Intrauterine insulin resistance in fetuses of overweight mothers

Abstract

Aim:  To investigate the relationship between maternal overweight and fetal insulin resistance.

Material and Methods:  Nineteen overweight and 30 lean pregnant women were recruited in the present study. Maternal and fetal insulin resistance were determined by measuring sex hormone binding globulin (SHBG) concentrations in maternal venous or umbilical cord serum, respectively. Maternal age, gestational age, height, pre-gravidity weight, pre-partum weight, as well as fetal gender, birth weight, birth height, and head circumference were collected as clinical data.

Results:  Fetuses of overweight mothers had larger birth weight (3.58 ± 0.55 kg vs 3.32 ± 0.42, adjusted P = 0.006) and lower SHBG concentrations (26.64 ± 3.65 vs 34.36 ± 7.84, adjusted P = 0.007) than those of lean mothers after values were adjusted for potential cofactors. Fetal SHBG level was negatively correlated with pre-gravidity body mass index (R = −0.392, adjusted P = 0.025) and weight gain during pregnancy (R = −0.332, adjusted P = 0.026) even with adjustment for potential cofactors. Among the 29 pregnant women with gestational diabetes mellitus, the overweight mothers had higher H1AC levels than their lean counterparts (6.47 ± 0.44 vs 5.74 ± 0.52, adjusted P = 0.004).

Conclusion:  Intrauterine insulin resistance is more prominent in fetuses of overweight mothers, an effect that is decreased by weight gain control during pregnancy.

Introduction

The intrauterine environment has a substantial influence on the health and disease state of a fetus in both the short- and long-term.^1–3 The long-term effects of the intrauterine environment on the offspring have been generalized in the ‘developmental origins of health and disease (DOHaD)’ theory.^4,5 Maternal obesity^6,7 and gestational diabetes mellitus (GDM)⁸ are both common pregnancy complications which are associated with various diseases of the offspring that may emerge during childhood or even adulthood.⁹

Recently, Catalano et al.¹⁰ reported that fetuses of obese mothers were more likely to develop insulin resistance (IR) in the uterus. In the study, 53 lean (body mass index [BMI] <25 kg/m²) and 68 obese (BMI>30 kg/m²) healthy women were recruited, and umbilical cord glucose and insulin concentrations were measured. Then, homeostasis model assessment (HOMA-IR) was used to compare between lean and obese groups. Results from their study showed that HOMA-IR was positively correlated with maternal pregravid BMI, and was consequently greater in the obese group.

Catalano's study¹⁰ was the initial investigation focusing on the relationship between maternal body mass and fetal insulin resistance. Discovery of this relationship is of great significance for both clinic and public health. Therefore, in the present study, we intend to extend this research in two aspects.

First, as mentioned above, fetuses of obese mothers are more prone to insulin resistance, compared with those of lean mothers. However, it remains unclear whether fetuses of overweight (in global population: 25–30 kg/m²; in the Asian population: 23–30 kg/m²) mothers also suffer from increased insulin resistance in the uterus. Although not as severe as obesity, being overweight is twice as common in the general population.¹¹ If overweight expecting mothers also have increased occurrence of fetal insulin resistance, the standard for clinical control of body weight should be stricter. Therefore, the primary purpose of the present study is to investigate the influence of maternal BMI on fetal insulin resistance, specifically the difference between overweight and lean mothers.

Second, along with lipid metabolic disorders, glucose metabolic disorders are also associated with insulin resistance. Maternal overweight and gestational diabetes mellitus (GDM) are both IR-related disorders that often present concurrently in the clinic. Therefore, in the present study, we also aim to investigate the influence of maternal hyperglycemia on fetal insulin resistance.

Methods

Patients and diagnosis

The study population was randomly chosen from term (37–42 weeks) singleton pregnant women with or without gestational diabetes mellitus (GDM) who received prenatal care and caesarean section in our center. Mothers or fetuses complicated with severe diseases (i.e. preeclampsia, prenatal infection, hyperthyreosis, and fetal anomaly) were excluded.

