Fetal sexual dimorphism and preeclampsia among twin pregnancies

Abstract

Background

In singleton pregnancies, fetal sexual dimorphism has been observed in hypertensive disorders of pregnancy (HDP), particularly preeclampsia, a morbid syndrome that increases risk of adult-onset cardiovascular disease for mothers and their offspring. However, few studies have explored the effect of fetal sex on HDP among twin pregnancies.

Methods

We conducted a retrospective cohort study of 1,032 twin pregnancies between 2011 – 2022 using data from a perinatal database that recruits participants from three hospitals in Houston, TX. We categorized pregnancies based on fetal sex pairings into female/female, male/male, and female/male. Pregnancies with a female/female pairs were used as our reference group. Our primary outcomes included gestational hypertension, preeclampsia, superimposed preeclampsia, and preeclampsia subtyped by gestational age of delivery. A modified Poisson regression model with robust error variance was used to calculate the relative risk (RR) and 95% confidence interval (CI) for the association between fetal sex pairs and HDP.

Results

Adjusted models of female/male pairs were associated with preterm preeclampsia (RR 2.01, 95% CI 1.15–3.53) relative to those with female/female pairs. No associations with other HDP were observed among pregnancies with male/male pairs compared to those with female/female fetal sex pairs.

Conclusions

We found some evidence of sexual dimorphism for preterm preeclampsia among female/male twin pairs. Additional research is needed to understand what biological mechanisms could explain these findings.

Graphical Abstract

Introduction

One in 32 pregnancies in the United States are multifetal¹ which increases the risk of severe subtypes of preeclampsia, possibly through elevated cardiovascular burden.² This is a significant public health concern. Preeclampsia affects 5–10% of pregnancies and is a leading cause of maternal mortality.³ The pathogenesis is not fully understood and the only treatment is delivery, which often results in preterm birth.⁴ Furthermore, women with preeclampsia have a 3-fold increased risk of cardiovascular disease⁴ and their infants also have increased risk of cardiovascular disease later in life.^{5, 6} Given the serious nature of preeclampsia, and the long-term impacts on cardiovascular health, further research on preeclampsia in multifetal pregnancies is warranted.

In singleton pregnancies, the sex of the fetus is thought to impact the risk of preeclampsia, possibly through differences in placental metabolic and immunological function or maternal systemic angiogenic and cytokine expression.^7–12 Studies in multifetal pregnancies align with trends seen in singleton pregnancies with male fetal sex being associated with shorter gestation and male/male pregnancies at the highest risk for preterm delivery.^{13, 14} Approximately three studies have examined fetal sex and preeclampsia in multifetal pregnancies but results were inconsistent, and only one study accounted for preeclampsia subtypes (i.e. early and late onset or term and preterm as a proxy) which are thought to have different pathophysiologies.⁴ In studies among singletons, male fetal sex has been associated with preeclampsia but there appears to be a female bias in preeclampsia that results in a preterm delivery.^{8, 15, 16} This suggests that in singletons fetal sex may influence severity of preeclampsia. As data is limited our objective was to examine the association between fetal sex and hypertensive disorders of pregnancy [HDP] including preeclampsia subtypes among twin pregnancies.

Methods

Data Availability

Data to support this study is available upon request and protocol approval from Peribank [ Baylor College of Medicine, Department of Obstetrics and Gynecology]

Study Design and Population

We conducted a retrospective cohort study of 1,032 twin deliveries using data from Peribank¹⁷, an obstetrical database affiliated with Baylor College of Medicine and Texas Children’s Hospital. Pregnant individuals are recruited when they present to labor and delivery or immediately after delivery in cases of emergency.¹⁷ Data is obtained through patient interviews and electronic medical records. To be included, participants must be pregnant, ≥ 18 years of age (or ≥16 years of age if emancipated), able to read and understand the consent form, and able to sign the consent form.¹⁷ There are no exclusion criteria. The Institutional Review Boards at Baylor College of Medicine and Texas Children’s Hospital approved Peribank. Informed consent was obtained from each participant. For our study, the Institutional Review Board at the University of Texas Medical Branch determined that this study was exempt. One author had full access to all the data in the Peribank database and takes full responsibility for its integrity and data analysis.

