Coronavirus disease 2019 (COVID-19) and the risk of hypertensive disorders of pregnancy: a retrospective cohort study

ABSTRACT Objective: To evaluate of COVID-19 disease in pregnant women and its association with hypertensive disorders of pregnancy. Design: Retrospective Cohort Study Setting: Multicenter study from a large metropolitan hospital system Methods: Patients who tested positive for COVID-19 during their pregnancy and delivered were compared to the three subsequent deliveries of patients who tested negative (controls). We evaluated the impact of COVID-19 on the development of hypertensive disorders of pregnancy. Results: Compared with pregnancies negative for SARs-CoV-2 infection, maternal SARs-CoV-2 infection was associated with an increased risk for hypertensive disorders of pregnancy (OR 3.68, 95% CI 1.67 − 8.10). Tweetable Abstract Patients who test positive for COVID-19 during their pregnancy are at increased risk of developing a hypertensive disorder of pregnancy. Earlier SARs-CoV-2 infection results in an increased risk of developing a hypertensive disorder.


Introduction
In December 2019, reports from China described a cluster of cases of pneumonia that were believed to be linked to the Huanan Seafood Wholesale Market in Wuhan, Hubei Province of China (1). By early January of 2020, Chinese health authorities confirmed that the outbreak was caused by a novel coronavirus that differed from both severe acute respiratory syndrome coronavirus (SARs-CoV) and Middle East respiratory syndrome coronavirus (MERS-CoV) (2). The virus would subsequently be named severe acute respiratory syndrome coronavirus-2 (SARs-CoV-2), the agent responsible for the human coronavirus disease-2019 . In March of 2021, COVID-19 cases surpassed 126 million globally and have been responsible for nearly 2.8 million deaths (3). Of the 223 countries with confirmed cases, the United States and its territories has become the epicenter for COVID-19, accounting for approximately one-fourth of cases globally.
Early reports of COVID-19 in pregnancy suggested an association between SARs-CoV-2 infection and increased risk of hypertensive disorders of pregnancy (HDP), specifically gestational hypertension, preeclampsia, and preeclampsia with severe features (Appendix A). These reports, however, were mainly confined to case reports (4)(5)(6) or systematic reviews that incorporated multiple coronavirus species, namely SARs-CoV and MERS-CoV in addition to SARs-CoV-2 (7,8). Although there have more recent publications of pregnancy outcomes in patients diagnosed with COVID-19 (9), including systematic reviews (10)(11)(12), few reports have tested the association between SARs-CoV-2 infection and HDP. Furthermore, most systematic reviews have relied on large case series (13). In studies that have examined relationship between COVID-19 and HDP, outcomes were based on composite indices (14,15), were part of large cohort studies not powered for the outcome (10,16,17), or lacked outcome data on a comparator group of individuals who tested negative for SARs-CoV-2 (18). In addition, in systematic reviews without an appropriate control group, assumptions regarding associations between SARs-CoV-2 infections and hypertensive disorders of pregnancy are plagued by confounding and biases.
In this historical cohort study, we compared pregnant women who tested positive for SARs-CoV-2 within the course of their pregnancy to pregnant women who tested negative. The primary objective of this study was to compare the incidence of HDP between the two groups. We hypothesized that there would be higher rates of HDP in pregnancies complicated by COVID-19. Secondary outcomes included pregnancy outcomes, maternal complications, and neonatal outcomes.

