Intended for healthcare professionals

Clinical Review State of the Art Review

Obesity and pregnancy: mechanisms of short term and long term adverse consequences for mother and child

BMJ 2017; 356 doi: https://doi.org/10.1136/bmj.j1 (Published 08 February 2017) Cite this as: BMJ 2017;356:j1
  1. Patrick M Catalano, professor and director1 2,
  2. Kartik Shankar, associate professor3 4
  1. 1Department of Obstetrics and Gynecology, Center for Reproductive Health/MetroHealth Medical Center, Cleveland, Ohio, USA
  2. 2Case Western Reserve University, Cleveland, Ohio, USA
  3. 3Arkansas Children’s Nutrition Center, University of Arkansas for Medical Sciences, Little Rock, Arkansas, USA
  4. 4Department of Pediatrics, University of Arkansas for Medical Sciences, Little Rock, Arkansas, USA
  1. Correspondence to: P M Catalano pcatalano{at}metrohealth.org
  • Accepted 5 December 2016

Abstract

Obesity is the most common medical condition in women of reproductive age. Obesity during pregnancy has short term and long term adverse consequences for both mother and child. Obesity causes problems with infertility, and in early gestation it causes spontaneous pregnancy loss and congenital anomalies. Metabolically, obese women have increased insulin resistance in early pregnancy, which becomes manifest clinically in late gestation as glucose intolerance and fetal overgrowth. At term, the risk of cesarean delivery and wound complications is increased. Postpartum, obese women have an increased risk of venous thromboembolism, depression, and difficulty with breast feeding. Because 50-60% of overweight or obese women gain more than recommended by Institute of Medicine gestational weight guidelines, postpartum weight retention increases future cardiometabolic risks and prepregnancy obesity in subsequent pregnancies. Neonates of obese women have increased body fat at birth, which increases the risk of childhood obesity. Although there is no unifying mechanism responsible for the adverse perinatal outcomes associated with maternal obesity, on the basis of the available data, increased prepregnancy maternal insulin resistance and accompanying hyperinsulinemia, inflammation, and oxidative stress seem to contribute to early placental and fetal dysfunction. We will review the pathophysiology underlying these data and try to shed light on the specific underlying mechanisms.

Introduction

Obesity is the most common problem in obstetrics that affects both the mother and her offspring.1 It causes short term and long term problems for the mother, such as increasing her risk of gestational diabetes (GDM) and pre-eclampsia.2 Because obese women are more likely to have excessive gestational weight gain (GWG),3 this further increases the risk of developing the metabolic syndrome in later life. The offspring have an increased risk of obstetric morbidity and mortality,4 and, consistent with the developmental origins of health and disease, a long term risk of childhood obesity and metabolic dysfunction.5

Options for the management of obesity in non-pregnant women include lifestyle interventions, drugs to achieve weight loss, and, because of sustained benefits, bariatric bypass surgery. During pregnancy anti-obesity drugs and bariatric surgery are not options. Unfortunately, a large number of randomized controlled trials (RCTs) have shown that lifestyle interventions have limited success in decreasing excessive GWG or improving maternal or neonatal short term perinatal outcomes.6

In this review we will address why medical and lifestyle interventions during pregnancy have failed to improve perinatal outcomes in obese women. We will focus on human data generated from epidemiologic, genetic, and molecular research over the past 30 years, since the onset of the obesity epidemic in developed countries. More than 10 700 PubMed citations (search terms obesity and pregnancy) have been generated since the 1990s. Although the quality of the research is generally good, the problem of obesity in pregnancy does not have its onset at conception or its resolution at delivery—it encompasses the entire life cycle.

Prevalence of obesity and overweight in pregnancy

Obesity is classified according to World Health Organization (WHO) criteria of body mass index (BMI=weight (kg)/height (m)2; table 1).7 BMI estimates of body fat in non-pregnant women explain 50-70% of the variance in fat mass. During pregnancy, the significant increase in total body water makes the correlation less robust.8 It is therefore not valid to use BMI to classify obesity during pregnancy. Because of the differences in body composition among various racial groups, WHO has discussed using different criteria for the classification of obesity in Asian women.9

Table 1

World Health Organization body mass index (BMI) categories7

View this table:

The prevalence of obesity began to increase in the last decades of the 20th century in developed countries. From 1999 to 2010, obesity increased from 28.4% to 34.0% in women of reproductive age (20-39 years) in the United States.10 According to the Centers for Disease Control and Prevention, the prevalence of obesity in women of reproductive age did not change significantly between 2003-2004 and 2011-2012.11 However, 7.5% of women in this age group have class III obesity or BMI greater than 40.1 Currently in the US, 31.9% of reproductive age women are obese and 55.8% are overweight or obese, with a higher prevalence in non-Hispanic black and Mexican American women (table 2).1

Table 2

Prevalence of obesity (measured by body mass index; BMI) by race or ethnicity in women of reproductive age1

View this table:

The global age standardized mean BMI in women increased from 22.1 (credible intervals 21.7 to 22.5) in 1975 to 24.4 (24.2 to 24.6) in 2014, with a doubling of the prevalence of obesity between 1980 and 2008.12 In 2000, WHO noted that obesity “is now so common that it is replacing the more traditional public health concerns, including undernutrition and infectious disease, as one of the most significant contributors to ill health.”7 In Europe, WHO estimates that more than 50% of men and women are overweight or obese and 23% of women are obese. In South East Asia, 14% of the population are overweight and 3% are obese.13 The prevalence of obesity in women is double that of men in Africa and South East Asia.14 The increase in obesity is not confined to adults—there have been increases in obesity in children13 and significant increases in birth weights among various populations.2 This review will discuss the potential mechanisms by which maternal obesity in pregnancy propagates the vicious cycle of obesity throughout the life course.

