Epidemiology and Mechanisms of De Novo and Persistent Hypertension in the Postpartum PeriodCLINICAL PERSPECTIVE
Background—The pathophysiology of hypertension in the immediate postpartum period is unclear.
Methods and Results—We studied 988 consecutive women admitted to a tertiary medical center for cesarean section of a singleton pregnancy. The angiogenic factors soluble fms-like tyrosine kinase 1 and placental growth factor, both biomarkers associated with preeclampsia, were measured on antepartum blood samples. We then performed multivariable analyses to determine factors associated with the risk of developing postpartum hypertension. Of the 988 women, 184 women (18.6%) developed postpartum hypertension. Of the 184 women, 77 developed de novo hypertension in the postpartum period, and the remainder had a hypertensive disorder of pregnancy in the antepartum period. A higher body mass index and history of diabetes mellitus were associated with the development of postpartum hypertension. The antepartum ratio of soluble fms-like tyrosine kinase 1 to placental growth factor positively correlated with blood pressures in the postpartum period (highest postpartum systolic blood pressure [r=0.29, P<0.001] and diastolic blood pressure [r=0.28, P<0.001]). Moreover, the highest tertile of the antepartum ratio of soluble fms-like tyrosine kinase 1 to placental growth factor was independently associated with postpartum hypertension (de novo hypertensive group: odds ratio, 2.25; 95% confidence interval, 1.19–4.25; P=0.01; in the persistent hypertensive group: odds ratio, 2.61; 95% confidence interval, 1.12–6.05; P=0.02) in multivariable analysis. Women developing postpartum hypertension had longer hospitalizations than those who remained normotensive (6.5±3.5 versus 5.7±3.4 days; P<0.001).
Conclusions—Hypertension in the postpartum period is relatively common and is associated with prolonged hospitalization. Women with postpartum hypertension have clinical risk factors and an antepartum plasma angiogenic profile similar to those found in women with preeclampsia. These data suggest that women with postpartum hypertension may represent a group of women with subclinical or unresolved preeclampsia.
The hypertensive disorders of pregnancy are a well-recognized cause of maternal morbidity and mortality worldwide. The World Health Organization, in a systematic review of global data, recently identified hypertensive disorders of pregnancy as a dominant cause of maternal death in developing countries, accounting for ≈18% of all maternal deaths worldwide.1 Even among Western countries, characterized by very low rates of maternal death in the peripartum and postpartum periods, hypertensive disorders of pregnancy remain significantly associated with the future development of maternal hypertension, ischemic heart disease, and stroke.2
Editorial see p 1690
Clinical Perspective on p 1733
Development of de novo hypertension in the immediate postpartum period (postpartum hypertension [PPHTN]) is an infrequently studied clinical syndrome. PPHTN is usually defined as hypertension appearing after delivery through 6 weeks postpartum.3 PPHTN can present either as a de novo condition following a normotensive pregnancy or as persistent hypertension following a pregnancy complicated by severe preeclampsia.4 PPHTN following a normotensive pregnancy may herald the development of seizures (postpartum eclampsia), which is associated with significant morbidity. A recent large retrospective study indeed found that 63% of patients readmitted with delayed postpartum preeclampsia had no antecedent diagnosis of a hypertensive disorders of pregnancy.5 Furthermore, PPHTN increases the risk of intracerebral hemorrhage,6 which is compounded by generally decreased medical surveillance in the postpartum period after hospital discharge.
Despite the potential danger that PPHTN poses to women, only limited data exist in the literature describing its incidence and risk factors.7 This is due partly to infrequent medical visits for most women during the immediate postpartum period; often, the first medical visit occurs 6 weeks after delivery. Furthermore, the often asymptomatic nature of PPHTN limits capture of these patients for study. Most data in the literature are retrospective in nature and usually represent only the most severe cases such as patients readmitted with hypertensive crisis or seizures/eclampsia.8 Even less is known about the possible underlying pathobiology of PPHTN. Whereas iatrogenic causes such as excess intravenous fluid administration and certain medications (nonsteroidal anti-inflammatory drugs, ergot alkaloids) have been associated with PPHTN,4,9 very little is known about the pathophysiology of PPHTN.10,11
We therefore studied women with singleton pregnancies for antepartum risk factors that may lead to the development of PPHTN in the immediate postpartum period. In addition to collecting detailed antepartum demographic, clinical, and laboratory data, we measured the antepartum circulating levels of angiogenic factors—the antiangiogenic soluble fms-like tyrosine kinase 1 (sFlt1) and the proangiogenic placental growth factor (PlGF) —that have been linked to the pathogenesis of preeclampsia.12–17
Study Design and Oversight
This retrospective, clinical study was performed at Beth Israel Deaconess Medical Center (Boston, MA) in accordance with all institutional policies and with approval of the institutional review board. Because discarded blood samples were used for analysis, informed patient consent was not required for this study; the institutional review board waived the need for written informed consent from the participants for both the use of discarded blood samples and review of medical records.
