A role for the early pregnancy maternal milieu in the intergenerational transmission of obesity

Abstract

Objective:

Maternal obesity increases the risks for adverse pregnancy and offspring outcomes but with large heterogeneity. This study examined changes to the maternal metabolic milieu across pregnancy in women with obesity. It identified differences between a metabolically unhealthy obesity (MUO) phenotype and a metabolically healthy obesity (MHO) phenotype, as well as the differences in offspring adiposity between the two metabolic phenotypes.

Methods:

In early pregnancy, women were classified with MHO (n = 13) or MUO (n = 9) based on the presence of zero or ≥2 risk factors for metabolic syndrome, respectively (systolic blood pressure > 130 mm Hg or diastolic blood pressure > 85 mm Hg, HDL cholesterol < 50 mg/dL, LDL cholesterol ≥ 100 mg/dL, triglycerides ≥ 150 mg/dL, and glucose ≥ 100 mg/dL). Area under the pregnancy concentration curve for glucose and triglycerides measured at early (13-16 weeks), mid- (24-27 weeks), and late (35-37 weeks) pregnancy, gestational weight gain (GWG), energy expenditure, maternal fat accretion, and infant body composition were compared.

Results:

Maternal BMI, GWG, and fat accretion did not differ between MUO and MHO. Women with MUO had a greater area under the pregnancy concentration curve for glucose (+2,170 [382] mg/dL·day, p < 0.001) and triglycerides (+12,211 [3,916] mg/dL·day, p < 0.001). There were no differences in late-pregnancy total daily energy expenditure, but activity energy expenditure was significantly lower in MUO (−403 [144] kcal). MUO offspring had greater weight (+621 [205] g, p = 0.01) and adiposity (+5.8% [2.1%], p = 0.02) at 1 week of life but showed no differences in fat-free mass.

Conclusions:

Independent of GWG, MUO resulted in heightened exposure of fetal fat-promoting substrates. Differing metabolic phenotypes may explain heterogeneity of offspring adiposity born to women with obesity.

INTRODUCTION

Maternal obesity coupled with increased gestational weight gain (GWG) poses a heightened risk for adverse pregnancy and infant outcomes (1). Infants born to mothers with obesity are larger and have higher adiposity, which are both risk factors for the development of obesity throughout the life-span (2). Prior research has considered obesity in women as a stand-alone risk factor for pregnancy complications (3); however, the metabolic expression of obesity varies greatly. Obesity without a metabolic comorbidity is a phenotype defined as metabolically healthy obesity (MHO) (4). In contrast, obesity alongside hypertension, hyperlipidemia, and hyperglycemia is defined as metabolically unhealthy obesity (MUO).

Pregnant women progressively develop insulin resistance, glucose intolerance, and hyperlipidemia. These maternal metabolic perturbations are a mechanism to shunt growth-promoting substrates to the developing fetus. Pregnant women with obesity and preexisting metabolic derangements such as the case of pregnancy with MUO may not be able to adequately adapt to the metabolic demand of pregnancy (5), further exacerbating pregravid hyperglycemia and hyperlipidemia. Differences in maternal obesity metabolic phenotypes have not been explored, to our knowledge, as a mechanism for heterogeneity in offspring adiposity.

The aim of this analysis was as follows: 1) to examine changes in the maternal metabolic milieu and GWG across pregnancy in women with two different obesity phenotypes (MUO compared with MHO); and 2) to identify whether differences in offspring adiposity are present between women with MUO compared with MHO. We hypothesized that women classified with MUO during early pregnancy would be characterized by sustained elevations in glucose and lipids, resulting in increased fetal size and adiposity at birth.

METHODS

Study design

This is an a priori planned secondary analysis of a prospective observational study designed to assess determinants of GWG in pregnant women with obesity (6,7). Maternal assessments were performed between 13 and 16 weeks’ (“early”), between 24 and 27 weeks’ (“mid”), and between 35 and 37 weeks’ (“late”) gestation, and body composition of offspring was measured. The study was approved by Pennington Biomedical Institutional Review Board and participants provided written informed consent prior to participating.

Participants

Pregnant women aged 18 to 40 years with obesity (BMI ≥ 30 kg/m²) measured at the screening visit (gestational age <15 weeks) were eligible. Medical history was obtained from prenatal records, and women with preexisting metabolic or cardiovascular disease were excluded. Exclusion criteria included hypertension (systolic blood pressure > 160 mm Hg, diastolic blood pressure > 90 mm Hg) or diabetes (hemoglobin A_1c > 6.5%) at screening and medications that may affect body weight or energy intake.