Sources and selection criteria

The references for this review were obtained from various sources including PubMed and ClinicalTrials.gov. We searched PubMed from 1990 to July 2016 using the search terms pregnancy, obesity, overweight, fertility, spontaneous abortion, congenital anomalies, stillbirth, placenta, maternal metabolism, gestational diabetes and hypertension, pre-eclampsia, gestational weight gain, macrosomia, neonatal body composition, lactation, diet composition, diet quality, lifestyle interventions, epigenetics, and microbiome. We prioritized RCTs, meta-analyses, longitudinal observational studies, and cohort studies. We included only full text English language peer reviewed publications. We reviewed Cochrane Reviews on preconception lifestyle for people with infertility, interventions for preventing excessive weight gain during pregnancy, exercise in pregnant women for prevention of GDM, and diet or exercise for weight reduction in women after childbirth. We used clinical guidelines from the American Congress of Obstetricians and Gynecologists (ACOG), the Royal College of Obstetricians and Gynaecologists (RCOG), the Royal Australian and New Zealand College of Obstetricians and Gynaecologists (RANZCOG), the Society of Obstetricians and Gynaecologists of Canada (SOGC) and the Institute of Medicine (IOM) for gestational weight guidelines in pregnancy. We prioritized the results of guidelines in this review for quality according to the method outlined by the US Preventive Services Task Force.

Effects of maternal metabolism on fertility and reproduction

Obesity and reproductive function

Obesity perturbs the hypothalamic-pituitary-ovarian axis and overweight women display shorter luteal phases and lower levels of follicle stimulating hormone, luteinizing hormone, and progesterone.15 A study of more than 45 000 assisted reproductive embryo transfers showed that a higher BMI correlated with a reduced likelihood of successful pregnancy when autologous oocytes were used, but not when oocytes from lean donors were used, suggesting a direct effect of obesity on the oocyte.16

Obesity is also associated with changes in ovarian granulosa cells and the follicular fluid surrounding the oocyte. Differences in follicular fluid insulin, triglycerides, free fatty acids, proinflammatory cytokines, oxidized low density lipoprotein, and fatty acid composition have been observed in obese women, suggesting that numerous mechanisms probably contribute to disruptions in oocyte development.171819 Data from women undergoing in vitro fertilization (IVF) from a donated oocyte have been used to isolate the role of the endometrium, but results have been conflicting. A meta-analysis of data from six centers found no difference in the rate of pregnancy, implantation, or miscarriage between obese versus non-obese women.20 A recent study of surrogate pregnancies also indicated that increasing BMI of the surrogate mother had no significant effect on IVF, embryo transfer, or neonatal outcomes.21 However, another study of 9587 women found that obesity in the recipient had negative effects on implantation, pregnancy success, and live births,22 suggesting that obesity significantly alters endometrial gene expression in women during the luteal phase.23 Thus, evidence suggests that both uterine and ovarian changes associated with obesity contribute to reproductive dysfunction.

Increased risk of spontaneous abortion and congenital anomalies

Overweight and obese women have an increased risk of spontaneous miscarriage. A meta-analysis reported that women with a BMI ≥25 had a higher risk of miscarriage (odds ratio 1.67, 95% confidence interval 1.25 to 2.25).24 A subgroup analysis suggested a higher risk of miscarriage after oocyte donation (1.52, 1.10 to 2.09) and induction of ovulation (5.11, 1.76 to 14.83). An observational cohort study of women with recurrent early pregnancy loss reported that obese women had a 58% risk of euploid miscarriage compared with 37% in non-obese women.25

Obese women are at increased risk of pregnancy affected by congenital anomalies. A systematic review and meta-analysis reported an increase in the following congenital anomalies in the offspring of obese compared with non-obese women: spina bifida (2.24, 1.86 to 2.69), neural tube defects (1.87, 1.62 to 2.15), limb reduction anomalies (1.34, 1.03 to 1.73), cardiovascular anomalies (1.30, 1.12 to 1.51), and cleft lip and palate (1.20, 1.03 to 1.40).26 A prospective observational cohort study reported a dose-response effect in relation to congenital heart defects with increasing BMI (overweight: 1.15, 1.01 to 1.32; obese: 1.26, 1.09 to 1.44; and morbidly obese: 1.34, 1.04 to 1.43).27 Although maternal folate status and glucose intolerance have been suggested, no definitive mechanism has been identified.28

Placental changes associated with maternal obesity

The placentas of obese women are significantly heavier at birth (mean 693 (standard deviation 184) g v 614 (152) g; P=0.002),29 and placental weight has a stronger correlation than maternal age, pre-pregnancy BMI, and GWG with neonatal birth weight and fat mass.30 The maternal metabolic environment affects early placental growth and gene expression, as well as later placental function, which become clinically manifest in late pregnancy.31 A potential mechanism is that insulin receptors are more abundant on the maternal trophoblast surface in early gestation than in later gestation.32 The hyperinsulinemia associated with obesity related insulin resistance results in differential responses in placental trophoblasts.31 One study reported a 30-fold decrease in insulin sensitive genes that regulate the cell cycle and cholesterol homeostasis in placental villous tissue from obese women during the first trimester. Maternal obesity, insulin resistance, and hyperinsulinemia together impaired global gene profiling related to mitochondrial dysfunction and decreased energy metabolism.31 For example, the maternal insulin secretory response in early pregnancy is strongly correlated with placental weight and neonatal adiposity at birth.33

The placentas of obese women at term are characterized by an increase in total lipid content and an accumulation of macrophages and proinflammatory mediators compared with normal weight women.3435 At a molecular level, the placental transcriptome of obese women with GDM shows activation of genes related to lipid metabolism (fig 1).36 Pregravid obesity is associated with a systemic low grade metabolic inflammatory state and subclinical endotoxemia, both of which may contribute to worsening of pregnancy associated insulin resistance.37 Maternal obesity is associated with greater placental cytokine expression and greater vessel muscularity but little difference in other histological features such as fibrin deposition or placental maturity.38 The precise mechanisms underlying the inflammatory responses and role in fetal growth, however, remain unclear.