Study Participants and Clinical Data
We retrospectively studied 988 women admitted for cesarean section (C-section) of a singleton pregnancy during the period of October 2012 through March 2014. We focused only on subjects who underwent C-section because they had routine blood draws before delivery. We recorded a number of antepartum characteristics such as maternal age, body mass index on admission, gestational age on admission, gestational age at delivery, parity, smoking status, race, history of any hypertensive disorder or diabetes mellitus in a previous pregnancy, highest systolic blood pressure (SBP) and diastolic blood pressure (DBP) during the antepartum period, and laboratory values such as alanine aminotransferase level, serum creatinine level, uric acid level, and platelet count. Each patient’s clinical course was recorded, with collection of additional clinical data such as infant birth weight, maternal estimated blood loss during delivery, sequential postpartum SBP and DBP, and intraoperative plus postpartum intravenous fluid therapy. The length of hospital admission, number of patients requiring readmission, and number of patients requiring antihypertensive medication were also recorded for all participants. PPHTN was defined as any SBP ≥140 mm Hg or DBP ≥90 mm Hg at least 48 hours after delivery and up to 6 weeks postpartum. Blood pressures were also recorded from the outpatient office visits, when available, up to 6 weeks postpartum.
Measurement of Antepartum Circulating Angiogenic Proteins
Discarded venous EDTA blood samples collected before C-section (between 12 and 96 hours before C-section) were obtained from the hospital laboratory, and plasma was divided into aliquots and stored at −70°C. Blood samples were missing in 7 subjects. A single operator performed quantitative sandwich ELISA for both sFlt1 and PlGF biomarkers on these discarded plasma samples from the remaining subjects (n=981) in duplicate with commercially available ELISA kits (R&D Systems, Inc, Minneapolis, MN), as described elsewhere.13 The interassay coefficients for sFlt1 and PlGF were 7% and 11%, respectively. The operator was blinded to clinical information. The clinicians treating the patients did not have any knowledge of the biomarker levels because the assays were in done in 1 batch after the outcomes had occurred.
Patient Cohort Classification
All subjects in our study were first classified into 2 groups: normotensive pregnancy (NL: SBP <140 mm Hg and DBP <90 mm Hg before delivery; n=774 patients) and hypertensive pregnancy (HTN: SBP ≥140 mm Hg or DBP ≥90 mm Hg once before delivery; n=214 patients). These 2 groups were then further categorized as either remaining/becoming normotensive (as defined above) or developing PPHTN (SBP ≥140 mm Hg or DBP ≥90 mm Hg) ≥48 hours after C-section. Therefore, the resultant 4 subgroups into which all subjects were categorized were NL/NL, NL/PPHTN, HTN/NL, and HTN/PPHTN.
Antepartum, delivery, and postpartum characteristics of normotensive and hypertensive patients are presented overall and by PPHTN status as mean±SD, medians (25th–75th percentiles), or numbers and percentages. Characteristics were compared between postpartum groups with independent-samples t tests or χ2 tests, as appropriate. Angiogenic factors were natural log transformed to meet the assumptions of parametric testing. Pearson correlation coefficients were used to determine the associations between natural log–transformed antepartum angiogenic factors and highest recorded antepartum and postpartum SBPs and DBPs. Separate univariate and multivariable logistic regression models were used to predict the risk of PPHTN by the overall tertile of each angiogenic factor among normotensive and hypertensive patients. Adjustment was made for variables significant in univariate models and for clinically known confounders: gestational age, body mass index, nulliparous, race, history of chronic hypertension, and history of diabetes mellitus. Odds ratios (ORs) and 95% confidence intervals (CIs) were used to summarize the results of logistic regression models. The average days of hospitalization were compared between the normotensive and hypertensive groups with independent-samples t tests (natural log–transformed average days of hospitalization) and χ2 tests. Statistical analyses were conducted with the use of SAS software version 9.4 (SAS Institute). Two-tailed values of P<0.05 were considered to indicate statistical significance.