Primary Exposure

Our primary exposure was multifetal pregnancy with female/female, male/male, or female/male fetal sex pairs as determined in the administrative database. Approximately 18 women were excluded due to missing data on fetal sex (1.74%), the remaining 1,014 participants were included in our analysis.

Primary Outcome

The primary outcomes were HDP which included preeclampsia, gestational hypertension, and superimposed preeclampsia diagnosed using the American College of Obstetricians and Gynecologists criteria.¹⁸ Preeclampsia is diagnosed by “systolic blood pressure of 140 mm Hg or more or diastolic blood pressure of 90 mm Hg or more on two occasions at least 4 hours apart after 20 weeks of gestation in a woman with a previously normal blood pressure” and proteinuria “at least 300 mg per 24-hour urine collection or protein/creatinine ratio of 0.3 mg/dL” or evidence of systemic organ dysfunction in the absence of protienuria.¹⁸ In addition, preeclampsia can be defined based on gestational age of delivery and whether preeclampsia results in preterm birth <37 weeks gestation (i.e. preterm preeclampsia and term preeclampsia). Gestational hypertension is defined as hypertension without proteinuria or severe features occurring after 20 weeks gestation with blood pressure levels returning to normal in the postpartum period.¹⁸ Superimposed preeclampsia is defined as preeclampsia superimposed upon previously diagnosed chronic hypertension.¹⁸

Covariates

Information was collected regarding the age of the mother (median, IQR), body mass index [BMI], racial/ethnic background (Non-Hispanic [NH] White, Hispanic, NH Asian/Other), foreign-born status (no, yes), marital status (married, single), education level (some college and above, high school, and less than high school), and insured status (private, Medicaid/Children’s health insurance program [CHIP], other/unknown/no insurance). Data on alcohol (no, yes), tobacco (no, yes), and drug use (no, yes) was obtained. Information on maternal comorbidities (endometriosis, chronic health conditions, thyroid disease, mental health issues, seizure disorder, sickle cell disease, hypospadias, prior gestational diabetes) (no, yes) was collected. Data was also available on the number of previous pregnancies (no pregnancy, 1–2 prior, >2 prior pregnancies) and the gestational age at first prenatal visit by trimester were collected (0–12 weeks, >12 weeks).

Statistical Analyses

We compared maternal demographic and clinical characteristics between fetal sex pairs (female/female, male/male, and female/male) using a modified Poisson regression model with robust error variance, which is an accepted method to directly measure relative risk [RR] and 95% confidence intervals [CI] risk rather than using odds ratios as estimates for risk.¹⁹ The same modeling approach was used to examine fetal sex pairs and risk of HDP. Female/female fetal sex pairs were used as our reference group as male fetal sex is most consistently linked with elevated risks of adverse pregnancy outcomes.^{13, 14} Our model covariates were selected a priori. They included maternal age, maternal comorbidity status (yes/no), education level, and insurance status. The missingness among the variables included in our model ranged from 1 – 6%. We used fully conditional multiple imputation with ten iterations to address missing data. All analyses were conducted using SAS version 9.4, Cary, North Carolina.

Results

Population and Characteristics

In our study population there were a total of 1,014 women with multifetal pregnancies, including 337 (33.2%) with female/female, 345 (34.0%) with male/male, and 332 (32.7%) with female/male pairs (Table 1). Foreign-born mothers were less likely to have female/male twins compared to female/female (RR 0.77, 95% CI 0.62–0.97). Mothers with a high school education or less (RR 0.65, 95% CI 0.47–0.91) and those insured through Medicaid/CHIP (RR 0.78, 95% CI 0.63–0.97) were less likely to have female/male pairs compared to female/female. Median BMI was slightly lower in mothers of male/male twin pairs compared to those with female/female (RR 0.97, 95% CI 0.96–0.99). We no found differences in maternal health variables and reproductive histories between fetal sex pairs (Table S1).