Figure1

Fig 1 Differential expression of placental metabolic genes in lean women with type 1 diabetes and obese women with gestational diabetes (gene expression increases as the color changes from blue (downregulated) to red (upregulated). Unsupervised hierarchical clustering of metabolic genes that are modified in the placenta of obese women with gestational diabetes (GDM) and lean women with type 1 diabetes (type 1 DM). When the placental transcriptome at term was compared between lean women with type 1 DM and obese women with GDM, genes related to lipid metabolism were preferentially activated in obese women with GDM. Eleven genes related to lipid transport and activation and seven genes related to lipid metabolism were enhanced36

The placenta has been hypothesized to function as a “nutrient sensor,” whereby the placenta integrates maternal and fetal nutritional cues, and through regulation of nutrient sensing pathways attempts to match fetal demand and maternal supply.39 In support of this model, maternal obesity is associated with increased placental expression of system A amino acid transporters 1 and 2 (SNAT1 and SNAT2) in conjunction with greater activity of the mTOR (mechanistic target of rapamycin) and insulin growth factor 1 (IGF-1) signaling pathways, which presumably contribute to fetal overgrowth.40 By contrast, a recent report found that maternal obesity decreased placental taurine transporter activity in villous explants, without changes in protein expression.41 Maternal obesity and GDM also affect lipid transport across the placenta. Placental expression of fatty acid binding protein FABP4 and endothelial lipase is raised in obese women with diabetes.4243 Similarly, placentas from obese women show increased lipid accumulation but lower concentrations of FABP5 and uptake of omega-6 polyunsaturated fatty acid (oleic acid).344445 Both increases and decreases in the long chain polyunsaturated fatty acid (PUFA) transporter, CD36, have been reported in placenta from obese women.4546 An important caveat of human placental responses to obesity is that most information comes from studying late gestation placentas.

In contrast to increased inflammatory markers, concentrations of placenta and plasma estradiol and progesterone are lower in obese than in lean women.47 During pregnancy, estradiol and progesterone are synthesized in placental mitochondria from cholesterol precursors. Although there are no significant differences in circulating or placental cholesterol concentrations, mitochondrial cholesterol concentrations are 40% lower in the placentas of obese women, and this is related to decreases in the mitochondrial cholesterol translocator protein.47 The clinical significance of the decrease in placental steroid production is unknown but may be related to the adverse perinatal outcomes seen in obese women.

In summary, in early pregnancy the human placenta is responsive to the high concentrations of maternal insulin found in obese women. This results in altered gene expression in relation to mitochondrial steroid hormone production and energy metabolism. Furthermore, placental size in early pregnancy is strongly correlated with subsequent fetal adiposity at birth.33 The inflammatory milieu of late pregnancy may be an important factor affecting maternal insulin resistance.48

Differences in maternal metabolism in obese and normal weight women

Pregnancy is a unique metabolic condition because of the changes in maternal metabolism needed to provide for fetal growth and increased maternal energy requirements. Insulin sensitivity decreases by 40-50% during the course of pregnancy but significantly improves within days of delivery.4950 The following data come from a secondary analysis of the longitudinal changes in metabolism in a small number of lean (n=6) and obese (n=10) otherwise healthy women before a planned pregnancy and during early and late gestation.515253

Lean and fat mass increased significantly in both lean and obese women but the increase in fat mass was greater in lean women (fig 2). Although there was a significant 23% increase in resting energy expenditure, over time there was no significant difference between groups. Basal carbohydrate oxidation increased 68% over time and was greater in obese women.

Figure2

Fig 2 Longitudinal changes in body composition in lean and obese women before pregnancy through to late gestation. (A) Changes in lean body mass (kg; mean and standard deviation) in lean (n=5) and obese (n=6) women: prepregnancy, early pregnancy (12-14 weeks), and late pregnancy (34-36 weeks). Change over time, P=0.0001 and between groups, P=0.34. (B) Longitudinal changes over time (kg; mean and standard deviation) of fat mass in lean (n=5) and obese (n=6) women: prepregnancy, early pregnancy (12-14 weeks) and late pregnancy (34-36 weeks). Change over time, P=0.0001 and between groups, P=0.0253

There was a 40% decrease in insulin sensitivity in lean and obese women over time and a trend (P=0.07) for a greater decrease in obese women (fig 3). Because of decreases in insulin sensitivity over time fat oxidation increased 220% in both lean and obese women (P=0.003).

Figure3

Fig 3 Longitudinal changes in insulin sensitivity in lean and obese women before pregnancy through to late gestation. Longitudinal changes over time in insulin sensitivity (mean and standard deviation) as estimated by the hyperinsulinemic-euglycemic clamp (mg/kg.ffm/min) in lean (n=5) and obese (n=6) women: prepregnancy, early pregnancy (12-14 weeks) and late pregnancy (34-36 weeks). Change over time, P=0.0001 and between groups, P=0.0753

There was decreased suppression of lipolysis during insulin infusion (hyperinsulinemic-euglycemic clamp), but no significant difference between groups (P=0.30). Basal free fatty acids decreased 16% over time (P=0.02) but free fatty acids increased by 62% during insulin infusion with the clamp over the course of pregnancy (P=0.0004), although there was no significant difference between groups (fig 4). There was a 61% increase in fasting cholesterol and 260% increase in triglycerides (P=0.0001) over time, but no significant difference between lean and obese women. Others, however, have reported that in late pregnancy obese women have increases in circulating triglycerides and very low density lipoprotein cholesterol but a decrease in high density lipoprotein compared with lean women.54555657

Figure4

Fig 4 Longitudinal changes in free fatty acid concentrations in lean and obese women before pregnancy through to late gestation. (A) Longitudinal changes over time in basal free fatty acid concentration (mE/L; mean and standard deviation) in lean (n=5) and obese (n=6) women: prepregnancy, early pregnancy (12-14 weeks), and late pregnancy (34-36 weeks). Change over time, P=0.02 and between groups, P=0.30. (B) Longitudinal changes over time in free fatty acid concentration (mE/L; mean and standard deviation) during the hyperinsulinemic-euglycemic clamp in lean (n=5) and obese (n=6) women: prepregnancy, early pregnancy (12-14 weeks) and late pregnancy (34-36 weeks). Change over time, P=0.0004 and between groups, P=0.8253

A longitudinal study of amino acid concentrations in obese and normal weight pregnant women found a decrease in leucine, methionine, phenylalanine, tyrosine, and valine from the second to the third trimester. However, there were no changes in branched chain amino acids between lean and obese women.58 At term there was a decrease in amino acid concentrations and fetal-maternal amino acid gradients in obese women who had small for gestational age neonates compared with normal or obese women with average for gestational age neonates.59 The decreased fetal-maternal amino acid gradients seen in some obese women may be related to altered placental amino acid transporter activity41 and decreased placental SNAT activity.60

In summary, significant decreases in insulin sensitivity (increased insulin resistance) occur during pregnancy, with obese women showing greater decreases in insulin sensitivity. In addition to affecting glucose metabolism, decreases in insulin sensitivity result in increases in various lipids and amino acids. Hence, uniform metabolic changes seem to occur during pregnancy, and differences between lean and obese women may relate more to the preconception metabolic condition of the mother rather than to any specific effect of pregnancy.