Maternal anthropometrics

Maternal weight, height, and body composition were measured after an overnight fast. GWG was expressed in grams per week between study visits. Body fat percentage was measured using a three-compartment model (fat mass, fat-free mass, and total body water) from body weight, body volume by plethysmography (BODPOD, COSMED, Concord, California), and body water (mean estimate of using zero-intercepts of hydrogen-2 and oxygen-18 isotopes) (8) during early and late-pregnancy outcome assessments. Body fat mass was then calculated using body fat percentage and the measured body weight. Visceral adipose tissue (VAT) was obtained through multisection magnetic resonance imaging (MRI).

Cardiometabolic biomarkers

Blood pressure was measured twice following a 5-minute rest, and an overnight fasted blood sample was drawn to measure insulin (Immulite 2000, Siemens, Munich, Germany), glucose, triglycerides, low-density lipoprotein (LDL) cholesterol (Beckman DXC600, Beckman Coulter Inc., Brea, California), and high-density lipoprotein (HDL) cholesterol (Trinity DXC600, Trinity Biotech pic, Wicklow, Ireland).

Energy expenditure and intake

As detailed previously (7), free-living energy expenditure (total daily energy expenditure, TDEE) was measured via doubly labeled water, and resting metabolic rate was measured via indirect calorimetry. Activity energy expenditure (AEE) was calculated as follows: AEE = TDEE – resting metabolic rate – (0.10 × TDEE), the latter accounting for the thermic effect of food. Energy intake was calculated using the intake-balance method (7). Participants captured their energy intake in real time, using remote food photography (9), and diet was analyzed for macronutrient content (percentage of fat, protein, and carbohydrates).

Substrate oxidation

Energy metabolism and the respiratory quotient (RQ) were measured in a metabolic chamber overnight during early and late pregnancy. After participants ate a standard dinner at 1900 (30% of energy requirements, consisting of 30% fat, 55% carbohydrates, and 15% protein), lights were turned off between 1130 and 0600 the next morning. Sleep RQ was the mean ratio of volume of carbon dioxide and volume of oxygen between 0200 and 0500.

Classification of metabolic health phenotypes

In early pregnancy, participants were phenotyped as MHO and MUO according to the National Cholesterol Education Program Adult Treatment Panel III guidelines for metabolic risk (10). Metabolic risk factors included the following: 1) systolic blood pressure > 130 mm Hg or diastolic blood pressure > 85 mm Hg; 2) HDL cholesterol < 50 mg/dL; 3) LDL cholesterol ≥ 100 mg/dL; 4) triglycerides ≥ 150 mg/dL; and 5) glucose ≥ 100 mg/dL. MUO was defined as obesity plus two or more risk factors. In order to create two dichotomous groups, women with only one risk factor (n = 29) were excluded from the analysis. Waist circumference above 88 cm is typically considered an additional risk factor for MUO; however, given the inherent difficulties of waist circumference measurements in pregnant women, we obtained VAT through MRI. VAT was not used as a criterion for metabolic risk.

Offspring anthropometrics

Offspring body weight (average 8 [4] d) was measured with the infant nude to the nearest 5 g on a calibrated scale (SCALE-TRONIC, White Plains, New York). Length was measured twice, with the infant’s head in the Frankfort position using an infantometer with a stationary headboard and a moveable footboard. Infant fat mass was assessed using air displacement plethysmography (PEA POD, COSMED).

Statistical analysis

All women with MHO and MUO were included in this analysis regardless of their development of gestational diabetes or preeclampsia. In the present study, one woman with MHO and two women with MUO developed gestational diabetes (data not shown) and they were retained within the analysis. Analyses were conducted using SPSS Statistics version 25 (IBM Corp., Armonk, New York). Data are reported as mean (SE), and α ≤ 0.05 was the predetermined level of significance. Cumulative exposure of glucose and triglycerides was calculated as area under the pregnancy concentration curve (AUC) using gestational age (days) as the time variable during early, mid-, and late pregnancy. Independent samples t tests were used to test for differences in pregnancy and offspring variables between MHO and MUO.

RESULTS

Participants

Of the 51 women included in the analysis, 13 (26%) had no metabolic comorbidities (MHO), 29 had one (excluded from analysis), and 9 (18%) had two or more (MUO). Women with MUO had two (n = 8) or three (n = 1) cardiometabolic disease risk factors. At early pregnancy, women with MUO had significantly higher glucose and triglycerides and they tended to have higher total cholesterol, LDL cholesterol, and VAT and lower HDL cholesterol. The two groups were otherwise similar and they did not differ with respect to age, BMI, or gestational age during early pregnancy (Table 1). There were no differences in the number of women with obesity class III (BMI > 40) between MHO (n = 2) and MUO (n = 2).

