ANGIOGENIC BIOMARKERS AND BLOOD PRESSURE IN PREGNANCY: WHAT CAN THIS COMBINATION REVEAL ABOUT PREGNANCY OUTCOMES AND HYPERTENSION LATER IN LIFE? By Megan Eagle A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Epidemiology – Doctor of Philosophy 2023 ABSTRACT Background: Maternal blood Pressure (BP) is closely monitored in pregnancy to identify the development of hypertensive disorders of pregnancy (HDPs), i.e. gestational hypertension (GH) and preeclampsia (PE). In addition to their association with increased risk for adverse perinatal outcomes, these hypertensive disorders in pregnancy are also been associated with increased risk of developing hypertension later in life. In 2017, the American Association of Cardiology/American Heart Association (ACC-AHA) introduced new BP criteria for non-pregnant populations, lowering the level at which elevated BP and hypertension are diagnosed, yet the previous cut off of 140/90 is still used during pregnancy. Imbalances of angiogenic biomarkers (elevated levels of soluble endoglin (sEng) and soluble Fms-like tyrosine kinase-1(sFlt-1) and lower levels of placental growth factor (Plgf)) in maternal blood are associated with PE. Less is known about associations between angiogenic biomarkers during pregnancy in pregnant individuals with elevated BP lower than 140/90. If angiogenic and anti-angiogenic biomarkers are correlated with BP in these ranges, they could provide clues to pregnancy health and to maternal vessel health later in life for these groups as well. Aims: This dissertation first explores associations between angiogenic biomarkers (Plgf, sFlt-1 and endoglin) and MAP (mean arterial pressure: the average pressure throughout one cardiac cycle) during pregnancy (Aim 1). It then examines whether the combination of MAP and angiogenic biomarkers are associated with adverse pregnancy outcomes, such as preterm or small for gestational age birth (Aim 2). The final aim of the dissertation explores associations between angiogenic biomarkers during pregnancy and the development of later life hypertension among women with normal BP, moderately elevated BP and hypertensive disorders during pregnancy (Aim 3). Methods: Secondary analysis of data from the POUCH (Pregnancy Outcomes and Community Health) Study which enrolled pregnant people at 16-27 weeks gestation followed them throughout pregnancy, and from the POUCHmoms follow up study 7-15 years later. First, multivariable logistic regression models are used to estimate the association between each angiogenic biomarker having MAP in the top quartile close to 20 weeks gestation, and at the time of the highest BP in pregnancy (both for the whole sample, and only those who do not have PE, GH or cHtn). Then associations between combinations of MAP and angiogenic biomarker levels and size for gestational age, as well as length of gestation are assessed. Finally associations between angiogenic biomarker levels in pregnancy and stage of hypertension 7-15 years later are estimated, and then associations between groups based on blood pressure category in pregnancy combined with biomarker level are used in polytomous regression models. Findings: Lower levels of Plgf at mid-pregnancy in this study are associated with having higher (MAP), even among participants without pre-eclampsia or other hypertensive disorders. Of the three biomarkers, ‘low’ Plgf levels associated with SGA and ‘high’ endoglin levels were associated with shorter length of gestation. Pregnancies with MAP ‘high’ and Plgf ‘low’ had greater than twice the odds of delivering an SGA infant. Given the lack of any association between sFlt-1 and pregnancy BP in the first two studies, we only examined Plgf and endoglin and found in our study they do not help determine which individuals with moderately elevated BP during pregnancy are at greatest risk of hypertension 7-15 years later. This dissertation is dedicated to my daughter Paola for inspiring me and bringing me joy. And to my family, for everything. iv ACKNOWLEDGEMENTS I would like to recognize my advisor, Dr. Claudia Holzman, the POUCH study team and the study participants for creating such a wealth of information in the interest of improving the health of future generations. I would also like to thank the members of my dissertation committee for bringing such breadth of expertise to our discussions: Dr. Claudia Holzman, DVM, MPH, PhD, Dr. Stephanie Watts, PhD, Dr. Lindsay Admon MD, MSc, Dr. Ahnalee Brincks, PhD, and Dr. Nicole Talge, PhD. v TABLE OF CONTENTS CHAPTER 1: INTRODUCTION……………………………………………………………………………………………………….1 REFERENCES………………………………………………………………………………………………………………….12 CHAPTER 2: ASSOCIATIONS BETWEEN ANGIOGENIC BIOMARKERS AND MEAN ARTERIAL PRESSURE IN PREGNANCY…………………………………………………………………………………………………………19 REFERENCES…………………………………………………………………………………………………………….......35 APPENDIX………………………………………………………………………………………………………………………39 CHAPTER 3: ELEVATED MEAN ARTERIAL PRESSURE AND PREGNANCY OUTCOMES, WHAT CAN ANGIOGENIC BIOMARKERS TELL US?.……………………………………………………………………………………….46 REFERENCES………………………………………………………………………………………………………………….61 APPENDIX………………………………………………………………………………………………………………........65 CHAPTER 4: PREGNANCY BLOOD PRESSURE, ANGIOGENIC BIOMARKERS AND HYPERTENSION 7-15 YEARS LATER……………………………………………………………………………………………………………………..69 REFERENCES………………………………………………………………………………………………………………….84 APPENDIX………………………………..…………………………………………………………………………………...88 CHAPTER 5: CONCLUSIONS……………………………………………………………………………………………………….95 REFERENCES……..…………………………………………………………………………………………………………100 vi CHAPTER 1: INTRODUCTION 1 Background Adverse pregnancy outcomes are important public health concerns worldwide, due to their impact on maternal and child health. In the United States, in 2020, over 10% of births were preterm, and there were notable disparities based on race and on socioeconomic status(1). Hypertensive disorders of pregnancy (gestational hypertension (GH), preeclampsia (PE) and related vascular disorders) along with chronic hypertension (cHTN) affect up to 10% of pregnancies worldwide(2). In addition to being associated with maternal morbidity and mortality, these disorders also are associated with adverse pregnancy outcomes, such as preterm birth and poor fetal growth. Impaired maternal vascular health and disturbances in the blood supply to the placenta and fetus are important causes of adverse pregnancy outcomes. Monitoring pregnancy vascular health early in pregnancy through non-invasive methods has the potential to aid early detection of pregnancies at high risk for adverse maternal and fetal outcomes, allowing for better surveillance, management and delivery at a suitable setting(3). An important component of prenatal care is monitoring maternal blood pressure (BP) to help identify both GH (defined as systolic BP greater than 140 mmHg and or diastolic BP greater than 90 mmHg first noted after 20 weeks gestation) and PE (which includes hypertension as previously defined in addition to proteinuria or other signs of target organ involvement)(4). Angiogenic biomarkers related to vascular development in the placenta and detectable in maternal blood, such as placental linked growth factor (Plgf), soluble fms-like tyrosine kinase-1 (sFlt-1) and soluble endoglin (sEng) have been associated with PE and accompanying adverse outcomes(5,6). 2 Recently, clinicians and researchers have proposed that the current clinical guidelines for diagnosing GH and PE may be missing some women who have higher risk of vascular-related adverse pregnancy outcomes but only have moderately elevated BP(7). Vascular placental abnormalities have been linked not only to PE but also to complications such as fetal growth restriction, placental abruption, preterm birth, stillbirth and with elevated risk for adverse pregnancy outcomes in a subsequent pregnancy(8). We hypothesize that levels of angiogenic markers are indicative of placental vascular problems and also may be associated with moderately elevated BP and be useful in determining which women with moderately elevated BP are at higher risk of adverse pregnancy outcomes. In addition to their association with adverse outcomes in pregnancy, both GH and PE have shown a strong association with elevated risk of Cardiovascular Disease (CVD) after pregnancy(9,10) and are also associated with higher odds of being hypertensive at midlife(11–13). Current clinical guidelines discuss the importance of postpartum cardiovascular risk factor assessment, counseling, and follow up for those with GH or PE, although there is still some inconsistency between guidelines (14–16). We and others have shown that even women with moderately elevated BP in pregnancy are at increased risk of higher BP later in life (17,18). We hypothesize that the same PE-related maternal blood angiogenic makers measured in pregnancy may be useful in determining which women with moderately elevated BP are at higher risk of later life elevations in BP and therefore merit closer follow-up. Definitions of Hypertensive Disorders in Pregnancy and in Non-pregnant Adults As noted above, the American College of Obstetricians and Gynecologists (ACOG) defines HDPs, and the diagnosis of chronic hypertension during pregnancy, using cut-off BP of 140 mmHg 3 systolic and/or 90 mmHg diastolic(4). Beginning in 2017, the American College of Cardiology/American Heart Association (ACC/AHA) revised their guidelines for non-pregnant populations, defining stage 1 hypertension as a systolic BP of 130-139 and or/ a diastolic BP of 80-89 mmHg, and elevated BP as systolic greater than 120 mmHg(19). If clinicians applied the ACC/AHA cutoffs to pregnant populations, it would greatly increase the number of people defined as hypertensive during pregnancy. One study found that changing the classification increased prevalence of hypertension in pregnancy from 10% to 28%(20), and another found 60% of participants in a community based cohort had either elevated BP or stage 1 hypertension during pregnancy(17). Since 2017, there has been considerable interest in determining whether or not BP elevations below the current 140/90 threshold are associated with adverse pregnancy outcomes, and several large cohort studies have found that stage 1 hypertension during pregnancy is associated with increased risk of preterm birth, delivering a small for gestational age infant, and developing PE (21–23). Mean Arterial Pressure While clinical guidelines typically provide separate diastolic (the pressure against the vessel walls when the heart is at rest) and systolic (pressure against the vessel walls when the heart contracts) cutoffs to define BP thresholds, there is another approach to characterizing BP that is used in both research and clinical practice. This alternative considers the average BP over the course of a single cardiac cycle (Mean Arterial Pressure (MAP)) which can be measured directly by invasive techniques or using newer automated BP cuffs, or estimated as diastolic BP x 2 plus systolic BP divided by 3(24). In epidemiological research, potential confounders of relationships between BP and health outcomes may be related to either systolic or diastolic BP(25). MAP 4 provides a single continuous measure that incorporates both components of BP. Outside of pregnancy, MAP is often used to ensure that perfusion to vital organs is adequate in septic shock, and elevated MAP been correlated with cardiovascular and cerebrovascular disease(26,27). When measured in the first and second trimester of pregnancy, MAP may be superior to systolic BP or diastolic BP alone in predicting PE (28,29). At BPs below 140/90 the risk of developing PE may increase with increasing levels of MAP. One recent study found that the risk of PE increases exponentially beginning at MAP of 85-95 mmHg (a level that would overlap the AHA/ACC category of elevated blood pressure) in mid pregnancy, and the risk of small for gestational age (SGA) and preterm birth (PTB) increases beginning at MAP of 90-100 mmHg (a level which corresponds to state 1 hypertension)(7). Angiogenic biomarkers in pregnancy Three angiogenic biomarkers measured in maternal blood, Plgf, sFlt-1 and sEng, have been investigated often as potential indicators of vascular health in pregnancy (6,30–32), During pregnancy, the placenta is the main source of all three biomarkers, and all three enter maternal circulation; however, they differ in their role in angiogenesis and patterns of change throughout pregnancy. In normal placentation, trophoblasts deeply penetrate the uterine endometrium and myometrium, and invade the maternal spiral arteries, changing these arteries to decrease resistance and increase blood flow to the placenta. In pregnancies that go on to develop PE, the trophoblast invasion is shallow, leading to inadequate spiral artery remodeling and causing placental hypoxia and stress. This hypoxia and stress then leads to changes in the expression of angiogenic and antiangiogenic factors in the placenta(6). Within the placenta, Plgf and sEng, 5 promote new blood vessel growth and stabilize vascular networks, while sFlt-1 binds to Plgf and vascular endothelial growth factor (VEGF)and acts as an antagonist, limiting the growth of new vessels(33). The placenta secretes these factors into maternal circulation, where they act together with hormonal changes in pregnancy to aid in maternal adaptation to the increased cardiac output and blood volume during pregnancy. Imbalances in these factors in maternal circulation has a negative impact on the endothelium. VEGF and Plgf play a role in maintaining the integrity of the maternal endothelium, particularly in the brain, kidney and liver (organs whose dysfunction are hallmarks of PE), and this important role is limited when sFlt1 binds to the free forms of these factors. Elevated levels of both sFlt1 and sEng impede vasodilation, increasing peripheral vascular resistance and raising maternal BP(33). PE is associated with lower maternal blood levels of Plgf (as early as 11-13 weeks gestation), higher levels of endoglin (usually beginning in the second trimester) and higher sFlt-1 levels (most noticeable close to delivery)(34,35). Angiogenic biomarkers and pregnancy outcomes Preeclampsia and gestational hypertension Plgf, sFlt1 and sEng have been studied extensively in the context of PE. Evidence for their role in the development of PE comes from the analysis of their expression in the human placenta and from animal models, as well as numerous case control and cohort studies examining associations between these maternal biomarkers and PE(5,36,37). The use of Plgf, sFlt-1 and sEng are beginning to be used in predictive models and in clinical practice to evaluate risk of PE among women already deemed at higher risk (38–40), thereby informing decisions around preventive use of aspirin and frequency of fetal monitoring. Lower levels of Plgf and higher 6 levels of SFlt-1 have been associated with GH in several studies, though levels are not as altered as they are in PE (41–43). The evidence for an association between endoglin levels and GH remains mixed(5,30). Even absent a diagnosis of hypertension in pregnancy, these biomarkers have been associated with vascular placental pathology and disorders related to abnormal placentation such as fetal growth restriction or small for gestational age (44–46) and as indicators of one of the pathways that can lead to preterm birth(47). Small for Gestational Age (SGA) and fetal growth restriction (FGR) One potential complication of hypertension in pregnancy (PE, GH and cHtn) is fetal growth restriction ((FGR)-also referred to as Intrauterine growth restriction or IUGR))(48). Diagnosis of FGR is based on ultrasound estimated fetal weight less than the 10 th percentile for gestational age(49); however, ultrasound based estimates of fetal growth are most often used in clinical practice only when IUGR is already suspected. Therefore, in epidemiologic research, SGA (birth weight below the 10th percentile for gestational age) is used as an outcome instead since birthweight and gestational age are available in cohort studies, prenatal records, and vital statistics. SGA calculations should use population specific references which ideally should be corrected for any implausible gestational age estimates (50). Lower Plgf and higher sEng have been associated with SGA in several studies, even in the absence of PE(3,44,46). The altered ratios between sFlt-1 and Plgf in maternal serum that have been noted in studies of both PE and SGA/FGR may not be reflective of the same imbalances in the placenta, with at least one recent study finding that in PE it is placental sFlt1 that is elevated, while in FGR Plgf is low, but sFlt1 is not elevated(51). 7 Preterm Birth Preterm birth (PTB) is defined by the World Health Organization as birth before 37 weeks of gestation. PTB can be classified by duration (e.g. early, extremely early, late PTB) or by delivery circumstance (spontaneous versus medically indicated or iatrogenic).The worldwide prevalence of PTB is estimated to be near 10%, but ranges from 5% in some countries to almost 20% in others.(52). As attention shifts to the impact of elevated BPs below the current criteria for diagnosis of HDPs, the potential that these moderate BP elevations in pregnancy might be linked to PTB must be considered. Angiogenic biomarkers might improve our understanding of which individuals with moderate BP elevations experience PTB. Several studies indicate that angiogenic biomarker levels in maternal blood in early to mid pregnancy do differ between pregnancies that deliver at term, and those that deliver early(53,54), yet others find that the addition of these biomarkers to models based on other risk factors does not improve prediction enough to be useful in clinical practice(55). Earlier research by our group found evidence that among non-smokers, low Plgf and high sEng were both associated with medically indicated PTB (mainly attributable to PE), and high sFlt1 was associated with medically indicated PTB in pregnancies not affected by PE(56). Angiogenic biomarkers and cardiovascular disease Multiple cohort studies and several recent meta-analysis demonstrate strong evidence that people with HDPs have continued risk for hypertension and cardiovascular disease after pregnancy(9,10,13,57–62). The links between lower PLGF and higher sFLt-1 during pregnancy have been investigated with both chronic (63) and new onset hypertension in the postpartum period(64), and with later cardiovascular and renal disease(60,65,66). Findings from this 8 research have been mixed, with a lack of consensus as to the utility of these biomarkers for long term prediction. Outside of pregnancy, higher sEng is associated with higher systolic BP, left ventricular hypertrophy, and diabetic retinopathy (67). Higher Plgf has been linked to lower HDL levels later in life and to the development of CVD in women(53). This seemingly unexpected direction of association may be related to a recent finding that in the postpartum period, higher rather than lower levels of Plgf are associated with hypertension(68). Dissertation Aims The discrepancy between the ACOG BP cut points used to define hypertension during pregnancy and AHA/ACC cut points used to define hypertension in non-pregnant populations motivates a need for greater understanding of risks for adverse pregnancy outcomes associated with BP elevations below the 140/90 threshold. Recent research using claims data and large institutional databases suggests that stage 1 hypertension does increase the odds of small for gestational age, preterm birth, or a need for neonatal intensive care (often as composite outcomes) (21,69). As many as 1 in 4 pregnant people would be classified as having a hypertensive disorder if the threshold was lowered, and up to 60% would have at least pre- hypertension(17,20,21). In previous prospective cohort studies, we found that BPs corresponding to pre-hypertension and stage 1 hypertension are associated with increased odd of developing stage 2 hypertension 7-15 years later(17), and that abnormalities of placental vasculature may aid risk stratification for this group(70). We are not aware of any studies that examine angiogenic biomarkers that have been linked to placental vascular pathology in combination with elevated BP below the 140/90 threshold and adverse pregnancy outcomes or later life hypertension. 9 This dissertation first explores associations between angiogenic biomarkers (Plgf, sFlt-1 and endoglin) and MAP during pregnancy (Aim 1). It then examines whether the combination of MAP and angiogenic biomarkers are associated with adverse pregnancy outcomes, such as preterm or small for gestational age birth (Aim 2). The final aim of the dissertation explores associations between angiogenic biomarkers during pregnancy and the development of later life hypertension among women with normal BP, moderately elevated BP and hypertensive disorders during pregnancy (Aim 3). These explorations are a ‘first step’ in determining whether these angiogenic biomarkers might aid in risk stratification among women with higher BP who do not have cHtn, GH or PE. During pregnancy, improved risk stratification would allow for increased monitoring of pregnancies at risk that are not identified by current criteria and inform decision-making about place of delivery, an issue that is critically important for those who live far from hospitals able to care for a low birth weight or preterm infant. For people facing barriers to care such as distance and insurance coverage, pregnancy may be one of the few times that they have consistent access to care(71). Improved identification of risk for future maternal cardiovascular disease also will allow health providers to emphasize cardiovascular disease prevention, monitoring, and intervention while patients are actively engaged with the health system. The three aims of this dissertation are as follows: Aim 1: Estimate the associations among maternal blood levels of sFlt-1, Plgf and endoglin measured at mid-pregnancy and MAP and assess whether any associations found remain after excluding those who develop PE, or any HDP. Aim 2: Assess the relationships among maternal blood levels of sEng, sFlt-1 and Plgf measured at mid-pregnancy, MAP at approximately 20 weeks gestation and risk of delivering preterm or a 10 small for gestational age (SGA) infant. Aim 3: Estimate associations among maternal levels of sEng and Plgf measured at mid pregnancy and hypertension 7-15 years later and determine if the odds of developing hypertension 7-15 years later among those with moderately elevated BP during pregnancy differs for those with less favorable biomarker levels. 