Effects of maternal obesity on maternal and neonatal outcomes during the index pregnancy

Maternal implications

Maternal metabolism

Subclinical metabolic dysfunctions in obese women, such as GDM and pre-eclampsia, which are associated with adverse pregnancy outcomes, become clinically manifest later in gestation.6162 These metabolic disorders have a higher prevalence in certain racial or ethnic groups, including African-Americans and Southern Asians.63

Premature delivery and stillbirth

Obese women have an increased risk of indicated preterm deliveries because of the aforementioned antepartum complications, but they also have an increased risk of spontaneous preterm births.64 The pathophysiology of preterm delivery is not well characterized, although inflammation is associated with both maternal obesity and preterm labor.65 In a systematic review and meta-analyses, the risk of perinatal mortality was related to increasing maternal BMI. The relative risk for each five unit increase in maternal BMI in overweight and obese women was 1.21 for fetal death (1.09 to 1.35), 1.24 for stillbirth (1.18 to 1.30), 1.16 for perinatal death (1.00 to 1.35), 1.15 for neonatal death (1.07 to 1.23), and 1.18 for infant death (1.09 to 1.28).4

Intrapartum

Obese women have an increased risk of a failed trial of labor, cesarean delivery, and endometritis; they also have double the risk of a composite measure of maternal morbidity and a fivefold increased risk of neonatal injury.66 The length of labor in nulliparous women is inversely proportional to maternal BMI.67 The unadjusted odds ratios of cesarean delivery are 1.46 (1.34 to 1.60) and 2.05 (1.86 to 2.27) in overweight and obese women, respectively, compared with normal weight women.68 The success rate of a trial of labor after cesarean is inversely related to BMI (BMI <19.8 (83.1%), BMI 19.8-26 (79.9%), BMI 26.1-29 (69.3%), and BMI >29 (68.2%); P<0.001). Obese women are also more likely to undergo a repeat cesarean delivery before active labor.69 Similarly, gaining more than 18.14 kg (40 lb) is associated with a decreased success rate of a trial of labor after cesarean (66.8% v 79.1%; P<0.001).70

Anesthesia

Maternal obesity significantly increases the risk of anesthetic complications. The Centre for Maternal and Child Enquiries and RCOG recommend that women with a pregravid BMI >40 should have an antenatal consultation with an obstetric anesthesiologist.71 The risk of epidural failure is greater in obese women compared with normal weight and overweight women.72 Severely obese women have significantly greater hypotension and prolonged fetal heart rate decelerations, after controlling for epidural bolus dosing and hypertensive disorders, compared with normal weight women.73 The combination of spinal anesthesia and obesity greatly impairs respiratory function for up to two hours.74 General anesthesia also poses a risk for obese pregnant women because of potential difficulties with endotracheal intubation75 and the increased prevalence of obstructive sleep apnea.76

Cesarean delivery

Broad spectrum antimicrobial prophylaxis is recommended for all cesarean deliveries.77 However, there is no consensus regarding proper dosing on the basis of BMI. Compared with normal weight women, there is an increased risk of surgical site infections after cesarean delivery in women who are overweight (odds ratio 1.6, 1.2 to 2.2), obese class I (2.4, 1.7 to 3.4), and obese classes II and III (3.7, 2.6 to 5.2).78 Different types of skin preparation, skin closure techniques, and supplemental oxygen have not helped decrease the rate of post-cesarean infectious morbidity in overweight and obese women.798081 However, closure of subcutaneous tissue greater than 2 cm can significantly decrease the incidence of wound disruption.82

Postpartum venous thromboembolism

In the postpartum period, maternal obesity is a risk factor for venous thromboembolism and a higher risk of pulmonary embolism (adjusted odds ratio 14.9, 3.0 to 74.8).83 The Pregnancy and Thrombosis Working Group in the US recommends considering thromboprophylaxis in selected patients only, noting that there are insufficient data to recommend routine pharmacologic prophylaxis with cesarean delivery.8384 Because cesarean delivery increases the risk of venous thromboembolism, ACOG recommends placement of pneumatic compression devices for all patients before and after cesarean delivery.85 By contrast, the RCOG recommends considering prophylactic low molecular weight heparin for seven days after delivery in obese women who have one or more additional risk factors, such as smoking, and considering low molecular weight heparin early in pregnancy until six weeks postpartum for those with two or more risk factors.71

Breast feeding

There are multiple barriers to successful breast feeding in obese women. These include the physical problems of large size, increased risk of cesarean delivery, and the fact that neonates often need to be evaluated in special care nurseries. Obese postpartum women may also have a decrease in their first phase of milk production. Despite these obstacles, breast feeding should be encouraged, not only because of the potential neonatal benefits, but because of potential maternal benefits relating to postpartum weight reduction and decreased risk of diabetes in obese women who developed GDM.86

Mental health

Postpartum depression affects one in seven women. ACOG recommends that all pregnant women should be screened at least once for depression in the postpartum period using a validated screening tool, because depressive disorders can affect both maternal and neonatal wellbeing. When clinically significant depression is diagnosed, the obstetric care provider should start treatment or refer the patient as needed.87

Neonatal implications

Neonates born to obese women have an increased risk of overgrowth. Fetal overgrowth has been classified using various criteria. Macrosomia is defined as birth weight greater than 4000 g or 4500 g without consideration for gestational age. Large for gestational age (LGA) is defined as birth weight greater than the 90th centile for gestational age. However, many of the population based references fail to include adjustment for factors that can affect fetal growth such as sex or race and ethnicity. Although the ponderal index (weight/length9) has been used as a surrogate for BMI in neonates,88 our group and others have looked at measures of fetal growth that use estimates of body composition—fat-free or lean body mass and fat mass. Although fat mass constitutes only 12-14% of birth weight, it accounts for about 50% of the variance in term birth weight.89

