Association between adverse pregnancy outcomes and awareness of cardiovascular risk factors: A nuMoM2b‐HHS secondary analysis

Abstract

Background

Adverse pregnancy outcomes (APOs) affect 10%–20% of pregnancies and are associated with increased risk of cardiovascular disease.

Objective

To determine whether the history of APO is associated with a person's awareness of their cardiovascular risk factors (CRF) 2–7 years postpartum.

Methods

This is a secondary analysis of a multicenter cohort study in which nulliparous pregnant people were enrolled in the first trimester and followed prospectively. Participants with a CRF (hypertension, dyslipidemia, and/or prediabetes/diabetes) at follow‐up were included. We excluded participants with: pre‐gestational diabetes, chronic hypertension, dyslipidemia during the index pregnancy or subsequent pregnancy, and non‐fasting blood draws. The exposure was APO during index pregnancy: hypertensive disorders of pregnancy, gestational diabetes, preterm delivery, small‐for‐gestational age, and stillbirth. The primary outcome is a composite of participant awareness of a CRF diagnosis (participant‐reported or prescription medication‐treated). Unadjusted and adjusted logistic regression models estimated associations between exposure and outcome.

Results

Among 980 participants, 77% had one CRF, 20% had two, and 3% had three. The most common CRF was hypertension (51%) and 35% of participants had an APO. Participant awareness of their CRF diagnosis was only 22% (220/980)—26% with an APO versus 20% without (p = 0.04). After multivariable adjustment, history of APO was not associated with participant awareness of CRFs.

Conclusion

Only 22% of the participants with a CRF reported knowing about or being treated for their CRF. Participants with and without an APO did not statistically differ in awareness of their CRFs 2–7 years after their first pregnancy.

Abbreviations

ACOG

American College of Obstetrics and Gynecology

APOs

adverse pregnancy outcomes

ASCVD

atherosclerotic cardiovascular disease

CRF

cardiovascular risk factors

CVD

cardiovascular disease

GDM

gestational diabetes

HDP

hypertensive disorders of pregnancy

nuMoM2b‐HHS

Nulliparous Pregnancy Outcomes Study: Monitoring Mothers‐to‐be Heart Health Study

PTB

preterm birth

SGA

small for gestational age

1. INTRODUCTION

Cardiovascular disease (CVD) is the leading cause of morbidity and mortality in women [1]. Adverse pregnancy outcomes (APOs) have consistently been associated with increased risk of CVD [2, 3] and coronary atherosclerosis [4, 5]. APOs are present in 10%–20% of all pregnancies and are associated with a 1.8–4‐fold risk of future CVD [1, 6]. Since APOs are associated with increased risk of CVD, early identification of high‐risk women via diagnosis of cardiovascular risk factors (CRF) is imperative to provide appropriate risk stratification and individualized healthcare to positively impact their future cardiovascular health.

Specific APOs associated with CVD risk include hypertensive disorders of pregnancy (HDP), gestational diabetes (GDM), preterm birth (PTB), small for gestational age (SGA) birthweight, and stillbirth. HDP includes diagnoses of gestational hypertension, preeclampsia, and eclampsia. HDP confers a particularly increased risk for future cardiovascular disease [7]. HDP rates have doubled from 2007 to 2014 to 77.8 per 1000 live births [8]. Women with HDP are diagnosed with chronic hypertension, CVD, diabetes, and hyperlipidemia up to 10 years earlier than their counterparts [9] and have evidence of early vascular aging [10, 11]. They experience an accelerated time‐to‐first cardiovascular event and are at a higher risk of death before the age of 70 [12].

Women with GDM have a 43% greater risk of developing CVD [13] and a twofold increased risk of coronary artery calcium at mid‐life independent of diabetes progression [9]. PTB affects roughly 10% of deliveries [14] and has consistently been associated with a higher prevalence of maternal CVD (OR 1.72–2.85) [15, 16, 17, 18] and higher blood pressure, total cholesterol, and triglycerides [19, 20]. There is also evidence of an independent association between PTB and ischemic heart disease (IHD) death with an HR of 1.38–2.25 [21, 22]. SGA and stillbirth have also been associated with increased risk of maternal CVD [9, 23].