11 REFERENCES 1. Osterman MJK, Hamilton BE, Martin JA, et al. Births : final data for 2020. Natl. Vital Stat. Reports. 2AD;70(17). (https://stacks.cdc.gov/view/cdc/112078) 2. Agrawal A, Wenger NK. Hypertension During Pregnancy. Curr. Hypertens. Rep. 2020;22(9). 3. Bonacina Manel ; Farràs, Alba ; Garcia‐Manau, Pablo ; Serrano, Berta ; Hurtado, Ivan ; Ferrer‐Oliveras, Raquel ; Illan, Lidia ; Armengol‐Alsina, Mireia ; Carreras, Elena E; M. Angiogenic factors for planning fetal surveillance in fetal growth restriction and small‐for‐ gestational‐age fetuses: A prospective observational study. BJOG. 129(11). 4. Gestational Hypertension and Preeclampsia: ACOG Practice Bulletin, Number 222. Obstet. Gynecol. 2020;135(6):e237–e260. 5. Helmo FR, Lopes AMM, Carneiro ACDM, et al. Angiogenic and antiangiogenic factors in preeclampsia. Pathol. Res. Pract. [electronic article]. 2017;214(1):7–14. (https://doi.org/10.1016/j.prp.2017.10.021) 6. Rana S, Burke SD, Karumanchi SA. Imbalances in circulating angiogenic factors in the pathophysiology of preeclampsia and related disorders. Am. J. Obstet. Gynecol. [electronic article]. 2022;226(2):S1019–S1034. (https://doi.org/10.1016/j.ajog.2020.10.022) 7. Reddy M, Rolnik DL, Harris K, et al. Challenging the definition of hypertension in pregnancy: a retrospective cohort study. Am. J. Obstet. Gynecol. [electronic article]. 2020;222(6):606.e1-606.e21. (https://doi.org/10.1016/j.ajog.2019.12.272) 8. Hauspurg A, Redman EK, Assibey-Mensah V, et al. Placental findings in non-hypertensive term pregnancies and association with future adverse pregnancy outcomes: a cohort study. Placenta [electronic article]. 2018;74(March):14–19. (https://linkinghub.elsevier.com/retrieve/pii/S0143400418307185) 9. Cain MA, Salemi JL, Tanner JP, et al. Pregnancy as a window to future health: maternal placental syndromes and short-term cardiovascular outcomes. Am. J. Obstet. Gynecol. [electronic article]. 2016;215(4):484.e1-484.e14. (http://dx.doi.org/10.1016/j.ajog.2016.05.047) 10. Canoy D, Cairns BJ, Balkwill A, et al. Hypertension in pregnancy and risk of coronary heart disease and stroke: A prospective study in a large UK cohort. Int. J. Cardiol. [electronic article]. 2016;222:1012–1018. (http://dx.doi.org/10.1016/j.ijcard.2016.07.170) 12 11. Parikh NI, Norberg M, Ingelsson E, et al. Association of Pregnancy Complications and Characteristics With Future Risk of Elevated Blood Pressure. Hypertension [electronic article]. 2017;HYPERTENSIONAHA.116.08121. (http://www.ncbi.nlm.nih.gov/pubmed/28137991%0Ahttp://hyper.ahajournals.org/look up/doi/10.1161/HYPERTENSIONAHA.116.08121) 12. Lindstrom L, Skj??rven R, Bergman E, et al. Chronic Hypertension in Women after Perinatal Exposure to Preeclampsia, Being Born Small for Gestational Age or Preterm. Paediatr. Perinat. Epidemiol. 2017;31(2):89–98. 13. Behrens I, Basit S, Melbye M, et al. Risk of post-pregnancy hypertension in women with a history of hypertensive disorders of pregnancy: nationwide cohort study. Bmj [electronic article]. 2017;j3078. (http://www.bmj.com/lookup/doi/10.1136/bmj.j3078) 14. Wu P, Park K, Gulati M. The fourth trimester: Pregnancy as a predictor of cardiovascular disease. Eur. Cardiol. Rev. . 2021;16. 15. Melchiorre K, Thilaganathan B, Giorgione V, et al. Hypertensive Disorders of Pregnancy and Future Cardiovascular Health. Front. Cardiovasc. Med. 2020;7. 16. Groenhof TKJ, van Rijn BB, Franx A, et al. Preventing cardiovascular disease after hypertensive disorders of pregnancy: Searching for the how and when. Eur. J. Prev. Cardiol. 2017;24(16):1735–1745. 17. Dunietz G, Strutz K, Holzman C, et al. Moderately elevated blood pressure during pregnancy and odds of hypertension later in life: the POUCHmoms longitudinal study. BJOG An Int. J. Obstet. Gynaecol. [electronic article]. 2017;1–8. (http://doi.wiley.com/10.1111/1471-0528.14556) 18. Boghossian NS, Albert PS, Mendola P, et al. Delivery Blood Pressure and Other First Pregnancy Risk Factors in Relation to Hypertensive Disorders in Second Pregnancies. Am. J. Hypertens. 2015;28(9):1172–1179. 19. Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: Executive summary: A report of the American college of cardiology/American Heart Association task . 2018 1269–1324. 20. Bello NA, Zhou H, Cheetham TC, et al. Prevalence of Hypertension among Pregnant Women When Using the 2017 American College of Cardiology/American Heart Association Blood Pressure Guidelines and Association with Maternal and Fetal Outcomes. JAMA Netw. Open. 2021;4(3). 13 21. Darwin KC, Federspiel JJ, Schuh BL, et al. ACC-AHA Diagnostic Criteria for Hypertension in Pregnancy Identifies Patients at Intermediate Risk of Adverse Outcomes. Am. J. Perinatol. 2021;38:E249–E255. 22. Sabol BA, Porcelli B, Diveley E, et al. Defining the risk profile of women with stage 1 hypertension: a time to event analysis. Am. J. Obstet. Gynecol. MFM. 2021;3(4). 23. Greenberg VR, Silasi M, Lundsberg LS, et al. Perinatal outcomes in women with elevated blood pressure and stage 1 hypertension. Am. J. Obstet. Gynecol. 2021;224(5):521.e1- 521.e11. 24. Poon LCY, Zymeri NA, Zamprakou A, et al. Protocol for Measurement of Mean Arterial Pressure at 11-13 Weeks’ Gestation. Fetal Diagn. Ther. [electronic article]. 2012;31(1):42–48. (https://www.karger.com/DOI/10.1159/000335366) 25. Flint AC, Conell C, Ren X, et al. Effect of Systolic and Diastolic Blood Pressure on Cardiovascular Outcomes. N. Engl. J. Med. 2019;381(3):243–251. 26. Melgarejo JD, Yang W-Y, Thijs L, et al. Association of Fatal and Nonfatal Cardiovascular Outcomes With 24-Hour Mean Arterial Pressure. Hypertension [electronic article]. 2021;77(1):39–48. (https://doi.org/10.1161/HYPERTENSIONAHA.120.14929) 27. Wang D, Wang J, Liu J, et al. The Role of Cumulative Mean Arterial Pressure Levels in First Stroke Events Among Adults with Hypertension: A 10-Year Prospective Cohort Study. Clin. Epidemiol. 2022;14:665–676. 28. Cnossen JS, Vollebregt KC, De Vrieze N, et al. Accuracy of mean arterial pressure and blood pressure measurements in predicting pre-eclampsia: Systematic review and meta- analysis. BMJ. 2008;336(7653):1117–1120. 29. Gasse C, Boutin A, Coté M, et al. First-trimester mean arterial blood pressure and the risk of preeclampsia: The Great Obstetrical Syndromes (GOS) study. Pregnancy Hypertens. 2018;12:178–182. 30. Kleinrouweler CE, Wiegerinck MMJ, Ris-Stalpers C, et al. Accuracy of circulating placental growth factor, vascular endothelial growth factor, soluble fms-like tyrosine kinase 1 and soluble endoglin in the prediction of pre-eclampsia: A systematic review and meta- analysis. BJOG An Int. J. Obstet. Gynaecol. 2012;119(7):778–787. 31. Nolasco-Leaños AG, Ramírez-Valenzuela KL, Carrillo-Juárez RI, et al. Circulating angiogenic factors are associated with progression to preeclampsia and the occurrence of adverse outcomes in women with gestational hypertension. J. Hypertens. 2021;39(3):573–580. 14 32. Anderson UD, Olsson MG, Kristensen KH, et al. Review: Biochemical markers to predict preeclampsia. Placenta [electronic article]. 2012;33(SUPPL.):S42–S47. (http://dx.doi.org/10.1016/j.placenta.2011.11.021) 33. Umapathy A, Chamley LW, James JL. Reconciling the distinct roles of angiogenic/anti- angiogenic factors in the placenta and maternal circulation of normal and pathological pregnancies. Angiogenesis [electronic article]. 2020;23(2):105–117. (https://doi.org/10.1007/s10456-019-09694-w) 34. Odibo AO, Patel KR, Spitalnik A, et al. Placental pathology, first-trimester biomarkers and adverse pregnancy outcomes. J. Perinatol. 2014;34(3):186–191. 35. Noori M, Donald AE, Angelakopoulou A, et al. Prospective study of placental angiogenic factors and maternal vascular function before and after preeclampsia and gestational hypertension. Circulation. 2010;122(5):478–487. 36. Phipps EA, Thadhani R, Benzing T, et al. Pre-eclampsia: pathogenesis, novel diagnostics and therapies. Nat. Rev. Nephrol. 2019;15(5). 37. Rana S, Burke SD, Karumanchi SA. Imbalances in circulating angiogenic factors in the pathophysiology of preeclampsia and related disorders. Am. J. Obstet. Gynecol. 2022;226(2):S1019–S1034. 38. Cerdeira AS, Agrawal S, Staff AC, et al. Angiogenic factors: potential to change clinical practice in pre-eclampsia? BJOG An Int. J. Obstet. Gynaecol. 2018;125(11). 39. Poon LC, Magee LA, Verlohren S, et al. A literature review and best practice advice for second and third trimester risk stratification, monitoring, and management of pre- eclampsia: Compiled by the Pregnancy and Non-Communicable Diseases Committee of FIGO (the International Federation of Gyneco. Int. J. Gynecol. Obstet. 2021;154(S1):3–31. 40. Tarca AL, Taran A, Romero R, et al. Prediction of preeclampsia throughout gestation with maternal characteristics and biophysical and biochemical markers: a longitudinal study. Am. J. Obstet. Gynecol. 2022;226(1):126.e1-126.e22. 41. Hirashima C, Ohkuchi A, Takahashi K, et al. Gestational hypertension as a subclinical preeclampsia in view of serum levels of angiogenesis-related factors. Hypertens. Res. [electronic article]. 2011;34(2):212–217. (http://www.nature.com/doifinder/10.1038/hr.2010.212) 42. Jacobs M, Nassar N, Roberts C, et al. Levels of soluble fms-like tyrosine kinase one in first trimester and outcomes of pregnancy: A systematic review. Reprod. Biol. Endocrinol. [electronic article]. 2011;9:1–8. 15 43. Ukah UV, Hutcheon JA, Payne B, et al. Placental Growth Factor as a Prognostic Tool in Women With Hypertensive Disorders of Pregnancy: A Systematic Review. Hypertens. (Dallas, Tex. 1979). 2017;70(6):1228–1237. 44. Benton SJ, McCowan LM, Heazell AEP, et al. Placental growth factor as a marker of fetal growth restriction caused by placental dysfunction. Placenta. 2016;42:1–8. 45. Johnson KM, Smith L, Modest AM, et al. Angiogenic factors and prediction for ischemic placental disease in future pregnancies. Pregnancy Hypertens. 2021;25:12–17. 46. Romero R, Nien JK, Espinoza J, et al. A longitudinal study of angiogenic (placental growth factor) and anti-angiogenic (soluble endoglin and soluble vascular endothelial growth factor receptor-1) factors in normal pregnancy and patients destined to develop preeclampsia and deliver a small for. J. Matern. Neonatal Med. 2008;21(1):9–23. 47. Senagore P, Rahbar MH, Tian Y, et al. Placental Vascular Pathology Findings and Pathways to Preterm Delivery. Am. J. Epidemiol. 2009;170(2):148–158. 48. Li F, Wang T, Chen L, et al. Adverse pregnancy outcomes among mothers with hypertensive disorders in pregnancy: A meta-analysis of cohort studies. Pregnancy Hypertens. [electronic article]. 2021;24(September 2020):107–117. (https://doi.org/10.1016/j.preghy.2021.03.001) 49. Martins JG, Biggio JR, Abuhamad A. Society for Maternal-Fetal Medicine Consult Series #52: Diagnosis and management of fetal growth restriction: (Replaces Clinical Guideline Number 3, April 2012). Am. J. Obstet. Gynecol. [electronic article]. 2020;223(4):B2–B17. (https://doi.org/10.1016/j.ajog.2020.05.010) 50. Talge NM, Mudd LM, Sikorskii A, et al. United states birth weight reference corrected for implausible gestational age estimates. Pediatrics. 2014;133(5):844–853. 51. Gaccioli F, Sovio U, Gong S, et al. Increased Placental sFLT1 (Soluble fms-Like Tyrosine Kinase Receptor-1) Drives the Antiangiogenic Profile of Maternal Serum Preceding Preeclampsia but Not Fetal Growth Restriction. Hypertension. 2023;80(2):325–334. 52. Vogel JP, Chawanpaiboon S, Moller AB, et al. The global epidemiology of preterm birth. Best Pract. Res. Clin. Obstet. Gynaecol. [electronic article]. 2018;52:3–12. (https://doi.org/10.1016/j.bpobgyn.2018.04.003) 53. Cassidy A, Chiuve SE, Manson JE, et al. Potential role for plasma placental growth factor in predicting coronary heart disease risk in women. Arterioscler. Thromb. Vasc. Biol. 2009;29(1):134–139. 16 54. Stepan H, Galindo A, Hund M, et al. Clinical utility of sFlt-1 and PlGF in screening, prediction, diagnosis and monitoring of pre-eclampsia and fetal growth restriction. Ultrasound Obstet. Gynecol. 2023;61(2):168–180. 55. Parry S, Carper BA, Grobman WA, et al. Placental protein levels in maternal serum are associated with adverse pregnancy outcomes in nulliparous patients. Am. J. Obstet. Gynecol. [electronic article]. 2022;227(3):497.e1-497.e13. (https://doi.org/10.1016/j.ajog.2022.03.064) 56. Mijal RS, Holzman CB, Rana S, et al. Mid-pregnancy levels of angiogenic markers as indicators of pathways to preterm delivery. J. Matern. Neonatal Med. 2012;25(7):1135– 1141. 57. Grand’Maison S, Pilote L, Okano M, et al. Markers of Vascular Dysfunction after Hypertensive Disorders of Pregnancy: A Systematic Review and Meta-Analysis. Hypertension. 2016;68(6):1447–1458. 58. Alonso-Ventura V, Li Y, Pasupuleti V, et al. Effects of preeclampsia and eclampsia on maternal metabolic and biochemical outcomes in later life: a systematic review and meta-analysis. Metabolism. 2020;102:154012. 59. Amor AJ, Vinagre I, Valverde M, et al. Preeclampsia is associated with increased preclinical carotid atherosclerosis in women with type 1 diabetes. J. Clin. Endocrinol. Metab. 2020;105(1):85–95. 60. Rangaswami J, Naranjo M, McCullough PA. Preeclampsia as a Form of Type 5 Cardiorenal Syndrome: An Underrecognized Entity in Women’s Cardiovascular Health. CardioRenal Med. 2018;19141:160–172. 61. Iino K, Higuchi T, Ogawa M, et al. Blood pressure during pregnancy is a useful predictive maker for hypertension and dyslipidemia later in life, a population-based, cross-sectional study. Maturitas [electronic article]. 2016;87:84–88. (http://dx.doi.org/10.1016/j.maturitas.2016.02.012) 62. Stuart JJ, Tanz LJ, Missmer SA, et al. Hypertensive disorders of pregnancy and maternal cardiovascular disease risk factor development: An observational cohort study. Ann. Intern. Med. 2018;169(4):224–232. 63. Neuman RI, Figaroa AMJ, Nieboer D, et al. Angiogenic markers during preeclampsia: Are they associated with hypertension 1 year postpartum? Pregnancy Hypertens. 2021;23:116–122. 64. Goel A, Maski MR, Bajracharya S, et al. Epidemiology and mechanisms of de novo and persistent hypertension in the postpartum period. Circulation. 2015;132(18):1726–1733. 17 65. Eastabrook G, Aksoy T, Bedell S, et al. Preeclampsia biomarkers: An assessment of maternal cardiometabolic health. Pregnancy Hypertens. [electronic article]. 2018;13(May):204–213. (https://doi.org/10.1016/j.preghy.2018.06.005) 66. Garrido-Gimenez C, Mendoza M, Cruz-Lemini M, et al. Angiogenic Factors and Long-Term Cardiovascular Risk in Women That Developed Preeclampsia during Pregnancy. Hypertension. 2020;1808–1816. 67. Blázquez-Medela AM, García-Ortiz L, Gómez-Marcos MA, et al. Increased plasma soluble endoglin levels as an indicator of cardiovascular alterations in hypertensive and diabetic patients. BMC Med. [electronic article]. 2010;8(1):86. (http://bmcmedicine.biomedcentral.com/articles/10.1186/1741-7015-8-86) 68. Hamza A, Gerlinger C, Radosa J, et al. Pilot study: placental biomarker predictive capability (sFlt-1, PlGF and their ratio) of postpartum maternal outcome. Arch. Gynecol. Obstet. 2019;299(6):1557–1566. 69. Sutton EF, Hauspurg A, Caritis SN, et al. Maternal outcomes associated with lower range stage 1 hypertension. In: Obstetrics and Gynecology. Lippincott Williams and Wilkins; 2018:843–849. 70. Holzman CB, Senagore P, Xu J, et al. Maternal risk of hypertension 7–15 years after pregnancy: clues from the placenta. BJOG An Int. J. Obstet. Gynaecol. 2021;128(5):827– 836. 71. Johnston EM, McMorrow S, Caraveo CA, et al. Post-aca, more than one-third of women with prenatal medicaid remained uninsured before or after pregnancy. Health Aff. 2021;40(4):571–578. 18 CHAPTER 2: ASSOCIATIONS BETWEEN ANGIOGENIC BIOMARKERS AND MEAN ARTERIAL PRESSURE IN PREGNANCY 19 Background Maternal blood Pressure (BP) is closely monitored in pregnancy to identify the development of hypertensive disorders of pregnancy (HDPs), i.e., gestational hypertension (GH) and preeclampsia (PE). In addition to their association with increased risk for adverse perinatal outcomes, hypertensive disorders of pregnancy have also been associated with increased risk of developing hypertension and/or cardiovascular disease later in life(1,2). In 2017, the American Association of Cardiology/American Heart Association (ACC-AHA) introduced new BP criteria for non-pregnant populations, lowering the level at which elevated BP (120-129 mmHg systolic with diastolic less than 80),and hypertension (stage 1: >=130 mmHg systolic or >= 80 mmHg diastolic, stage 2: >=140 mmHg or >=90 mmHg diastolic) are diagnosed, yet the previous cut off of 140/90 is still used during pregnancy(3). A few studies have examined the association between the 2017 ACC-AHA criteria and risk for pre-eclampsia (PE), finding an intermediate level of risk for PE for those with elevated BP or stage 1 hypertension (4–6). Multiple studies have demonstrated that imbalances of angiogenic and anti-angiogenic biomarkers (elevated levels of soluble endoglin (sEng) and soluble Fms-like tyrosine kinase- 1(sFlt-1) and lower levels of placental growth factor(Plgf)) in mid to late pregnancy in maternal blood are associated with PE (7–12). During pregnancy, these biomarkers are produced mainly by placental tissue, although production by other maternal cells contributes to overall levels. Lower levels of Plgf and higher levels of SFlt-1 also have been associated with GH in several studies, though levels are not as altered as seen in PE (13–15). The evidence for an association between endoglin levels and GH remains mixed(11,12). Women with GH are likely a 20 heterogeneous group with some undergoing the same pathophysiological process as PE but not meeting the PE diagnostic criteria before the end of pregnancy (13). Increasingly, angiogenic and anti-angiogenic biomarkers are being used in clinical practice to evaluate risk of developing PE in women with elevated BP (16,17). Angiogenic biomarkers may also be linked to the increased risk of later life CVD that has been observed among those with a history of HDPs (18– 20). Less is known about associations between angiogenic and anti-angiogenic biomarkers during pregnancy in women with elevated BP lower than the 140/90 used to diagnose gestational hypertension and preeclampsia. If angiogenic and anti-angiogenic biomarkers are correlated with BP in these ranges, they could provide clues to pregnancy health and to maternal vessel health later in life for these groups as well. Alternatively, these biomarkers might be specific to PE-related pathology and only help inform PE risk. Beyond the consideration of systolic and diastolic BP separately, some studies focus on mean arterial pressure (MAP), the average pressure throughout one cardiac cycle. MAP can be measured directly using invasive testing and with newer BP monitoring equipment, but often it is estimated in both research and practice using the formula of systolic pressure plus 2 x diastolic/3, with best practice being using the 2nd of 3 blood pressure readings(21). MAP provides a single continuous measure that incorporates both components of BP. A systematic review noted that in the first and second trimester of pregnancy, MAP may be superior to systolic BP or diastolic BP alone in predicting preeclampsia (22). The association of maternal blood sEng, sFlt-1, and Plgf levels with maternal MAP, however, is unclear. In this study, we hypothesize that lower levels of Plgf and higher levels of sflt-1 and endoglin 21 will be associated with elevated MAP. We consider both MAP measured at approximately 20 weeks’ gestation (when BP would be at its lowest-and before the onset of pregnancy related BP elevations) and the maximum across pregnancy. Additionally, we explore characteristics associated with BP in pregnancy (i.e. timing in pregnancy and maternal demographic, anthropometric, and behavioral characteristics) and whether these factors impact any biomarker-MAP associations observed. Finally, we consider whether any identified associations are driven by individuals who develop PE. Methods Recruitment and Study Sample The POUCH (Pregnancy Outcomes and Community Health) Study recruited pregnant participants at 52 clinics in 5 Michigan communities between 1998 and 2004(23). Inclusion criteria were: singleton pregnancy; maternal serum alpha-fetoprotein (MSAFP) screening at 15- 22 weeks, at least 15 years of age, able to English, no pre-pregnancy diabetes. A total of 3019 participants were enrolled in the study at 16-27 weeks’ gestation. The study protocol included maternal questionnaires and interview, collection of biologic samples, and follow-up after delivery(23). A sub-cohort participants received more detailed data collection including measurement of multiple biomarkers, medical records abstraction, and placental examinations. This sub-cohort included all those who delivered preterm or had elevated MSAFP and a random sample (with oversampling of Black individuals) of those who delivered at term (n=1,371). For our analysis, we excluded 30 participants who did not have blood pressure measurements and an additional 63 who did not have serum collected for biomarker measurement, leaving an analytic sample of 1,278 women (Figure A). 