Patient Characteristics Associated With PPHTN
Among 774 normotensive pregnancies, 9.9% of these women (n=77) went on to develop de novo PPHTN (Table 1). Black women (20.8% versus 7.0%), women with a higher body mass index at delivery (mean±SD, 34.1±7.3 versus 30.0±5.2 kg/m2), and women with a history of diabetes mellitus (13.0% versus 3.9%) demonstrated increased risk of developing de novo PPHTN compared with normotensive women who remained normotensive (all P<0.001). The highest recorded antepartum SBPs and DBPs, although still in the normotensive range, were significantly higher in those women developing de novo PPHTN (SBP, 127±8 versus 119±11 mm Hg; DBP, 78±7 versus 74±8 mm Hg; all P<0.001; Table 1).
Among 214 hypertensive pregnancies of any cause, 50.0% (n=107) remained hypertensive in the postpartum period. Similar to patients with de novo PPHTN, patients with persistent hypertension had a higher body mass index at delivery (mean±SD, 35.1±7.3 versus 32.3±5.3 kg/m2), history of hypertension (29.9% versus 3.7%), and history of diabetes mellitus (21.5% versus 5.6%) compared with women who were normotensive postpartum (all P<0.01). The highest recorded antepartum SBPs and DBPs were significantly higher in those women developing persistent PPHTN (SBP, 155±15 versus 141±9 mm Hg; DBP, 94±9 versus 89±7 mm Hg; all P<0.001) than among initially hypertensive women who became normotensive in the postpartum period (Table 2).
There was no difference in intraoperative and postoperative intravenous fluid volumes administered between women with PPHTN and those without. More than 90% of all patients received a nonsteroidal medication, and there was no difference in doses between these 2 groups. The majority of patients (62.0%) developed PPHTN between 48 and 72 hours after delivery; another 27.2% developed the disorder on day 4; and 5.4% developed PPHTN on day 5. The remaining patients developed PPHTN after the first week. Of 184 patients with PPHTN, 39 patients had severe hypertension (defined as SBP ≥160 mm Hg or DBP ≥110 mm Hg).
Angiogenic Factors and PPHTN
Women who developed de novo PPHTN had significantly higher antepartum sFlt1 levels compared with women who remained normotensive postpartum (median, 10 189 pg/mL [25th–75th percentiles, 5655–16650 pg/mL] versus 7721 pg/mL [25th–75th percentiles, 5088–12199 pg/mL]) and a higher sFlt1/PlGF ratio (52.1 [25th–75th percentiles, 22.0–84.6] versus 33.1 [25th–75th percentiles, 14.1–65.9]; all P<0.05; Table 1). Similar results were seen in women with persistent PPHTN compared with subjects who were normotensive postpartum among the initially hypertensive group (sFlt1/PlGF ratio, 93.9 [25th–75th percentiles, 47.3–199.3] versus 57.2 [25th–75th percentiles, 26.0–128.8]; P<0.001; Table 2).
Those in the highest tertile of antepartum circulating sFlt1 concentration (OR, 1.89; 95% CI, 1.07–3.31; P=0.03), lowest PlGF concentration (OR, 2.38; 95% CI, 1.26–4.49; P=0.008), and highest tertile of sFlt1/PlGF ratio (OR, 2.23; 95% CI, 1.24–4.04; P=0.008) were at a significantly increased risk of developing de novo PPHTN in univariate analyses, and these effects persisted after multivariable adjustment (highest sFlt1: OR, 2.29; 95% CI, 1.22–4.33; P=0.01; lowest PlGF: OR, 2.15; 95% CI, 1.10–4.20; P=0.02; and highest sFlt1/PlGF ratio: OR, 2.25; 95% CI, 1.19–4.25; P=0.01; Table 3). Similar results were seen among the initially hypertensive group. Multivariable analysis for the highest tertile of sFlt1/PlGF ratio revealed a significantly increased OR favoring the persistence of hypertension postpartum (OR, 2.61; 95% CI, 1.12–6.05; P=0.02; Table 4). Of the 988 patients, 263 patients had labor before the C-section. Excluding patients with labor before the C-section did not change the results (data not shown).
