Exploring Fetal Sex Dimorphism in the Risk Factors of Gestational Diabetes Mellitus—A Prospective Cohort Study

Gestational diabetes mellitus (GDM) is a common pregnancy complication. Its etiology remains incompletely understood. Studies in recent years suggest that fetal sex may affect maternal metabolic milieu during pregnancy. We sought to assess whether there is fetal sex dimorphism in the risk factors of GDM. In a prospective pregnancy cohort in Shanghai, China, we studied 2,435 singleton pregnant women without pre-existing diabetes. GDM was diagnosed according to the International Association of Diabetes and Pregnancy Study Groups (IADPSG)' criteria. Log-binomial models were applied to obtain the adjusted relative risk (aRR). A total of 380 (15.6%) women developed GDM. Family history of diabetes was associated with an increased risk of GDM in women bearing a female fetus [aRR 1.74 (1.27–2.40), p < 0.001], but not in women bearing a male fetus (p = 0.68) (test for interaction, p = 0.03). Alcohol drinking was associated with an increased risk of GDM in women bearing a male fetus only (p = 0.023), although the test for interaction did not reach statistical significance (p = 0.055). In conclusion, family history of diabetes was associated with an increased risk of GDM in women bearing a female fetus only in this Chinese pregnancy cohort. There may be a need to consider fetal sex dimorphism in evaluating the risk factors of GDM.


INTRODUCTION
Gestational diabetes mellitus (GDM) is a common pregnancy complication characterized by impaired glucose tolerance with first-time recognition during pregnancy (1). Major risk factors of GDM include pre-pregnancy overweight and obesity, excessive gestational weight gain (GWG) during early/mid pregnancy, certain ethnicity (e.g., Asian), advanced maternal age, multiple pregnancy, history of GDM and family history of diabetes (2)(3)(4)(5). It has been recognized in recent years that fetal sex may influence glucose metabolism in the mother during pregnancy. Women bearing a female fetus may have a higher insulin resistance, but those bearing a male fetus may have a lower beta cell function during pregnancy (6,7). A meta-analysis has shown that pregnant women carrying a boy have a 4% higher relative risk of GDM than those carrying a girl (8). However, among women with GDM, those who carried a girl have about 6% higher relative risk of developing type 2 diabetes in the future (9,10). During pregnancy, the placenta secretes "diabetogenic" hormones (e.g., placental lactogen) into the maternal circulation, and higher maternal circulating concentrations of such diabetogenic hormones have been reported in women bearing a female fetus (11,12). Taken together, available evidence suggests a possible fetal sex dimorphism in the maternal endocrine environment that may allow for differential effects of certain GDM risk factors. We sought to test this hypothesis in a prospective pregnancy cohort.