22 Mean Arterial Pressure (MAP) in pregnancy We abstracted blood pressure measurements (systolic & diastolic) from prenatal/hospital medical records at two time points during pregnancy: the time closest to 20 weeks’ gestation and on the date of highest diastolic BP in pregnancy. We then calculated mean Arterial Pressure (MAP) using the formula of (2 x diastolic + systolic)/3. We included both the MAP closest to 20 weeks when pregnancy-associated increases in BP would be less apparent(24), and the MAP at the highest diastolic BP measure during pregnancy, here after referred to as ‘highest MAP’. MAP was divided into quartiles with cut points based on the MAP distribution among women without cHtn, GH or PE. The top quartile of MAP at approximately 20 weeks was ≥ 84.7 mmHg (n=356 participants), and the top quartile for the highest MAP was ≥ 95.3 mmHg (n=407 participants). As a point of reference, blood pressure of 110/70 corresponds to MAP of 83.3 mmHg, 120/80 corresponds to an MAP of 93.3 mmHg, and 140/90 corresponds to an MAP of 106.7 mmHg Assays of Angiogenic Factors Endoglin (sEng), Placental growth factor (Plgf) and Soluble FMS-like tyrosine kinase-1 (sFlt-1) were measured in the maternal specimens collected at the study intake visit (16-27 weeks’ gestation). Measurements were made with commercially available kits and conducted by the Karumanchi Laboratory by personnel who were blinded to clinical information(25). Two assays were run for each angiogenic factor and the average was used in analysis. Inter-assay coefficients of variation were 7.6%, 10.9%, and 3.0% for sFlt-1, Plgf, and sEng, respectively, and intra-assay coefficients were 3.3%, 5.6%, and 6.3% (25). 23 Covariates Covariates were considered based on their documented relation to BP in pregnant and non- pregnant populations (26,27). Less is known about associations between covariates and maternal blood levels of sEng, sFlt-1, and Plgf (25,28). Maternal age, parity (primiparous vs. multiparous), education (high school education or less vs post-secondary education), health insurance status (Medicaid yes/no), and race/ethnicity were obtained through self-report. Maternal self-report of smoking history was dichotomized as those who never smoked or quit by mid-pregnancy vs. those who smoked throughout pregnancy. Due to the small number of women from race/ethnic groups other than Caucasian/white or African American/Black, self- reported race/ethnicity was dichotomized as Black or white/other. Gestational age was based on last menstrual period (LMP) if the due date from ultrasound at less than 25 weeks’ gestation was consistent with that calculated from LMP; otherwise, gestational age was based on ultrasound. Pre-pregnancy BMI was calculated from self-reported weight and corroborated with data from medical records. Hypertensive Disorders of Pregnancy Diagnoses of cHtn, GH and PE were abstracted from prenatal and labor and delivery records. GH was based on minimal criteria of diastolic BP >= 90 mmHg or systolic BP >= 140 mmHg on 2 different days after 20 weeks gestation without evidence of proteinuria. Preeclampsia was defined according to clinical criteria in use at the time of the study(29), as GH plus proteinuria that was defined as 2 + protein on urine dipstick once or 1 + protein on 2 occasions after 20 weeks’ gestation. (The current criteria include blood BP >=140/90 and proteinuria or signs of end organ damage)(30).Chronic hypertension (cHtn) was either pre-pregnancy diagnosis of 24 hypertension or > 90 diastolic or >140 systolic on 2 different days before 20 weeks gestation. Women with cHtn who developed superimposed PE were classified as PE (25). Statistical Analysis All analyses were performed with survey procedures in SAS v 9.4 (SAS Institute Inc., Cary, NC) and weighted to account for the oversampling of African American women, and women with preterm delivery or elevated MSAFP into the sub-cohort. Biomarkers were not normally distributed and therefore log-transformed for all analyses; the geometric means of the log- transformed biomarkers were then transformed back to normal scale. Bivariate associations between covariates and MAP, as well as between covariates and each biomarker, were evaluated using regression models. In adjusted multivariable logistic regression models, we estimated the odds of being in the top quartile of MAP at 20 weeks’ gestation, and for the highest MAP in pregnancy, in relation to a 1 unit increase in the log transformed sEng, 1 unit increase in log transformed sFlt-1, or 1 unit decrease in log transformed Plgf. After examining the unadjusted odds ratio, we adjusted for gestational age at biomarker measurement and maternal demographics. Of note, since educational level was not associated with differences in any of the biomarkers and education was closely correlated with maternal age and Medicaid insurance status, we elected not to use it in the multivariable linear regression models. Finally, a fully adjusted model including BMI and smoking as well as maternal demographics and gestational aged at measurement was examined. To explore whether any associations found were driven by PE or by any hypertension in pregnancy (PE, GH or cHtn), another group of regression models POUCH (Pregnancy Outcomes 25 and Community Health) Study enrolled a total of 3019 pregnant people at 16-27 weeks gestation from 52 clinics in 5 Michigan communities between 1998 and 2004. The potential for interactions between biomarkers and gestational age at blood collection was explored using interaction terms, and by stratifying on median gestational age at enrollment, i.e. at or before 22.4 weeks’ gestation vs. after. To assess whether any observed associations were consistent between Black, white and participants who identified as other races, we examined models stratified by race. Results Descriptive statistics Table 2.1 provides numbers and weighted percentages of covariates and the mean of MAP at 20 weeks and the mean highest MAP for each level of the covariate in the analytic sample. Participants had their biomarkers measured between 15 and 27.5 weeks of gestation (mean 22.4 weeks)-with 899 (71.3%) measured between 20 and 24.9 weeks. Maternal age ranged from 15.3-47.3 years with a mean of 26.6 years; 13.3% were over 29 years old. In our sample, 535 participants (weighted percent 40.2%) were primiparous. The weighted percent of Black participants was 23.4%, 48.4% of participants were insured through Medicaid, and 45.5% had a high school education or less. 236 (weighted mean 17.4%) of the participants continued to smoke after mid-pregnancy. The mean body mass index of participants in the study was 26.8, and 47.3% of the participants were overweight or obese (BMI >25). Biomarkers and covariates The overall geometric mean of sEng was higher in those women whose levels were measured after 25 weeks gestation (p=.01), while levels of sFlt-1 did not vary by gestational age at 26 measurement in our sample (Table 2). Younger women had higher levels of sFlt-1, but not significantly higher levels of endoglin. Participants with Medicaid insurance had lower sEng than those who did not have Medicaid (p=.02), and participants who smoked throughout pregnancy had lower sEng compared to those who quit or never smoked (p=.003). Multiparous women had lower levels of sEng, but higher levels of Sft-1. African American participants had higher levels of both of these biomarkers than white/other participants, and obese (BMI >30) participants had higher levels than those of normal weight (Table 2.2). Mean Plgf increased with gestational age in our study and decreased with maternal age. Smokers had higher levels of Plgf than non-smokers, and Black participants had higher levels of Plgf compared to white/other participants. Obese women had the lowest levels of Plgf, while neither Medicaid insurance status nor education were associated with Plgf. Pre-pregnancy BMI as a continuous variable had a small but significant negative association with all 3 biomarkers (data not shown). Bivariate Analysis of Mean Arterial Pressure (MAP) The weighted average MAP at approximately 20 weeks' gestation was 80.3 mmHg, and the average highest MAP was 92.2 mmHg. In bivariate analyses (shown in Table 2.1), MAP at 20 weeks did not significantly differ by the gestational age at which biomarkers were measured or by parity, but did differ by age at enrollment, with older participants having higher MAP. Mean MAP at ~20 weeks was lower in participants who were Black (p<.0001), were insured by Medicaid (p<.0001), had less education (p<.0001), and in those who continued smoking throughout pregnancy compared with their counterparts (p=.001). MAP at ~20 weeks was strongly associated with BMI, with overweight and obese women having higher MAP than 27 underweight and normal weight women. The highest MAP showed similar associations; however, it also varied by parity (93.5 mmHg in primiparous participants, 91.3 mmHg in multiparous participants women, p=.001). Associations between biomarkers and hypertension in pregnancy When comparing biomarker levels across the four hypertension groups (normotensive, cHtn, GH and PE/cHtn with superimposed PE), we found that compared to participants with normal blood pressure throughout pregnancy, those who developed PE had higher mean sEng (p<.0001) and sFlt-1 (p=.003), and lower mean Plgf (<.0001) (Table 2.2). Of note, there were no significant differences in mean biomarker levels between participants without hypertension in pregnancy and women with GH or cHtn (without superimposed PE) for any of the three biomarkers. Mean Plgf increased with gestational age in our study and decreased with maternal age. Smokers had higher levels of Plgf than non-smokers, and Black participants had higher levels of Plgf compared to white/other participants. Obese women had the lowest levels of Plgf, while neither Medicaid insurance status nor education were associated with Plgf. Pre-pregnancy BMI as a continuous variable had a small but significant negative association with all 3 biomarkers (data not shown). Associations between biomarkers and elevated MAP Increases in of 1 unit in log sEng were not associated with odds of being in the highest quartile of MAP at 20 weeks (OR= 0.76, 95% CI 0.45, 1.26) nor with the highest MAP (OR =0.80, 95%CI 0.49, 1.31). This did not change when we removed the 44 women with PE or the 144 women with any HDP. Increase in log sFlt-1 decreased the odds of being in the highest quartile of MAP 28 at 20 weeks (OR=0.68, 95% CI 0.53, 0.87) and for the highest MAP (OR=0.75, 95%CI 0.60, 0.95). However, these associations attenuated to non-significance following adjustment for smoking and BMI. . Results were similar when we excluded women with PE and women with cHtn, GH or PE (Table 2.3). A one unit decrease in log Plgf increased the odds of being in the upper quartile of MAP at both time points. The unadjusted odds of being in the top quartile of MAP at 20 weeks’ gestation associated with in a 1 unit increase in log Plgf were 2.01 (95% Cl 1.59, 2.56) and at the time of the highest MAP were 2.04 (95% Cl 1.62, 2.58). (Table 3). In the adjusted models, effect sizes were slightly attenuated but remained statistically significant. Further reductions in effect size were observed after removing participants with PE, aOR of MAP in the highest quartile at 20 weeks associated with a 1 unit decrease in log Plgf was 1.49 (95% Cl 1.09, 2.04) and 1.35 (95% Cl .93, 1.97) when we eliminated all 144 women with PE/GH or cHTN. We tested models including an interaction term between each biomarker and race and found no statistically significant interactions (data not shown). Because this analysis might be underpowered,, stratified associations by race using 3 groups (Black, White, Other) and found that a 1 unit decrease in log Plgf was associated with an aOR of MAP >= 84.7 mmHg at 20 weeks of 1.38 (95% Cl 0.92, 2.06) among Black participants, 1.52 (95% Cl 1.00, 2.32) among white participants, and 1.52 (95% Cl 0.18, 2.06) for other races. For highest MAP (>= 95.3), the aORs were 1.84 (95% Cl 1.23, 2.75), 1.39 (95% Cl 0.93, 2.07) and 1.67 (95% Cl 0.43, 6.67) respectively. The odds ratios associated with increases in sFlt-1 and sEng were not significant (Table 2.4). We also examined effect modifications by gestational age at biomarker measurement and found that there was an interaction between sEng and gestational age at measurement, no 29 significant interaction between sFlt-1 and gestational age and an interaction approaching significance for Plgf (data not shown). In a stratified analysis, biomarker measurements taken after 22.4 weeks (the median in our study) was in the expected directions for Plgf and sEng, while the odds remained close to 1 for sFlt_1. These associations were not significant with the exception of PLGF when measured after 22.4 weeks. (Table 2.5). Discussion In this prospective cohort of 1278 pregnant individuals who had angiogenic biomarkers measured in mid-pregnancy, higher levels of Plgf, in unadjusted and adjusted analyses, were associated with decreased odds of MAP in the top quartile, both at approximately 20 weeks and at the time of the highest diastolic BP in pregnancy. For MAP at 20 weeks, the association between higher Plgf and lower odds of MAP >= 84.7 mmHg (the upper quartile among normotensive participants) remained even after excluding participants with PE and was of a similar magnitude but no longer significant when we excluded participants with cHtn, or who went on to develop GH or PE. For the upper quartile of highest MAP (>= 95.3 mmHg), the aOR remained significant when we excluded participants with PE and when we excluded participants with any cHtn, GH or PE. Sflt-1 and sEng were not significantly associated with MAP. Our consideration of higher maternal MAP in the absence of PE during pregnancy follows recent concerns that BP cut points historically used for PE diagnosis (which includes two episodes of bp ≥140/90) may miss some women with PE-like vascular pathophysiology. Biomarkers such as Plgf that are linked to PE could be useful in identifying subgroups of women with PE-like pathology who do not fit the current criteria for high BP in pregnancy. Lowering the 30 BP cutoffs for HDPs may be another strategy to consider; however, the number of women labeled at risk by using new BP cut-points alone would be large and likely not highly specific. Research examining associations between ACA/AHA Stage 1 hypertension and maternal and fetal outcomes does find elevated risk for PE among this group, and some increase in fetal/neonatal adverse events(4–6,31). Our upper quartile of highest MAP would capture both AHA/ACA stage 1 and the elevated BP group. Reclassifying hypertension in pregnancy using the 2017 ACC/AHA definitions would lead to an increase in the number of people classified as having an HDP or cHtn. In one large cohort, use of the ACC/AHA definition increased the percentage of women classified as having a HDP or cHtn from 10.3 to 28.1% (4). Changing the classifications in the first trimester could lead to more people receiving aspirin therapy to prevent PE, while later in pregnancy, more people might be delivered before 38 weeks’ gestation (potentially preventing adverse events, but also potentially increasing inductions and late preterm delivery) (3). Future studies could examine the potential role of biomarkers such as Plgf in identifying higher risk individuals within the AHA/ACA stage one and elevated BP groups in order to preferentially select those who are most likely to benefit from interventions or close monitoring during pregnancy and the post- partum period. Our biomarker findings in relation to PE and HDP overall are similar to other studies. In one prospective study of placental angiogenic factors measured throughout pregnancy, PE and GH were associated with lower PLGF levels measured at 18-25 weeks’ gestation; associations with sEng and sFlt-1 typically were not observed until later in pregnancy (32). More recent research suggests that Plgf can be a useful marker of risk for PE as early as the first 31 trimester(33), and multiple studies have shown that sFlt-1 and endoglin differ between individuals who go on to develop PE and those who do not when measured closer to term(17,34). Our own analysis stratified on earlier or later gestational age at measurement suggests that one of the reasons we found no association between MAP and endoglin or sFlt-1 may be that the majority of participants in our study had biomarkers measured before 24 weeks’ gestation. In our study, pre-pregnancy BMI had an inverse association with all three biomarkers. Several other studies also demonstrate inverse associations between Sflt-1 and sEng and BMI(25,35). One study notes obese women with PE do not demonstrate abnormal angiogenic biomarker profiles to the extent that normal weight women do(36). While this may be partially due to hemodilution, adipose tissue is highly vascularized and secretes pro and antiangiogenic factors (potentially in ways that differ between lean and obese women). The correlation between BMI and higher BP, and between BMI and lower levels of biomarkers in maternal circulation makes adjustment for BMI critically important. Limitations and strengths Biomarkers were measured once in this study; thus, we are not able to assess whether biomarkers lead to changes in MAP or vice versa, nor can we determine biomarker trends or whether the observed associations are causal. MAP was calculated from BP documented in medical records. Research indicates clinical BP correlates with BP measured in research protocols, but may overestimate systolic and to a lesser degree diastolic blood pressure(37). It is not known whether this measurement error is more likely for certain groups of women, but it is not likely to be correlated with biomarkers. 32 BMI, which is strongly associated with both BP and serum levels of biomarkers is not an ideal measure of adiposity which may be relevant to the secretion of angiogenic and antiangiogenic biomarkers. However, BMI is widely used in clinical practice where other measures of adiposity are less easily available. Waist circumference, which can be helpful clinically, is not useful in pregnancy and not likely to have been assessed by patients themselves before pregnancy. As mentioned earlier, our ability to assess whether sFlt-1 and sEng are associated with elevations in MAP is likely limited by the gestational age at which the biomarkers were measured in the POUCH Study cohort. Given the growing recognition that individuals with BP elevations lower than 140/90 (current cutoff used to define HDPs) are at greater risk for adverse maternal and fetal outcomes, it is important to understand how potential biomarkers of risk for adverse outcomes are related to milder elevations in MAP. This racially diverse prospective study included detailed socio- demographic information, data from clinical records and biomarkers, allowing us to examine potential confounding of any associations between biomarkers and MAP. While several recent studies have considered the role of MAP and angiogenic biomarkers in the prediction of pre- eclampsia (33,34,38), ours is one of the only studies we are aware of to examine associations between angiogenic/antiangiogenic biomarkers (sEng, Sflt1 and Plgf) and MAP in women without hypertensive disorders of pregnancy. Conclusion Lower levels of Plgf at mid-pregnancy in our study are associated with having higher Mean Arterial Pressures (MAP), even among participants without pre-eclampsia or other hypertensive disorders. As we learn more about the large group of individuals who have ACA/AHA elevated 33 or stage 1 hypertension during pregnancy, it is possible that PLGF would help identify individuals who may be phenotypically similar to individuals who develop PE during pregnancy and merit closer monitoring for adverse infant/fetal outcomes, postpartum, and midlife CVD complications. 34 REFERENCES 1. Black MH, Zhou H, Sacks DA, et al. Hypertensive disorders first identified in pregnancy increase risk for incident prehypertension and hypertension in the year after delivery. J. Hypertens. 2016;34(4):728–735. 2. Parikh NI, Norberg M, Ingelsson E, et al. Association of Pregnancy Complications and Characteristics With Future Risk of Elevated Blood Pressure. Hypertension [electronic article]. 2017;HYPERTENSIONAHA.116.08121. (http://www.ncbi.nlm.nih.gov/pubmed/28137991%0Ahttp://hyper.ahajournals.org/look up/doi/10.1161/HYPERTENSIONAHA.116.08121) 3. Garovic VD, Dechend R, Easterling T, et al. Hypertension in Pregnancy: Diagnosis, Blood Pressure Goals, and Pharmacotherapy: A Scientific Statement From the American Heart Association. Hypertension. 2022;79(2):E21–E41. 