Recruited patients were divided into two groups, according to World Health Organization (WHO) criteria on BMI for the Asian population.¹² In detail, pregnant women whose pre-gravidity BMI was over 23 kg/m² were enrolled in the overweight group, while those with a pre-gravidity BMI in the range of 18.5–23 kg/m² were placed in the lean group. The diagnosis of GDM was determined according to the oral glucose tolerance test (OGTT) result performed within 24–28 weeks of pregnancy following American Diabetes Association (ADA) criteria (two-step approach).

The study was approved by the research ethical committee of The First Affiliated Hospital of Sun Yat-sen University and informed consent was provided by all volunteers.

Clinical data

Clinical data, including maternal age, gestational age, height, pre-gravidity weight, pre-partum weight, as well as fetal gender, birth weight, birth height, and head circumference were obtained from subjects or measured by researchers.

Blood sample collection

Maternal venous blood samples were collected within 37–38 weeks of gestation and umbilical cord venous blood samples were collected immediately after placenta delivery. The samples were collected in two sterile vacuum blood collection tubes, one with anticoagulant and the other without. Samples in the anticoagulative tubes for measuring H1Ac (in total blood) and glucose (in plasma) were sent to the clinical laboratory within 1 h following collection. The other samples were placed at room temperature (25°C) for 1 h to allow the blood to clot. After blood coagulation, sera were separated by centrifugation at 720 g for 15 min at 4°C, frozen in aliquots of 100µL, and stored at −80°C until analysis.

Laboratory procedures

Sex hormone binding globulin (SHBG) from maternal and fetal sera was detected using an Architect SHBG detection kit (Abbott Laboratory, Princeton, NJ, USA) using an i2000 immunoanalysis array machine (Abbott Laboratory). For the GDM group only, the fasting plasma glucose level was determined using a GOD-PAP kit (Human, Wiesbaden, Germany); the fasting insulin level was examined using a chemiluminescent microparticle immunoassay (CMIA) kit (Abbott Laboratory); the plasma C-peptide was tested by an ARCHITECT C-Peptide Reagent kit (Abbott Laboratory, Princeton, NJ, USA); the cholesterol and triglyceride (TG) levels were measured using the enzymatic colorimetric test kit (Human, Wiesbaden, Germany), high-density lipoprotein (HDL) and low-density lipoprotein (LDL) were determined by homogeneous assay using Chestiest N HDL kits (Human, Wiesbaden, Germany); the H1Ac were tested by Bio-Rad Variant II H1Ac kit (Bio-Rad Laboratories, Hercules, CA, USA).

Statistical analysis

Data were shown as mean ± SD and statistical analyses were performed using SPSS version 13.0 software. Maternal and fetal SHBG, clinical data, and laboratory data were compared between overweight and lean groups. Differences were adjusted for potential cofactors by ANCOVA. Correlations between SHBG and other test indexes were analyzed and positive findings were adjusted for potential cofactors by partial correlation.

Results

Baseline characteristics

Among the 49 pregnant women, 19 were diagnosed as overweight according to the WHO criteria of BMI for the Asian population,¹² while the remaining 30 mothers were placed in the control (lean) group. Clinical data including age, gestational age, height, pre-gravidity and pre-partum weight, weight gain during pregnancy, pre-gravidity and pre-partum BMI, morbidity of gestational diabetes mellitus (GDM) were compared between overweight and lean mothers (Table 1). Gender, birth weight, height, and head circumference were compared between fetuses of overweight and lean mothers (Table 1). As shown in Table 1, after adjusting for maternal age, gestational age, fetal gender and morbidity of GDM, the overweight group had higher maternal weight and BMI both before and late in pregnancy, while no differences in height, weight gain during pregnancy, and blood glucose levels (as measured by the OGTT test) were observed between groups. Fetuses of overweight mothers were heavier at birth (after adjusted by the morbidity of GDM, maternal age, gestational age, and fetal gender) and had a shorter gestational age than those of lean mothers.