Although women with APOs have increased risk for CVD, it is unknown whether experiencing an APO leads to increased identification of CRFs after the affected pregnancy, improved education and management of CRFs, or increased healthcare utilization [24, 25]. In this study we aim to determine whether having an APO is associated with increased patient awareness of CRF diagnosis in the years after an affected pregnancy.

2. METHODS

2.1. Study design

This is a secondary analysis of a multicenter prospective cohort study. Participants were eligible for enrollment if they were nulliparous with a viable singleton pregnancy in the first trimester between October 2010 and September 2013 (n = 10,038 enrolled) [26]. Participants who agreed to follow‐up contact were seen for a subsequent study visit 2–7 years after delivery of the index pregnancy (n = 4,508 seen for a subsequent study visit) [27].

For this secondary analysis, participants in the follow‐up study were included if they had an incident CRF at the time of the study visit (2–7 years after delivery of the index pregnancy). These CRFs included chronic hypertension, dyslipidemia, and diabetes or prediabetes. Participants were considered to have a CRF if they (1) met criteria for those conditions based on blood pressure or laboratory values at the time of the study visit, (2) if they reported taking medication for any of these conditions, or (3) were told by a healthcare provider that they have any of these conditions.

Blood pressure criteria for chronic hypertension at the time of the study visit included systolic blood pressure (SBP) of ≥130 or diastolic blood pressure (DBP) ≥80. Laboratory criteria for dyslipidemia at the time of the study visit included total cholesterol ≥240 mg/dL, LDL‐c ≥160 mg/dL, HDL‐c < 35 mg/dL, or triglycerides ≥200 mg/dL. Laboratory criteria for diabetes or prediabetes at the time of the study visit included fasting glucose ≥126 mg/dL or ≥100 mg/dL, respectively.

Participants were excluded if they had pregestational diabetes, chronic hypertension, or dyslipidemia at the time of their index pregnancy as these were related to the outcome of interest; if they had any subsequent pregnancy ≥12 weeks gestation since the index pregnancy (to avoid confounding from subsequent pregnancies); or if they were not fasting for their blood draw at the follow‐up study visit.

The exposure of interest was APO during the index pregnancy. APOs included HDP (gestational hypertension, preeclampsia, eclampsia, or new onset intrapartum/postpartum hypertension), GDM [based on glucose tolerance test criteria (Appendix S1) or chart abstraction], PTB (before 37 weeks gestation) for any reason, SGA (birthweight < 10th percentile) [28], or stillbirth. Participants with APOs were compared to those who did not experience APOs.

The outcome of interest was participant awareness of their CRF diagnosis at the time of the follow‐up study visit. Those who reported taking medication or having been diagnosed by a healthcare provider were considered aware of their CRF diagnosis, while those who did not report being diagnosed or treated by a healthcare provider at the time of the study visit were considered unaware of their CRF diagnosis. Relevant questions regarding CRF awareness from the study questionnaire are listed in Appendix S2.

Access to healthcare and healthcare utilization were assessed via participant interviews at the time of the follow‐up study visit 2–7 years after delivery of the index pregnancy. These questions are detailed in Appendix S2.

As study participants were only asked “have you been told you have diabetes” (not prediabetes), we completed a sensitivity analysis of those participants with only prediabetes as their CRF diagnosis.

2.2. Statistical considerations

Based on the sample size of N = 980 participants meeting inclusion criteria for this secondary analysis with 344 (35%) exposed to an APO, we have 80% power to detect a difference between group proportions as small as 0.08 when the proportion among those with the history of APO is 0.2 under the null hypothesis, and the proportion among the group without APO is 0.2, using a two‐sided likelihood ratio test and alpha = 0.05. Such a difference is plausible and clinically meaningful.