4. Bello NA, Zhou H, Cheetham TC, et al. Prevalence of Hypertension among Pregnant Women When Using the 2017 American College of Cardiology/American Heart Association Blood Pressure Guidelines and Association with Maternal and Fetal Outcomes. JAMA Netw. Open. 2021;4(3). 5. Darwin KC, Federspiel JJ, Schuh BL, et al. ACC-AHA Diagnostic Criteria for Hypertension in Pregnancy Identifies Patients at Intermediate Risk of Adverse Outcomes. Am. J. Perinatol. 2021;38:E249–E255. 6. Sutton EF, Rogan SC, Lopa S, et al. Early Pregnancy Blood Pressure Elevations and Risk for Maternal andNeonatal Morbidity. Obstet. Gynecol. 2020;136(1):129–139. 7. Ohkuchi A, Hirashima C, Takahashi K, et al. Prediction and prevention of hypertensive disorders of pregnancy. Hypertens. Res. [electronic article]. 2017;40(1):5–14. (http://www.nature.com/doifinder/10.1038/hr.2016.107) 8. Herraiz I, Llurba E, Verlohren S, et al. Update on the Diagnosis and Prognosis of Preeclampsia with the Aid of the sFlt-1/PlGF Ratio in Singleton Pregnancies. Fetal Diagn. Ther. 2018;43(2). 9. Agrawal S, Cerdeira AS, Redman C, et al. Meta-Analysis and Systematic Review to Assess the Role of Soluble FMS-Like Tyrosine Kinase-1 and Placenta Growth Factor Ratio in Prediction of Preeclampsia: The SaPPPhirE Study. Hypertens. (Dallas, Tex. 1979) [electronic article]. 2018;71(2):306–316. (http://hyper.ahajournals.org/lookup/doi/10.1161/HYPERTENSIONAHA.117.10182%0Aht tp://www.ncbi.nlm.nih.gov/pubmed/29229743) 35 10. Wu P, Van Den Berg C, Alfirevic Z, et al. Early pregnancy biomarkers in pre-eclampsia: A systematic review and meta-analysis. Int. J. Mol. Sci. 2015;16(9). 11. Helmo FR, Lopes AMM, Carneiro ACDM, et al. Angiogenic and antiangiogenic factors in preeclampsia. Pathol. Res. Pract. [electronic article]. 2017;214(1):7–14. (https://doi.org/10.1016/j.prp.2017.10.021) 12. Kleinrouweler CE, Wiegerinck MMJ, Ris-Stalpers C, et al. Accuracy of circulating placental growth factor, vascular endothelial growth factor, soluble fms-like tyrosine kinase 1 and soluble endoglin in the prediction of pre-eclampsia: A systematic review and meta- analysis. BJOG An Int. J. Obstet. Gynaecol. 2012;119(7):778–787. 13. Hirashima C, Ohkuchi A, Takahashi K, et al. Gestational hypertension as a subclinical preeclampsia in view of serum levels of angiogenesis-related factors. Hypertens. Res. [electronic article]. 2011;34(2):212–217. (http://www.nature.com/doifinder/10.1038/hr.2010.212) 14. Jacobs M, Nassar N, Roberts C, et al. Levels of soluble fms-like tyrosine kinase one in first trimester and outcomes of pregnancy: A systematic review. Reprod. Biol. Endocrinol. [electronic article]. 2011;9:1–8. (http://www.embase.com/search/results?subaction=viewrecord&from=export&id=L514 69367%5Cnhttp://www.rbej.com/content/9/1/77%5Cnhttp://dx.doi.org/10.1186/1477- 7827-9- 77%5Cnhttp://rug.on.worldcat.org/atoztitles/link/?sid=EMBASE&issn=14777827&id=doi: 10.1186%2F1) 15. Ukah UV, Hutcheon JA, Payne B, et al. Placental Growth Factor as a Prognostic Tool in Women With Hypertensive Disorders of Pregnancy: A Systematic Review. Hypertens. (Dallas, Tex. 1979). 2017;70(6):1228–1237. 16. Cerdeira AS, Agrawal S, Staff AC, et al. Angiogenic factors: potential to change clinical practice in pre-eclampsia? BJOG An Int. J. Obstet. Gynaecol. 2018;125(11). 17. Poon LC, Magee LA, Verlohren S, et al. A literature review and best practice advice for second and third trimester risk stratification, monitoring, and management of pre- eclampsia: Compiled by the Pregnancy and Non-Communicable Diseases Committee of FIGO (the International Federation of Gyneco. Int. J. Gynecol. Obstet. 2021;154(S1):3–31. 18. Eastabrook G, Aksoy T, Bedell S, et al. Preeclampsia biomarkers: An assessment of maternal cardiometabolic health. Pregnancy Hypertens. [electronic article]. 2018;13(May):204–213. (https://doi.org/10.1016/j.preghy.2018.06.005) 36 19. Suvakov S, Bonner E, Nikolic V, et al. Overlapping pathogenic signalling pathways and biomarkers in preeclampsia and cardiovascular disease. Pregnancy Hypertens. [electronic article]. 2020;20(August 2019):131–136. (https://doi.org/10.1016/j.preghy.2020.03.011) 20. Garrido-Gimenez C, Mendoza M, Cruz-Lemini M, et al. Angiogenic Factors and Long-Term Cardiovascular Risk in Women That Developed Preeclampsia during Pregnancy. Hypertension. 2020;1808–1816. 21. Poon LCY, Zymeri NA, Zamprakou A, et al. Protocol for Measurement of Mean Arterial Pressure at 11-13 Weeks’ Gestation. Fetal Diagn. Ther. [electronic article]. 2012;31(1):42–48. (https://www.karger.com/DOI/10.1159/000335366) 22. Cnossen JS, Vollebregt KC, De Vrieze N, et al. Accuracy of mean arterial pressure and blood pressure measurements in predicting pre-eclampsia: Systematic review and meta- analysis. BMJ. 2008;336(7653):1117–1120. 23. Holzman C, Bullen B, Fisher R, et al. Pregnancy outcomes and community health: the POUCH study of preterm birth. Paediatr. Perinat. Epidemiol. 2001;15(s2):136–158. 24. Loerup L, Pullon RM, Birks J, et al. Trends of blood pressure and heart rate in normal pregnancies: A systematic review and meta-analysis. BMC Med. 2019;17(1):1–12. 25. Mijal RS, Holzman CB, Rana S, et al. Midpregnancy levels of angiogenic markers in relation to maternal characteristics. Am. J. Obstet. Gynecol. [electronic article]. 2011;204(3):244.e1-244.e12. (http://dx.doi.org/10.1016/j.ajog.2010.10.001) 26. Kurian AK, Cardarelli KM. Racial and ethnic differences in cardiovascular disease risk factors: a systematic review. Ethn. Dis. 2007;17(1):143–152. 27. Salles GF, Schlüssel MM, Farias DR, et al. Blood pressure in healthy pregnancy and factors associated with no mid-trimester blood pressure drop: A prospective cohort study. Am. J. Hypertens. 2015;28(5):680–689. 28. Metz TD, Allshouse AA, Euser AG, et al. Preeclampsia in high risk women is characterized by risk group-specific abnormalities in serum biomarkers. Am. J. Obstet. Gynecol. [electronic article]. 2014;211(5):512.e1-512.e6. (http://dx.doi.org/10.1016/j.ajog.2014.04.027) 29. ACOG Practice Bulletin No. 33: Diagnosis and Management of Preeclampsia and Eclampsia. Obstet. Gynecol. 2002;99(1):159–167. 30. Gestational Hypertension and Preeclampsia: ACOG Practice Bulletin, Number 222. Obstet. Gynecol. 2020;135(6):e237–e260. 37 31. Greenberg VR, Silasi M, Lundsberg LS, et al. Perinatal outcomes in women with elevated blood pressure and stage 1 hypertension. Am. J. Obstet. Gynecol. 2021;224(5):521.e1- 521.e11. 32. Noori M, Donald AE, Angelakopoulou A, et al. Prospective study of placental angiogenic factors and maternal vascular function before and after preeclampsia and gestational hypertension. Circulation. 2010;122(5):478–487. 33. Wright D, Tan MY, O’Gorman N, et al. Serum PlGF compared with PAPP-A in first trimester screening for preterm pre-eclampsia: Adjusting for the effect of aspirin treatment. BJOG An Int. J. Obstet. Gynaecol. 2022;(December 2021):1–10. 34. Tayyar A, Krithinakis K, Wright A, et al. Mean arterial pressure at 12, 22, 32 and 36 weeks’ gestation in screening for pre-eclampsia. Ultrasound Obstet. Gynecol. 2016;47(5). 35. Hauspurg A, Countouris ME, Jeyabalan A, et al. Risk of hypertension and abnormal biomarkers in the first year postpartum associated with hypertensive disorders of pregnancy among overweight and obese women. Pregnancy Hypertens. [electronic article]. 2019;15(November 2018):1–6. (https://doi.org/10.1016/j.preghy.2018.10.009) 36. Jääskeläinen T, Heinonen S, Hämäläinen E, et al. Impact of obesity on angiogenic and inflammatory markers in the Finnish Genetics of Pre-eclampsia Consortium (FINNPEC) cohort. Int. J. Obes. 2019;43(5):1070–1081. 37. Lane-Cordova AD, Wilcox S, Fernhall B, et al. Agreement between blood pressure from research study visits versus electronic medical records and associations with hypertensive disorder diagnoses in pregnant women with overweight/obesity. Blood Press. Monit. 2021;341–347. 38. Tarca AL, Taran A, Romero R, et al. Prediction of preeclampsia throughout gestation with maternal characteristics and biophysical and biochemical markers: a longitudinal study. Am. J. Obstet. Gynecol. 2022;226(1):126.e1-126.e22. 38 APPENDIX Figure A: Flow Chart of POUCH Study and Analytic Sample • Not sampled Pouch Cohort into subcohort (n=3019) (n=1648) • 30 women missing blood pressure Pouch Subcohort measurements (n=1371) • 63 no blood sample for biomarker measurement Analytic Sample (n=1278) 39 Table 2.1: Mean Arterial Pressure by Maternal Characteristics, Pregnancy Outcomes and Community Health Study Frequency & Mean Arterial Pressure mmHg at 20 weeks Highest Mean Arterial Pressure weighted% N=1261 mmHg N=1278 N=1278 N (%)* Mean 95%CL p** Mean 95%CL p** Gestational Age 15-19.9 204 (15.4%) 80.2 (78.6, 81.8) ref 92.2 (90.5, 93.9) ref 20-24.9 899 (71.3%) 80.3 (79.6, 81.0) .92 92.4 (91.6, 98.2) .83 25-27.9 175 (13.3%) 80.3 (78.9, 81.8) .92 92.3 (90.0, 92.7) .44 Maternal Age <20 221 (14.4%) 77.3 (75.8, 78.8) ref 89.9 (88.1, 91.7) ref 20-29 722 (56.9%) 80.0 (79.3, 80.8) .001 91.8 (91.0, 92.7) .05 >=30 335 (28.7%) 82.3 (81.1, 83.4) <.0001 94.2 (93.1, 95.4) <.0001 Parity Primiparous 535 (41.2%) 80.3 (79.4, 81.3) ref 93.5 (92.5, 94.6) ref Multiparous 742 (58.8%) 80.3 (79.5, 81.1) .89 91.3 (90.5, 92.1) .001 Race White/other 760 (76.6%) 81.0 (80.2, 81.7) ref 93.3 (92.5, 94.0) ref Black 518 (23.4%) 78.0 (77.2, 78.9) <.0001 88.9 (87.8, 89.9) <.0001 Medicaid no 564 (51.6%) 81.9 (81.1, 82.8) ref 93.9 (93.0, 94.7) ref yes 713 (48.4%) 78.5 (77.7, 79.4) <.0001 90.5 (89.6, 91.4) <.0001 Education ≤High School 643 (45.5%) 78.9 (78.1, 79.8) ref 90.6 (89.6, 91.5) ref >High School 635 (54.5%) 81.4 (80.6, 82.2) <.0001 93.6 (92.7, 94.5) <.0001 Smoking Nonsmoker/quit 1042(82.7%) 80.7 (80.0, 81.3) ref 92.7 (92.0, 93.4) ref Smoked 236 (17.3%) 78.5 (77.0, 80.0) <.0001 89.9 (88.4, 91.4) .001 PrePregBMI Underwt 58 (3.8%) 72.5 (70.2, 74.8) <.0001 83.5 (80.7, 86.2) <.0001 Normal 574 (46.9%) 77.8 (77.1, 78.6) ref 90.5 (89.7, 91.4) ref Overwt 283 (23.1%) 82.6 (81.4, 83.7) <.0001 93.3 (91.9, 94.7) .001 Obese 363 (26.3%) 83.9 (82.6, 85.2) <.0001 95.6 (94.3, 96.8) <.0001 *percent weighted to reflect sampling frame. **p values from differences of least squares means in proc survey reg 40 Table 2.2: Mean Angiogenic Biomarker levels by Maternal Characteristics, Pregnancy Outcomes and Community Health Study N (%)* Endoglin N=1278 SFLT_1 N=1278 PLGF N=1275 Mean* 95%CL p** Mean* 95%CL p** Mean* 95%CL p** Gestational Age 15-19.9 204 5.29 (5.05, 5.55) ref 1670 (1500, 1859) ref 225.9 (205.6, 248.3) ref (15.4%) 20-24.9 899 5.28 (5.17, 5.40) .95 1664 (1588, 1745) .95 379.0 (361.0, 397.9) <.0001 (71.3%) 25-27.9 175 5.80 (5.40, 6.12) .01 1700 (1525, 1896) .82 561.4 (514.2, 613.0) <.0001 (13.3%) Maternal Age <20 221 5.57 (5.25, 5.90) ref 1981 (1798, 2183) ref 427.5 (384.3, 475.6) ref (14.4%) 20-29 722 5.29 (5.17, 5.41) .11 1588 (1504, 1676) <.001 355.1 (336.3, 375.0) .003 (56.9%) >=30 335 5.37 (5.17, 5.58) .31 1693 (1574, 1821) .01 368.3 (337.3, 402.2) .03 (28.7%) Parity Primiparous 535 5.60 (5.43, 5.77) ref 1948 (1839, 2063) ref 377.9 (353.8, 403.8) ref (41.2%) 742 5.18 (5.01, 5.31) <.0001 1499 (1421, 1580) <.0001 362.2 (342.3, 383.3 .34 Multiparous (58.8%) Race White/oth 760 5.30 (5.18, 5.43) ref 1598 (1520, 1679) ref 348.3 (330.0, 367.3) ref (76.6%) Black 518 5.50 (5.32, 5.65) .04 1931 (1832, 2036) <.0001 443.8 (418.8, 470.4) <.0001 (23.4%) Medicaid no 564 5.47 (5.32, 5.62) ref 1704 (1610, 1804) ref 353.3 (332.4, 375.4) ref (51.6%) yes 713 5.23 (5.09, 5.36) .02 1633 (1543, 1727) .29 385.7 (362.9, 409.9) .05 (48.4%) 41 Table 2.2 (cont’d) Education <= 12 years 643 5.35 (5.20, 5.50) ref 1655 (1560, 1756) ref 384.4 (360.5, 410.0) ref (45.5%) >12 years 635 5.35 (5.22, 5.49) .96 1682 (1593, 1776) .69 355.9 (335.9, 377.2) .08 (54.5%) Smoking Non Smoker 1042(82.7 5.42 (5.30, 5.53) ref 1687 (1614, 1764) ref 344.6 (329.3, 361.3) ref %) Smokers 236 5.04 (4.83, 5.26) .003 1588 (1446, 1744) .25 507.5 (461.7, 557.7) <.0001 (17.3%) PrePregBMI Underwt 58 (3.8%) 5.33 (4.77, 5.96) .45 1926 (1587, 2337) .82 559.3 (473.2, 661.0) .001 Normal 574 5.57 (5.42, 5.73) ref 1883 (1780, 1992) ref 409.2 (383.8, 436.2) ref (46.9%) Overwt 283 5.43 (5.25, 5.62) .26 1629 (1507, 1761) .003 377.1 (346.3, 410.7) .13 (23.1%) Obese 363 4.91 (4.73, 5.10) <.0001 1350 (1247, 1463) <.0001 282.4 (262.5, 303.8) <.0001 (26.3%) *geometric mean of log transformed angiogenic biomarkers have been back transformed to normal scale, sEng and sFlt-1 pg/ml, Plgf ng/ml **p values from differences of least squares means in proc survey reg 42 Table 2.3: Associations between change in angiogenic biomarkers at mid-pregnancy and odds being in top quartile of MAP close to 20 weeks (MAP ≥ 84.7), and top quartile of highest MAP in pregnancy (≥95.3). POUCH Study 1278 participants 1 unit incease in Log Endoglin 1 unit increase in Log SFLT_1 1 unit decrease in Log PLGF (N=1278) (N=1278) (N=1275) MODELS OR 95% CI OR 95% CI OR 95% CI OR 95% CI OR 95% CI OR 95% CI MAP high MAP high MAP high ≥84.7 MAP ≥84.7 MAP ≥84.7 MAP at 20 ≥95.3 at 20 ≥95.3 at 20 ≥95.3 weeks weeks weeks Unadjusted 0.76 (0.45, 0.80 (0.49, 0.68* (0.53, 0.87) 0.75* (0.60, 2.01* (1.59, 2.04* (1.62, 1.26) 1.31) 0.95) 2.56) 2.58) Adjusted 0.71 (0.42, 0.76 (0.45, 0.63* (0.48, 0.83) 0.72* (0.56, 2.16* (1.63, 2.19* (1.65, Gestational 1.21) 1.28) 0.92) 2.86) 2.91) Age and Demographic ** Fully 1.10 (0.66, 1.10 (0.66, 0.87 (0.66, 1.16) 0.97 (0.74, 1.49* (1.09, 1.50* (1.11, Adjusted*** 1.85) 1.84) 1.28) 2.04) 2.02) Excluding 44 1.12 (0.63, 0.90 (0.52, 0.88 (0.65, 1.18) 0.95 (0.71, 1.45* (1.03, 1.41* (1.03, women with 1.98) 1.56) 1.26) 2.04) 1.94) PE*** Excluding 0.92 (0.49, 0.82 (0.44, 0.84 (0.60, 1.17) 0.98 (0.72, 1.35 (0.93, 1.41* (1.00, 144 women 1.73) 1.53) 1.34 1.97) 1.98) with PE, GH or cHtn*** *p<.05 from proc survey logistic. **Demographics = maternal age, parity, race, Medicaid status ***Full Model = Gestational age, demographics, smoking, pre- pregnancy BMI. Note-Cut points for quartiles based on quartiles of women not diagnosed with PE/GH/cHtn. 43 Table 2.4: Associations between angiogenic biomarkers at mid-pregnancy and odds being in top quartile of MAP at 20 weeks, and top quartile of highest MAP in pregnancy. POUCH Study 1278 pregnant women, stratified by race 1 unit increase Log Endoglin 1 unit increase Log SFLT_1 1 unit decrease Log PLGF MODELS OR 95% CI OR 95% CI OR 95% CI OR 95% CI OR 95% CI OR 95% CI MAP high MAP high MAP high ≥84.7 MAP ≥84.7 MAP ≥84.7 MAP at 20 ≥95.3 at 20 ≥95.3 at 20 ≥95.3 weeks weeks weeks Full 1.08 (0.65, 1.04 (0.62, 1.04 (0.62, 0.92 (0.70, 1.52* (1.12, 1.60* (1.20, model 1.79) 1.71) 1.72 1.20) 2.04) 2.14) All women (N=1278) Full 1.22 0(.61, 1.08 (0.54, 1.09 (0.54, 0.97 (0.69, 1.52* (1.00, 1.39 (0.93, model 2.47) 2.17) 2.17) 1.37) 2.32) 2.07) white (N=664) Full 1.00 (0.48, 0.89 (0.46, 0.89 (0.46, 0.94 (0.64, 1.38 (0.92, 1.84* (1.23, Model 2.12) 1.72) 1.71) 1.39) 2.06) 2.75) AA/Black (N=518) Full 1.61 (0.23, 1.91 (0.38, 1.91 (0.38, 0.64 (0.14, 1.52 (0.18, 1.67 (0.43, Model- 11.35) 9.57) 9.57) 2.95) 12.68) 6.47 others (N=96) Full Model = Gestational age, maternal age, parity, Medicaid, smoking, pre-pregnancy BMI. *p<.05 from proc survey logistic. 44 Table 2.5: Associations between angiogenic biomarkers at mid-pregnancy and odds being in top quartile of MAP at 20 weeks, and top quartile of highest MAP in pregnancy. POUCH Study 1278 pregnant women. Stratified by gestational age at of measurement 1 unit increase Log Endoglin 1 unit increase Log SFLT_1 1 unit decrease Log PLGF MODELS OR 95% CI OR 95% CI OR 95% CI OR 95% CI OR 95% CI OR 95% CI MAP High MAP High MAP High ≥84.7 MAP ≥84.7 MAP ≥84.7 MAP at 20 ≥95.3 at 20 ≥95.3 at 20 ≥95.3 weeks weeks weeks Full .69 (0.30, 0.66 (0.29, 0.95 (0.64, 0.69 (0.46, 1.37 (0.85, 1.28 (.80, 2.03) Model*, < 1.59) 1.50) 1.46) 1.03) 2.21) 22.4 wks gestation (N= 621) Full 1.41 (0.73, 1.54 (0.78, 0.97 (0.66, 1.09 (.74, 1.53* (1.01, 1.74* (1.18, Model*, >= 2.79) 3.03) 1.43) 1.62) 2.30) 2.57) 22.4 wks gestation (N = 657) Full Model = Gestational age, demographics, smoking, pre-pregnancy BMI. *p<.05 from proc survey logistic. 45 CHAPTER 3: ELEVATED MEAN ARTERIAL PRESSURE AND PREGNANCY OUTCOMES, WHAT CAN ANGIOGENIC BIOMARKERS TELL US? 46 Introduction In 2017, the American Heart Association (AHA) and the American Cardiology Association (ACA) released new blood pressure (BP) guidelines defining stage 1 hypertension as a BP of 130-139 mmHg with a diastolic of 80-89 mmHg and elevated BP as a systolic of 120-129 mmHg(1). The American College of Obstetrics and Gynecology (ACOG) however continues to use the cutoff of 140/90 mmHg to define chronic hypertension (cHtn-hypertension that is identified before 20 weeks of gestation), gestational hypertension (gHtn-new onset of hypertension after 20 weeks of gestation), and in the definition of preeclampsia (PE-new onset hypertension with either new onset proteinuria or evidence of new renal, cerebrovascular, hematologic, or liver involvement). In one large cohort study, the application of the ACC/AHA 2017 definitions during pregnancy more than doubled the prevalence of hypertension (from 10.3 % to 28.1%) (2). If this is true in general, more than 1 of 4 pregnant people in the US would be considered to have hypertension in pregnancy. Since 2017, there have been several studies examining the impact of ACC/AHA stage 1 hypertension on pregnancy complications and outcomes- including preeclampsia (PE), preterm delivery before 37 weeks gestation, and poor fetal growth resulting in a small for gestational age (SGA) neonate. Multiple studies have documented the association between ACA/AHA stage 1 hypertension during pregnancy and the development of preeclampsia (2–7). While associations between ACA/AHA stage 1 hypertension and other pregnancy outcomes are not as strong as for stage 2 hypertension (i.e., BP >140/90), stage 1 hypertension has been associated with intrauterine growth restriction/IUGR and small for gestational age/SGA (5), preterm delivery(3,5,6), and composite neonatal outcomes(2,4). Increases in mean arterial pressure 47 (MAP;- estimated as systolic BP + 2x diastolic BP/3) also have been associated with increased risk of placental vascular complications including PE, IUGR and stillbirth(8). Due in part to concerns that lower BP or MAP in pregnancy could lower placental perfusion, the ACOG guidelines have not yet adopted the BP cutoffs now recommended by the ACA/AHA for non-pregnant populations; continued research in this area is needed in order to provide sound clinical guidance (9). In addition, the application of AHA/ACA guidelines to BP during pregnancy would result in a large number of pregnancies being re-classified as higher risk; thus, more research is needed to identify which of these pregnancies truly are at greater risk for outcomes such as poor fetal growth or preterm birth. Both fetal growth and preterm delivery are impacted by the health of the placenta, and PE has often been described as a disorder of placental origin(10,11) . In a previous study, we found that elevated maternal BP in pregnancy (with or without a PE diagnosis) in combination with vascular placental pathology was associated with increased odds of developing hypertension later in life (12). While we cannot directly observe the placenta during pregnancy, there are maternal blood biomarkers that have been linked to both placental health and PE; these include placental growth factor (Plgf), soluble fms-like tyrosine kinase (sFlt-1) and soluble endoglin (sEng) (13). The sflt-1 and Plgf already in use in research and some clinical settings as markers of elevated risk for developing PE(14), and can be helpful in identifying individuals with cHtn who are at greatest risk for developing PE and/or other adverse outcomes(15,16). In this study, we assess relationships among maternal blood levels of sEng, sFlt-1 and Plgf and MAP measured at mid-pregnancy and pregnancy outcomes (preterm birth; SGA birth). This is a ‘first step’ in determining whether these biomarkers might aid in risk stratification among 48 women with and without higher MAP. We first consider all pregnancies and then only those without a diagnosis of PE, gHtn, or cHtn. The data are from the Pregnancy Outcomes and Community Health (POUCH) Study, a prospective cohort that includes biomarker measurements, detailed demographics, and BP measures in over 1000 participants Methods Study sample The POUCH (Pregnancy Outcomes and Community Health) Study enrolled a total of 3019 pregnant people at 16-27 weeks gestation from 52 clinics in 5 Michigan communities between 1998 and 2004. Pouch inclusion criteria were singleton pregnancy, having had maternal serum alpha-fetoprotein (MSAFP) screening at 15-22 weeks, being at least 15 years of age, speaking English, and having no diagnosis of pre-pregnancy diabetes. The study protocol included maternal questionnaires and an interview, collection of biologic samples, and follow-up to delivery (18). A sub-cohort of 1,371 participants, which included all those who delivered preterm or had elevated MSAFP along with a random sample of term births (with oversampling of African Americans), had their medical records abstracted, additional biomarkers measured, and placentas examined. Our analysis excludes the 64 participants who did not have sufficient blood samples collected at study intake and an additional 46 who did not have BP measurements close to 20 weeks gestation in the medical records (n = 1261). Exposure measures Mean Arterial Pressure (MAP) was calculated from the systolic and diastolic blood pressures recorded in the prenatal medical record at the visit closest to 20 weeks gestation (range 16-24 weeks for 98% of the sample), estimated by the formula MAP = systolic blood pressure + 2 x 49 diastolic blood pressure/3. During the second trimester (13-26 weeks) blood pressure is usually at its lowest, and by using blood pressure measured during this time, rather than the highest blood pressure recorded over the course of pregnancy, we avoid differences in blood pressure linked to differences in length of gestation. Endoglin (sEng), Placental growth factor (Plgf) and Soluble FMS-like tyrosine kinase-1 (sFlt-1) were measured in the maternal serum specimens collected at the study visit (16-27 week’s gestation) and stored at -80 degrees Celsius. Measurements were made with commercially available kits and conducted by the Karumanchi Laboratory while blinded to clinical information (17). Two assays were run for each biomarker, and the average was used in analysis. A sample consisting of 778 subcohort members (those who delivered a term infant that was greater than 10th percentile of weight for gestational age and were not diagnosed with hypertension or PE) was used to determine the cut points for upper quartile of MAP, endoglin and sFlt-1 and the lower quartile of Plgf. Women in the upper quartile of MAP, endoglin and sFlt-1 are referred to as ‘high’ in their respective categories; the lower three quartiles are considered ‘not high’. For Plgf we use the lower quartile as ‘low’ and the upper three quartiles ‘not low.’ We then created a combination variable for each biomarker with MAP, i.e., MAP ‘not high’/ biomarker ‘not high’ (or in the case of Plgf, ’not low’), MAP ‘not-high’/biomarker ‘high’, MAP ‘high’/biomarker ‘not high’, MAP ‘high/biomarker ‘high’. Hypertensive Disorders Hypertensive disorders were determined by review of prenatal records by trained abstractors. Participants with blood pressures greater than or equal to 140 systolic or 90 diastolic before 20 weeks gestation, and those with a pre-pregnancy diagnosis of hypertension were defined as 50 having chronic hypertension (cHtn). Gestational hypertension (gHtn) was defined as at least two blood pressure readings greater than or equal to 140 systolic or 90 diastolic first noted after 20 weeks gestation in the absence of proteinuria. Preeclampsia (PE) was defined as gestational hypertension plus proteinuria (2+ on urine dipstick on one occasion or 1+ on two occasions) or new onset or worsening proteinuria after 20 weeks. Outcome measures Pregnancy outcomes were extracted from medical records by trained study staff. Gestational age at delivery was calculated either by date of last menstrual period (LMP), or in cases where gestational age estimated by ultrasound differed by more than 2 weeks, the ultrasound estimate was used (of note, ultrasound results <25 weeks were available for 93% of participants). Birth weights were obtained from delivery records, and size for gestational age was calculated using a national reference sample that is sex specific and corrected for implausible gestational age estimates (18). Small for gestational age (SGA) was defined as <10th percentile and large for gestational age (LGA) >90th percentile. Covariates Covariates considered for adjusted models were maternal characteristics that: 1) have documented relations to BP in pregnant and non-pregnant populations (19,20); 2) are associated with biomarker levels in the POUCH study(17); 3) have documented associations with preterm delivery and/or SGA; and 4) pre-date the pregnancy. Maternal age (continuous), parity (primiparous vs. multiparous), health insurance (Medicaid yes/no-used as a proxy measure for family income), education, smoking status and race/ethnicity were obtained through self-report. Because birth outcomes among pregnant people not categorized as non- 51 Hispanic White or Black/African American most closely resembled those of non-Hispanic White participants, these individuals were grouped with the non-Hispanic White participants Maternal self-report of smoking history was dichotomized as those who never smoked or quit by mid- pregnancy vs. those who smoked throughout pregnancy. Pre-pregnancy BMI (kg/m2) was calculated from self-reported weight and corroborated data from medical records. Analytic plan All analyses were performed with survey procedures in SAS v 9.4 (SAS Institute Inc., Cary, NC) and weighted to account for the POUCH Study design and sampling scheme. Distributions of all three biomarkers were skewed; therefore, we used log transformed biomarkers in all analyses and back transformed the geometric means (and 95% confidence intervals) in the tables. Survey multinomial logistic techniques were used to compare the odds of having an SGA or LGA infant for each of the MAP/biomarker groups, with the ‘not high’ MAP-‘not high’ biomarker group as the referent. Contrast statements were used to investigate differences between the groups with MAP ‘high’/biomarker ‘not high’ (or in the case of Plgf, ‘low’/‘not low’) and those with MAP ‘high’/ biomarker ‘high’ (or ‘low’/‘low’ for Plgf) biomarkers in order to test our hypothesis that biomarkers could be of use in identifying a subset of those with ‘high’ MAP with more likely to have PTD or deliver SGA. After examining the unadjusted associations, the analysis was repeated adjusting for maternal age, gestational age at which the biomarker was measured, parity, smoking, Medicaid Insurance status, race, and pre-pregnancy BMI. Education was not used in final analysis as it was not associated with our exposures or outcomes once we adjusted for maternal age and Medicaid insurance status. We then repeated analyses two times: first, after removing those with PE, and then removing those with gHtn, cHtn or PE. 52 Survey regression analysis was used to examine differences in gestational age at delivery between the groups, following the same approach. Results Participants in the analytic sample, (n=1261) had endoglin, sFlt_1 and Plgf blood levels measured at 15-27.5 weeks gestation with a mean of 22.4 weeks (table 3.1). The mean age of participants was 26.6 years, and the mean pre-pregnancy BMI was 26.8 kg/m2. In this sample, 41% were primiparous, 24% were Black 41% were insured through Medicaid, and 17% were current smokers. Overall, maternal characteristics of the analytic sample were similar to that of the whole subcohort (n=1371). Mean MAP in the analytic sample was 80.2 mmHg. The geometric means for endoglin, sflt-1, and Plgf were 5.37 ng/ml, 1669.03 ng/ml, and 368.71 pg/ml, respectively (table 3.1). Of the 10% with a hypertensive disorder, 3% developed PE, 4% developed gHtn and 3% had cHtn. As mentioned above, a sub-sample consisting of the778 participants with no cHtn, gHtn, PE, SGA infants or delivery before 37 weeks was used to set the quartile cut points for MAP and biomarkers. The mean MAP for the this group was 79.8 mmHg and the upper quartile of MAP > 85.1 mmHg (for reference the MAP equivalent of a blood pressure of 120/80 is 93.3 mmHg, and of 120/60 is 80 mmHg, and 20 weeks gestation corresponds to the lowest MAP for most pregnancies). The cut points for the upper quartiles of endoglin and sFlt1 were 6.11 ng/ml and 2,514.93 ng/ml respectively. The cut point for lowest quartile Plgf was 247.15 pg/ml. For all the MAP/biomarker combinations, close to 50% of the participants fell into the ‘not high’ MAP/’not high’ biomarker group (table 3.2). The smallest groups were those with ‘high’ MAP and ‘high’ (or ‘low’ for Plgf) biomarkers, e.g. ‘high’ MAP/ ‘high’ sFlt1 included 88 women and ‘high’ 53 MAP/’low’ Plgf included 133 women. As expected, when we removed participants who developed PE, and then all who were diagnosed with any type of hypertension, the ‘high’ MAP/’high’ biomarker groups became even smaller, the latter ranging from 53 to 82 participants . Delivery of an SGA infant The proportion of SGA infants in our analytic sample was 9% (table 3.1). In unadjusted models, participants with a MAP ‘not high’ but endoglin ‘high’ (defined as 4 th quartile for normal pregnancies) had an OR of 1.88 (95% CL 1.11, 3.20) for the delivery of an SGA infant compared to those with both MAP ‘not high’ and endoglin ‘not high’ (table 3.3). The association was of a similar magnitude after adjustment for maternal age, parity, Medicaid status, race, smoking and pre-pregnancy BMI, but attenuated to non-significance once we removed pregnancies with PE or any hypertensive disorder. The group with both MAP ‘high’ and endoglin ‘high’ also had higher odds of SGA than those with MAP ‘not high’/endoglin ‘not high’, although neither this nor the contrast between the MAP ‘high’/endoglin ‘high’ group and the MAP ‘high’/endoglin ‘not high’ group were statistically significant. There were no significant associations between the endoglin/MAP groups and delivery of an LGA (>10th percentile weight for gestational age) infant. The analysis looking at groups based on sFlt-1 did not show any significant associations. Analyses based on combinations of MAP and Plgf revealed no significant differences in odds of delivering an SGA infant in unadjusted models. In adjusted models, however, compared to those with MAP ‘not high’/Plgf ‘not low’ two groups had an elevated odds of delivering an SGA infant: those with MAP ‘not high’/Plgf ‘low’ (OR=2.26, 95% CI 1.22, 4.21), and those with MAP high/Plgf low (OR= 2.45, 95% CI 1.07, 5.42). In addition, the odds of having an SGA infant among 54 those with MAP ‘high’/Plgf ‘low’ were 2.89 (95% CI 1.16, 7.19) times the odds of having an SGA infant among those with MAP ‘high’/Plgf ‘not low’. These associations remained statistically significant when we excluded those who developed PE, and when we excluded those who developed any hypertensive disorders. Additionally, in adjusted models the group with MAP ‘not high’, but Plgf ‘low’ had lower odds of delivering an LGA infant than those who had MAP ‘not high’ and Plgf, even after excluding women with gestational hypertension, chronic hypertension or preeclampsia (OR = 0.46, 95% CI 0.22, 0.98). Gestational age at delivery The mean gestational age at delivery in our analytic sample was 39.1 weeks (table 3.1). Endoglin ‘high’ (i.e. in the upper quartile) was associated with shorter length of gestation regardless of MAP (table 3.4). In the adjusted model, those with MAP ‘not high’/endoglin ‘high’ delivered 0.42 weeks earlier (95% CI -0.77, -0.19) than those with MAP ‘not high’/endoglin ‘not high’, and those with MAP ‘high’/endoglin ‘high’ delivered 0.85 weeks (95% CI -1.36, -0.34) earlier. These associations attenuated but remained significant when we removed those who developed PE and approached statistical significance when we removed those with any hypertensive disorders during pregnancy. The contrast between those with MAP ‘high’/endoglin ‘high’ and those with MAP ‘high’/endoglin ’not high’ was significant in the adjusted model, but not once we removed those with hypertensive disorders. We did not find any statistically significant differences in length of gestation between the different sFlt-1 groups. For groups based on MAP and Plgf, the group with MAP ‘high’/Plgf ‘low’ delivered 0.28 (95% CI -0.67, 0.09) weeks earlier in the unadjusted model than those with MAP ‘not high’, and Plgf 55 ‘low’ Plgf, and 0.43 weeks earlier (-0.85, 0.00) in the adjusted model. These findings appear to have been driven mainly by women with PE since the statistically significant associations were no longer present after we removed those participants from the analysis. The contrast comparing individuals with MAP ‘high’/Plgf ‘low’ and individuals with MAP ‘high’/Plgf ‘not low’ was not statistically significant. Discussion We used a prospective cohort of 1261 pregnancies to assess outcomes (SGA, LGA, and length of gestation) in relation to maternal MAP and blood levels of endoglin, sFlt1 and Plgf measured at mid-pregnancy. Of the three biomarkers, ‘low’ Plgf levels stood out as being most strongly related to SGA and ‘high endoglin’ levels most strongly related to lower gestational age at delivery. Pregnancies with MAP ‘high’ (upper quartile) and Plgf ‘low’ (lower quartile) had greater than twice the odds of delivering an SGA infant compared with women who had a more favorable profile (i.e. both MAP ‘not high’ and Plgf ‘not low’). Focusing just on women with MAP ’high’, those with Plgf ‘low’ had greater than threefold increased odds of delivering an SGA infant compared to those with Plgf ‘not low’. Thus, maternal Plgf levels showed promise for risk stratifying women with ‘high’ MAP. Women with the combination MAP ‘not high’ and Plgf ‘low’ also were at greater risk of delivering an SGA infant. Importantly, our findings persisted when we removed women with hypertensive disorders developed before or during pregnancy. We found no associations between sFlt1 and SGA. We also examined MAP/biomarker combinations in relation to length of gestation. Women with ‘High’ endoglin levels had shorter gestations irrespective of MAP being high or not. This association was markedly reduced after removing women with hypertensive disorders. Plgf and 56 sFlt1 appeared unrelated to length of gestation. The placenta is the main source of all three biomarkers analyzed in this study, and each is associated with PE(21–24); however, they differ in their role in angiogenesis and patterns of change throughout pregnancy. Within the placenta, Plgf and endoglin, along with other factors, promote new blood vessel growth and stabilize vascular networks, while sFlt-1 binds to Plgf and Vascular endothelial growth factor (VEGF)and acts as an antagonist(25). In maternal circulation, higher levels of both sFlt1 and soluble endoglin are linked to maternal vascular dysfunction (25). In normal pregnancies, Plgf increases until 30 weeks gestation and then begins to decrease(26), and differences in Plgf between those who go on to develop PE and those who do not may be noticeable as early as 11-13 weeks gestation (27). Endoglin and sFlt1 increase throughout gestation and differences in levels of sFlt1 and sEng between women who go on to develop PE and those who do not become significant after 19 weeks or later (21,28). Serum concentrations of all three biomarkers have been found to be lower in those with higher BMI and higher among African American women, while multiparity is associated with lower sFlt1, and smoking is associated with lower Plgf and higher endoglin (17). Our study is the first that we know of to examine how these three mid-pregnancy maternal biomarkers in combination with MAP might relate to pregnancy outcomes, even among women without hypertensive disorders. The association between low Plgf and SGA has been noted in other cohorts(29,30), and a recent meta-analysis of Plgf as a predictor of outcomes for women with hypertensive disorders in pregnancy concluded that there was moderate to high evidence that the low Plgf identifies pregnancies at risk for preterm delivery or adverse neonatal outcomes(16). In our study, low Plgf was associated with SGA even among women who had 57 elevated MAP but no diagnoses of a hypertensive disorder. We did not find an association between low Plgf and shorter length of gestation. Preterm delivery is a complex and multifactorial phenomenon(31). Lower Plgf has been associated with infants who are small due to placental pathology (which might then lead to either medically indicated or spontaneous preterm birth)(32) yet might also be associated with constitutionally smaller infants with a smaller placenta that is otherwise healthy (which might not lead to preterm delivery) . Other common causes of preterm delivery (such as infections or short cervix) most likely are not associated with Plgf nor with MAP and would not be expected to differ by groups based on quartiles of MAP or Plgf. In one recent study, both Plgf and MAP were important components of first trimester predictive models of a composite outcome of placental vascular conditions (defined as gHtn, PE, intrauterine growth restriction, placental abruption or still birth) (8). Our study used the lowest quartile of Plgf, but in a larger cohort study the lowest 2.5th quantile of the ratio of Plgf/VEGF-1 at 20-23 weeks was associated with placental pathology(30). The association between high endoglin and delivery of an SGA infant has been noted before(33,34), even in those who do not develop PE or gestational hypertension. In our data this association was not predicated on MAP levels and did not rise to the level of statistical significance though odds ratios were slightly elevated .While our finding that sFlt-1 was not associated with SGA or length of gestation this may be because sFlt-1 does not differ between normal and abnormal pregnancies until later in gestation(28). However, other research has reported that while sFlt-1 can be helpful in identifying risk of PE it is not correlated with SGA(34). 58 Limitations and strengths Our study did not include repeated measures of biomarkers, and thus we are not able to determine when in gestation levels of Plgf begin to differ between those who go one to have an SGA infant and those who do not. This could be important if early interventions such as aspirin, which may improve placental circulation(35), are to be considered. We abstracted blood pressures from medical records, and blood pressure measurements in clinical taken in clinical settings are thought to overestimate both systolic and diastolic BP(36). However, it is unlikely overestimates would be related to our biomarkers or outcomes of interest. In subgroups defined by quartiles of MAP and biomarkers, sample sizes were limited; thus, we could not conduct meaningful analyses of effect modifications by maternal characteristics such as BMI or racial/ethnic groups. On the plus side, the cohort was drawn from a multi-community, socio- economically diverse sample that was recruited early in pregnancy. The sample size was large enough for us to examine MAP and three different biomarkers in relation to risk of SGA and preterm delivery after removing women with hypertensive disorders. This provided important clues as to how we might risk stratify women with higher BP who do meet current BP cutoffs used to define hypertension in pregnancy. Conclusion The obstetric community has been grappling with the potential risks and benefits of lowering the threshold of MAP or blood pressure to identify pregnancies at risk of adverse outcomes. Maternal blood biomarkers have the potential to improve risk stratification along with BP. If the criteria for defining chronic and gestational hypertension in pregnancy is lowered, our finding that women with higher MAP and but normal Plgf at mid pregnancy do not appear to be at 59 greater risk for SGA than those with normal MAP may be helpful in distinguishing which women should be monitored more closely. In our study, this relationship became more apparent once we adjusted for maternal characteristics- suggesting that any clinical application may need to utilize approaches such as multiples of the mean as has been done in larger cohorts(37) and established prenatal screening. Studies in larger cohorts are needed to more clearly identify thresholds of Plgf that are associated with delivery of an SGA infant. 60 REFERENCES 1. Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: Executive summary: A report of the American college of cardiology/American Heart Association task . 2018 1269–1324 . 2. Bello NA, Zhou H, Cheetham TC, et al. Prevalence of Hypertension among Pregnant Women When Using the 2017 American College of Cardiology/American Heart Association Blood Pressure Guidelines and Association with Maternal and Fetal Outcomes. JAMA Netw. Open. 2021;4(3). 3. Greenberg VR, Silasi M, Lundsberg LS, et al. Perinatal outcomes in women with elevated blood pressure and stage 1 hypertension. Am. J. Obstet. Gynecol. 2021;224(5):521.e1- 521.e11. 4. Sutton EF, Rogan SC, Lopa S, et al. Early Pregnancy Blood Pressure Elevations and Risk for Maternal andNeonatal Morbidity. Obstet. Gynecol. 2020;136(1):129–139. 5. Darwin KC, Federspiel JJ, Schuh BL, et al. ACC-AHA Diagnostic Criteria for Hypertension in Pregnancy Identifies Patients at Intermediate Risk of Adverse Outcomes. Am. J. Perinatol. 2021;38:E249–E255. 6. Sutton EF, Hauspurg A, Caritis SN, et al. Maternal outcomes associated with lower range stage 1 hypertension. In: Obstetrics and Gynecology. Lippincott Williams and Wilkins; 2018:843–849. 7. Sabol BA, Porcelli B, Diveley E, et al. Defining the risk profile of women with stage 1 hypertension: a time to event analysis. Am. J. Obstet. Gynecol. MFM. 2021;3(4). 8. Monari F, Spano’ Bascio L, Banchelli F, et al. First-trimester prediction model for placental vascular disorders: An observational prospective study. Pregnancy Hypertens. [electronic article]. 2022;28(January):35–40. (https://doi.org/10.1016/j.preghy.2022.02.002) 9. Agrawal A, Wenger NK. Hypertension During Pregnancy. Curr. Hypertens. Rep. 2020;22(9). 10. Koenen S V., Nikkels PGJ, Veerbeek JHW, et al. Spiral artery remodeling and maternal cardiovascular risk. J. Hypertens. 2016;34(8):1570–1577. 11. Karumanchi SA, Granger JP. Preeclampsia and Pregnancy-Related Hypertensive Disorders. Hypertension. 2016;67(2):238–242. 61 12. Holzman CB, Senagore P, Xu J, et al. Maternal risk of hypertension 7–15 years after pregnancy: clues from the placenta. BJOG An Int. J. Obstet. Gynaecol. 2021;128(5):827– 836. 13. Widmer M, Cuesta C, Khan KS, et al. Accuracy of angiogenic biomarkers at ≤20 weeks’ gestation in predicting the risk of pre-eclampsia: A WHO multicentre study. Pregnancy Hypertens. 2015;5(4). 