Table 1.

Maternal demographics and clinical variables by fetal sex pairs

Variables	Female/Female	Male/Male	*Risk Ratio (RR) RR, 95% CI	Female/Male	*Risk Ratio (RR) RR, 95% CI
N	337 (33.2)	345 (34.0)		332 (32.7)
Age (Median, IQR)	32 (27 – 25)	31 (27 – 35)	0.99, 0.98 – 1.01	32 (28 – 35)	1.01, 1.00 – 1.02
Foreign-born Status (n, %)
No	181 (54.4)	190 (55.6)	Ref	207 (65.9)	Ref
Yes	152 (45.7)	152 (44.4)	0.98, 0.79 – 1.21	107 (34.1)	0.77, 0.62 – 0.97
Education, (n, %)
Some College and above	186 (57.6)	197 (60.4)	Ref	214 (69.5)	Ref
High School	66 (20.4)	79 (24.2)	1.06, 0.82 – 1.36	57 (18.5)	0.87, 0.66 – 1.15
Less than High School	71 (22.0)	50 (15.3)	0.81, 0.60 – 1.09	37 (12.0)	0.65, 0.47 – 0.91
Insurance (n, %)
Private	134 (40.2)	143 (41.7)	Ref	170 (52.0)	Ref
Medicaid / CHIP	189 (56.8)	188 (54.8)	0.98, 0.79 – 1.21	145 (44.2)	0.78, 0.63 – 0.97
Other / Unknown / No Insurance	10 (3.1)	12 (3.7)	1.06, 0.59 – 1.92	12 (3.7)	0.98, 0.54 – 1.75
Body Mass Index (BMI), Median (IQR)	32.9, (29.8–37.3)	31.3 (28.1–35.5)	0.97, 0.96–0.99	32.4 (28.7–37.1)	0.99, 0.98 – 1.01
Smoking, (n, %)
No	285 (83.3)	295 (85.3)	Ref	275 (83.3)	Ref
Yes	57 (16.7)	51 (14.7)	0.93, 0.70 – 1.25	55 (16.7)	1.0, 0.75 – 1.34

^*Risk ratio was calculated using a modified Poisson regression model with robust estimates.

Within our cohort, approximately 104 (10.8%) were diagnosed with preeclampsia, 35 (3.4%) were diagnosed with gestational hypertension, and 24 (2.7%) were diagnosed with superimposed preeclampsia (Table 2). Participants carrying female/male pairs were more likely to have preeclampsia compared to female/female pairs (RR 1.71, 95% CI 1.04–2.81) but this was not observed for male/male pairs (RR 1.45, 95% CI 0.88–2.41). There was no association between male/male (RR 0.99, 95% CI 0.41–2.38) and female/male pairs (RR 1.50, 95% CI 0.66–3.42) and gestational hypertension when compared to female/female pairs. There was no difference in the risk of superimposed preeclampsia between male/male (RR 2.06, 95% CI 0.70–6.07) and female/male pairs (RR 1.80, 95% CI 0.60–5.44) compared to female/female twin pairs.

Table 2.