Table 1. Baseline clinical characteristics and biochemical parameters of pregnant women and their offspring (mean ± SD)

	Overweight	Lean	P-value	Adjusted P-value
BMI, body mass index; GDM, gestational diabetes mellitus; OGTT, oral glucose tolerance test.
Maternal parameter
Number	19	30
Age (years)	30.89 ± 3.14	30.73 ± 3.57	0.873
Gestational age (days)	269.97 ± 5.8	273.78 ± 4.7	0.021
Height (m)	1.58 ± 0.04	1.60 ± 0.04	0.471	0.188
Pre-gravidity weight (kg)	62.23 ± 7.94	51.85 ± 4.19	0.000	0.000
Pre-partum weight (kg)	77.47 ± 6.85	65.28 ± 5.13	0.000	0.000
Weight gain during pregnancy (kg)	15.24 ± 4.05	13.43 ± 3.98	0.131	0.595
Pre-gravidity BMI (kg/m^2)	24.38 ± 1.51	20.55 ± 1.38	0.000	0.000
Pre-partum BMI (kg/m^2)	30.34 ± 1.82	25.88 ± 1.88	0.000	0.000
Complicated with GDM
Yes	12	17	0.769
No	7	13
OGTT-fast (mmol/L)	4.92 ± 0.97	4.52 ± 0.57	0.091	0.082
OGTT-1h (mmol/L)	10.40 ± 2.65	9.51 ± 1.77	0.184	0.086
OGTT-2h (mmol/L)	9.02 ± 2.68	8.12 ± 1.88	0.201	0.236
OGTT-3h (mmol/L)	6.19 ± 1.22	5.94 ± 1.42	0.560	0.594
Fetal parameters
Gender
Male	9	17	0.569
Female	10	13
Birth weight (kg)	3.58 ± 0.55	3.32 ± 0.42	0.106	0.006
Birth height (cm)	50.00 ± 2.22	49.70 ± 2.09	0.676	0.503
Head circumference (cm)	34.39 ± 1.69	34.50 ± 1.63	0.845	0.316

Insulin resistance in overweight mothers and fetuses

To investigate the insulin resistance status of mothers and fetuses, serum SHBG levels were detected and compared between groups, after adjusting for potential cofactors including maternal age, gestational age, fetal gender and morbidity of GDM. As shown in Figure 1, there was no statistically significant difference in SHBG levels between overweight and lean women (542.58 ± 110.46 nmol/L vs 565.51 ± 123.77 nmol/L, P = 0.514, adjusted P = 0.188) during late pregnancy (Figure 1A). On the contrary, fetuses of overweight mothers had lower SHBG levels than those of lean mothers even after adjusting for potential cofactors (26.64 ± 3.65 nmol/L vs 34.36 ± 7.84 nmol/L, P < 0.001, adjusted P = 0.007), indicating that they were more likely to develop insulin resistance in the uterus (Figure 1B).

Figure 1. Maternal and fetal sex hormone binding globulin (SHBG) levels in overweight and lean mothers. Parts A and B show the differences in maternal (A) and fetal (B) SHBG levels between overweight and lean groups. No significant difference in SHBG was found in mothers of each group, while fetuses of overweight mothers have lower SHBG levels in their umbilical cord sera. Parts C and D show the correlation between maternal pre-gravidity body mass index (BMI) and SHBG levels in maternal venous serum (C) or umbilical cord serum (D). Pre-gravidity BMI did not correlate with maternal SHBG levels, while it had a negative correlation with fetal SHBG level (R = −0.392, P = 0.005, adjusted P = 0.025). Closed rings in Parts C and D: subjects with gestational diabetes mellitus; Open rings in Parts C and D: subjects with normal glucose tolerance. Line in Part D: Fit line for whole cohort.

To analyze the relationship between maternal pre-gravidity BMI and insulin resistance in both mothers and fetuses, we further investigated the correlation between pre-gravidity BMI and SHBG levels, adjusting for the potential cofactors mentioned above. As shown in Figure 1, pre-gravidity BMI has a negative correlation with fetal SHBG levels (R = −0.392, P = 0.005, adjusted P = 0.025) but not with maternal SHBG levels (R = −0.113, P = 0.440, adjusted P = 0.537). As SHBG levels are negatively correlated with fetal insulin resistant status, this result indicates that offspring of mothers with higher pre-gravidity BMI have increased insulin resistance.

Maternal weight gain control during pregnancy improves fetal insulin resistance

As higher pre-gravidity BMI is associated with increased fetal insulin resistance, it would be interesting to examine whether weight gain control of pregnant women will improve fetal insulin resistance. We first studied the correlation between maternal weight gain and fetal SHBG, and found they had a negative correlation (R = −0.327, P = 0.022). Next, we performed a partial correlation analysis on maternal weight gain and fetal SHBG, taking GDM morbidity, fetal gender, maternal age, and pre-gravidity BMI as control factors. Results showed that after adjusting for the cofactors mentioned above, maternal weight gain still had a negative correlation (R = −0.332, adjusted P = 0.026) with fetal SHBG. This result suggests that weight gain control during pregnancy will improve fetal insulin resistance.