14. Poon LC, Magee LA, Verlohren S, et al. A literature review and best practice advice for second and third trimester risk stratification, monitoring, and management of pre- eclampsia: Compiled by the Pregnancy and Non-Communicable Diseases Committee of FIGO (the International Federation of Gyneco. Int. J. Gynecol. Obstet. 2021;154(S1):3–31. 15. Verlohren S, Brennecke SP, Galindo A, et al. Clinical interpretation and implementation of the sFlt-1/PlGF ratio in the prediction, diagnosis and management of preeclampsia. Pregnancy Hypertens. [electronic article]. 2022;27:42–50. (https://linkinghub.elsevier.com/retrieve/pii/S2210778921005535) 16. Ukah UV, Hutcheon JA, Payne B, et al. Placental Growth Factor as a Prognostic Tool in Women With Hypertensive Disorders of Pregnancy: A Systematic Review. Hypertens. (Dallas, Tex. 1979). 2017;70(6):1228–1237. 17. Mijal RS, Holzman CB, Rana S, et al. Midpregnancy levels of angiogenic markers in relation to maternal characteristics. Am. J. Obstet. Gynecol. [electronic article]. 2011;204(3):244.e1-244.e12. (http://dx.doi.org/10.1016/j.ajog.2010.10.001) 18. Talge NM, Mudd LM, Sikorskii A, et al. United states birth weight reference corrected for implausible gestational age estimates. Pediatrics. 2014;133(5):844–853. 19. Kurian AK, Cardarelli KM. Racial and ethnic differences in cardiovascular disease risk factors: a systematic review. Ethn. Dis. 2007;17(1):143–152. 20. Salles GF, Schlüssel MM, Farias DR, et al. Blood pressure in healthy pregnancy and factors associated with no mid-trimester blood pressure drop: A prospective cohort study. Am. J. Hypertens. 2015;28(5):680–689. 21. Kleinrouweler CE, Wiegerinck MMJ, Ris-Stalpers C, et al. Accuracy of circulating placental growth factor, vascular endothelial growth factor, soluble fms-like tyrosine kinase 1 and soluble endoglin in the prediction of pre-eclampsia: A systematic review and meta- analysis. BJOG An Int. J. Obstet. Gynaecol. 2012;119(7):778–787. 62 22. Nolasco-Leaños AG, Ramírez-Valenzuela KL, Carrillo-Juárez RI, et al. Circulating angiogenic factors are associated with progression to preeclampsia and the occurrence of adverse outcomes in women with gestational hypertension. J. Hypertens. 2021;39(3):573–580. 23. Anderson UD, Olsson MG, Kristensen KH, et al. Review: Biochemical markers to predict preeclampsia. Placenta [electronic article]. 2012;33(SUPPL.):S42–S47. (http://dx.doi.org/10.1016/j.placenta.2011.11.021) 24. Rana S, Burke SD, Karumanchi SA. Imbalances in circulating angiogenic factors in the pathophysiology of preeclampsia and related disorders. Am. J. Obstet. Gynecol. [electronic article]. 2022;226(2):S1019–S1034. (https://doi.org/10.1016/j.ajog.2020.10.022) 25. Umapathy A, Chamley LW, James JL. Reconciling the distinct roles of angiogenic/anti- angiogenic factors in the placenta and maternal circulation of normal and pathological pregnancies. Angiogenesis [electronic article]. 2020;23(2):105–117. (https://doi.org/10.1007/s10456-019-09694-w) 26. Saffer C, Olson G, Boggess KA, et al. Determination of placental growth factor (PlGF) levels in healthy pregnant women without signs or symptoms of preeclampsia. Pregnancy Hypertens. [electronic article]. 2013;3(2):124–132. (http://dx.doi.org/10.1016/j.preghy.2013.01.004) 27. Wright D, Tan MY, O’Gorman N, et al. Serum PlGF compared with PAPP-A in first trimester screening for preterm pre-eclampsia: Adjusting for the effect of aspirin treatment. BJOG An Int. J. Obstet. Gynaecol. 2022;(December 2021):1–10. 28. Noori M, Donald AE, Angelakopoulou A, et al. Prospective study of placental angiogenic factors and maternal vascular function before and after preeclampsia and gestational hypertension. Circulation. 2010;122(5):478–487. 29. Odibo AO, Patel KR, Spitalnik A, et al. Placental pathology, first-trimester biomarkers and adverse pregnancy outcomes. J. Perinatol. 2014;34(3):186–191. 30. Korzeniewski SJ, Romero R, Chaiworapongsa T, et al. Maternal plasma angiogenic index-1 (placental growth factor/soluble vascular endothelial growth factor receptor-1) is a biomarker for the burden of placental lesions consistent with uteroplacental underperfusion: A longitudinal case-cohort study. Am. J. Obstet. Gynecol. [electronic article]. 2016;214(5):629.e1-629.e17. (http://dx.doi.org/10.1016/j.ajog.2015.11.015) 63 31. Doi K. Multiple Mechanisms of Preterm Labor Other Than Intrauterine Infection BT - Preterm Labor and Delivery. In: Sameshima H, ed. . Singapore: Springer Singapore; 2020:89–94.(https://doi.org/10.1007/978-981-13-9875-9_8) 32. Benton SJ, McCowan LM, Heazell AEP, et al. Placental growth factor as a marker of fetal growth restriction caused by placental dysfunction. Placenta. 2016;42:1–8. 33. Dirlikov E, Torres-Aponte J, Thomas D, et al. Local Transmission of Zika Virus — Puerto Rico ,. Morb. Mortal. Wkly. Rep. 2016;65:1–5. 34. Romero R, Nien JK, Espinoza J, et al. A longitudinal study of angiogenic (placental growth factor) and anti-angiogenic (soluble endoglin and soluble vascular endothelial growth factor receptor-1) factors in normal pregnancy and patients destined to develop preeclampsia and deliver a small for. J. Matern. Neonatal Med. 2008;21(1):9–23. 35. Roberge S, Bujold E, Nicolaides KH. Meta-analysis on the effect of aspirin use for prevention of preeclampsia on placental abruption and antepartum hemorrhage. Am. J. Obstet. Gynecol. 2018;218(5):483–489. 36. Lane-Cordova AD, Wilcox S, Fernhall B, et al. Agreement between blood pressure from research study visits versus electronic medical records and associations with hypertensive disorder diagnoses in pregnant women with overweight/obesity. Blood Press. Monit. 2021;341–347. 37. Tarca AL, Taran A, Romero R, et al. Prediction of preeclampsia throughout gestation with maternal characteristics and biophysical and biochemical markers: a longitudinal study. Am. J. Obstet. Gynecol. 2022;226(1):126.e1-126.e22. 64 APPENDIX Table 3.1: Demographic Characteristics and Biomarkers, 1261 Pregnancies POUCH Study, Michigan 1998-2004 Characteristic Gestational Age enrollment Mean (95% CL) 22.4 weeks (22.2, 22.5) Range 15.0-27.5 weeks Maternal Age Mean (95%CL) 26.6 years (26.2, 26.9) Range 15.3-47.3 years Parity N (%)* Primiparous 530 (41%) Multiparous 730 (59%) Race/Ethnicity N (%)* White/other 757 (75%) African American 504 (25%) Medicaid N (%)* No 561 (51%) Yes 699 (49%) Smoking N (%)* Nonsmoker/quit before 20 weeks 1029 (83%) Smoked 232 (17%) Pre-Pregnancy BMI Mean (95% CL) 26.8 (26.3, 27.3) Range 14.1-63.6 Hypertensive Disorder N (%)* None 1117 (90%) Preeclampsia 44 (3%) Gestational Htn 56 (4%) Chronic Htn 44 (3%) MAP at 20 weeks Mean (95% CL) 80.2 mmHg (79.7, 80.8) Range 50.7 – 118.7 Biomarker Mean** (95% CL) log Endoglin 5.37 ng/ml (5.26, 5.47) log Sflt-1 1669.03 ng/ml (1619.71, 1737.15) log Plgf 368.71 pg/ml (354.25, 387.61) Size for Gestational Age N (%)* Appropriate for Gest Age 1000 (79%) Small for Gestational Age 131 (9%) Large for Gestational Aga 128 (13%) Length of Gestation Mean (95% CI) 39.1 weeks (39.0, 39.2) Range 21.4-44.9 weeks * % weighted to reflect sampling frame, **geometric mean 65 Table 3.2: Biomarker/MAP groups: All Pregnancies, Pregnancies without Preeclampsia, Pregnancies without Chtn, Ghtn or Preeclampsia All Pregnancies Without PE Without Chtn, Ghtn, or PE Biomarker Group-Endoglin** Frequency (%)* Frequency (%)* Frequency (%)* MAP not high/Endoglin not high 595 (48.6%) 592 (48.5%) 581 (47.9%) MAP not high/Endoglin high 284 (19.4%) 278 (18.9%) 274 (18.6%) MAP high/Endoglin not high 261 (23.5%) 248 (22.6%) 197 (17.9%) MAP high/Endoglin high 121 (8.5%) 99 (7.2%) 65 (5.3%) Total 1261 (100%) 1217 (97.3%) 1117 (89.9%) Biomarker Group-sFlt_1** MAP not high/sFlt_1 not high 622 (49.6%) 618 (49.5%) 607 (49.0%) MAP not high/sFlt_1 high 257 (18.4%) 252 (18.0%)) 248 (17.6%) MAP high/sFlt_1 not high 294 (25.4%) 270 (23.9%) 209 (18.8%) MAP high/sFlt_1 high 88 (6.6%) 77 (6.0%) 53 (4.5%) Total 1261 (100%) 1217 (97.3) 1117 (89.9%) Biomarker Group-Plgf*** MAP not high/Plgf not low 695 (52.2%) 692 (52.0%) 679 (51.2 %) MAP not high/ Plgf low 182 (15.6%) 176 (15.3%) 174 (15.2%) MAP high/Plgf not low 248 (20.7%) 235 (19.9%) 179 (15.5%) MAP high/Plgf low 133 (11.4%) 111 (10.1%) 82 (7.9%) Total 1258 (100%) 1214 (97.3%) 1114 (89.8%) *Percent weighted to reflect sampling frame **Not high = Lower 3 quartiles, High= upper quartile ***Not low = Upper 3 quartiles, Low = lowest quartile. 66 Table 3.3: Biomarker/MAP groups and size for gestational age-Odds of SGA** or LGA** vs AGA. Pouch Subcohort 1261 Pregnancies with MAP at ~20 weeks, and biomarker measurements available. All Pregnancies All Pregnancies Pregnancies Pregnancies Unadjusted Adjusted * w/out PE w/out PE, cHtn, Adjusted* gHtn, Adjusted* Endoglin Group OR (95% CL) aOR (95% CL) aOR (95% CL) aOR (95% CL) #1 MAP not high/ Reference Reference Reference Reference Endoglin not high #2 MAP not high/ SGA 1.88 (1.11, 3.20) 1.90 (1.10, 3.23) 1.54 (0.87, 2.73) 1.48 (0.83, 2.64) Endoglin high LGA 1.21 (0.69, 2.13) 1.23 (0.69, 2.24) 1.32 (0.73, 1.39) 1.35 (0.74, 2.47) #3 MAP high/ SGA 0.74 (0.38, 1.47) 1.12 (0.56, 2.24) 1.19 (0.58, 2.43) 1.31 (0.64, 2.71) Endoglin not high LGA 1.08 (0.62, 2.97) 0.71 (0.38, 1.31) 0.73 (0.39, 1.37) 0.82 (0.44, 1.54) #4 MAP high/ SGA 1.28 (0.61, 2.67) 1.77 (0.85, 3.71) 2.03 (0.93, 4.46) 1.72 (0.66, 4.50) Endoglin high LGA 1.35 (0.62, 2.97) 1.08 (0.48, 2.39) 1.16 (0.52, 2.61) 0.88 (0.34, 2.27) Contrast #4 vs #3 SGA 1.72 (0.71, 4.14) 1.59 (0.66, 3.83) 1.71 (0.68, 4.33) 1.31 (0.44, 3.90) LGA 1.24 (0.52, 2.97) 1.53 (0.62, 3.74) 1.59 (0.64, 3.93) 1.07 (0.38, 3.02) sFlt_1 Group OR (95% CL) aOR (95% CL) aOR (95% CL) aOR (95% CL) #1 MAP not high/ Reference Reference Reference Reference sFlt_1 not high #2 MAP not high/ SGA 1.50 (0.86, 2.60) 1.17 (0.65, 2.10) 0.86 (0.48, 1.57) 0.82 (0.45, 1.52) sFlt_1 high LGA 1.07 (0.60, 1.91) 1.37 (0.75, 2.51) 1.40 (0.76, 2.57) 1.48 (0.79, 2.77) #3 High MAP/ SGA 0.77 (0.41, 1.43) 1.12 (0.58, 2.18) 1.28 (0.65, 2.53) 1.40 (0.70, 2.78) sFlt_1 not high LGA 1.05 (0.61, 1.81) 0.73 (0.41, 1.23) 0.76 (0.43, 1.34) 0.83 (0.46, 1.49) #4 MAP high/ SGA 0.95 (0.46, 1.99) 1.01 (0.48, 2.13) 0.91 (0.41, 2.05) 0.60 (0.21, 1.67) sFlt_1 high LGA 1.36 (0.57, 3.22) 1.10 (0.40, 3.05) 1.12 (0.40, 3.13) 0.87 (0.24, 3.16) Contrast #4 vs #3 SGA 1.24 (0.53, 2.93) 0.90 (0.38, 2.12) 0.71 (0.29, 1.78) 0.43 (0.14, 1.33) LGA 1.29 (0.51, 3.23) 1.51 (0.53, 4.27) 1.48 (0.52, 4.23) 1.05 (0.28, 3.88) Plgf Group OR (95% CL) aOR (95% CL) aOR (95% CL) aOR (95% CL) #1 MAP not high/ Reference Reference Reference Reference Plgf not low #2 MAP not high/ SGA 1.21 (0.66, 2.23) 2.26 (1.22, 4.21) 2.37 (1.22, 4.60) 2.26 (1.16, 4.42) Plgf low LGA 0.78 (0.39, 0.61 (0.22, 0.97) 0.49 (0.23, 1.03) 0.46 (0.22, 0.98) 1.53) #3 MAP high/ SGA 0.58 (0.31, 1.09) 0.85 (0.44, 1.64) 0.98 (0.50, 1.94) 1.00 (0.50, 1.03) Plgf not low LGA 0.80 (0.43, 1.48) 0.59 (0.30, 1.14) 0.60 (0.31, 1.16) 0.58 (0.29, 1.18) #4 MAP high/ SGA 1.00 (0.47, 2.12) 2.45 (1.07, 5.62) 3.08 (1.27, 7.47) 3.24 (1.21, 8.68) Plgf low LGA 1.41 (0.74, 2.69) 0.61 (0.30, 1.27) 0.68 (0.32, 1.43) 0.66 (0.29, 1.46) Contrast #4 vs #3 SGA 1.71 (0.71, 4.14) 2.89 (1.16, 7.19) 3.13 (1.19, 8.23) 3.23 (1.10, 9.43) LGA 1.76 (0.80, 3.91) 1.04 (0.45, 2.41) 1.14 (0.48, 2.67) 1.12 (0.44, 2.88) * Adjusted for gestational age at measurement of biomarker, maternal age, parity, Medicaid status, race, smoking during pregnancy, pre-pregnancy BMI. **SGA = less than 10th percentile birthweight for gestational age, LGA = greater than 90th percentile birthweight for gestational age. 67 Table 3.4: Biomarker groups and length of gestation in weeks. Pouch Sub cohort 1261 Pregnancies with MAP at ~20 weeks, and biomarker measurements available All Pregnancies All Pregnancies Pregnancies Pregnancies Unadjusted Adjusted * w/out PE w/out PE, cHtn, Adjusted* gHtn, Adjusted* Endoglin Group Weeks Weeks Weeks Weeks (95% CL) (95% CL) (95% CL) (95% CL) #1 MAP not high/ Reference Reference Reference Reference Endoglin not high #2 MAP not high/ -0.53 -0.42 -0.32 -0.28 Endoglin high (-0.82, -0.24) (-0.77, -0.19) (-0.61, -0.04) (-0.57, 0.00) #3 MAP high/ -0.12 -0.20 -0.13 -0.07 Endoglin not high (-0.39, 0.14) (-0.48, 0.07) (-0.40, 0.15) (-0.36, 0.22) #4 MAP high -0.62 -0.85 -0.59 -0.37 /Endoglin high (-1.14,-0.11) (-1.36, -0.34) (-1.07, -0.11) (-0.93, 0.18) Contrast #4 vs #3 -0.50 -0.65 -0.47 -.31 (-1.03, 0.03) (-1.18, -0.11) (-0.95, 0.03) (-0.88, 0.27) sFlt_1 Group #1 MAP not high/ Reference Reference Reference Reference sFlt_1 not high #2 MAP not high/ -0.26 -0.17 -0.08 -0.04 sFlt_1 high (-0.56, 0.04) (-0.48, 0.13) (-0.38, 0.22) (-0.35, 0.26) #3 MAP high/ -0.13 -0.25 -0.17 (-0.45, -0.09 sFlt_1 not high (-0.40, 0.14) (-0.53, 0.04) 0.11) (-0.40, 0.21) #4 MAP high/ sFlt_1 -0.35 -0.43 -0.23 -0.01 high (-0.83, 0.14) (-0.91, 0.05) (-0.65, 0.20) (-0.49, 0.46) Contrast #4 vs #3 -0.21 -0.19 -.06 0.08 (-0.73, 0.31) (-0.69, 0.32) (-0.51, 0.40) (-0.44, 0.59) Plgf Group #1 Normal Reference Reference Reference Reference MAP/Normal Plgf #2 Normal MAP/Low -0.04 -0.02 0.14 0.15 Plgf (-0.39, 0.31) (-0.40, 0.36) (-0.23, 0.51) (-0.23, 0.52) #3 High MAP/Normal 0.16 -0.12 -0.10 -0.00 Plgf (-0.26, 0.29) (-0.41, 0.17) (-0.38, 0.19) (-0.32, 0.31) #4 High MAP/Low -0.28 -0.43 -0.15 -0.04 Plgf (-0.67, 0.09) (-0.85, 0.00) (-0.55, 0.25) (-0.48, 0.41 Contrast #4 vs #3 -0.30 -0.31 -0.05 -0.03 (-0.73, 0.13) (-0.76, 0.14) (-0.48, 0.38) (-0.52, 0.45 *Adjusted for gestational age at measurement of biomarker, maternal age, parity, Medicaid status, race, smoking during pregnancy, pre-pregnancy BMI 68 CHAPTER 4: PREGNANCY BLOOD PRESSURE, ANGIOGENIC BIOMARKERS AND HYPERTENSION 7-15 YEARS LATER 69 Introduction Hypertensive disorders of pregnancy (HDPs)-gestational hypertension (gHtn) and pre- eclampsia(PE) are associated with an increased risk of cardiovascular disease(CVD) later in life(1–5). The link between HDPs and later development of hypertension has been observed repeatedly in cross-sectional(6) and cohort studies(2,6–9). In one large Danish registry based cohort with over 1 million pregnancies, those with a HDP had a 12 to 25 fold higher risk of developing hypertension in the first year after pregnancy compared to those with no HDP, and close to 10 fold higher risk 10 years after an affected pregnancy(2). Others estimate that up to 1/3 of people diagnosed with any type of hypertension during pregnancy go on to develop hypertension within 10 years(7); the increased risk includes those with gestational hypertension without preeclampsia(9). The strength and consistency of the findings in both retrospective and prospective cohorts in a variety of populations led the American College of Obstetricians and Gynecologists (ACOG) to include counseling on lifetime risk of CVD in their recommendations for care of women with HDPs(10). The ACOG defines HDPs, and the diagnosis of chronic hypertension during pregnancy, using cut- off blood pressures (BP) of 140 mmHg systolic and/or 90 mmHg diastolic(10). Beginning in 2017, the hypertension guidelines outside of pregnancy changed, defining stage 1 hypertension as a systolic BP of 130-139 and or/ a diastolic BP of 80-89 mmHg, and elevated blood pressure as systolic BP greater than 120 mmHg(11). Considerably more people have elevated BP or stage 1 hypertension in pregnancy than have HDPs as currently defined. Therefore re-classifying BP during pregnancy would greatly increase the prevalence of hypertension during pregnancy. For example, one study found that changing the classification 70 increased prevalence of hypertension in pregnancy from 10% to 28%(12), and another found 60% of participants in a community based pregnancy cohort had either elevated BP or stage 1 hypertension(13). Recent evidence suggests that stage 1 hypertension during pregnancy is associated with increased risk of preterm birth, delivering a small for gestational age infant and developing pre-eclampsia (14–17). Given the known association between HDPs and later life CVD, those who have stage 1 hypertension and elevated BP during pregnancy also are likely to be at greater risk for developing stage 2 hypertension and later CVD. BPs between 120 and 140 mmHg systolic and 80-90 mmHg diastolic during pregnancy have been associated with twice the odds of hypertension in subsequent pregnancies (18) and approximately 3 times the odds of developing stage 2 hypertension 7 to 15 years after pregnancy(13). Maternal blood levels of angiogenic biomarkers such as soluble FMS-like Tyrosine kinase-1 (sFlt- 1), placental growth factor (PLGF) and endoglin, have been helpful in identifying risk for PE, the HDP most strongly associated with later maternal cardiovascular and cerebrovascular disease) (19). These same biomarkers also might be helpful in identifying an increased risk of later life hypertension among those with BP elevations below the ACOG threshold of 140/90 mmHg during pregnancy. There is a growing body of evidence linking angiogenic biomarkers with CVD risk factors both during (20–22) and after pregnancy(5,23). Unrelated to pregnancy, patients with type 2 diabetes and/or hypertension who had elevated blood endoglin levels are more likely to have signs of end organ damage(24). Even Plgf, which has been studied mostly in pregnancy, recently was found to be altered in patients with metabolic syndrome(25) and was identified as a potential predictor for coronary heart disease among women in the Nurse’s Health Study(26). 71 In our previous work with the POUCH Study pregnancy cohort, we observed that participants with lower mid-pregnancy blood Plgf levels more likely to have Mean Arterial Pressure (MAP) in the upper quartile of our sample. We also found that among participants with moderately elevated blood pressure (SBP >= 120-139 mmHg and/or DBP 80-89 mmHg) those mid- pregnancy Plgf in the lower quartile were more likely to deliver an infant who was small for gestational age (SGA). We also found endoglin in the upper quartile was associated with preterm delivery in all blood pressure groups. Of note we did not find any associations between sFlt-1 measured at mid-pregnancy and MAP, preterm delivery or SGA in our previous studies. Other studies have found sFft-1 levels do not begin to differ between normal pregnancy and pregnancy’s with GH or PE until later in the pregnancy(27)The current study analyzes data from a subgroup of POUCH Study participants followed for their cardiovascular health (POUCHmoms Study). Here we examine whether the Plgf and endoglin measured at mid pregnancy are associated with hypertension 7 to 15 years later. (Given that higher sFlt-1 was not associated with MAP or pregnancy outcomes in our previous work, and may have been measured too early we elected not to include it here). Since clinicians already use history of HDPs for risk stratification of later CVD, the pregnancy Plgf and endoglin levels in these individuals would not alter the recommendation for closer surveillance. However, among people with moderately elevated BP in pregnancy, maternal Plgf and endoglin levels might help identify a subset within this group who are at greater risk of developing later life hypertension. Therefore, this latter group was of particular interest. 72 Methods Study design and analytic sample We used data from the Pregnancy Outcomes and Community Health (POUCH) cohort study, which enrolled 3019 prenatal patients from 52 clinics in Michigan from 1998-2004, in order to examine pathways to preterm birth(28), and the POUCH-moms follow up study which was designed to examine early evidence of cardiovascular disease 7-15 years after the POUCH Study pregnancy. Inclusion criteria for the POUCH Study were singleton pregnancy, no pre-pregnancy diabetes, maternal serum alpha-fetoprotein (MSAFP) screening at 15-22 weeks, age greater than 15 and able to read English. A subcohort of POUCH Study participants was created in order to maximize resources for more detailed evaluations. The subcohort included all those who delivered preterm, or had elevated MSAFP and a random sample (with oversampling of AfricanAmericans) of those who delivered at term. Inverse probability weighting is used in analysis of POUCH Study data to reflect this sampling scheme(29). Subcohort participants (N=1,371) had biomarkers measured in blood collected at 16-27 weeks gestation (N=1301). Participants in the subcohort who agreed to be contacted for follow up studies were invited to participate in the POUCHmoms Study 7-15 years after the POUCH Study pregnancy(13). The POUCHmoms Study enrolled 678 of the original subcohort participants. The analytic sample for our study includes the 637 participants in POUCHmoms Study who had both biomarker and BP measurements available (figure A). Angiogenic biomarkers Maternal serum samples were collected at 16-27 weeks gestation. Measurements of Plgf and endoglin were made using commercially available enzyme linked-linked immunofluorescent 73 assay kits by Karumanchi lab personnel who were blinded to participant’s clinical information. Each assay was measured twice and results were averaged and log transformed to create normal distributions for use in general linear models. In order to create groups for analysis based on low Plgf or elevated endoglin, we determined quartile cut points based on the distribution of the log transformed biomarkers among participants who did not have cHtn (BP >=140 mmHg Systolic/90 mmHg diastolic before 20 weeks gestation) and did not develop gHtn (BP >=140/90 developing after 20 weeks gestation and resolving post-partum) or PE (BP >=140/90 and proteinuria consistent with the definition in use at the time of the POUCH study). The lowest quartile of Plgf distribution in those without hypertension in pregnancy was log Plgf less than or equal to 5.486 (Plgf <= 241.29 pg/ml), the upper quartile of log endoglin was greater than 1.816 (6.15 mg/dl) (data not shown). Blood pressure during pregnancy Eight BP measurements were abstracted from prenatal records by trained study staff-the two highest SBP measures before 20 weeks, the two highest DBP measures before 20 weeks gestation, and the two highest SBP and DBP measures after 20 weeks gestation. Participants with chronic hypertension, gHtn or PE according to diagnosis in their medical records, or having at least 2 SBP reading >= 140 mmHg or at least 2 DBP readings >= 90 were categorized as having hypertension in pregnancy. Participants with SBP 120-139 mmHg on at least 2 occasions and/or DBP 80-89 on at least 2 occasions would not be considered hypertensive according to pregnancy hypertension guidelines, and were categorized as having “moderately elevated BP”. Participants who did not have 2 or more SBP readings above 120 and did not have 2 or more DBP readings above 80 were categorized as normotensive. 