Methods
This is a historical cohort study of all pregnant patients who tested positive for SARs-CoV-2 by molecular or polymerase chain reaction (PCR) based tests at four participating community hospitals located in a large metropolitan city and who delivered between 1 February and 24 November 2020. Patients with positive antibody tests alone were excluded from the study. Index cases were defined as any patient who tested positive at any point during their pregnancy, regardless of symptoms, and who delivered within the study period. Undelivered patients who tested positive for SARs-CoV-2 were not included. This study was approved by the Ascension St. John and Ascension Providence Institutional Review Boards (IRB) (Reference #1,617,953). No funds were received for this work.
Index COVID-19 cases were compared to the subsequent three deliveries (controls) of patients who were asymptomatic or tested negative. If any of the three subsequent deliveries following an index case tested positive for SARs-CoV-2, that patient was treated as a new index case and the next negative case was used as a control. The subsequent three SARs-CoV-2 negative deliveries were then matched to the new index case. Ultimately, a 3:1 ratio of controls-to-index cases was achieved.
Given the severe shortage of test kits in the initial stage of the study, only symptomatic patients and patients undergoing a cesarean delivery were tested for SARs-CoV-2. Starting in July of 2020, participating hospitals implemented universal SARS-CoV-2 testing for all patients admitted to the labor and delivery units. Thus, 32.1% of patients included in the control group were asymptomatic patients without suspicion for COVID-19.
Descriptive data were collected on index cases and controls. Data were abstracted from the electronic medical record (EMR) common to the four participating institutions. Abstracted data included patient demographics, medical histories and comorbid conditions, social history and recreational drug/substance use, gravidity, parity, obstetrical history, order of pregnancy (e.g. singleton v. higher-order pregnancies), past pregnancy complications, current pregnancy complications, body mass index (BMI), COVID-19 self-reported symptoms, clinical signs of COVID-19, vital signs, laboratory values and inflammatory markers, imaging results, severity of disease, treatments, interventions and maternal and fetal outcomes. Race and ethnicity were self-reported at the time of admission. Gestational age (GA) was calculated based on the best obstetric estimate of the expected date of delivery (earliest ultrasound or sure last menstrual period). GA at COVID-19 diagnosis was determined by the date of the first positive SARs-CoV-2 test. Due to the limited imaging in the control group, analysis of imaging findings was omitted the from final analysis.
Maternal outcomes included gestational hypertension, preeclampsia without severe features, preeclampsia with severe features, eclampsia, hemolysis, elevated liver enzymes, and low platelets (HELLP) syndrome, gestational diabetes, intrauterine growth restriction, preterm labor, preterm birth, prelabor rupture of the membranes (PROM), preterm prelabor rupture of the membranes (PPROM), cervical insufficiency, venous thromboembolism, and placental abruption. HDP were define according to the American College of Obstetricians and Gynecologist criteria for gestational hypertension, preeclampsia with and without severe features, eclampsia, chronic hypertension with superimposed preeclampsia and HELLP syndrome (Appendix A). Neonatal outcomes were limited to the need for neonatal intensive care, neonatal demise, need for higher acuity care (i.e. special care nursery), or no intervention.
Data collection was coordinated by the principal investigator at each participating site. Data quality checks were done on a continuous basis, with reabstraction of data randomly assigned to cases across all sites. Demographic characteristics of index cases and controls were compared using descriptive statistics, including frequencies, mean and standard deviation for normally distributed continuous variables and median and interquartile range for nonnormal continuous data. Race and ethnicity were based on the National Institute of health (NIH) categories. All patients who identified as Hispanic ethnic were grouped as "Hispanic." For analysis purposes, a "race reduced" or "all other" category included Native Hawaiian, Other Pacific Islander, American Indian or Alaskan Native, or self-reported other. These categories were combined to enable comparisons between groups, as these groups had lower frequencies of patients.
Univariable analysis was conducted with Student's t-test, the chi-squared test, and analysis of variance (ANOVA). Multivariable analyses was done using logistic regression. From the univariable analysis, all variables associated with HDP or with COVID-19 status, with a p-value of 0.09 or less, were entered into a multivariable logistic regression equation using a forward likelihood ratio algorithm. The model with lowest −2 log likelihood value was taken to be the best model for the data. Logistic regression results are reported as the odds ratio (OR) and 95% confidence interval. All data were analyzed using SPSS v. 27.0 and a p-value less than 0.05 was considered to indicate statistical significance.

Results
Our primary analysis included 5,962 patients who delivered at four participating hospitals during the study period. Of these, 70 patients (who delivered) tested positive for SARs-CoV-2 during their pregnancy and 210 patients who were asymptomatic (prior to universal testing) or had a negative SARs-CoV-2 test were matched to the index cases. Of the 70 patients who tested positive for SARs-CoV-2, 22 (32.4%) were diagnosed with a HDP. Of the 210 patients with negative tests or who were not tested but asymptomatic, 38 (13.3%) were diagnosed with HDP ( Figure S1).
Baseline demographics were significantly different for race and social history. Patients who tested positive for COVID-19 were more likely to be Black/African American and have a history of recreational drug use during pregnancy (Table 1). Patients who tested positive for COVID-19 also delivered at a significantly earlier mean gestational age than patients who tested negative, 38w0d ± 2.4 vs. 38w5d ± 1.8, respectively (p = 0.05). Gravidity, parity, number of prior vaginal and Cesarean sections, medical comorbidities, and alcohol use and/or tobacco use, did not significantly differ between groups. Overall, patients who reported a previous pregnancy complication were more likely to have a negative COVID-19 test. Except for HDP, there were no significant differences in pregnancy complications between index cases and controls (Table S1).
Patients with COVID-19 diagnosed within their pregnancy were significantly more likely to be diagnosed with a hypertensive disorder of pregnancy (p< 0.01; Table 2). Overall rates of gestational hypertension (5.7% vs 3.8%, p = 0.500) and preeclampsia without severe features (7.1% vs 2.4%, p = 0.061) were not statistically significant, while preeclampsia with severe features (18.6% vs 7.1%, p = 0.006) was significantly higher in patients with a known COVID-19 diagnosis in pregnancy (Table S2). One patient had HELLP syndrome and one patient had eclampsia. Both were COVID-19 negative. The effect of SARs-CoV-2 infection on overall rate of HDP was modified by BMI, parity, preeclampsia in a prior pregnancy, comorbid chronic hypertension, asthma, prior cesarean section, recreational drug use during pregnancy, and race (Table S3). Laboratory values and inflammatory markers were not included in the analysis secondary to lack of a standard laboratory order set across institutions. In adjusted analyses (  (Figure 1).

