/Fnt index was shown as a significant factor associated with birth weight (OR 19.95, 95% CI 12.27 27.64). Birth weight = -2307.558 + 131.378 GA 221.619 female offspring + 0.683 placental weight + 19.955 sTfR/Ftn index + 4.884 Pre-pregnancy BMI. Pre-pregnancy BMI resulted non significant, but a positive correlation with birth weight was found. The model explained a 50% of the variability found of birth weight in this population. When count only with pre-pregnancy BMI in the model, this variable explained only 2.8% of the variability (OR: 15.65 CI: 4.44-26.85, P= 0.006). In contrast, sTfR/sFnt alone in the model explained 14% of the variability (OR: 22.58 CI:12.79-32.39, P=0.000). Birth weight was directly correlated to pre-pregnancy BMI (r=0.168, P=0.006, N=261) and to sTfR/sFnt index (r=0.374, P=0.000, N= 130). Placental TfR expression depending on maternal pre-pregnancy BMI and gestational diabetes condition and maternal iron status classification (Table 13.1-2) TfR1 is the main iron uptake protein in the placenta. Non statistical differences were shown in the protein expression of placental TfR1 depending on maternal pre-pregnancy BMI, nor depending on gestational diabetes. In contrast, significant differences were found in placental TfR expression and the ratio placental TfR/placental weight at 24th week of pregnancy among mother's iron status classification. Both parameters measured were higher in IDA women but not significantly different from others two groups, whereas significant differences were found between iron sufficient women and ID without anemia, with the highest values for the iron sufficient women (Table 13). Women with depleted total body iron stores (TBI<0) had greater placental expression of TfR (0.50±0.88, N=89) than those with total body iron stores greater than zero (0.39±0.79, N=36), but the difference was not significant (P=0.561). Newborn with cord serum ferritin concentrations <60¿g/L (N=3) had greater expression of placental TfR (0.67±0.73) compared to newborn with cord serum ferritin values >60¿g/L (0.59±0.50, N=10), but the difference was not significant (P=0.858). A multiple linear regression model with placental TfR as dependent variable and maternal Hb, Hc, RBC, ferritin, TBI, at delivery and neonatal iron satus indicators such as sTfR, and placental and newborn weight were included in the model to predict maternal and neonatal iron status regulation of placental TfR. The model included as predictors maternal Hb, neonatal sTfR and placental weight (R2=0.18, P=0.021). Placental weight acts as a confounder. 4.9. Genetic polymorphisms of mothers and neonates (Table 14.1) No significant differences between the genotype distribution of LEP and LEPR polymorphisms in the different groups were found. In a general note, GA and GG alleles in these two polymorphisms of the leptin (LEP) and leptin receptor (LEPR) genes in mothers and their offspring were predominant compare with AA alleles. 4.10. Genetic polymorphisms related to iron status, leptin levels and neonatal birth weight (Table 14.2) Our results revealed a significant difference between women from control group with the different maternal genotypes of LEP 19 GA polymorphism regarding to neonatal birth weight. The genotype GG showed the highest birth weght. compared to AA with the lowest birth weight. In contrast, in obese group, the neonatal genotype AA of the LEP 19 GA polymorphism showed significant higher birth weight compared to GG genotype. However, AA genotype of the LEP 25486 gene in the neonate had the lowest birth weight compared to GG. Moreover, the presence of homozygosis A/A for LEPR Q223R (GA) polymorphism in the neonate born from normal mothers was related to lower placental/birth weight ratio. This effect was also seen in overweight mothers homozygous for LEP G2548A polymorphisms which showed heaviest placentas than those with the ancestral genotype. Transferrin decreases in overweight mothers as the mutation of the LEP G2548A gene polymorphism worsens from G/G to homozygosis A/A at 24th weeks. Serum iron at 34th weeks of pregnancy resulted significantly lower in the obese mothers which had the G/A polymorphism of the LEP G2548A gene. Moreover, mothers' LEPR G/A and A/A alleles also determined a decrease in sTfR and sTfR/sFnt index at 34th weeks of pregnancy in the control group, being significant in the A/A homozygous forms. In overweight and obese women the transferrin at 34th weeks and the ferritin at delivery, increases as much as the LEPR polymorphism worsens. Differences in serum iron levels were found in overweight group between GA and AA genotype of the maternal LEP 19 gene, GA genotype with the highest serum iron levels, and AA with the lowest. Hemoglobin in control group was higher in AA and GG genotype of maternal LEP25486 gene and lower in GA. In a general note, it seems as a tendency of the GA genotype in maternal LEP25486 to iron deficiency. This tendency to iron deficiency is found in GG genotype of control of maternal LEPR polymorphism. 7. Conclusions 1. Obesity in pregnancy is related to lower study levels, unemployement and less oportunities for an estable job. Also has been related to adverse pregnancy outcomes such as previous aborptions, cesarean delivery, impaired glucose tolerance, gestational diabetes, and high blood pressure during pregnancy. 2. Serum iron has been shown as a biomarker of body fat mass and body mass index during pregnancy. 3. Obesity has been related to iron deficiency and poor iron stores, inflammation and a poor nutrition state with deficiency in vitamins and minerals (vitamin B12 and folic acid) related to iron metabolism. 4. Gestational diabetes has been correlated positively to pre-pregnancy body mass index and ferritin levels. The impact of high serum ferritin on the risk of gestational diabetes is, at least in part, mediated by obesity. A diet rich in iron and iron supplements use could be the reason of the high ferritin levels and an excess of iron can lead to insulin resistance. In contrast, iron-deficiency pregnant women had a reduced risk of gestational diabetes. 5. Obese pregnant women should be advised to mantein an adequate iron state to prevent iron defficiency and at the same time, avoiding iron overload that had been demonstrated to have serious consequences for risk of develop gestational diabetes during pregnancy, specially with the increase in body mass index. 6. Maternal iron deficiency adversely affects fetal iron status. Newborn from obese mother extract iron in amounts proportional to the levels available in the mother and are in disadvantage in term of iron status comparing to newborn from non-obese mothers. Gestational diabetes affects negatively fetal iron stores and its effects could be related to the degree of erytropoiesis but not maternal iron status. 7. Preobe study demonstrates that both maternal and neonatal iron status up-regulates the placental TfR expression, since neonatal and maternal iron parameters has been correlated to placental TfR protecting the newborn from iron deficiency with detriment for the mother. 8. The increased placental TfR1 expression found in placentas from iron deficiency anemia and iron sufficients mothers and by extension, obese and gestational diabetes mothers, could attempt to compensate the highest iron demands from neonates with increased requirements, due to iron deficiency situation and/or augmented erytropoiesis. 9. Iron deficiency anemia in pregnancy has been considered as a risk factor for pre-term delivery and subsequent low birth weight. However, in Preobe study did not find a lower birth weight in children of anemic mothers, actually, the results showed an association between iron defficiency during the second trimester of pregnancy and an increased birth weight and non association with preterm birth. It is possible that the demands on the mother are increased in the case of a newborn infant of greater size. 10. There is an association between LEP and LEPR polymorphism, maternal iron status, and parameters of neonatal growth. How these associations are inter-related to iron metabolism and their impact on the future neonatal health need further research. However, our preliminary results show a tendency to iron deficiency in those who are homozygous GG for maternal LEPR polymorphism.