TABLE 1.

Maternal characteristics in early pregnancy^*

		Metabolic health status
	All (N = 22)	MHO (n = 13)	MUO (n = 9)	p value
BMI (kg/m2)	35.2 ± 1.0	34.6 ± 1.3	36.0 ± 1.7	0.51†
Body weight (kg)	94.1 ± 3.1	92.8 ± 4.4	96.0 ± 4.6	0.63†
Height (cm)	163.8 ± 1.9	163.8 ± 2.2	163.9 ± 3.5	0.95†
Age (y)	28.3 ± 1.0	27.6 ± 1.4	29.3 ± 1.4	0.42†
Race (n, %)
White	14 (63.6%)	8 (61.5%)	6 (66.7%)	0.89#
Black or African American	6 (27.3%)	4 (30.8%)	2 (22.2%)
Other	2 (9.1%)	1 (7.7%)	1 (11.1%)
Gestational age (wk, d)	14,5 ± 0,1	14,5 ± 0,1	14,5 ± 0,1	0.83†
Parity (n, %)				0.64#
0	10 (45.5%)	6 (46.2%)	4 (44.4%)
≥1	12 (54.5%)	7 (57.4%)	5 (55.6%)
SBP (mm Hg)	102 ± 2	101 ± 2	102 ± 3	0.85†
DBP (mm Hg)	60 ± 1	60 ± 2	62 ± 2	0.50†
Glucose (mg/dL)	94.5 ± 1.8	87.7 ± 1.3	96.9 ± 3.3	0.03†
Triglycerides (mg/dL)	115.4 ± 11.1	90.6 ± 7.7	151.2 ± 19.7	0.004†
Total cholesterol (mg/dL)	174.2 ± 7.1	163.8 ± 3.7	189.2 ± 15.8	0.08†
LDL (mg/dL)	90.5 ± 5.2	82.3 ± 3.0	102.3 ± 11.0	0.11†
HDL (mg/dL)	60.6 ± 1.9	63.3 ± 1.5	56.7 ± 3.9	0.09†
Fat mass (kg)	42.4 ± 2.0	41.6 ± 2.9	43.6 ± 2.7	0.62†
Fat mass percentage (%)	44.8 ± 0.8	44.4 ± 1.0	45.4 ± 1.4	0.56†
VAT (kg)	0.9 ± 0.1	0.7 ± 0.1	1.1 ± 0.2	0.07†
Sleep RQ (AU)	0.86 ± 0.01	0.84 ± 0.01	0.89 ± 0.01	0.005†

Abbreviations: AU, arbitrary units; DBP, diastolic blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein; MHO, metabolically healthy obesity; MUO, metabolically unhealthy obesity; SBP, systolic blood pressure; RQ, respiratory quotient; VAT, visceral adipose tissue.

^#χ² analysis.

^†Independent samples t test.

^*Early pregnancy assessments occurred between 13 and 16 weeks’ gestation. Data are presented as mean ± SEM unless otherwise noted.

The maternal metabolic milieu across pregnancy

GWG (MUO: −111.3 [79.1] g/wk vs. MHO: 111.7 [125.8] g/wk, p = 0.15) and fat mass gain (MUO: 0.3 [1.4] kg vs. MHO: 2.4 [0.7] kg, p = 0.20) were not different between MUO and MHO. At the end of pregnancy, fat mass remained similar between groups (MUO: 43.9 [2.3] kg vs. MHO: 43.9 [2.5] kg, p = 0.99). At the end of pregnancy, women with MUO had significantly higher fasting glucose (MUO: 95.6 [3.1] mg/dL vs. MHO: 82.8 [1.8] mg/dL, p = 0.001) and tended to have higher triglycerides (MUO: 238.8 [23.8] mg/dL vs. MHO: 182.1 [18.3] mg/dL, p = 0.07). In women with MUO, AUC from ~13 to ~37 weeks of pregnancy was significantly greater for glucose (MUO: 14,259 [359] vs. MHO: 12,089 [200] mg/dL·day, p < 0.001) and triglycerides (MUO: 32,074 [3,962] vs. MHO: 19,863 [1,792] mg/dL·day, p = 0.005; Figure 1).