The antepartum ln (sFlt1) positively correlated with the highest postpartum SBP (r=0.26, P<0.001) and DBP (r=0.26, P<0.001); the ln (PlGF) negatively correlated with the highest postpartum SBP (r=−0.23, P<0.001) and DBP (r=−0.22, P<0.001); and the ln (sFlt1/PlGF ratio) positively correlated with highest postpartum SBP (r=0.29, P<0.001) and DBP (r=0.28, P<0.001). The predictive risk of PPHTN increased with rising levels of antepartum sFlt1/PlGF ratio among all women (Figure).
Women developing PPHTN also experienced a significantly longer hospitalization than those who did not (6.5±3.5 versus 5.7±3.4 days), and a higher proportion of women developing PPHTN were treated by antihypertensive medications (22.8% versus 0.5%; both P<0.001; Table 5). The sFlt1/PlGF ratio was also elevated among women who were treated with antihypertensive medications compared with those who were not (89.5 [95% CI, 37.5–165.6] versus 38.5 [95% CI, 17.1–80.4]; P<0.001; Table 6).
In this large, observational, clinical study at a single tertiary care hospital, we have demonstrated that the development of de novo PPHTN following normotensive pregnancies was common (9.9%) and that ≈50% of women with hypertensive pregnancies demonstrated persistence of hypertension in the postpartum period. The clinical risk factors for the eventual appearance of de novo PPHTN strongly resembled those for preeclampsia. Moreover, the antepartum angiogenic factor levels central to the pathogenesis of preeclampsia—the antiangiogenic sFlt1 and the proangiogenic PlGF—followed the same pattern (higher sFlt1, lower PlGF, and higher sFlt1/PlGF ratio) and independently predicted the development of de novo and persistent PPHTN. Among all affected patients in our study, the highest tertile of the sFlt1/PlGF ratio increased the odds of developing PPHTN (both de novo and persistent) in univariate and multivariate analyses. Finally, women developing PPHTN had longer hospital stays than those who were normotensive after delivery and required more antihypertensive medications.
Limited data describing the overall incidence and risk factors for PPHTN exist in the literature.7,18,19 Most studies have been retrospective,20 have not focused on PPHTN per se,5,21 or have had a small cohort.22 To the best of our knowledge, the present work represents one of the largest clinical studies designed to specifically study PPHTN in the literature. Our finding that the development of PPHTN was common, with an incidence of nearly 18% in our cohort, was quite notable. Considering the associated maternal morbidity (and potential mortality) of unrecognized (and therefore untreated) PPHTN, this higher-than-anticipated frequency suggests that the initial postpartum medical visit may need to occur sooner than is typical, particularly for those women with risk factors. From our data, these high-risk features appear to be identical to many of those of preeclampsia: elevated body mass index, history of diabetes mellitus, and antepartum blood pressure elevation. Taken together with prior observations, our data suggest that de novo or persistent hypertension can occur even after the placenta has been delivered. Therefore, healthcare providers should alert women to and continue to be vigilant for the signs and symptoms of this disorder and to manage them promptly to avoid serious complications such as pulmonary edema or eclampsia in the postpartum period.
Indeed, evaluation of antepartum circulating angiogenic factors revealed a pattern identical to that observed in preeclampsia in these women. These data therefore suggest that women who develop de novo PPHTN may actually represent a group of women with subclinical preeclampsia that manifests as hypertension postpartum. Similarly, patients with persistent PPHTN had higher antepartum sFlt1/PlGF ratios than those who were normotensive in the postpartum period, suggesting that persistent PPHTN may represent a subgroup of patients who either may have unresolved preeclampsia resulting from delayed clearance of antiangiogenic factor or may have a severe form of preeclampsia. Our findings also extend the biology of these factors, sFlt1 and PlGF, beyond preeclampsia and into another hypertensive disorder of pregnancy. Antepartum measurement of these angiogenic factors, in combination with a woman’s other clinical risk factors for PPHTN, may better enable clinicians in the future to determine which patients require closer surveillance after delivery and closer follow-up after hospital discharge.