Various techniques are available to estimate neonatal body composition including anthropometrics, densitometry, magnetic resonance imaging, and dual energy x ray absorptiometry. It is difficult to estimate the various compartments of adiposity (such as visceral adipose tissue) in neonates and measures of brown adipose tissue are available only as research tools. Infants of obese women are heavier at birth because they have greater amounts of adipose tissue than those born to non-obese women.9091 In our population in Cleveland, Ohio, US, maternal pregravid BMI had the strongest correlation with neonatal adiposity.92 This may relate to the large proportion of obese women in our population. By contrast, GWG greater than the IOM guidelines recommend was significantly related to increased neonatal adiposity in normal weight women.92 Other factors associated with fetal adiposity and cord blood markers of metabolic dysfunction include maternal free fatty acids, insulin, cholesterol, and gestational age.939495

Newborns of obese mothers have higher concentrations of umbilical cord leptin and interleukin 6 (IL-6) than infants of lean mothers. Furthermore, neonates of obese mothers have greater insulin resistance than those of lean mothers. Insulin resistance in obese infants significantly correlated with maternal insulin resistance and neonatal body fat.29 In both male and female neonates, placental weight had the strongest correlation with neonatal fat mass (r2=0.20-0.39). However, in male infants, maternal BMI and GWG were significant predictors of both lean mass and fat mass. By contrast, maternal plasma markers of inflammation (IL-6 and C reactive protein) were independently associated with female body fat and lean body mass.30

In summary, at term infants of obese mothers are significantly heavier than those of normal weight women because of an increase in body fat, with pregravid maternal BMI and placental size being the strongest correlates of infant weight in obese women. On the basis of variations in maternal pregravid insulin sensitivity and early placental response to maternal insulin, we propose a model of fetal overgrowth in obese women (fig 5). The decrease in insulin sensitivity in obese women leads to an increase in the insulin response, which affects early placental growth and gene expression. This results in the release of placental factors (cytokines, human placental lactogen, and others), which crosstalk with maternal insulin sensitive tissue (skeletal muscle, liver, and adipose tissue), decreasing maternal insulin sensitivity in these tissues. The result is an increase in nutrient availability, which contributes to the fetal adiposity that becomes manifest in late gestation.

Figure5

Fig 5 Maternal-placental crosstalk and fetal growth. Because of the reduced insulin sensitivity in obese compared with lean women, there is an increase in the insulin response in early pregnancy which affects early placental growth and gene expression. This then results in the release of placental factors that decrease insulin sensitivity in maternal tissue (skeletal muscle, liver, and adipose tissue), thereby resulting in increased nutrient availability for feto-placental growth. Increased availability of nutrients such as glucose and lipids thereby contributes to fetal adiposity, which becomes manifest only in late gestation

Long term effects of maternal obesity on the mother and offspring

Maternal weight gain, postpartum weight retention, and childhood adiposity

Compared with normal weight women, childbearing in obese women increases postpartum weight retention, because a greater proportion of obese women than normal weight women have excessive GWG and excessive GWG is related to postpartum weight retention.9697 GWG over the course of pregnancy has sharply increased over the past decades, along with the increase in the prevalence of obesity. Recent studies suggest that postpartum weight gain is most strongly associated with weight gain during the first trimester.9899 Substantial postpartum weight gain starts at six to 12 months after delivery. Findings from the Danish National Birth Cohort show that in addition to GWG, weight retention at six months and weight gain from six to 18 months postpartum contribute equally to adverse maternal weight status seven years after delivery.100 Common adiposity associated variants of genes such as FTO (fat mass and obesity associated protein) or MC4R (melanocortin 4 receptor) in either mother or child are not associated with GWG.101 Mathematical models of dynamic weight change have been developed that predict GWG on the basis of changes in energy intakes.102 Owing to the lack of high quality studies, an extensive systematic review of the effect of dietary composition on GWG concluded that although energy intake is associated with GWG, the contribution of specific macronutrients is unclear.103

Research into psychosocial aspects of behavior indicates that weight and healthy eating related self efficacy, or an individual’s perceived ability to undertake goal oriented behaviors despite potential barriers, was inversely associated with body weight in both early pregnancy and two years postpartum.104 Several RCTs have evaluated physical activity and dietary interventions to reduce GWG. Meta-analyses of these RCTs have reported conflicting evidence, with some studies finding significant beneficial effects of physical activity on reducing maternal GWG105 but others reporting minimal effects.106 A recent systematic review found that diet or exercise or both interventions reduced the risk of excessive GWG by 20% (average risk ratio 0.8, 0.73 to 0.87; n=7096 over 24 studies). Interestingly, dietary interventions such as low glycemic diets and supervised or unsupervised exercise by itself or combined with dietary changes led to similar reductions in the number of women with excessive GWG.107

Excessive GWG has also been proposed as a mediator of adverse health outcomes in a child’s long term health and risk of obesity. In population based cohorts, greater GWG has been associated with greater offspring weight in childhood,108 as well as higher BMI and systolic blood pressure in early adulthood.109 Greater weight gain during pregnancy is associated with higher BMI z-score, sum of subscapular and triceps skinfold thicknesses, and systolic blood pressure in childhood.110 Similar associations were seen between increased maternal BMI, greater GWG, and higher parity with greater waist circumference, higher BMI, and greater fat mass index in the offspring.111

Efforts to improve nutrition and physical activity during pregnancy for parents and children during and after delivery require not only a concerted effort on the part of the individuals but potentially considerable fiscal resources and commitments of time. Populations in greatest need are often the ones least able to afford the socioeconomic burden. Hence, obesity during pregnancy needs to be recognized as not only an individual problem but also one of the larger community and society.

Obesity and metabolic dysfunction in the offspring

Observational studies provide a strong link between maternal body weight status and obesity and cardiometabolic risk factors in the offspring. LGA at birth significantly increases the risk of obesity in adolescence and adulthood. In addition, children of obese mothers who are born LGA and who have childhood obesity have double the risk of developing insulin resistance.5 Meta-analyses report that high maternal BMI is unequivocally associated with LGA.112

These findings have been independently confirmed in the larger Hyperglycemia and Adverse Pregnancy Outcomes (HAPO) study, which showed that, independent of glycemia, maternal BMI is strongly associated with excess fetal growth and adiposity.113 An American cohort of 854 participants showed that parental obesity doubles the risk of adult obesity.114 Greater adiposity associated with maternal obesity persists into childhood and adolescence, as shown by the Avon Longitudinal Study of Parents and Children (ALSPAC) and Mater-University studies. Recently gestational fasting plasma glucose has been positively associated with birth weight. Gestational fasting plasma glucose and prepregnancy obesity interacted significantly with the association of offspring growth and overweight status in the first 36 months of life.115 Therefore, there is considerable evidence that maternal obesity during pregnancy increases adiposity at multiple life stages in the offspring—at birth29113 and during childhood, adolescence,116117 and adulthood.101 These data suggest that the effects of a short period of overnutrition during gestation can affect health throughout life.