FIGURE 1.

Glucose (A) from early to late pregnancy and (B) AUC. Triglycerides (C) from early to late pregnancy and (D) AUC. Error bars represent standard error (SEM). *Denotes significant differences in AUC between MHO and MUO. AUC, area under the pregnancy curve; MHO, metabolically healthy obesity; MUO, metabolically unhealthy obesity

Maternal energy expenditure and intake

Between MUO and MHO, no differences were observed in energy expenditure during early (MUO: 2,520 [155] kcal/d vs. MHO: 2,589 [75] kcal/d) and late (MUO: 3,300 [64] kcal/d vs. MHO: 3,448 [40] kcal/d) pregnancy, energy intake across pregnancy (MHO: 3,015 [73] kcal/d vs. MUO: 2,756 [482] kcal/d, p = 0.13), and the percentage of energy from protein, carbohydrate, and fat intake. AEE was lower in MUO compared with MHO in late pregnancy (MUO: 400 [129] kcal vs. MHO: 802 [74] kcal, p = 0.02).

Substrate oxidation

Despite clear differences in substrate oxidation (sleep RQ) in early pregnancy, which demonstrates impaired fat oxidation in women with MUO, there were no differences in substrate oxidation at the end of pregnancy between women with MUO and MHO (MUO: 0.86 [0.02] vs. MHO: 0.86 [0.01], p = 0.69).

Offspring anthropometrics

Gestational age at delivery and infant age at assessment were similar between infants born to mothers with MUO and MHO. Infants born to mothers with MUO (n = 7) weighed more (+620.8 [204.6] g, p = 0.01) and had more fat mass (+0.27 [0.09] kg, p = 0.01) and percentage of fat (+5.8% [2.1%], p = 0.02) compared with infants born to mothers with MHO (n = 7). There was no significant difference in the amount of fat-free mass (Table 2).

TABLE 2.

Infant anthropometrics

	Maternal health status in early pregnancy
	MHO	MUO	p value
Infant sex			0.45#
Male	5	3
Female	2	4
Gestational age at delivery (wk, d)	39, 3 ± 0, 2	39, 5 ± 0, 2	0.62†
Infant age at assessment (d)	6.7 ± 0.4	9.7 ± 2.3	0.24†
Infant weight (g)	3,095.6 ± 77.9	3,716.4 ± 220.7	0.01†
Infant fat-free mass (kg)	2.87 ± 0.08	3.17 ± 0.16	0.12†
Infant fat mass (kg)	0.28 ± 0.05	0.55 ± 0.07	0.01†
Infant fat mass (%)	8.9 ± 1.7	14.6 ± 1.3	0.02†

Data are presented as mean ± SEM.

Abbreviations: MHO, metabolically healthy obesity; MUO, metabolically unhealthy obesity.

^†Independent samples t test.

^#χ² analysis.

DISCUSSION

This is the first study, to our knowledge, to report that, independent of maternal obesity, the maternal metabolic milieu in early pregnancy may contribute to infant adiposity. Despite comparable GWG and maternal fat mass gain, infants born to mothers with an MUO phenotype (i.e., obesity and at least 2 risk factors for metabolic syndrome) had greater fat mass at approximately 1 week of age.

We hypothesized that excess maternal substrates during pregnancy have three metabolic fates. Metabolic fates include oxidation by the maternal-fetal unit, increased energy storage by the mother, and increased energy storage by the developing fetus. Differences in substrate oxidation in early pregnancy demonstrate clear metabolic impairments in fat oxidation during early pregnancy in women with an MUO phenotype. By the end of pregnancy, these differences are no longer evident, indicating similar maternal fat oxidation between phenotypes. Maternal fat accrual and GWG were similar between groups, elucidating that differences in early pregnancy metabolites did not influence energy deposition in the mother. Instead, we hypothesized that a prolonged elevation of maternal substrates, indicated by the increased glucose and triglyceride AUC across pregnancy, manifested in downstream implications, which likely increased fetal fat accrual. Natural progressive insulin resistance with pregnancy promotes higher fasting and postprandial glucose and increased lipolysis, resulting in greater glucose, triglyceride, and free-fatty acid availability for the developing fetus (11). Thus, excess glucose and lipids in early pregnancy, such as in the case of MUO, likely result in an overflow of growth-promoting substrates to the placenta throughout gestation, accelerating fetal fat accretion.