Association between fetal sex pairs and hypertensive disorders of pregnancy

Outcome	Fetal Sex Pairs	n, %	*Crude RR	Age-adjusted RR	*†Adj RR, 95 CI
Gestational Hypertension	Female/Female	10, 2.9%	Ref	Ref	Ref
	Male/Male	10, 2.9%	0.99, 0.41 – 2.37	0.98, 0.41 – 2.35	0.99, 0.41 – 2.38
	Female/Male	14, 4.2%	1.45, 0.64 – 3.26	1.49, 0.66 – 3.35	1.50, 0.66 – 3.42
Preeclampsia	Female/Female	26, 7.6%	Ref	Ref	Ref
	Male/Male	37, 10.7%	1.41, 0.85 – 2.32	1.41, 0.85 – 2.33	1.45, 0.88 – 2.41
	Female/Male	41, 12.4%	1.63, 1.00 – 2.66	1.63, 0.99 – 2.66	1.71, 1.04 – 2.81
Superimposed Preeclampsia	Female/Female	5, 1.7%	Ref	Ref	Ref
	Male/Male	10, 3.4%	1.96, 0.67 – 5.73	2.00, 0.68 – 5.87	2.06, 0.70 – 6.07
	Female/Male	9, 3.3%	1.89, 0.63 – 5.64	1.85, 0.62 – 5.51	1.80, 0.60 – 5.44
ǂPreterm Preeclampsia	Female/Female	19, 5.6%	Ref	Ref	Ref
	Male/Male	29, 8.4%	1.51, 0.85 – 2.69	1.51, 0.85 – 2.70	1.56, 0.87 – 2.79
	Female/Male	36, 10.8%	1.96, 1.12 – 3.41	1.95, 1.12 – 3.40	2.01, 1.15 – 3.53
§Term Preeclampsia	Female/Female	7, 2.1%	Ref	Ref	Ref
	Male/Male	8, 2.3%	1.13, 0.41 – 3.12	1.12, 0.41 – 3.10	1.17, 0.42 – 3.23
	Female/Male	5, 1.5%	0.74, 0.23 – 2.33	0.75, 0.24 – 2.35	0.84, 0.26 – 2.68

^*Relative risk was calculated using a modified Poisson regression model with robust estimates

^†Adjusted for maternal age, maternal comorbidities, education level, and payment method.

^‡Preterm preeclampsia was defined as preeclamptic birth occurring after less than 37 gestation weeks.

^§Term preeclampsia was defined as a preeclamptic birth occurring after 37 or more gestation weeks.

Of those with preeclampsia, 84 (8.1%) were diagnosed with preterm preeclampsia while 20 (1.9%) were diagnosed with term preeclampsia (Table 2). Female/male pairs were more likely to have preterm preeclampsia compared to female/female pairs (RR 2.01, 95% CI 1.15–3.53), while no difference was noted in male/male pairs (RR 1.45, 95% CI 0.88–2.41). There was no difference in term preeclampsia for the male/male (RR 1.56, 95% CI 0.42–3.23) and female/male pairs (RR 0.84, 95% CI 0.26–2.68) compared to female/female pairs.

Lastly, as a sensitivity analysis, we explored if associations between female/male pairs were consistent if male/male pairs were the reference group. We found no association between female/male fetal sex pairs and gestational hypertension, preeclampsia, superimposed preeclampsia, or preeclampsia subtypes (Table S2). However, effect estimates were in a similar direction as our analysis using female/female pairs as the reference group.

Discussion

We observed that individuals carrying female/male fetal sex pairs were more likely to have preterm preeclampsia compared to those carrying female/female pairs. Funaki et al., among a Japanese population, found that pregnancies with male/male pairs have higher risk of preterm birth but lower risk of preeclampsia when compared to female/female pairs, although associations were only significant among dichorionic twins.²⁰ Relative risks were similar for preeclampsia when they examined male/female dichorionic twin pairs but confidence intervals overlapped one.²⁰ In contrast, Lučovnik et al., in a study conducted in Slovenia found no associations between monozygotic fetal sex pairs and preeclampsia compared to female/male dizygotic pairs, suggesting no influence of zygosity.²¹ Only one study examined preeclampsia subtypes. Steen et al.,¹⁴ in 16,045 twin pregnancies in Sweden, found that preterm preeclampsia was more common in pregnant individuals carrying female/male or female/female twin pairs compared to those with male/male pairs. We observed no significant difference in preterm preeclampsia risk for female/male pairs compared to male/male pairs, although the effect estimates were in a similar direction.