METHODS
This was a prospective cohort study based on the recently developed Shanghai Birth Cohort (SBC) (13). Briefly, the SBC cohort recruited women in pre-conception care (1/4) or early antenatal care (11-17 weeks of gestation; 3/4) clinics in 6 tertiary obstetric care hospitals in Shanghai between 2013 and 2015. The study was approved by the research ethics boards of Shanghai Xinhua Hospital (the coordination center, approved on August 23, 2013, ref no. M2013-010) and all participating hospitals. All study participants signed an informed consent form. As part of the SBC cohort, the present study sought to assess sex dimorphism in the risk factors of GDM in women bearing a singleton fetus with complete data on major risk factors of GDM (age, pre-pregnancy overweight/obesity status, weight gain in early pregnancy, history of GDM, and family history of diabetes) and without pre-gestational diabetes (normal fasting blood glucose levels during the first antenatal care visit). A total of 2,435 singleton pregnant women were eligible and included.
The primary outcome was GDM. All women received a 75 g oral glucose tolerance test (OGTT) in screening for GDM during 24-28 weeks of gestation. GDM was diagnosed according to the criteria proposed by the International Association of Diabetes and Pregnancy Study Groups (IADPSG): if any one of the glucose values was at or above the following thresholds: fasting 5.1 mmol/L, 1 h 10.0 mmol/L and 2 h 8.5 mmol/L (14).
Pre-pregnancy weight (kg) was based on self-reports, while weight at the early pregnancy visit (11-17 weeks) was measured to the nearest 0.1 kg using the routinely available electronic weighing device in the hospital's prenatal care clinics. Weight at the delivery visit was abstracted from the patient's medical chart. Height (cm) was measured to the nearest 0.1 cm using the routinely available electronic stadiometer in the prenatal care clinics. Pre-pregnancy BMI (kg/m 2 ) was used to define weight status: underweight (<18.5), normal weight (18.5-23.9), overweight (24.0-27.9) and obese (≥28) according to Chinese standards. GWG in early pregnancy was calculated as the difference between weight at the first antenatal visit (around 14 weeks; 11-17 weeks) and pre-pregnancy weight. GWG in early pregnancy and its velocity (kg/week) were standardized into Z scores, using BMI category-specific mean and SD values derived from the study cohort (because there were no standards for GWG in early pregnancy). The Z scores were obtained by subtracting the mean from an individual's raw value and then dividing by the SD.
Because weight gain in late gestation in women with GDM would have been affected by clinical management after the diagnosis of GDM, the analysis on GWG as a risk factor was restricted to GWG in early pregnancy (before the clinical manifestation of the disease). Exploratory analyses observed similar findings for GWG z score and GWG velocity z score in early pregnancy; we presented the main results for GWG velocity in early pregnancy.
Ethnicity and family history of diabetes (first-degree relatives) were based on self-reports. Obstetric histories (parity, history of GDM) were based on self-reports and cross-verified in medical charts. Maternal age at recruitment was based on the date of birth and verified by personal identification card (with date of birth), and was classified as >35 years or not since maternal age >35 years is a classical risk factor of GDM (7). Smoking and alcohol drinking were based on self-reports in the early pregnancy (11-17 weeks) visit (did you smoking during this pregnancy? yes/no; did you drink alcohol during this pregnancy: yes/no).
Data were presented as mean±SD or number (%). Student t-test and Chi-square test were used to examine the differences in continuous and categorical variables, respectively. Log-binomial regression models were fitted to obtain the crude and adjusted relative risk (aRR) with 95% confidence intervals (CI) for the associations between risk factors and GDM. In the adjusted analyses, only co-variables with p < 0.2 were retained in the final parsimonious regression models to obtain more stable effect estimates. Interaction between two variables were assessed by examining the significance (p-value) of their multiplicative term in regression models. The study had a power of 93% to detect a significant interaction of fetal sex with a GDM risk factor of >=10% prevalence if there was a RR of >=2.0 in one stratum vs. the absence of relative risk increase in the other stratum, and a power of 47% if the RR was >=1.5, assuming a baseline event rate of 15% in GDM. All analyses were performed using Statistical Analysis System, Version 9.4 (SAS Institutes, Cary, North Carolina). P < 0.05 was considered statistically significant.

RESULTS
Of the 2,435 singleton pregnant women in the study cohort, 1,271 (52.2%) women had a male fetus, 1,164 (47.8%) had a female fetus, and 380 (15.6%) developed GDM. A greater proportion of women bearing a male fetus were at advanced maternity age (>35 years) (p = 0.002), or had gestational hypertension disorders (p = 0.013) than women bearing a female fetus ( Table 1). All other risk factors (such as maternal overweight and obesity, history of GDM, family history of diabetes) were similar in women bearing a male vs. a female fetus. Average GWG and average GWG velocity in early pregnancy were all similar in women bearing a male fetus vs. those bearing a female fetus.
In the whole study cohort, faster GWG in early pregnancy was associated with an increased risk of GDM (aRR = 1.12 per SD increase in GWG velocity, 95% CI 1.03-1.23, p = 0.010) ( Table 2). Advanced maternal age (>35 years) was associated with an increased risk of GDM in the whole study cohort without the adjustment for co-variables (RR = 1.41, P = 0.024), but the association became statistically non-significant after the adjustment (P = 0.098). Compared to normal weight women, overweight and obese women were at an increased risk of  Table 3 presents the risk factors of GDM by fetal sex, and Table 4 presents the estimates (95% confidence intervals) of interaction effects of fetal sex with GDM risk factors. Family history of diabetes was associated with an increased risk of GDM in women bearing a female fetus [aRR = 1.74 (1.27-2.40), p < 0.001], but not in women bearing a male fetus (p = 0.680); the test for interaction was statistically significant (p = 0.030). Faster GWG velocity in early pregnancy was associated with an increased risk of GDM in women bearing a female fetus [aRR = 1.17 (1.03-1.33); p = 0.017], but not    Table 3. GWG, gestational weight gain.