Demographics at the time of the index pregnancy and at the time of the follow‐up study visit were compared between participants with and without APOs (Table 1). Access to healthcare and nature of healthcare utilization at the time of the follow‐up study visit were also compared between participants with and without APOs (Table 2). We report unadjusted differences in CRF awareness by APO history. For these comparisons, categorical measures are compared with chi‐square and continuous measures compared with two‐sample t‐test. To further explore participant factors associated with awareness of cardiovascular risk factors, univariate analyses of participant demographics by awareness status and by reported contact with a healthcare provider were conducted.

TABLE 1.

Participant characteristics at the time of the index pregnancy and at the time of the HHS visit (2–7 years postpartum).

Variable	Value	APO n = 344	No APO n = 636	p
From the time of first pregnancy
Age (years) at enrollment	Mean ± SD	27.3 ± 6.3	26.8 ± 5.3	0.292
Race/ethnicity	Asian	7 (2.0)	18 (2.8)	0.051
	Hispanic	55 (16.0)	105 (16.5)
	Non‐Hispanic Black	67 (19.5)	79 (12.4)
	Non‐Hispanic White	197 (57.3)	403 (63.4)
	Other	18 (5.2)	31 (4.9)
Pre‐pregnancy BMI (kg/m2)	Mean ± SD	27.5 ± 6.6	25.7 ± 6.1	<.001
Early pregnancy BMI (kg/m2)	Mean ± SD	28.3 ± 6.7	26.4 ± 6.0	<.001
Marital status	Married	193 (56.3)	401 (63.1)	0.035
Education status attained	Associates degree or more	202 (58.7)	394 (61.9)	0.323
Tobacco	Ever used	144 (42.0)	234 (36.8)	0.112
	Smoked 3 m < pregnancy	71 (20.7)	97 (15.3)	0.031
Healthcare payment	Government insurance	113 (33.0)	179 (28.2)	0.118
Weight change during pregnancy	Below guidelines	42 (12.9)	77 (12.8)	0.988
	Within guidelines	60 (18.5)	109 (18.1)
	Above guidelines	223 (68.6)	416 (69.1)
From the time of HHS visit 1 (2–7 years postpartum)
Years since delivery	1	24 (7.0)	47 (7.4)	0.955
	2	155 (45.1)	277 (43.6)
	3	114 (33.1)	220 (34.6)
	4	42 (12.2)	72 (11.3)
	5	9 (2.6)	20 (3.1)
Age in years at visit 5	Mean ± SD	30.8 ± 6.3	30.4 ± 5.4	0.273
Postpartum weight retention	>10 lbs below pp wt	17 (5.0)	31 (4.9)	0.333
	within 10 lbs of pp wt	135 (39.4)	278 (44.2)
	>10 lbs above PP wt	191 (55.7)	320 (50.9)
Postpartum weight retention	Mean ± SD	17.6 ± 24.2	14.7 ± 27.1	0.101
BMI at visit 5	Mean ± SD	30.5 ± 8.1	28.0 ± 7.7	<.001
BMI category at visit 5	<25 (normal weight)	104 (30.3)	280 (44.2)	<.001
	25 to < 30 (overweight)	82 (23.9)	163 (25.8)
	≥30 (obese)	157 (45.8)	190 (30.0)
Education status attained	Associates degree or more	205 (59.6)	437 (68.8)	0.004
Tobacco	previously	60 (17.4)	117 (18.4)	0.711
	currently	55 (16.0)	64 (10.1)	0.007
Healthcare payment	Government insurance	91 (26.5)	154 (24.3)	0.433
Testing since index pregnancy	Cholesterol	120 (34.9)	228 (35.8)	0.763
	Glucose	49 (14.3)	65 (10.3)	0.06

Note: Participant characteristics are divided by those with and without APOs. Values are reported as frequency and percent unless noted.