74 Follow Up Blood Pressures Blood pressure at the POUCHmoms visit was assessed by trained study staff, who took three BP readings at least 1 minute apart with the participant seated with the arm supported at the level of the heart, using automated BP cuffs that had been compared with manual readings before the study visit. The second two readings were averaged and recorded. For our analyses we categorized BP based on the ACC/AHA 2017 guidelines(11). BP below 120 mmHg systolic and below 80 mmHg diastolic were categorized as normal, BP 120-129 mmHg systolic and below 80 diastolic were categorized as elevated, BP of 130-139 mmHg systolic or 80-89 mmHg diastolic were categorized as stage 1 hypertension and BP >=140 systolic or >=90 diastolic were categorized as stage 2 hypertension. Given the small number of participants with elevated BP they were combined with the normotensive group for analysis. Covariates Covariates were selected a priori based on their relationships to BP or to biomarker levels during pregnancy(29). Maternal demographic information, reproductive history, lifestyle factors, height and pre-pregnancy weight were obtained from the POUCH Study enrollment questionnaire or interview. Self-reported race was dichotomized for this study as African American or non-African American, parity was dichotomized as nulliparous or multiparous, Medicaid enrollment was used as a proxy for socio-economic status. Smoking during pregnancy was dichotomized as smoking after 20 weeks’ gestation or quit/never-smoked. Body Mass Index (BMI) (kg/meters squared) was calculated from self-reported height and pre-pregnancy weight and used as a continuous variable in analyses. Gestational age at the enrollment visit (when biomarkers were measured) was based on last menstrual period if within 2 weeks of 75 date estimated from an ultrasound at less than 25 weeks gestation, and if not, ultrasound dating was used. The interval between the participant’s age at the beginning of the index pregnancy and their age at the POUCHmoms follow up visit was also included in analyses. Analytic Approach All analyses were performed with survey procedures in SAS v 9.4 (SAS Institute Inc., Cary, NC) and weighted to account for the POUCH Study design and sampling scheme. In order to describe the analytic sample we used weighted and unweighted frequencies of demographic, lifestyle, health and BP characteristics during the POUCH Study pregnancy and age, follow up interval and blood pressure group at follow up visit. We used survey frequency procedures and survey means to generate descriptive statistics and compare characteristics during the POUCH Study pregnancy among three groups defined by their BP at the POUCHmoms Study follow-up, i.e. Normal/elevated blood pressure, stage 1 hypertension and stage 2 hypertension at follow up (Chi-squared statistic). We then went on to use multivariable logistic regression to examine associations between log transformed biomarkers (as continuous exposures and comparing the lowest quartile of Plgf or upper quartile of endoglin) and stage of hypertension at follow up in adjusted and unadjusted models. Finally we created 6 groups that combined BP during pregnancy (not elevated, moderately elevated, hypertensive in pregnancy) and biomarker in pregnancy quartile group (‘low’ vs ‘not low’ for Plgf or ‘high’ vs ’not high’ for endoglin) based on the distribution of the log transformed biomarkers in those without PE/gHtn or cHtn). We used these groups in weighted polytomous regression models as exposure variables to estimate the unadjusted and adjusted odds of having stage 1 or stage 2 hypertension at follow up. We also ran unweighted models, and models that excluded women with chronic hypertension as 76 sensitivity analysis. Results The majority, 55% (56% weighted), of participants included in our sample were 20-29 years old during their POUCH Study pregnancy (table 4.1) and this is similar to the weighted percent of 57% in the full sub cohort(29). Almost three quarters (N=473, weighted percent 74%) of the 637 participants had their POUCHmoms Study follow up visit 11-15 years after enrolling in the POUCH Study. At follow-up36% (38% weighted) were over the age of 40. African Americans made up 36% (25% weighted) of our analytic sample, slightly less than the percent in the original sub cohort (40% unweighted). Similarly, our analytic sample had fewer women insured by Medicaid during the POUCH Study pregnancy (46% versus 49% in the full sub cohort). Based on a priori knowledge, we expected smoking during pregnancy, being insured by Medicaid, and length of time between index pregnancy and follow to be associated with follow up BP. However, none of these characteristics showed a statistically significant difference among BP groups in our study (table 4.2). Mean pre-pregnancy BMI did differ between groups, with participants who went on to develop stage 2 hypertension having the highest BMI (p < 0.0001). Participants with stage 2 hypertension 7-15 years after their POUCH Study pregnancy also were older (p=0.04). Frequency of stage 1 and stage 2 hypertension at follow up also differed by race (p= 0.0009); the weighted percent of African Americans with stage 2 hypertension was 29% versus 16% for non-African American participants. BP group in pregnancy (not elevated, moderately elevated and cHtn/gHtn/PE) was also significantly associated with stage of hypertension at follow up (p<0.0001). Mean Log PLGF was highest in women who are NOT hypertensive at follow up, and lowest in those who go on to have stage 2 77 hypertension, which is in the expected direction- however the difference in means was not statistically significant (p = 0.11). Mean endoglin did not show an increasing trend from NOT hypertensive to Stage 1 and then stage 2 hypertension. In unadjusted logistic regression models with no hypertension as the reference group, a one unit increase on in Plgf on the logarithmic scale was associated with decreased odds of having stage 2 hypertension at follow up with an OR of 0.69 (95% CL 0.48, 0.99), but this was not significant after adjustment for age during pouch pregnancy, pre-pregnancy BMI, gestational age at measurement, follow up interval, parity, race and smoking (table 4.3). There was no difference in odds of stage 1 or stage 2 hypertension between the lowest quartile of Plgf and the upper 3, nor in any of the models for endoglin. Subsequent analyses combined the three level BP group in pregnancy with a dichotomized Plgf level, i.e. ‘low’ (quartile 1) versus ‘not low’ (quartiles 1-3) or a dichotomized endoglin level, i.e. ‘high’ (quartile 4) versus ‘not high’ endoglin. For the Plgf based groups the number and weighted percent for each group were as follows: pregnancy BP not elevated/Plgf not ‘low’ N=150 (21%), pregnancy BP not elevated/Plgf ‘low’ N=30 (4%), pregnancy BP moderately elevated/Plgf not ‘low’ N=283 (46%), pregnancy BP moderately elevated/Plgf ‘low’ N= 95 (18%) Meanwhile 51 (8%) had hypertension in pregnancy / Plgf not ‘low’, and 26 (3%) had hypertension in pregnancy/Plgf ‘low’. For the endoglin groups, 60 (7%) of participants were in the group with pregnancy BP not elevated endoglin ‘high’, 100 (15%) were in the group with moderately elevated BP/endoglin ‘high’ and 37 (4%) were in the group with hypertension in pregnancy and ‘high’ endoglin. Statistical models with the combined BP and biomarker variable were used to estimate the 78 odds of having stage 1 or 2 hypertension at follow-up; in each the referent group was those with normal BP in pregnancy and ‘normal’ (Plgf ‘not low’ or endoglin ‘not high’) biomarker.. Overall we found that women with hypertension in pregnancy had 10-20 times greater odds of having stage 2 hypertension 7-15 years later, and women with moderately elevated BP had 2-3 times the odds of having stage 2 hypertension later (tables 4.4 and 4.5). For the group with moderately elevated BP pressure in pregnancy and ‘low’ Plgf the unadjusted odds ratio of having stage 2 hypertension at follow up were 3.21 (95% CL 1.32, 7.84), while for those with moderately elevated blood PF and Plgf ‘not low’ the odds ratio were 2.31 (95% Cl 1.06, 5.01) (table 4.4). In the adjusted models the odds ratios were slightly attenuated, with aOR for Stage 2 hypertension for the moderately elevated BP, ‘low’ Plgf group being 3.00 (95% CL 1.00, 9.5). Odds ratios of stage 1 or 2 hypertension for the group with normal BP in pregnancy but ‘low’ Plgf was close to 1.00 in unadjusted and adjusted models. The odds of Stage 1 or 2 hypertension at follow up was high for those with cHtn or HDPs, regardless of whether or not Plgf was low. The group with moderately elevated BP in pregnancy and ‘high’ endoglin had an odd ratio of 2.65 (95% Cl 0.80, 8.83) for stage 2 hypertension at follow-up and these results were similar in adjusted models (table 4b). While the moderately elevated BP group that did not have ‘high’ endoglin had a higher odds ratio for stage 2 hypertension, OR 3.92 (95% Cl 1.59, 9.64), aOR 3.77 (95% Cl 1.33, 10.54), these ORs were well within the confidence intervals of the high endoglin group described above and therefore considered not statistically significantly different For the groups with cHtn or HDP the odds of stage 1 hypertension were again similar between the group with and without ‘high’ endoglin and not statistically significant, while the estimated 79 aOR of stage 2 hypertension was 15.16 (95% Cl 3.16, 72.64) for those with ‘high’ endoglin and 22.62 (95% Cl 5.82, 87.99) for those without, again each odds ratio estimate was contained within the confidence interval of the other . In sensitivity analysis we removed participants with chronic hypertension (e.g. hypertension diagnosed before 20 weeks gestation) and obtained very similar results. When we repeated our analyses without using survey weighting procedures we also obtained similar results although with somewhat narrower confidence intervals (data not shown). Discussion Our findings suggest that Plgf and endoglin measured in maternal blood at mid-pregnancy do not help determine which individuals with moderately elevated BP during pregnancy are at greatest risk of hypertension 7-15 years later. Nor do these biomarkers distinguish future hypertension risk among women with normal BP or among women with hypertension in pregnancy. Regardless of biomarker levels, we did show that individuals with moderately elevated BP in pregnancy have higher odds of later life hypertension than women whose BP remains below 120/80 throughout pregnancy; this is consistent with prior analysis from the POUCH Study (13). It seemed plausible that maternal blood levels of Plgf and endoglin, which are associated with PE-related placental pathology(30),might serve as indicators of later life hypertension risk. Placental pathologies, such as impaired spiral artery remodeling and acute atherosis, which are often seen in PE, have been linked to later maternal CVD (31,32). Recent research from our group using data from the POUCH Study found evidence that certain vascular placental abnormalities were associated with greater odds of developing hypertension 7-15 years later 80 among those with moderately elevated BP in pregnancy (33). However, detailed placental pathology is not available for most pregnancies, therefore we sought to identify easily measured biomarkers linked to both placental abnormalities and later life CVD, particularly for risk stratification of people who experience BP elevations below the 140/90 threshold used to diagnose HDPs. There are multiple pathways common to PE and CVD including angiogenic, metabolic, inflammatory and the renin-angiotensin system (20,34). While some have reported that maternal serum Plgf was inversely associated with later risk of CVD (26), others concur with our study showing maternal Plgf levels unrelated to BP after pregnancy(35). Several studies have found links between Plgf and lipid profiles after pregnancy(36–38), hence the link between low Plgf and later CVD may not be through a direct effect on BP. While elevated endoglin levels are often seen in PE, associations are more mixed with gHtn(21,30,39). Endoglin measured outside of pregnancy has been associated with microvascular complications in people with hypertension and diabetes(25) yet, there is at least one report of an inverse association between endoglin levels outside of pregnancy and coronary artery disease(40). This study has several limitations to consider. We were unable to follow all participants from the POUCH Study subcohort, therefore our POUCHmoms Study sample size was more limited resulting in small cell sizes and wide confidence intervals for some groups in our analyses. Regarding selection bias of those followed, it was reassuring that the age distribution of the POUCHmoms sample was similar to that of the full POUCH subcohort, as was the distribution of hypertensive disorders in pregnancy. Furthermore, it seems unlikely that levels of Plgf and endoglin in the POUCH pregnancy were related to loss to follow up. BP measurements during 81 pregnancy were abstracted from the prenatal record, and hence there may have been variability in measurement technique and accuracy. Biomarker measurement in the POUCH study captured maternal blood levels between 16 and 27 weeks gestation, which may have been too early to pick up variations in endoglin levels related to placental pathology and later risk of hypertension. Changes in PE-related endoglin levels typically occur later in pregnancy(27,41). We did not include diagnosis of HDPs in subsequent pregnancies in our models, nor changes in smoking habits or BMI during the follow up interval. However, this was a conscious choice as our intent was to assess whether maternal Plgf and endoglin levels could be used to determine who among the large group of participants with moderately elevated BP in pregnancy might be candidates for closer BP monitoring after pregnancy due to their greater risk of developing hypertension later. One additional point, most participants in the POUCHmoms follow-up Study were not at the age when hypertension risk climbs, thus longer follow-up periods might alter our findings. Strengths of this study include the use of clinically-trained study staff who followed best practices for BP measurement at the POUCHmoms Study follow up. The POUCH Study included participants from multiple diverse communities, and collected detailed information about demographics, anthropomorphic and behavioral characteristics, as well as measurement of angiogenic biomarkers. Ours is the only study that we are aware of to explore associations of Plgf and endoglin with hypertension years after pregnancy among those without HDPs. Contrary to our expectations, maternal Plgf and endoglin levels did not help identify future risk of hypertension in our study sample. While it is possible that we would have observed an association between one or both biomarkers we analyzed if we had measured them later in 82 pregnancy, it is also possible that neither one is associated with blood pressure years after pregnancy. This “negative” finding is nevertheless helpful as may help narrow future hypotheses and possibly encourage investigators to consider other potentially useful biomarkers of risk for later life hypertension. Future studies might consider other angiogenic biomarkers, such as PAPP-A, or biomarkers along the other pathways common to placental disorders and CVD. Cardiovascular risk assessment and counseling during pregnancy and in the post-partum period is an important intervention not only for people diagnosed with HDPs as is currently recommended, but for everyone with elevated blood pressure or stage 1 hypertension by non-pregnant criteria. 83 REFERENCES 1. Cain MA, Salemi JL, Tanner JP, et al. Pregnancy as a window to future health: maternal placental syndromes and short-term cardiovascular outcomes. Am. J. Obstet. Gynecol. [electronic article]. 2016;215(4):484.e1-484.e14. (http://dx.doi.org/10.1016/j.ajog.2016.05.047) 2. Stuart JJ, Tanz LJ, Missmer SA, et al. Hypertensive disorders of pregnancy and maternal cardiovascular disease risk factor development: An observational cohort study. Ann. Intern. Med. 2018;169(4):224–232. 3. Rangaswami J, Naranjo M, McCullough PA. Preeclampsia as a Form of Type 5 Cardiorenal Syndrome: An Underrecognized Entity in Women’s Cardiovascular Health. CardioRenal Med. 2018;19141:160–172. 4. Melchiorre K, Thilaganathan B, Giorgione V, et al. Hypertensive Disorders of Pregnancy and Future Cardiovascular Health. Front. Cardiovasc. Med. 2020;7. 5. Demartelly VA, Dreixler J, Tung A, et al. Long-Term Postpartum Cardiac Function and Its Association With Preeclampsia. J. Am. Heart Assoc. 2021;10(5):e018526. 6. Canoy D, Cairns BJ, Balkwill A, et al. Hypertension in pregnancy and risk of coronary heart disease and stroke: A prospective study in a large UK cohort. Int. J. Cardiol. [electronic article]. 2016;222:1012–1018. (http://dx.doi.org/10.1016/j.ijcard.2016.07.170) 7. Behrens I, Basit S, Melbye M, et al. Risk of post-pregnancy hypertension in women with a history of hypertensive disorders of pregnancy: nationwide cohort study. Bmj [electronic article]. 2017;j3078. (http://www.bmj.com/lookup/doi/10.1136/bmj.j3078) 8. Lindstrom L, Skjorven R, Bergman E, et al. Chronic Hypertension in Women after Perinatal Exposure to Preeclampsia, Being Born Small for Gestational Age or Preterm. Paediatr. Perinat. Epidemiol. 2017;31(2):89–98. 9. Parikh NI, Norberg M, Ingelsson E, et al. Association of Pregnancy Complications and Characteristics With Future Risk of Elevated Blood Pressure. Hypertension [electronic article]. 2017;HYPERTENSIONAHA.116.08121. (http://www.ncbi.nlm.nih.gov/pubmed/28137991%0Ahttp://hyper.ahajournals.org/look up/doi/10.1161/HYPERTENSIONAHA.116.08121) 10. Gestational Hypertension and Preeclampsia: ACOG Practice Bulletin, Number 222. Obstet. Gynecol. 2020;135(6):e237–e260. 84 11. Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: Executive summary: A report of the American college of cardiology/American Heart Association task . 2018 1269–1324 p. 12. Bello NA, Zhou H, Cheetham TC, et al. Prevalence of Hypertension among Pregnant Women When Using the 2017 American College of Cardiology/American Heart Association Blood Pressure Guidelines and Association with Maternal and Fetal Outcomes. JAMA Netw. Open. 2021;4(3). 13. Dunietz G, Strutz K, Holzman C, et al. Moderately elevated blood pressure during pregnancy and odds of hypertension later in life: the POUCHmoms longitudinal study. BJOG An Int. J. Obstet. Gynaecol. [electronic article]. 2017;1–8. (http://doi.wiley.com/10.1111/1471-0528.14556) 14. Darwin KC, Federspiel JJ, Schuh BL, et al. ACC-AHA Diagnostic Criteria for Hypertension in Pregnancy Identifies Patients at Intermediate Risk of Adverse Outcomes. Am. J. Perinatol. 2021;38:E249–E255. 15. Sabol BA, Porcelli B, Diveley E, et al. Defining the risk profile of women with stage 1 hypertension: a time to event analysis. Am. J. Obstet. Gynecol. MFM. 2021;3(4). 16. Greenberg VR, Silasi M, Lundsberg LS, et al. Perinatal outcomes in women with elevated blood pressure and stage 1 hypertension. Am. J. Obstet. Gynecol. 2021;224(5):521.e1- 521.e11. 17. Reddy M, Rolnik DL, Harris K, et al. Challenging the definition of hypertension in pregnancy: a retrospective cohort study. Am. J. Obstet. Gynecol. [electronic article]. 2020;222(6):606.e1-606.e21. (https://doi.org/10.1016/j.ajog.2019.12.272) 18. Boghossian NS, Albert PS, Mendola P, et al. Delivery Blood Pressure and Other First Pregnancy Risk Factors in Relation to Hypertensive Disorders in Second Pregnancies. Am. J. Hypertens. 2015;28(9):1172–1179. 19. Nolasco-Leaños AG, Ramírez-Valenzuela KL, Carrillo-Juárez RI, et al. Circulating angiogenic factors are associated with progression to preeclampsia and the occurrence of adverse outcomes in women with gestational hypertension. J. Hypertens. 2021;39(3):573–580. 20. Eastabrook G, Aksoy T, Bedell S, et al. Preeclampsia biomarkers: An assessment of maternal cardiometabolic health. Pregnancy Hypertens. [electronic article]. 2018;13(May):204–213. (https://doi.org/10.1016/j.preghy.2018.06.005) 21. Ramadan H, Rana S, Mueller A, et al. Myocardial performance index in hypertensive 85 disorders of pregnancy: The relationship between blood pressures and angiogenic factors. Hypertens. Pregnancy [electronic article]. 2017;36(2):161–167. (https://doi.org/10.1080/10641955.2017.1280048) 22. Shahul S, Medvedofsky D, Wenger JB, et al. Circulating antiangiogenic factors and myocardial dysfunction in hypertensive disorders of pregnancy. Hypertension. 