DISCUSSION
Our study is the first to explore fetal sex dimorphism in the risk factors of GDM. In this Chinese pregnancy cohort, an increased risk of GDM was observed for family history of diabetes in women bearing a female fetus only. There is also some weak evidence of fetal sex dimorphism for alcohol drinking in relation to the risk of GDM; a significant association was observed for one fetal sex only, but the tests for interactions did not reach statistical significance. The findings call for attention to fetal sex dimorphism in evaluating the risk factors of GDM. Family history of diabetes is a well-known risk factor of GDM. We observed an increased risk of GDM in women bearing a female fetus only. Circulating placental lactogen and oestriol levels have been reported to be higher in women bearing a female fetus than in those bearing a male fetus (11,12). We speculate that higher circulating levels of such diabetogenic hormones may "unmask" the genetic susceptibility to glucose intolerance in these women with a family history of diabetes bearing a female fetus.
Excessive or higher GWG during early pregnancy and advanced maternal age have been associated with an increased risk of GDM (15,16). A significant risk increase was observed in women bearing a female fetus only, but the test for interaction did not reach statistical significance (p = 0.25), probably due to the study's lack of power to detect moderate interactions. There is insufficient evidence in sex dimorphic association for maternal age or GWG.
There have been conflicting data concerning the association between alcohol drinking and GDM. Some studies reported that alcohol drinking might decrease the risk of GDM (17,18), while others reported an increased risk (19). In contrast, alcohol drinking was a significant risk factor for GDM in women bearing a male fetus only in our cohort, although the test for interaction did not reach statistical significance (p = 0.055). More studies in other cohorts/populations are needed to clarify the association.
Recent studies have shown that fetal sex may affect glucose metabolism in the mother. We did not detect the reported about 4% higher relative risk of GDM in women carrying a boy in a meta-analysis (8). This is not a surprise; ad hoc power calculation indicated that our study's power was only 8% to detect such a small risk difference. Women bearing a male fetus might be slightly and significantly more likely to develop GDM than women bearing a female fetus (7), but women with GDM in the first pregnancy were more likely to develop type 2 diabetes before second pregnancy if they delivered a female infant (9,10). The placenta produces placental lactogen and other diabetogenic hormones (20,21), and circulating levels of these diabetogenic hormones have been reported to be higher in women bearing a female fetus (11,12). Insulin resistance has been reported to be higher in women bearing a female fetus (6), yet beta-cell function appears to be lower in women bearing a male fetus (7), although there is a lack of confirmative studies. Taken together, available data suggest that the fetal sex may have a complex impact on the maternal endocrine milieu which may allow sex dimorphism in the effects of certain risk factors of GDM.
Our study has strengths and limitations. The study population was a large prospective cohort of relatively homogenous ethnicity (>98% Han Chinese) with low likelihood of confounding by ethnic difference in the susceptibility to GDM. Limitations are the crude measure of alcohol drinking (yes/no only), and limited power to detect risk factors of relatively small effects. Secondly, our study did not cover some possible GDM risk factors such as physical activity, dietary intake, the use of micronutrient or other supplements, passive smoking and sleeping habit. Lastly, we should caution that p-values in different strata in stratified analyses might be not comparable due to different sample sizes. Also, the generalizability of the findings to other ethnic groups/populations is unclear and cannot be assumed without confirmation studies.
In conclusion, family history of diabetes was associated with an increased risk of GDM in women bearing a female fetus only in this Chinese pregnancy cohort. Fetal sex dimorphism should be considered in evaluating the risk factors of GDM.

DATA AVAILABILITY STATEMENT
The datasets generated for this study are available on request to the corresponding author.

ETHICS STATEMENT
This study complies with the guidelines of the Declaration of Helsinki. Written informed consent has been obtained from each study participant. The study was approved by the research ethics boards of Shanghai Xinhua Hospital (the coordination center, approved on August 23, 2013, ref no. M2013-010).