Discussion
We found that SARs-CoV-2 infection in pregnancy is associated with significantly higher incidence of HDP. After controlling for other variables, patients with SARs-CoV-2 infections were nearly 4 times more likely to have a diagnosis of HDP compared to patients without a history of SARs-CoV-2 infection in pregnancy. SARs-CoV-2 infection was the third highest contributor to HDP diagnoses, only behind previous history of preeclampsia and comorbid chronic hypertension ( Figure 1). Overall, 31.4% of patients with a COVID-19 during their pregnancy developed a HDP, compared to 13.3% of controls (19)(20)(21). Our results are consistent with the available literature (16,22). While one meta-analysis (10) did not demonstrate this association, the studies included had heterogenous designs, different primary outcomes, and most lacked a control group. The one study within the metaanalysis with matched controls found a higher incidence of preeclampsia in pregnant patients with COVID-19 (   gestational hypertension of 9.0% and 3.6% (p < 0.001), preeclampsia without severe features of 3.6% and 1.4% (p = 0.034), and preeclampsia with severe features 5.4% and 3.6% (p = 0.12) for patients with and without COVID-19, respectively (24). They did not report an aggregate risk, however, and they did not control for other variables. Metz et al. (2021) found that for patients with COVID-19, HDP was the most common indication among preterm births. In comparison to the COVID negative cohort, in our study patients diagnosed with COVID- 19 during pregnancy delivered at a significantly earlier gestational age, however, both groups largely delivered at term. In contrast to our study, Metz et al. found a relative risk of 1.61 (95% CI 1.18-2.20) for severe COVID-19, with rates of HDP increasing with the severity of COVID-19 manifestation. However, the impact of confounders was not ascertained. In our study, patients with a positive SARs-CoV-2 test and those who were asymptomatic had the highest likelihood of developing HDP. This is an important finding as patients with positive screening test may be at a higher risk of developing HDP, even if asymptomatic.
While COVID-19 conferred a significant increased likelihood of having a diagnosis of HDP in our study, a diagnosis of chronic hypertension and having preeclampsia in a previous pregnancy conferred the greatest risk for HDP. However, COVID-19 remained a significant risk factor for HDP after controlling for these confounders.
Patients with COVID-19 were more likely to have lower absolute lymphocyte count and elevated CRP and uric acid. There was a trend toward a higher ferritin level. However, it should be noted that laboratory values were not analyzed as a predictor of HDP in our study given the relatively small proportion of control patients with inflammatory markers and the lack of a standard COVID-19 laboratory order-set across institutions. However, these results are consistent with the inflammatory state associated with COVID-19. Inflammation has been shown to play a significant role in the pathogenesis of HDP (25)(26)(27). Furthermore, HDP has been associated with alterations in proinflammatory cytokines (25,26,28). The inflammatory response is a central clinical characteristic of SARs-CoV infection, with well-established extrapulmonary systemic hyperinflammation secondary to a proinflammatory state driven by the release of host cytokines (29)(30)(31)(32)(33). Thus, it is likely that the acute inflammatory state, even if subclinical, may provoke or exacerbate HDP.
It has been reported that COVID-19 modulates angiotensin-converting enzyme 2 (ACE-2) within the placenta, which may provide one component in the pathogenesis of HDP in pregnancies complicated by SARs-CoV-2 infection (34). ACE-2 has been shown to be heavily expressed in the placenta, including cells involved in decidualization, trophoblast invasion, vascular remodeling, and placental blood flow (35). ACE2 is upregulated in pregnancy, and alterations in the levels of angiotensin II (vasoconstrictor) and angiotensin-(1-7) refers to the active heptapeptide of angiotensin and the "1-7" (vasodilator) caused by SARs-CoV-2 can lead to vasoconstriction, inflammation and procoagulopathic effects seen in HDP (36). Furthermore, Mendoza et al (2020) showed abnormalities in angiogenic factors related to COVID-19 infection that resulted in increased risk of a preeclampsia-like syndrome (37). Finally, it has been shown that placentas of women infected with SARs-CoV-2 have higher rates of decidual arteriopathy when compared to controls (38). Together, these reports provide compelling evidence for the pathogenesis behind the observed increased risk of HDP in pregnant patients with COVID-19.
The timing of SARs-CoV-2 infection appears to be important. Early infection (before 32 weeks' gestation) was associated with a significantly increased risk of HDP compared to negative controls. Infections occurring after 32 weeks' gestation showed a trend toward increased risk of HDP, however was not statistically significant. Our study was not powered for this outcome and it is likely that a larger sample would have generated a significant increased risk for late SARs-CoV-2 infection compared to COVID-19 negative controls. When comparing only COVID-19 positive pregnancies, early infection conferred a significantly higher risk for HDP than late infection. This observation is consistent with the inflammatory explanation of increased HDP risk in COVID-19. Through modulation of angiogenic factors and inflammatory cytokines, it is likely that COVID-19 exerts maximal impact on placental physiology at earlier gestations, enabling these physiologic changes to manifest as HDP over time.
Our study has several strengths. Most studies to date that have reported increased risk of HDP associated with COVID-19 have lacked a control group. Our study incorporated a control group of unaffected patients. We compared COVID-19 cases to three controls that delivered in temporal proximity to the index case, thus eliminating inherent temporal biases as a result of fluctuations in the prevalence of COVID-19 cases. Furthermore, a large control group allowed us to conduct detailed statistical analyses to elucidate an association between SARs-CoV-2 infections and HDP. Other studies to date have lacked a larger control group.
The main limitation of this study is the retrospective design. With any retrospective study, there is potential misclassification based on nonreported or incorrectly recorded ICD-10 codes. Additionally, given the scarcity of SARs-CoV-2 diagnostic tests early in the study period, approximately one-third of patients did not have a diagnostic test. This increases the potential for misclassification (e.g. including an asymptomatic COVID-19 positive case in the control group). However, we believe that this would potentially decrease the strength of association of COVID-19 and HDP, therefore the association found in this study may be an underestimate. There is also the possibility that patients could have tested positive at an outside institution during their pregnancy but did not disclose their positive test. Additionally, our study was underpowered for analysis of important subcategories, such as the effects of inflammatory markers on HDP within COVID-19 positive cases. It is recognized that there are overlapping signs (e.g. transaminitis) and symptoms (e.g headache) associated with COVID-19 and HDP. Accordingly, clinicians should take these overlapping features into account when making clinical decisions regarding HDP in COVID-19 positive pregnant patients.