Abbreviation: pp, postpartum.

TABLE 2.

Access to healthcare among participants at 2–7 years after first pregnancy.

		Whole cohort	APO	No APO
Access to and Content of Care	Category (values presented as frequency and column percentage)	N = 980	n = 344	n = 636	p
Is there a place you usually go when sick/need health advice?	Yes, there is one place only	702 (72.1)	251 (73.2)	451 (71.6)	0.808
	Yes, there is more than one place	178 (18.3)	59 (17.2)	119 (18.9)
	No, there is no place	93 (9.6)	33 (9.6)	60 (9.5)
What kind of place?	None/nowhere	93 (9.6)	33 (9.6)	60 (9.6)	0.183
	Walk‐in clinic	87 (9.0)	24 (7.0)	63 (10.1)
	Doctors office	712 (73.5)	254 (74.1)	458 (73.2)
	Hospital emergency room	40 (4.1)	20 (5.8)	20 (3.2)
	Some other place, specify	37 (3.8)	12 (3.5)	25 (4.0)
Is that the same place that you go when you need routine or preventive care such as a physical examination or check up?	Yes	723 (82.6)	264 (85.4)	459 (81.1)	0.105
	No	152 (17.4)	45 (14.6)	107 (18.9)
if no… then: where do you go when you need routine or preventive care	None / nowhere	43 (17.5)	17 (21.8)	26 (15.5)	0.73
	Walk‐in clinic	14 (5.7)	4 (5.1)	10 (6.0)
	Doctors office	175 (71.1)	54 (69.2)	121 (72.0)
	Hospital emergency room	9 (3.7)	2 (2.6)	7 (4.2)
	Some other place, specify	5 (2.0)	1 (1.3)	4 (2.4)
(V5A) During the past 24 months, have you seen or talked with ____.	a doctor who specializes in women's health	837 (85.5)	278 (81.0)	559 (87.9)	0.004
	a general doctor who treats a variety of illnesses?	702 (71.7)	260 (75.6)	442 (69.6)	0.048

Note: This table describes the access to care of our patient population, both with and without APOs.

Associations between APO history and subsequent CRF awareness were estimated using three models. Model 1 was an unadjusted logistic regression model. Model 2 was a multivariable logistic regression model adjusted for seven characteristics (insurance, age at the follow‐up study visit, BMI at the follow‐up study visit, education at the follow‐up study visit, postpartum weight retention, time from index pregnancy to follow‐up study visit, and whether participants reported having been seen by a “general doctor who treats a variety of illnesses”). Model 3 was a multivariable logistic regression model adjusted for characteristics identified using backwards selection to yield a parsimonious model, with APO history forced to remain. In a sensitivity analysis, we excluded people with only prediabetes and no other CRFs.

This data analysis was generated using SAS software, Version 9.4 of the SAS System for Windows; SAS Institute Inc. Cary, NC. Graphics were created using GraphPad Prism version 10.0.3 for Windows, GraphPad Software, La Jolla California.

3. RESULTS

There were 980 participants meeting inclusion criteria for this secondary analysis (Figure 1). Most people (77%) met only one risk factor for inclusion, 20% had two risk factors, and 3% had all three. The most common risk factor for inclusion was HTN (51%) followed by dyslipidemia (38%) and prediabetes/diabetes (37%). Half (50%) of participants with prediabetes/diabetes were people with prediabetes alone but no other risk factors.

FIGURE 1.

Enrollment and inclusion cascade. Flowchart of the included and excluded participants in the final cohort available for analysis. nuMoM2b, Nulliparous Pregnancy Outcomes Study: Monitoring Mothers‐to‐be; HHS: Heart Health Study.