The pathogenesis of preeclampsia is not fully understood and complicated by its heterogeneous nature. Preeclampsia consists of varying subtypes (e.g., early vs late onset) hypothesized to have different biological pathways leading to the clinical symptoms of preeclampsia, however, there are no biomarkers or clinical indicators that can distinguish subtypes.^{4, 22, 23} Abnormal placentation due to failed spiral artery remodeling and placental ischemia is thought to occur in majority of cases of early onset preeclampsia or preterm preeclampsia.²⁴ In contrast, late onset or term preeclampsia may be a consequence of maternal factors such as obesity, or diabetes.²⁵ Preeclampsia subtypes also have varying risks of long-term complications, with those with early onset or preterm preeclampsia having the highest risk of cardiovascular disease.²³ However, twin pregnancies are not frequently included in preeclampsia studies. Bergman et al²⁶ recently found that cardiovascular risk patterns in multifetal pregnancies with preeclampsia do not mimic singleton pregnancies, suggesting that perhaps preeclampsia is driven by different mechanisms in multifetal pregnancies. Our findings did not closely mimic findings from singleton pregnancies where female fetal sex is consistently linked to preterm preeclampsia.^{8, 15, 27}

Although the impact of fetal sex on maternal outcomes is well documented, the mechanisms explaining these associations are understudied. Female fetal sex has been associated with lower first trimester pro-inflammatory markers (IFNγ and IL-12) and increased second trimester pro-inflammatory (TNFβ and IL1β), anti-inflammatory (IL4r), and regulatory cytokines (IL5 and IL10).⁸ Male fetal sex has been associated with a pro-inflammatory state in the first trimester, which may explain reported higher rates of early pregnancy loss.^{8, 28, 29} Steen et al., postulated that testosterone may act to protect male fetuses against preeclampsia by dampening the maternal immunological response.¹⁴ However, these findings do not fully explain our observed associations between female/male twin pairs and preeclampsia when compared to female/female twin pairs.

Individuals carrying twin pregnancies are at an increased risk of preeclampsia compared to those carrying singleton pregnancies.^{20, 30–32} Studies have suggested that pregnancies with dichorionic twins are more likely to have preeclampsia due to an increase of placental mass in the womb which activates the maternal immune system as well as increases soluble FMS like tyrosine kinase receptor [sFLT-1] production.²⁰ However, studies examining associations between zygosity and HDP have been inconclusive.¹³ Our results however may be consistent with this rationale. Female/male twin pairs are dizygotic, possessing an increased placental mass compared to monozygotic twins.³² However, zygosities and chorionicities were not evaluated for our entire cohort, therefore we are unable to draw a definitive conclusion.

One of our study strengths includes using a diverse study population with low frequencies of missing data. The dataset we used also undergoes frequent quality checks and contains extensive information on each pregnancy. Thus, we were able to explore several hypertensive disorders and different preeclampsia subtypes. Some of our limitations include our small sample size, due to the low prevalence of twin pregnancies in the general population. Another limitation is that we could not capture early fetal losses. Lastly, we could not rule out the potential of residual confounding.

Perspectives

Our study demonstrated an increased risk of preterm preeclampsia in pregnant individuals with female/male fetal sex pairs compared to female/female pairs. This area is understudied and there is not sufficient evidence to suggest that fetal sex is a predictive marker of preeclampsia risk. The mechanisms that may explain our observed association have not been fully explored, especially in twin pregnancies. There is a need for additional research on preeclampsia in multifetal pregnancies more generally, particularly to understand preeclampsia subtypes within these populations. Studies should consider fetal sex when exploring biological mechanisms driving preeclampsia in multifetal pregnancies.

Novelty and Relevance.

1. What is New?

   1. We found that fetal sex pairs in twin pregnancies may influence risk of preeclampsia.
2. What is Relevant? (How the study related to hypertension)

   1. Preeclampsia is a maternal hypertensive disorder that increases risk of cardiovascular disease, but multifetal pregnancies are often excluded from studies, thus identifying risk factors in this population is important.
3. Clinical/Pathophysiological Implications

   1. Sexual dimorphism exists in preeclampsia risk among twin pregnancies, but mechanisms are unexplored.

Non-standard Abbreviations and Acronyms

HDP

Hypertensive disorders of pregnancy

RR

Relative risk

CI

Confidence interval

BMI

Body mass index

NH

Non-Hispanic