Some limitations of our study require mention. First, our cohort was limited to women undergoing C-section only, which may introduce bias and limit the generalizability of our findings. However, this study design was chosen to increase the time period over which we were able to collect data on postpartum blood pressure measurements because women undergoing C-section are typically hospitalized longer than those delivering vaginally. In addition, all consecutive women with singleton pregnancies undergoing C-section were included in our study, regardless of the indication for C-section, to minimize this potential bias. Furthermore, we would have preferred to have access to longitudinal blood pressure measurements beyond 1 week after delivery in all women to determine the duration of PPHTN among the affected women in our cohort. However, ambulatory blood pressure monitoring is not routinely used to monitor for postpartum preeclampsia; hence, these data were not available. Furthermore, data were not available for all patients at the 6 weeks’ follow-up in the outpatient clinic because only a fraction of our cohort delivering at our center had their postpartum care there. Another limitation of our study is that not all subjects had blood drawn routinely for preeclampsia laboratory assessment during the antepartum period. This may have led to a lack of statistical significance for some of the routine clinical laboratory tests such as uric acid or serum creatinine. Finally, we did not find significant overall morbidity in our study other than a modest increase in the duration of hospitalization. Future studies should determine whether the antepartum angiogenic profile among high-risk patients correlates with postpartum morbidity such as postpartum eclampsia or postpartum HELLP (hemolysis-elevated liver enzymes-low platelets counts) syndrome. It would also be important to evaluate whether women with PPHTN are at risk for the development of cardiovascular disease in the long term.
We prospectively identified the major clinical risk factors for the development of PPHTN and for the first time implicated antepartum plasma levels of the angiogenic factors sFlt1 and PlGF in the pathogenesis of this disorder. Moreover, our description of the striking clinical similarities between the factors predicting the development of de novo PPHTN and those predicting preeclampsia, coupled with the similarity of the angiogenic profile that precedes de novo PPHTN and accompanies preeclampsia, suggests that the development of de novo PPHTN may in fact represent subclinical preeclampsia.
Sources of Funding
Dr Thadhani is supported by a K24 award from National Institute of Diabetes and Digestive and Kidney Diseases. Dr Karumanchi is an investigator of the Howard Hughes Medical Institute. Dr Rana is supported by a KO8 award from the Eunice Kennedy Shriver National Institute of Child Health and Human Development. Dr Seely is supported by an R56 award from the National Heart, Lung, and Blood Institute.
Drs Thadhani and Karumanchi are coinventors on patents related to preeclampsia biomarkers that are held at Massachusetts General Hospital and Beth Israel Deaconess Medical Center. They have financial interest in Aggamin LLC. Drs Thadhani, Karumanchi, and Rana report serving as consultants to Roche Diagnostics. Dr Karumanchi reports serving as a consultant to Siemens Diagnostics and has received research funding from Thermofisher. The other authors report no conflicts.
Guest Editor for this article was Suzanne Oparil, MD.
- Received January 26, 2015.
- Accepted August 6, 2015.
- © 2015 American Heart Association, Inc.
- Bellamy L,
- Casas JP,
- Hingorani AD,
- Williams DJ
- Maynard SE,
- Min JY,
- Merchan J,
- Lim KH,
- Li J,
- Mondal S,
- Libermann TA,
- Morgan JP,
- Sellke FW,
- Stillman IE,
- Epstein FH,
- Sukhatme VP,
- Karumanchi SA
- Rana S,
- Powe CE,
- Salahuddin S,
- Verlohren S,
- Perschel FH,
- Levine RJ,
- Lim KH,
- Wenger JB,
- Thadhani R,
- Karumanchi SA
- Romero R,
- Nien JK,
- Espinoza J,
- Todem D,
- Fu W,
- Chung H,
- Kusanovic JP,
- Gotsch F,
- Erez O,
- Mazaki-Tovi S,
- Gomez R,
- Edwin S,
- Chaiworapongsa T,
- Levine RJ,
- Karumanchi SA
- Noori M,
- Donald AE,
- Angelakopoulou A,
- Hingorani AD,
- Williams DJ
Postpartum hypertension occurring ≥48 hours after delivery is a relatively common complication that is associated with significant morbidity to the mother. However, there are limited data describing the epidemiology and pathogenesis of this disorder. In this single-center clinical study of 988 women who underwent delivery by cesarean section, we demonstrate that postpartum hypertension is a relatively common complication (≈18%). Women with postpartum hypertension use more healthcare resources, as evidenced by prolonged hospitalization and greater use of antihypertensive drugs. We further demonstrate that women with postpartum hypertension have the same clinical risk factors and antepartum circulating plasma angiogenic profile as women with preeclampsia. We conclude that women with postpartum hypertension may represent a group of women with subclinical preeclampsia or unresolved preeclampsia. Future efforts should be directed at early identification of these patients and evaluation of interventions to decrease morbidity associated with this condition.