The exposure, APO in index pregnancy, was present in 35% of participants (344/980), with the most common being HDP (25%) followed by PTB (8.5%), SGA (4.5%), GDM (4.0%), and stillbirth (0.3%). Most participants had only one APO (28%). 6.5% had two APOs, and 0.31% had three APOs. History of APO was associated with higher BMI and tobacco use both at the time of the 1st study visit early in the index pregnancy and at the follow‐up study visit 2–7 years after the pregnancy (Table 1).

When comparing access to care and where participants received their medical care by history of APO, groups had similar responses when asked “Is there a place you usually go when sick/need healthcare advice?”; 72% responded yes—one place only, 18% reported more than one place, and 10% reported they had no place. Of those seeking care for illness, 74% went to a doctor's office, 9% a walk‐in clinic, 4% an emergency room, 4% other, and 10% reported not seeking care. When asked about any visits in the past 24 months, 81% and 88% (APO vs. no APO, p = 0.004) had seen a provider who specializes in women's health (obstetrician or gynecologist) and 76% versus 70% (p = 0.048) had seen a provider who treats a variety of illnesses (general practice, family medicine, or internal medicine) (Table 2). There were only 3 stillbirth deliveries included in this secondary analysis (versus 54 in the parent cohort study out of N = 10,038). The 70% exclusion rate among stillbirth deliveries (versus 55% in live birth deliveries) was driven by non‐participation in the follow‐up study.

Comparisons of participant factors by awareness of CV risk factor diagnosis and by engagement with a healthcare provider are reported in Appendix S3. Participants who were aware of their CV risk factors were older (28.3 ± 5.9 vs. 26.6 ± 5.6 years at initial pregnancy, p < 0.001), had more years of education at baseline (66.8% vs. 59.1% completed at least an undergraduate degree, p = 0.038) and at 2–7 year follow‐up (73.2% vs. 63.4% completed at least an undergraduate degree, p = 0.007), and were less likely to have government insurance at 2–7 year follow‐up (19.5% vs. 26.6%, p = 0.032) compared to participants who were unaware of their CV risk factor diagnosis. Participants who engaged with a healthcare provider in the 24 months before the 2–7 year follow‐up visit were older (27.3 ± 5.7 vs. 26.2 ± 5.6 years at initial pregnancy, p = 0.008), more likely to have ever used tobacco (41.8% vs. 30.7%, p = 0.001), more likely to have had gestational weight gain within the recommended range (20.3% vs. 13.0%, p < 0.001), and more likely to have undergone screening for cholesterol (41.6% vs. 20.2%, p < 0.001) and glucose (13.1% vs. 8.3%, p = 0.037) compared to participants who had not engaged with a healthcare provider in the 24 months before the study visit.

The primary outcome (self‐awareness of a diagnosis of hypertension, dyslipidemia, or prediabetes/diabetes) at the study visit 2–7 years after pregnancy was present in 22% of participants (220/980). 26% of participants with an APO were aware of their diagnosis versus 20% of participants without an APO (p = 0.04). Table 3 details these diagnoses and the corresponding questionnaire data.

TABLE 3.

Primary outcome of self‐awareness of CRF diagnosis.

Self‐reported health history	APO	No APO	p value
	n = 344	n = 636
Composite: (self‐report) hypertension, high cholesterol, or diabetes	90 (26.2)	130 (20.4)	0.04
Been told you have hypertension	43 (12.5)	33 (5.2)	<0.001
Been told you have high cholesterol	49 (14.2)	97 (15.3)	0.672
Been told you have diabetes	4 (1.2)	7 (1.1)	0.93
Hypertension Diagnosis	206 (59.9)	292 (45.9)	<0.001
Been told you have hypertension	43 (20.9)	33 (11.3)	0.003
• Told on two or more healthcare visits that you had hypertension	27 (62.8)	17 (51.5)	0.324
• Currently prescribed medication for high blood pressure	9 (20.9)	3 (9.1)	0.161
High cholesterol diagnosis (any Hyperlipidemia)	109 (31.7)	262 (41.2)	0.003
Been told you have high cholesterol	49 (45.0)	97 (37.0)	0.154
Diabetes diagnosis (prediabetes or diabetes)	138 (40.1)	225 (35.4)	0.143
Been told you have diabetes	4 (2.9)	7 (3.1)	0.909
Taking medication for glucose at V5 (HHS1)	10 (7.2)	16 (7.1)	0.961
Been told you have diabetes or taking medication for glucose at V5	12 (8.7)	20 (8.9)	0.95