Gestational diabetes mellitus, overweight and fetal insulin resistance

Gestational diabetes mellitus is a common complication during pregnancy. As they can both be a result of insulin resistance, GDM and obesity are likely to complicate with each other. Several studies have shown that the offspring of GDM mothers developed insulin resistance either before or after birth. Therefore, we studied the influence of GDM on fetal insulin resistance, either as an independent factor or as a cofactor of overweight. Results showed that fetal SHBG levels were not different between GDM and control groups (31.53 ± 7.86 nmol/L vs 31.12 ± 7.20 nmol/L, P = 0.852). Then, to investigate the coeffect of overweight and GDM on fetal IR, we performed a two-way ANOVA on the total population. Results demonstrated that these two IR-related disorders did not have a coeffect on fetal IR development (F = 0.385, P = 0.538). Taken together, these results indicate that GDM has little relation with fetal insulin resistance in the present cohort, no matter as an independent factor or as a cofactor of being overweight.

In addition, we analyzed the potential influence of overweight on GDM pregnancy. As shown in Table 2, compared with lean GDM mothers, overweight GDM mothers had higher hemoglobin 1Ac (H1Ac) levels, indicating that higher pre-gravidity weight is related to decreased glucose control in GDM mothers. Other laboratory characteristics, including fasting insulin, C-peptide, cholesterol, triglyceride, HDL, and LDL were not different between overweight and lean GDM mothers.

Table 2. Biochemical parameters of pregnant women with gestational diabetes mellitus (mean ± SD)

	Overweight	Lean	P-value	Adjusted P-value
HbA1c, hemoglobin A1c; HOMA-IR, homeostasis model assessment of insulin resistance.
Number	12	17
Fast insulin (µU/mL)	14.78 ± 6.84	11.18 ± 3.97	0.134	0.967
HOMA-IR	3.26 ± 1.26	2.02 ± 0.68	0.009	0.399
C-peptide (pmol/L)	1.28 ± 0.47	0.94 ± 0.34	0.060	0.438
Total cholesterol (pmol/L)	5.93 ± 1.13	5.96 ± 1.00	0.951	0.983
Triglyceride (mmol/L)	6.41 ± 7.15	3.23 ± 0.95	0.125	0.751
High-density lipoprotein (mmol/L)	1.31 ± 0.40	1.54 ± 0.40	0.174	0.995
Low-density lipoprotein (mmol/L)	2.59 ± 0.77	3.13 ± 0.78	0.126	0.972
HbA1c (%)	6.47 ± 0.44	5.74 ± 0.52	0.000	0.004

Discussion

More than one billion adults from all over the world are either obese or overweight,¹¹ including a large number of reproductive-age women. Maternal obesity has already been identified as a risk factor for fetal insulin resistance.^10,13 Being overweight, a condition between obesity and normal-range BMI, is twice as common as obesity.¹¹ Therefore, in the present study, we mainly focus on whether being overweight could also be associated with fetal insulin resistance. Our results indicate that, compared to a fetus from a lean mother, insulin resistance is more likely to develop in the fetus of an overweight mother, an effect that could be reduced by weight gain control during pregnancy.

Sex hormone binding globulin level is negatively correlated with insulin resistance and has been taken as a biomarker for IR.^14–17 Our findings show that maternal pre-gravidity BMI is negatively associated with maternal and fetal SHBG, although no statistically significant difference was observed for maternal SHBG. This indicates that higher pre-gravidity BMI is associated with increased fetal IR. Next, maternal and fetal SHBG levels were compared between overweight and lean pregnancies. As expected, overweight mothers and their offspring had lower SHBG levels than lean mothers, although no statistical significance was found between mothers in the two weight groups. These results indicate that the influence of overweight on IR is more remarkable in the developing fetus than in the mother. The relationship between maternal pre-gravid BMI and fetal SHBG found in our study (R = −0.392) for overweight and lean mothers is consistent with previous studies on pre-gravid BMI and fetal HOMA-IR (R = 0.31) among obese and lean mothers during pregnancy^10,13 and confirm the relationship between maternal BMI and fetal insulin resistance. Further, because our data focuses on overweight mothers, our results help to fill the gap between obese and lean pregnant women. Thus, we can draw a conclusion that fetuses from both obese and overweight mothers are more likely to develop insulin resistance in the uterus.