Potential mechanisms linking maternal obesity to offspring susceptibility

Obesity is characterized by metabolic and endocrine derangements. Underlying these, lay a host of genetic, epigenetic, and lifestyle differences that directly or indirectly impinge on weight gain, adiposity, and metabolism. Specific themes have been examined as mechanisms that potentially link maternal obesity to offspring adiposity and metabolic health. Broadly, these studies point to a model whereby dietary changes or exposure to one or more obesity associated metabolites or hormones during critical stages of development can alter developmental pathways in the offspring and lead to greater accrual of fat mass.118119120121122

Studies in experimental models and limited human data suggest that the earliest stages of development (both preconception and periconception) are crucial.123124125 Although, developmental programming associated with obesity probably occurs throughout gestation and neonatal life, understanding the crucial periods during which the major changes are initiated is key to devising effective interventions. These persistent effects could be in part mediated through reprogramming of organ and tissue structure and could function through epigenetic mechanisms in utero. Furthermore, developmental trajectories are also likely to be modified in the postnatal period through signals in milk, changes in the microbiome conferred by the mother, complementary feeding choices, and the home environment, which further act on developmental changes (fig 6).

Figure6

Fig 6 Overview of potential mechanisms influencing the risk of obesity in the offspring of obese mothers. Pregravid obesity has important effects on reproductive function and the development of the oocyte. Persistent effects of maternal obesity can be mediated through resetting of epigenetic marks in early developmental stages, which may influence the commitment and renewal of stem cells. Maternal obesity and high energy diets also alter the growth of the offspring by altering nutrient transport through the placenta, perhaps as a result of altered placental development caused by placental inflammation and alterations in nutrient sensing pathways. Effects of maternal obesity and poor diets may also be perpetuated in the offspring through programming of neonatal growth, with alterations in milk composition and restructuring of the infant’s microbiome having pleiotropic effects on the offspring’s health

Dietary patterns and nutrient changes

Recent observational studies have examined the association between the composition and quality of the maternal diet and offspring outcomes (summarized in table 3).118126127128129130131132133 Overall, although poor quality diets seem to have detrimental effects in the offspring, they are unlikely to be mediators of changes associated with maternal obesity but probably represent an independent risk factor. Studies have also evaluated the association between type of dietary fat and child health outcomes. In the Project VIVA cohort, maternal omega-3 PUFA intake was associated with lower childhood adiposity at 3 years of age.134 However, in a recent report from the Southampton’s Women’s Survey, maternal plasma omega-6 PUFA levels at 34 weeks’ gestation were positively correlated with childhood fat mass at 4 (β=0.14 standardized β coefficients, P=0.01) and 6 (β=0.11 standardized β coefficients, P=0.04) years of age, but no associations between maternal omega-3 PUFA levels were seen.135 Variance in the nature of the cohorts, mean omega-3 PUFA and omega-6 PUFA levels, and age of assessment of children may contribute to the differences.

Table 3

Summary of studies examining maternal dietary patterns in pregnancy on offspring growth and obesity risk*

View this table:

The HAPO studies confirmed that there is a continuous association between maternal glucose levels in pregnancy and neonatal fat mass.113 Glucose concentrations in pregnancy also predict fat and lean mass of offspring in late childhood.136 Studies using continuous glucose monitoring report that despite controlled diets, obese women have a higher 24 hour glucose area under the curve than normal weight women.137 Intake of carbohydrates may be associated with greater fetal overgrowth and adiposity. In the Treatment of Obese Pregnant Women study, lower carbohydrate intake was associated with lower neonatal fat mass. Conversely, high glycemic load during pregnancy was associated with greater GWG, higher birth weight, and increased risk of LGA birth.138

Inflammation

Low grade subclinical systemic and tissue specific inflammation is a hallmark of obesity and is causally linked to insulin resistance. Several studies have examined the association between circulating maternal inflammatory markers and fetal, neonatal, and long term outcomes of adiposity in the offspring. The studies found positive associations36139140 and a lack of association141142 between these parameters, depending on the markers assessed, time of assessment during pregnancy, and sample size. In the Project Viva cohort, higher second trimester C reactive protein was associated with higher mid-childhood adiposity.140

The mechanisms through which inflammatory signals in pregnancy lead to increased adiposity in the offspring are not known. Recent studies indicate that the inflammatory state extends to the fetal side. A gene expression study of umbilical cord tissue in term infants from healthy lean and obese women identified a proinflammatory signature characterized by upregulation of acute phase response genes, including EGR1 and FOSB.143 However, the contribution of inflammation to developmental programming is mainly associative and definitive mechanistic studies are lacking.

Epigenetic alterations in offspring

A widely hypothesized overarching mechanism that links in utero exposure to health in later life is through epigenetic modification of offspring DNA or modification of regulators of chromatin structure, such as histones, during crucial periods of development.144 In the context of observational studies examining epigenetic mechanisms, DNA methylation is the most well characterized epigenetic process. Although the epigenetic effects of fetal undernutrition have been widely studied, fewer studies have focused on the effects of overnutrition or maternal obesity during pregnancy.