2016;67(6):1273–1280. 23. Akhter T, Wikström AK, Larsson M, et al. Association between angiogenic factors and signs of arterial aging in women with pre-eclampsia. Ultrasound Obstet. Gynecol. 2017;50(1):93–99. 24. Blázquez-Medela AM, García-Ortiz L, Gómez-Marcos MA, et al. Increased plasma soluble endoglin levels as an indicator of cardiovascular alterations in hypertensive and diabetic patients. BMC Med. [electronic article]. 2010;8(1):86. (http://bmcmedicine.biomedcentral.com/articles/10.1186/1741-7015-8-86) 25. Siervo M, Ruggiero D, Sorice R, et al. Angiogenesis and biomarkers of cardiovascular risk in adults with metabolic syndrome: Original Article. J. Intern. Med. 2010;268(4):338–347. 26. Cassidy A, Chiuve SE, Manson JE, et al. Potential role for plasma placental growth factor in predicting coronary heart disease risk in women. Arterioscler. Thromb. Vasc. Biol. 2009;29(1):134–139. 27. Romero R, Nien JK, Espinoza J, et al. A longitudinal study of angiogenic (placental growth factor) and anti-angiogenic (soluble endoglin and soluble vascular endothelial growth factor receptor-1) factors in normal pregnancy and patients destined to develop preeclampsia and deliver a small for. J. Matern. Neonatal Med. 2008;21(1):9–23. 28. Holzman C, Bullen B, Fisher R, et al. Pregnancy outcomes and community health: the POUCH study of preterm birth. Paediatr. Perinat. Epidemiol. 2001;15(s2):136–158. 29. Mijal RS, Holzman CB, Rana S, et al. Midpregnancy levels of angiogenic markers in relation to maternal characteristics. Am. J. Obstet. Gynecol. [electronic article]. 2011;204(3):244.e1-244.e12. (http://dx.doi.org/10.1016/j.ajog.2010.10.001) 30. Helmo FR, Lopes AMM, Carneiro ACDM, et al. Angiogenic and antiangiogenic factors in preeclampsia. Pathol. Res. Pract. [electronic article]. 2017;214(1):7–14. (https://doi.org/10.1016/j.prp.2017.10.021) 31. Koenen S V., Nikkels PGJ, Veerbeek JHW, et al. Spiral artery remodeling and maternal cardiovascular risk. J. Hypertens. 2016;34(8):1570–1577. 32. Benagiano M, Mancuso S, Brosens JJ, et al. Long-term consequences of placental vascular 86 pathology on the maternal and offspring cardiovascular systems. Biomolecules. 2021;11(11). 33. Holzman CB, Senagore P, Xu J, et al. Maternal risk of hypertension 7–15 years after pregnancy: clues from the placenta. BJOG An Int. J. Obstet. Gynaecol. 2021;128(5):827– 836. 34. Suvakov S, Bonner E, Nikolic V, et al. Overlapping pathogenic signalling pathways and biomarkers in preeclampsia and cardiovascular disease. Pregnancy Hypertens. [electronic article]. 2020;20(August 2019):131–136. (https://doi.org/10.1016/j.preghy.2020.03.011) 35. Neuman RI, Figaroa AMJ, Nieboer D, et al. Angiogenic markers during preeclampsia: Are they associated with hypertension 1 year postpartum? Pregnancy Hypertens. 2021;23:116–122. 36. Garrido-Gimenez C, Mendoza M, Cruz-Lemini M, et al. Angiogenic Factors and Long-Term Cardiovascular Risk in Women That Developed Preeclampsia during Pregnancy. Hypertension. 2020;1808–1816. 37. Charlton F, Tooher J, Rye KA, et al. Cardiovascular risk, lipids and pregnancy: Preeclampsia and the risk of later life cardiovascular disease. Hear. Lung Circ. 2014;23(3):203–212. 38. Catov JM, McNeil RB, Marsh DJ, et al. Early Pregnancy Atherogenic Profile in a First Pregnancy and Hypertension Risk 2 to 7 Years After Delivery. J. Am. Heart Assoc. 2021;10(5):e017216. 39. Gastrich MD, Faro R, Rosen T. Markers of preeclampsia and the relationship to cardiovascular disease: Review of the twenty-first century literature. J. Matern. Neonatal Med. 2010;23(8):751–769. 40. Saita E, Miura K, Suzuki-Sugihara N, et al. Plasma Soluble Endoglin Levels Are Inversely Associated With the Severity of Coronary Atherosclerosis—Brief ReportHighlights. Arterioscler. Thromb. Vasc. Biol. [electronic article]. 2017;37(1):49–52. (http://atvb.ahajournals.org/lookup/doi/10.1161/ATVBAHA.116.308494) 41. Tarca AL, Taran A, Romero R, et al. Prediction of preeclampsia throughout gestation with maternal characteristics and biophysical and biochemical markers: a longitudinal study. Am. J. Obstet. Gynecol. 2022;226(1):126.e1-126.e22. 87 APPENDIX Figure A: Pouch Study subcohort and Analytic Sample Pouch subcohort N = 1371 Pouchmoms study of these 1280 agreed N=678 to further contact 11 missing data for BP 642 were out of area, measurement during Analytic Sample N=637 deceased, declined or pregnancy not located 30 missing data for 60 did survey only, no biomarker blood pressure measurement during measures pregnancy 88 Table 4.1: Characteristics of the Study Sample during the POUCH Study Pregnancy (1998- 2004) and at the POUCHmoms Study (7-15 years later) Characteristics N % Weighted %* During POUCH Study Pregnancy Age (years) <20 102 16% 15% 20-29 350 55% 56% ≥30 185 29% 29% Gestational Age at Enrollment 15-19 weeks 99 16% 17% 20-24 weeks 456 72% 73% 25-27 weeks 82 12% 12% Body mass index Pre Pregnancy <18 31 5% 4% 18-25 285 45% 47% >25-30 135 21% 21% ≥30 186 29% 28% Parity 0 276 43% 44% ≥1 361 57% 56% Medicaid Insurance Yes 322 51% 46% No 315 49% 54% Race Black/African American 227 36% 25% White/Other 410 64% 75% Smoking during pregnancy None/quit before 20 weeks’ gestation 528 83% 83% Smoked 109 17% 17% Blood Pressure Group in Pregnancy** Not elevated 181 28% 26% Moderately elevated 378 60% 63% Hypertension in pregnancy 77 12% 11% At follow up Age(years) <30 47 8% 7% 30-39 358 56% 55% ≥40 232 36% 38% Follow up interval 7-10 years 164 26% 26% 11-15 years 473 74% 74% Hypertension Stage at follow up*** Normotensive or prehypertensive 381 60% 61% Stage 1 122 19% 20% Stage 2 134 21% 19% 89 Table 4.1 (cont’d) *weighted to reflect sampling frame for Pouch subcohort. **Blood Pressure Group in Pregnancy: Not elevated = SBP <120 mmHg and DBP < 80 mmHg; Moderately elevated = SBP 120-139 mmHg and/or DBP 80-89 mmHg; Hypertensive = diagnosis of chronic hypertension (cHtn), gestational hypertension (GH) or preeclampsia (PE). ***Hypertension stage at follow up: Normotensive or prehypertensive = SBP 120-129 mmHg and DBP < 80 mmHg, Stage 1 = SBP 130-139 mmHg and/or DBP 80-89 mmHg. Stage 2 = SBP >= 140 mmHg and/or DBP >= 90 mmHg. 90 Table 4.2: Bivariate Associations between characteristics during the POUCH Study pregnancy and hypertension stage 7-15 years later in the POUCHmoms Study (N= 637) Hypertension stage at follow up Normal or Stage 1 Stage 2 elevated Hypertension Hypertension N=381 N=122 N=134 Characteristics during the p* POUCH Study Pregnancy Mean (SD) Age 26.20 (0.33) 26.77 (0.59) 27.88 (0.57) 0.04 Interval to follow up 10.91 (0.08) 11.17 (0.15) 11.12 (0.13) 0.19 Week of gestation at 22.37 (0.13) 22.31 (0.23) 22.17 (0.23) 0.76 enrollment Pre-Pregnancy body mass 25.53 (0.41) 27.80 (0.76) 31.99 (0.92) <0.0001 index Log Plgf 5.95 (0.04) 5.86 (0.06) 5.79 (0.07) 0.11 Log endoglin 1.69 (0.01) 1.62 (0.04) 1.64 (0.03) 0.12 Mean arterial pressure 78.57 (0.46) 80.12 (1.01) 83.61 (0.93) <0.0001 (MAP) at 20 week’s gestation Characteristics during the POUCH Study Pregnancy, p** N (weighted %) Parity 0.27 Nulliparous 173 (64%) 54 (20%) 49 (16%) One or more 208 (59%) 68 (19%) 85 (22%) Medicaid Insurance 0.19 Yes 183 (57%) 64 (20%) 75 (23%) No 198 (64%) 58 (19%) 59 (17) Race/Ethnicity 0.0009 African American 116 (51%) 46 (20%) 65 (29%) White/other 265 (65%) 76 (19%) 69 (16%) Blood Pressure Group in <0.0001 Pregnancy Not elevated 124 (69%) 39 (23%) 18 (8%) Moderately elevated 233 (63%) 63 (17%) 83 (19%) PH/GH/cHtn 24 (31%) 20 (26%) 33 (44%) Smoking in Pregnancy 0.61 None/Quit before 20 319 (62%) 99 (19%) 110 (19%) weeks Yes 62 (56%) 23 (21%) 24 (22%) *p value from Proc Survey Means-Chi squared statistic, **p value from Proc Survey Reg Chi squared statistic 91 Table 4.3: Maternal blood Plgf and Endoglin levels at mid-pregnancy in the POUCH Study and Odds of Hypertension 7-15 later in the POUCHmoms Study (N= 637) Hypertension stage at Follow-up Hypertension stage at Follow-up Unadjusted Model Adjusted Model* Normal/e Stage 1 Stage 2 Normal/ Stage 1 Stage 2 levated N = 121 N= 133 elevated N=121 N=133 N=381 OR OR N=381 aOR aOR (95% CL) (95% CL) Ref (95% CL) (95% CL) Log Plgf Ref 0.82 0.69 Ref 0.84 0.79 (0.57, 1.18) (0.48, 0.99) (0.52, 1.36) (0.46, 1.37) Lowest Ref 1.00 1.37 Ref 0.88 1.12 quartile Plgf vs (0.56, 1.78) (0.81, 2.34) (0.46, 1.68) (0.59, 2.14) upper three quartiles Log Endoglin Ref 0.44 0.52 Ref 0.50 0.97 (0.16, 1.19) (0.21, 1.27) (0.17, 1.43) (0.35, 0.68) Highest Ref 1.12 0.84 Ref 1.20 1.03 quartile (0.66, 1.92) (0.50, 1.41) (0.68, 2.12) (0.58, 1.83) Endoglin vs lower three quartiles * Adjusted for maternal characteristics during the POUCH Study (pre-pregnancy BMI, age, race, Medicaid insurance, smoking, parity)) and interval between the POUCH Study and the POUCHmoms Study 92 Table 4.4: Maternal blood Plgf levels at mid-pregnancy and blood pressure (BP) grouping in the POUCH study and Odds of Hypertension 7-15 later in the POUCHmoms study (N= 637) Hypertension stage at Follow-up Hypertension stage at Follow-up Unadjusted Odds Ratio Adjusted Odds Ratio* Normal Stage 1 Stage 2 Normal/e Stage1 Stage2 /elevated N = 121 N= 133 levated N=121 N=133 N=381 OR OR N=381 aOR aOR (95% CL) (95% CL) Ref (95%CL) (95%CL) Pregnancy BP and Plgf group** Normal BP Ref 1.0 1.0 Ref 1.0 1.0 Plgf ‘not low’ N=150 Normal BP Ref 1.33 1.38 Ref 1.19 1.07 Plgf ‘low’ (0.43, 4.06) (0.38, 5.05) (0.39, 3.63) (0.27, 4.29) N=30 Moderately Ref 0.97 2.31 Ref 0.90 2.28 elevated BP (0.53, 1.80) (1.06, 5.01) (0.47, 1.72) (0.93, 5.62) Plgf ‘not low’ N=283 Moderately Ref 0.75 3.21 Ref 0.61 3.00 elevated BP (0.32, 1.72) (1.32, 7.84) (0.23, 1.61) (1.00, 9.05) Plgf ‘low’ N=195 Hypertension Ref 2.49 12.89 Ref 2.41 16.25 Plgf ‘not low’ (0.85, 7.27) (4.42, 37.59) (0.74, 7.83) (4.47, 59.12) N=51 Hypertension Ref 3.54 9.52 Ref 2.86 9.19 Plgf ‘low’ (0.92, (2.48, 36.56) (0.68, 12.03 (2.15, 39.23) N=26 13.59) * Adjusted for maternal characteristics during the POUCH Study (pre-pregnancy BMI, age, race, Medicaid insurance, smoking, parity) and interval between the POUCH Study and the POUCHmoms Study (years) **Low Plgf is lowest quartile of log Plgf, cut-point derived from Plgf distribution in participants without hypertensive disorders in pregnancy 93 Table 4.5: Maternal blood Endoglin levels at mid-pregnancy and blood pressure (BP) grouping in the POUCH study and Odds of Hypertension 7-15 later in the POUCHmoms study (N= 637) Hypertension stage at Follow-up Hypertension stage at Follow-up Unadjusted Odds Ratio Adjusted Odds Ratio* No Stage 1 Stage 2 No Stage1 Stage2 Htn N = 122 N= 134 Htn N=122 N=134 N=381 OR OR N=381 aOR aOR (95%CL) (95% CL) (95% CL) (95%CL) Pregnancy BP and Endoglin group** Normal BP Ref 1.0 1.0 Ref 1.0 1.0 Endoglin ‘not high’ N=121 Normal BP Ref 1.01 2.65 Ref 1.01 2.84 Endoglin ‘high’ (0.39, 2.62) (0.80, 8.83) 0.37, 2.73) (0.76, 10.61) N=60 Moderately Ref 0.80 3.92 Ref 0.70 3.77 elevated BP (0.42, 1.52) (1.59, 9.64) (0.36, 1.36) (1.35, 10.54) Endoglin ‘not high’ N=279 Moderately Ref 0.90 2.06 Ref 0.89 2.69 elevated BP (0.40, 2.02) (0.71, 5.95) (0.38, 2.09) (0.81, 8.89) Endoglin High N=100 Hypertension Ref 2.97 20.87 Ref 2.61 22.62 Endoglin ‘not (0.94, 9.42) (6.08, 71.61) (0.80, 8.86) (5.82, 87.99) high’ N=40 Hypertension Ref 2.13 10.77 Ref 2.09 15.16 Endoglin ‘high’ (0.61, 7.44) (2.95, 39.33) (0.50, 8.81) (3.16, 72.64) N=37 * Adjusted for Prepregnancy BMI, maternal age Pouch, interval between Pouch/Pouch Moms, race, Medicaid status during pouch, smoking during pouch pregnancy, and parity at pouch pregnancy. **High endoglin is upper quartile of log endoglin, cut-point derived from endoglin distribution in participants without hypertensive disorders in pregnancy. 94 CHAPTER 5: CONCLUSIONS 95 Summary The three papers in this dissertation used data from a multi-community longitudinal cohort (1) to examine associations among three angiogenic biomarkers (endoglin, sFlt1 and Plgf) and BP or MAP during pregnancy, as well as pregnancy outcomes and later life hypertension(2). The first paper asks whether each biomarker is associated with MAP in pregnancy, and whether these associations are driven only by PE or other forms of hypertension. The potential association between combinations of each biomarker with elevated MAP in mid pregnancy and SGA or PTB are examined in the second paper. The third paper considers the possibility that angiogenic biomarkers measured during pregnancy would be helpful in identifying which of the large group of people with moderately elevated BP in pregnancy go on to develop hypertension 7 to 15 years later. In the first paper we found that higher levels of Plgf, in unadjusted and adjusted analyses, were associated with decreased odds of MAP in the top quartile, both at approximately 20 weeks gestation and at the time of the highest diastolic BP recorded in the pregnancy (referred to as ‘highest” MAP). Sflt-1 and sEng were not statistically significantly associated with MAP. For MAP at 20 weeks, the association between higher Plgf and lower odds of MAP >= 84.7 mmHg (the upper quartile among normotensive participants) remained even after excluding participants with PE and was of a similar magnitude but no longer statistically significant when we excluded participants with cHtn, or who went on to develop GH or PE. For the upper quartile of ‘highest’ MAP (>= 95.3 mmHg), the aOR remained significant when we excluded participants with PE and when we excluded participants with any cHtn, GH or PE. As we learn more about the large group of individuals who have elevated BP or stage 1 96 hypertension during pregnancy yet do not meet the diagnostic cut off for PE or GH, it is possible that PLGF would help identify individuals who may be phenotypically similar to individuals who develop PE during pregnancy and merit closer monitoring for adverse fetal/infant outcomes, and maternal postpartum and midlife hypertension. Measurement of maternal serum biomarkers in our study took place between 20 and 25 weeks gestation for the majority of participants, which may be too early to observe any associations between blood pressure and sFlt1 or sEng (3,4) In the second paper we assessed several pregnancy outcomes (SGA, LGA, and length of gestation) in relation to maternal MAP and blood levels of endoglin, sFlt1 and Plgf measured at mid-pregnancy. Of the three biomarkers, ‘low’ Plgf levels again stood out as being most strongly correlated with SGA. We also found that ‘high endoglin’ levels were most strongly related to lower gestational age at delivery in our sample. When we created groups based on combining angiogenic biomarker and MAP information we found that pregnancies with MAP ‘high’ (upper quartile) and Plgf ‘low’ (lower quartile) had greater than twice the odds of delivering an SGA infant compared with women who had a more favorable profile (i.e. both MAP ‘not high’ and Plgf ‘not low’). Focusing just on women with MAP ’high’, those with Plgf ‘low’ had greater than threefold increased odds of delivering an SGA infant compared to those with Plgf ‘not low’. Thus, maternal Plgf levels showed promise for risk stratifying women with ‘high’ MAP. This finding persisted when we removed women with hypertensive disorders developed before or during pregnancy. We found no associations between sFlt1 and SGA. We also examined MAP/biomarker combinations in relation to length of gestation. Women with ‘High’ endoglin levels had shorter gestations irrespective of MAP 97 being high or not. This association was markedly reduced after removing women with hypertensive disorders. Plgf and sFlt1 appeared unrelated to length of gestation. In the third paper we observed that Plgf or endoglin measured in maternal blood at mid- pregnancy were not helpful in determining which individuals with moderately elevated BP during pregnancy are at greatest risk of hypertension 7-15 years later. Nor did these biomarkers distinguish future hypertension risk among women with normal BP or among women with hypertension in pregnancy. Regardless of biomarker levels, we did show that individuals with moderately elevated BP in pregnancy have higher odds of later life hypertension than women whose BP remains below 120/80 throughout pregnancy; this is consistent with prior analysis from the POUCH Study(5). While it is possible that we would have observed an association between one or both biomarkers we analyzed if we had measured them later in pregnancy, it is also possible that neither one is associated with BP years after pregnancy. Clinical Implications and future directions In the first paper, we found that the top quartile of MAP at either mid-pregnancy or the highest level in pregnancy was associated with more “adverse” angiogenic biomarker profiles (particularly with lower levels of Plgf). The commonly used clinical cut point of 140 mmHg systolic and/or 90 mmHg diastolic in pregnancy may miss people who are at risk for developing complications such as PE or FGR later in pregnancy. Other research has found that MAP levels similar to the boundary of the upper quartiles in mid pregnancy and later in our study sample are associated with higher odds of adverse pregnancy outcomes(6). Clinicians and researcher are grappling with the potential risks and benefits of lowering the threshold of diastolic/systolic BP or MAP to identify pregnancies at risk of adverse outcomes (7– 98 10). If the criteria for defining chronic and gestational hypertension in pregnancy are lowered, our finding, i.e. that women with higher MAP and but normal Plgf at mid pregnancy do not appear to be at greater risk for SGA than those with normal MAP, may be helpful in distinguishing which women should be monitored more closely. In our second study, this relationship became more apparent once we adjusted for maternal characteristics- suggesting that any clinical application may need to utilize population based multiples of the mean which take into account variables such as BMI. Studies in larger cohorts are needed to more clearly identify thresholds of Plgf that are associated with delivery of an SGA infant. Our findings in the third paper re-demonstrated that individuals who develop elevated BP in pregnancy below the current ACOG defined cut points are at risk for going on to develop stage 2 hypertension. Cardiovascular risk assessment and counseling during pregnancy and in the post-partum period is an important intervention not only for people diagnosed with HDPs as is currently recommended, but for everyone with elevated pregnancy BP or stage 1 hypertension by non-pregnant criteria. Even our “negative” finding in the third paper may help narrow future hypotheses and encourage the consideration of other potentially useful biomarkers of risk for later life hypertension. There are other angiogenic biomarkers, such as PAPP-A, which have been linked to both HPDs and cardiovascular CVD, and there are other important pathways (such as metabolic or inflammatory) which are common to both HDPs and CVD(11) which could be evaluated in future studies. 99 REFERENCES 1. Holzman C, Bullen B, Fisher R, et al. Pregnancy outcomes and community health: the POUCH study of preterm birth. Paediatr. Perinat. Epidemiol. 2001;15(s2):136–158. 2. Holzman CB, Senagore P, Xu J, et al. Maternal risk of hypertension 7–15 years after pregnancy: clues from the placenta. BJOG An Int. J. Obstet. Gynaecol. 2021;128(5):827– 836. 3. Honigberg MC, Cantonwine DE, Thomas AM, et al. Analysis of changes in maternal circulating angiogenic factors throughout pregnancy for the prediction of preeclampsia. J. Perinatol. 2016;36(3). 4. Romero R, Nien JK, Espinoza J, et al. A longitudinal study of angiogenic (placental growth factor) and anti-angiogenic (soluble endoglin and soluble vascular endothelial growth factor receptor-1) factors in normal pregnancy and patients destined to develop preeclampsia and deliver a small for. J. Matern. Neonatal Med. 2008;21(1):9–23. 5. Dunietz G, Strutz K, Holzman C, et al. Moderately elevated blood pressure during pregnancy and odds of hypertension later in life: the POUCHmoms longitudinal study. BJOG An Int. J. Obstet. Gynaecol. [electronic article]. 2017;1–8. (http://doi.wiley.com/10.1111/1471-0528.14556) 6. Reddy M, Rolnik DL, Harris K, et al. Challenging the definition of hypertension in pregnancy: a retrospective cohort study. Am. J. Obstet. Gynecol. [electronic article]. 2020;222(6):606.e1-606.e21. (https://doi.org/10.1016/j.ajog.2019.12.272) 7. Darwin KC, Federspiel JJ, Schuh BL, et al. ACC-AHA Diagnostic Criteria for Hypertension in Pregnancy Identifies Patients at Intermediate Risk of Adverse Outcomes. Am. J. Perinatol. 2021;38:E249–E255. 8. Bello NA, Zhou H, Cheetham TC, et al. Prevalence of Hypertension among Pregnant Women When Using the 2017 American College of Cardiology/American Heart Association Blood Pressure Guidelines and Association with Maternal and Fetal Outcomes. JAMA Netw. Open. 2021;4(3). 9. Sabol BA, Porcelli B, Diveley E, et al. Defining the risk profile of women with stage 1 hypertension: a time to event analysis. Am. J. Obstet. Gynecol. MFM. 2021;3(4). 10. Garovic VD, Dechend R, Easterling T, et al. Hypertension in Pregnancy: Diagnosis, Blood Pressure Goals, and Pharmacotherapy: A Scientific Statement From the American Heart Association. Hypertension. 2022;79(2):E21–E41. 100 11. Suvakov S, Bonner E, Nikolic V, et al. Overlapping pathogenic signalling pathways and biomarkers in preeclampsia and cardiovascular disease. Pregnancy Hypertens. [electronic article]. 2020;20(August 2019):131–136. (https://doi.org/10.1016/j.preghy.2020.03.011) 101