According to the developmental origins of health and disease (DOHaD) theory, maternal environment affects the long-term health of a fetus, especially concerning energy metabolism. Previous studies demonstrated that macrosomia or intrauterine growth restriction (IUGR), results of GDM or maternal obesity, tend to develop insulin resistance in childhood and adulthood.^18–20 Here, we presume that mothers who are overweight could also cause long-term effects on the fetus as similar trends in fetal insulin resistance were observed.

As maternal overweight has such relevance to fetal insulin resistance, which can have a negative impact on fetal health, it is necessary to investigate whether weight gain control during pregnancy will improve fetal IR or not. Results in the present study show that after adjusting for several cofactors, maternal weight gain is negatively associated with fetal SHBG. This result indicates that besides pre-gravidity BMI, greater weight gain during pregnancy will also increase fetal insulin resistance. That is, even fetuses from lean mothers can develop IR during pregnancy if the mother does not control her body weight within an appropriate range. In addition, weight gain control in pregnancy is also beneficial for reducing maternal insulin resistance,²¹ fetal overgrowth, shoulder dystocia, cesarean delivery, hypertensive disorders,²² brachial plexus injury and hypertensive disorders.²³ Therefore, it is of great importance for a pregnant woman to manage her weight in order to prevent disease, especially IR-related disease, in herself and her offspring.

Besides overweight, we also investigated GDM, another IR-related disease, for its relation to overweight and its potential influence on fetal insulin resistance. Results show that, in the present cohort, GDM did not increase the risk for developing fetal insulin resistance, either as an independent factor or as co-effecter with maternal BMI. This result is consistent with recently published data,²⁴ revealing that maternal SHBG is no longer an independent predictor for GDM where obesity is concerned. Interestingly, an overweight GDM mother is more prone to have higher H1Ac levels than their lean counterparts, indicating that overweight GDM patients should pay more strict attention to their long-term blood glucose control than their lean counterparts.

Measuring fetal insulin resistance still remains a difficult task. HOMA-IR is a commonly used method for measuring levels in adults, while its accuracy in fetal samples is questionable as the energy intake rhythm of a fetus is still unclear. The glucose clamp, the gold standard for measuring insulin resistance in adults, also appears to be ineffective for measuring levels in fetal/neonatal samples.²⁵ In the present study, we used sex hormone binding globulin (SHBG) as an index of insulin resistance. SHBG, a glycoprotein mainly produced by the liver, has a major biological function to bind sex hormones (testosterone and estradiol) in circulation.²⁶ Insulin resistance inhibits the production of SHBG in the liver,²⁷ and hence is negatively associated with SHBG levels.²⁸ In the clinic, patients with disorders associated with insulin resistance, such as diabetes mellitus²⁹ and polycystic ovary syndrome (PCOS)³⁰ typically have lower SHBG concentrations in circulation. Recently, SHBG has been taken as a sensitive biomarker for insulin resistance in adults²⁸ and fetuses.³¹ As SHBG is easy to measure with commercially available kits and stable, but is not influenced by food intake, we chose it as the IR marker in the present study.

In conclusion, being overweight during gestation is relevant to fetal IR, whereas weight gain control during pregnancy improves fetal IR. As the overweight population is much larger than the obese population, the necessity is now apparent to make overweight mothers a major public health focus for the prevention of IR disease developed in the next generation. Further studies based on a larger population would be beneficial for uncovering multiple factors associated with fetal insulin resistance and providing more rigorous evidence for clinical practice.

Acknowledgments

This work was supported by the Science and Technology Planning Project of Guangdong Province, China (No. 2009B03081099, to Zi-lian Wang) and Scholarship Award for Excellent Doctoral Student (to Bin Liu and Zi-lian Wang as supervisor) granted by the Ministry of Education of China.

Disclosure

The authors declare that they have no competing interests.