The Epigenetic Birth Cohort from Boston studied 319 newborns and found no correlation between maternal prepregnancy BMI and methylation of LINE-1 (long interspersed nuclear elements-1; a global measure of DNA methylation) in placenta or umbilical cord blood DNA.145 However, in another cohort maternal BMI positively correlated with DNA methylation of the PPARGC1A (peroxisome proliferator activated receptor γ co-activator 1α) promoter in umbilical cord DNA.146 Infants born to obese mothers with GDM have reduced methylation near the glucocorticoid receptor and the imprinted gene MEST (encoding mesoderm specific transcript).147 Cord blood DNA of neonates from obese mothers showed altered methylation of imprinted genes, MEG3 (maternally expressed 3) and PLAGL1.148 Associations between DNA methylation and risk of disease in later life are primarily associative. One study found that umbilical cord DNA methylation of the RXRA gene (retinoid X receptor α) was inversely correlated with maternal carbohydrate consumption in early pregnancy and positively correlated with the child’s adiposity at age 9.149 Furthermore, in a cohort of children born to obese mothers before or after bariatric surgery, distinct changes in the DNA methylation of glucoregulatory genes was seen in children born after bariatric surgery. These changes resulted in improved fasting insulin and homeostatic models of insulin resistance compared with siblings born before maternal bariatric surgery.150 Collectively, these studies suggest that maternal obesity can cause persistent loci and tissue specific epigenetic disruption in the offspring.

Another area of research is the description of nutritional and environmental influences on metastable epialleles (MEs)—loci where DNA methylation occurs stochastically early in embryogenesis and is maintained and stably propagated during differentiation.144151 This stochastic establishment of epigenotype leads to systemic variability in DNA methylation between individuals that is neither cell type specific nor genetically mediated. These MEs are sensitive to environmental influences, such as maternal nutrition in early pregnancy. In humans, candidate MEs have been identified by comparing genome scale methylation profiles of DNA between different tissue samples from the same individual and from monozygotic twin studies. Moreover, nutritional status early in pregnancy has been linked to persistent epigenetic changes in MEs in the offspring, as was elegantly shown in a study from rural Gambia where seasonal fluctuations in dietary intake occur.152 However, the influence of maternal obesity, GWG, and dietary patterns on these MEs remains to be evaluated.

Alterations in neonatal nutrition and infant microbiome

Long term associations between maternal obesity and infant adiposity may also be mediated by signals that alter the offspring’s physiology through early nutritional cues in breast milk. Breast milk from overweight and obese mothers has been reported to have higher levels of insulin and leptin, lower levels of omega-3 PUFA, and higher omega-6:omega-3 PUFA ratios,153 while showing no differences in macronutrient composition.154155156 Immunological modulators transforming growth factor β2 and sCD-14 have also been reported to be lower in breast milk from obese women.157 Several other hormones and cytokines present in milk have as yet uncharacterized biological effects on the gastrointestinal and immune systems of the offspring.

In addition to nutrients, human breast milk contains a variety of unconjugated glycans from the structurally diverse family known as human milk oligosaccharides (HMOs) which have specific structure related bioactivity. HMOs are natural prebiotics that serve as an energy source for a diverse array of colonic bacteria and hence shape and sustain the infant gut microbiome. Although studies have shown that the composition, diversity, and abundance of specific HMOs is associated with infant growth and adiposity,158 the effect of maternal obesity on the diversity of HMOs in milk is poorly understood.

The gut microbiota also plays a crucial role in regulating metabolism and adiposity, as shown by studies in germ-free mice and studies of human twins who are discordant for obesity. Obesity and diets high in fat are associated with a change in microbiome structure, most often an increase in Firmicutes and a decrease in Bacteroidetes.159

Pregnancy has been suggested to induce shifts in the gut microbiome, but reports are conflicting.65160 Importantly, the maternal microbiome is crucial in “seeding” the infant microbiome and the development of the infant’s immune system.161162 Few studies have evaluated the influence of obesity on the maternal microbiome. More recently, the microbiome of exclusively breastfed infants from lean and obese women was characterized at 2 weeks of age.156 These studies showed that both maternal obesity status and milk insulin concentrations were associated with early changes in the infant’s microbiome. The persistence of the changes in later life remains to be ascertained. However, it is clear that aspects of developmental programming associated with maternal obesity may be mediated through interactions between changes in the early diet (through alterations in breast milk) and reconfiguration of the microbiome.

Why have lifestyle interventions had limited success in decreasing perinatal morbidity?

Several RCTs have examined lifestyle interventions for obese women during pregnancy. The primary outcomes of these trials included avoiding excessive GWG and decreasing adverse perinatal outcomes, such as GDM and LGA neonates.163164165166167 Unfortunately, other than decreasing GWG by 1-2 kg, these studies have not improved perinatal outcomes. Recent RCTs (UPBEAT and the Norwegian Fit for Delivery Trial) suggest that behavioral lifestyle interventions do not prevent GDM or reduce the incidence of LGA babies despite increasing physical activity and reducing maternal dietary glycemic load, GWG, and maternal skinfold measures.6168 Investigators in the UK Pregnancies Better Eating and Activity Trial (UPBEAT) noted that increasing the intensity and duration of the lifestyle intervention would probably be impractical for most obese women.6 At least five meta-analyses concluded that, although lifestyle interventions started during pregnancy had some success in reducing excessive GWG, they had little effect on adverse pregnancy outcomes including fetal overgrowth.169170171172173

Data from our group and others suggest that improvement in maternal metabolic function should begin before conception to improve perinatal outcomes. We hypothesize that aspects of maternal metabolism often associated with obesity, such as increased insulin resistance and hyperinsulinemia, affect placental gene expression and function in early pregnancy before lifestyle interventions begin, although the results of these changes are not clinically manifest until late gestation. Ongoing preconception lifestyle intervention trials have a decrease in the incidence of GDM and neonatal adiposity as their primary outcomes.174

Although some individuals advocate weight loss in obese women during pregnancy, these recommendations are based on epidemiological data.175176177 In a perinatal database study in an affluent European population, investigators reported that weight loss in obese pregnant women was associated with reduced perinatal risk but not a higher risk of low birth weight or small for gestational age (SGA) neonates.178 By contrast, a recent meta-analysis of six cohort studies reported that although women who lost weight during pregnancy had a lower risk of having a LGA neonate, they also had a higher risk of having an SGA neonate below than the 10th and third centiles.179 Furthermore, our group reported an increased risk of SGA and decreased lean body mass (head circumference and length) in neonates of overweight or obese women with inadequate GWG.180 Therefore, unless new compelling data become available, overweight and obese women should comply with IOM guidelines on GWG.