In the clinical setting, all pregnancies in women with obesity are managed equally with the prescription of adherence to GWG guidelines set forth by the National Academy of Medicine. A meta-analysis of clinical intervention trials including 10,291 women has shown that, despite lower rates of GWG (0.3-2.4 kg), there is no beneficial effect on infant size at birth (12). Lack of effects could be attributed to heterogeneity in infant outcomes, including offspring size and adiposity at birth. This may stem from differing intrauterine exposures resulting from the maternal metabolic profile starting early in pregnancy. For example, among women with obesity, Boyle et al. have demonstrated distinct metabolic phenotypes in infant umbilical cord-derived mesenchymal stem cells (13). Stem cells that displayed impaired fatty acid metabolism corresponded to greater concentrations of maternal substrates and increased infant adiposity, highlighting that maternal obesity metabolic phenotypes may differentially influence downstream fetal metabolism and growth. The differentiation of MUO from MHO has been previously used to define health status in nonpregnant populations. MUO presents in 6% to 60% of obesity, depending on criteria used for classification (4), and, generally, the presence of MUO increases the risk for cardiometabolic complications, including type 2 diabetes, cardiovascular disease, and all-cause mortality (4). In the present study, one woman with MHO and two women with MUO developed gestational diabetes. A sensitivity analysis showed that the difference in infant adiposity between groups remained evident when women with gestational diabetes were excluded (data not shown). Overall, our data suggest that obesity coupled with at least two metabolic comorbidities might also have differing impacts on infant adiposity and, therefore, requires a more specialized therapeutic approach throughout prenatal care. Interventions that alter the maternal metabolic milieu, not just GWG, may be necessary to improve downstream offspring outcomes, particularly in women with MUO. In the present study, AEE was 50% higher in MHO, exposing the avenue of prenatal exercise interventions to reduce excess circulating glucose and lipids. Exercise is a well-known and powerful modulator of metabolic risk, even in the absence of weight change (14).

In the present study, we identified metabolic health risk in early pregnancy. Ideally, metabolic health risk would be identified during the preconception life stage. The latest Task Force Recommendations acknowledge that clinicians and prenatal care providers may lack the time necessary to administer all components of prenatal behavioral interventions to all patients (15). Thus, screening for individuals with an unhealthy metabolic phenotype prior to conception or early in pregnancy would allow for precise preventive strategies to be deployed to those at highest risk for transmitting obesity to their offspring.

We acknowledge the small sample study in the present analysis, which likely limits observed power. However, the present study is strengthened by the conservative delineation of metabolic phenotypes in women with obesity during early pregnancy and the rigorous measurement of potential confounding variables, including maternal energy intake and expenditure. Our gold standard measures allow us to exclude influence of these variables on infant body composition.

We are the first, to our knowledge, to show that maternal obesity coupled with risk factors for cardiometabolic disease likely results in prolonged fetal exposure to excess growth-promoting substrates. Future studies should examine the influence of the preexisting maternal metabolic milieu on adverse maternal and infant outcomes on a large scale. This study identifies a population of women with obesity who are highly vulnerable to adverse offspring outcomes and highlights the importance for prenatal or preconception interventions that alter the metabolic milieu in this population, who are most at need. It is possible that different obesity phenotypes may also need to be considered when evaluating prenatal intervention effects on offspring outcomes. Evaluating metabolic health in conjunction with BMI screening at the start of pregnancy may be clinically relevant to understanding the intergenerational transmission of obesity.

Study Importance.

What is already known?

• The intergenerational transmission of obesity begins in utero. Excess gestational weight gain (GWG) increases the risk of adverse outcomes; however, prenatal interventions that reduce GWG have no effect on infant size at birth.

• Offspring born to women with obesity have greater risk; however, the metabolic expression of obesity varies greatly.

• Fetal growth and offspring size are driven by maternal substrates (e.g., glucose, lipids).

What does this study add?

• Despite similar GWG and fat mass accrual, women with obesity and additional metabolic comorbidities (metabolically unhealthy obesity) have heightened levels of glucose and triglycerides throughout pregnancy compared with women with obesity and no additional metabolic comorbidities (metabolically healthy obesity).

• Offspring of women with metabolically unhealthy obesity weigh more and have higher levels of adiposity compared with offspring of women with metabolically healthy obesity.

How might these results change the direction of research or the focus of clinical practice?

• Metabolic evaluation needs to occur in early pregnancy in women with obesity to identify additional comorbidities.

• Precision approaches to prevent obesity risk in offspring are needed and should consider maternal obesity together with existing metabolic comorbidities to reach women and their offspring at high risk and to have the potential for greatest benefit.