Conclusion
The results from our study suggest that pregnant patients with a history of COVID-19 disease in pregnancy are at increased risk for developing a HDP, regardless of symptomatology. HDP has been reported to account for a significant increase in iatrogenic preterm birth in patients with COVID-19 (18). Clinicians should be aware of this association and closely monitor patients with a history of SARs-CoV-2 infection for potential complications associated with HDP. Collectively, we should aim to develop strategies to mitigate iatrogenic complications related HDP in pregnant patients with COVID-19.

Acknowledgments
This study is dedicated to the late Dr. David M. Zalenski who was integral to the COVID-19 response as the OB/GYN department chair at Ascension St. John Hospital. In addition to practicing obstetrics and gynecology, Dr. Zalenski was an educator and leader for the St. John OB/Gyn residency program. His kind presence is sorely missed.

Disclosure statement
No potential conflict of interest was reported by the authors.

Author Contributions
MB conceived of the idea, designed the protocol and wrote the manuscript. MB, FA, MT, KG, and ML curated the data. SS verified the protocol was methodologically sound and analyzed the data. MA was integral in the design and execution of the project. All authors discussed the final results and contributed to the final manuscript.

Ethics Approval
This study was approved by the Ascension St. John and Ascension Providence Institutional Review Boards (IRB) on 06/12/2020 (Reference #1,617,953).