Interventions to mitigate the effects of maternal obesity on neonatal health

In retrospective cohort studies, 10-20% of overweight or obese women lost weight between pregnancies.181182183184 Inter-pregnancy weight loss in overweight and obese women was associated with a decreased risk of GDM or pre-eclampsia in a subsequent pregnancy.178179180181182183184185 Three studies reported a decreased incidence of LGA in overweight and obese women who had an inter-pregnancy weight loss of as little as 1-2 BMI units.178182186 The inter-pregnancy decrease in weight was not associated with an increased risk of SGA infants.183 Four RCTs reported that a postpartum lifestyle intervention was effective in facilitating weight loss compared with a control group.187188189190 A recent meta-analysis of 11 studies with 769 women reported a significant mean −2.57 kg (−3.6 to −1.5) weight loss in the intervention group. In a subgroup analysis the most effective interventions were those with objective goals (use of heart rate monitors, pedometers, and exercise combined with intensive dietary intervention), which produced a −4.1 kg (−5.2 to −3.5) weight loss.191 Therefore, supervised intensive lifestyle intervention in overweight and obese postpartum women is feasible, efficacious, and safe even in lactating women. The idea that lifestyle intervention before pregnancy is necessary to improve placental function and fetal development is gaining traction as a viable paradigm to improve perinatal metabolic outcomes.174

Emerging treatments

Currently, 310 trials are listed on ClinicalTrials.gov under the heading of obesity and pregnancy. The studies primarily use lifestyle interventions to decrease infertility, excessive GWG, and pregnancy complications such as GDM and pre-eclampsia. Nutrient supplements such as vitamin D, probiotics, and omega-3 fatty acids are being tested to decrease the risk of GDM as well as allergic disease in infants. Ongoing studies are using drugs such as orlistat to improve reproductive fitness in obese women and metformin to decrease excessive weight gain and the risk of fetal macrosomia. Lastly, there are proposals that discuss the potential of prepregnancy lifestyle interventions to decrease the risk of gestational diabetes and fetal adiposity.

Guidelines

As noted previously, current guidelines for the management of obesity in pregnancy include those from the ACOG, RCOG, SOGC, RANZGOG, and IOM. All recommend that obese women try to achieve a healthy weight (based on BMI) before pregnancy. Obese women should receive counseling about GWG based on good nutrition, food choices, and physical activity. The management of pregnancy complicated by obesity should recognize the increased risk of medical and obstetric complications as outlined here. All of the guidelines, however, acknowledge that the recommendations for management are largely based on less than level A data (data based on good and consistent scientific evidence). The most recent ACOG guidelines were published in December 2015.192 The following recommendations were considered level A:

  • BMI calculated at the first antenatal visit should be used to provide diet and exercise counseling guided by IOM recommendations for GWG during pregnancy

  • Subcutaneous drains increase the risk of postpartum cesarean complications and should not be used routinely

  • Behavioral interventions using diet and exercise can improve postpartum weight reduction in contrast to exercise alone.

The RCOG guidelines were published in 2010 and the only level A evidence was that women undergoing cesarean section who have more than 2 cm of subcutaneous fat should have suturing of the subcutaneous tissue space to reduce the risk of wound infection and separation.71

Conclusions

The problems associated with obesity during pregnancy can be improved only with a comprehensive approach. Behavioral lifestyle interventions of improved nutrition and increased physical activity during pregnancy have helped modestly to reduce excessive GWG in overweight and obese women but have not decreased the risks of maternal metabolic dysfunction and fetal overgrowth. The perturbed metabolic environment of obese women exerts its effect on reproductive function before conception.

To date, interventions aimed at improving perinatal outcomes in obese women have been started after the first trimester, well after the feto-placental unit has been exposed to the adverse metabolic environment. Preliminary data have reported changes in maternal and placental gene expression and function before any phenotypic changes. Hence, just as optimal glucose control before conception is necessary to decrease the risk of congenital anomalies in the offspring of women with pre-existing diabetes, the metabolic environment of obese women—whether inflammation, insulin resistance, lipotoxicity, or hyperinsulinemia—requires improvements to reduce the adverse obesogenic mediated effects on pregnancy. Importantly, improvement does not necessarily imply a return to the non-obese metabolic condition. A 5% improvement in weight has been reported to be sufficient to exert positive metabolic effects in non-pregnant obese people.193 Although ideally these measures should be initiated before a planned pregnancy, it is still worth limiting excessive GWG and trying to improve metabolic conditioning postpartum.

Glossary of abbreviations

  • ACOG=American Congress of Obstetricians and Gynecologists

  • BMI=body mass index

  • FABP=fatty acid binding protein

  • GDM=gestational diabetes

  • GWG=gestational weight gain

  • HAPO=Hyperglycemia and Adverse Pregnancy Outcomes

  • HMO=human milk oligosaccharide

  • IL-6=interleukin 6

  • IOM=Institute of Medicine

  • IVF=in vitro fertilization

  • LGA=large for gestational age

  • ME=metastable epiallele

  • PUFA=polyunsaturated fatty acid

  • RANZCOG=Royal Australian and New Zealand College of Obstetricians and Gynaecologists

  • RCOG=Royal College of Obstetricians and Gynaecologists

  • RCT=randomized controlled trial

  • SGA=small for gestational age

  • SNAT=system A amino acid transporters

  • SOGC=Society of Obstetricians and Gynaecologists of Canada

Further research is needed to improve our understanding of how obesity in pregnancy is related to short term and long term adverse outcomes in the offspring. Meanwhile, only through the concerted and sustained efforts of healthcare providers and the community, as well as public health endeavors, will we begin to improve the long term health of obese women and their children.

Questions for future research

  • What is the paternal epigenetic contribution to metabolic dysfunction in the offspring?

  • Is there a differential metabolic response of fetal sex to the environment of obese pregnancy?

  • At term, do umbilical cord stem cells reflect the crosstalk between the maternal metabolic environment and the fetus in relation to developmental programming?

  • What are the mechanisms in the preconception and periconception window that affect oocyte development and have long term effects on developmental programming?

Footnotes

  • The authors are supported in part by the following grants: Eunice Kennedy Shriver National Institute of Child Health and development HD-11089-19 (PMC) and USDA Agricultural Research Service CRIS 6251-51000-010-05S (KS).

  • Contributors: PMC and KS both performed the literature search, wrote the draft article, and revised the manuscript. They are both guarantors.

  • Competing interests: The authors have read and understood BMJ policy on declaration of interests and have no conflicts to declare.

  • Provenance and peer review: Commissioned; externally peer reviewed.

References

View Abstract