Note: The primary outcome of awareness of CRF diagnosis (hypertension, hyperlipidemia, and/or diabetes) is divided by APO versus non‐APO as described above. The top portion of this table, as well as the three grey rows, uses the N at the top as the denominator. Rows beneath each grey row use the frequency in the grey row as the denominator. Bulleted questions about hypertension use the preceding “been told you have hypertension” frequency as the denominator.

Abbreviation: CRF, cardiovascular risk factor, HHS1/V5, heart health study visit 1.

Prior to adjustment, the odds ratio (OR) for the association between history of APO and subsequent CRF awareness was statistically significant at 1.38 (95% CI 1.01–1.88), indicating that women with APOs were more likely to report awareness of CRFs than women without APOs. However, after full adjustment in Model 2 and the use of a parsimonious model in Model 3, the association was similar in magnitude but no longer significant (Figure 2). Backwards selection removed all characteristics except for two that were positively associated with awareness of CRF diagnosis in the final model—seeing “a provider who treats a variety of illnesses” (aOR: 1.56, 1.08–2.24) and age (aOR: 1.05, 1.02–1.08).

FIGURE 2.

Forest plot of participant awareness of CRFs among participants with versus without history of APOs. Forest plot of the odds ratio and 95% confidence interval for participant awareness of CRFs among participants with versus without history of APOs, utilizing three models. Model 1 is unadjusted. Model 2 is fully adjusted for seven characteristics: (1) insurance (ever had government insurance at either V1 or V5, vs. otherwise), (2) age at visit 5 (continuous), (3) BMI at visit 5 (three categories: < 25, 25–29, 30+), (4) education at visit 5 (associates or more vs. less), (5) postpartum weight retention (three categories), (6) time (years) from index pregnancy to HHS visit (continuous), and (7) seen a general doctor who treats a variety of illnesses? (yes/no). Model 3 is a parsimonious model achieved through backward variable selection: (1) age at visit 5 (continuous) and (2) seen a general doctor who treats a variety of illnesses? (yes/no).

In the sensitivity analysis in which the 182 participants with prediabetes as their only risk factor were excluded, none were self‐aware of a CRF diagnosis. For the remaining 798 participants, self‐awareness of a diagnosis of hypertension, dyslipidemia, or diabetes increased to 28%. The unadjusted odds ratio was slightly greater in magnitude than that of the main analysis (OR: 1.40, 1.02–1.93); however, results from adjusted models were overall similar (Table 4).

TABLE 4.

Sensitivity analysis: OR for the association between APO history and participant awareness of CRFs 2–7 years postpartum (excluding prediabetes).

	Original results (prediabetes included)			Sensitivity results (prediabetes excluded)
Odds ratio	OR estimate	95% confidence limits		OR estimate	95% confidence limits
Model 1—unadjusted	1.379	1.013	1.877	1.403	1.019	1.931
Model 2—fully adjusted	1.346	0.974	1.859	1.364	0.974	1.910
Model 3—parsimonious	1.313	0.960	1.797	1.353	0.978	1.870

Note: This is the sensitivity analysis excluding patients with prediabetes as their only CRF using all three models. Model 1 is unadjusted. Model 2 is fully adjusted for seven characteristics: (1) insurance (ever had government insurance at either V1 or V5, vs. otherwise), (2) age at visit 5 (continuous), (3) BMI at visit 5 (three categories: < 25, 25–29, 30+), (4) education at visit 5 (associates or more vs. less), (5) postpartum weight retention (three categories), (6) time (years) from index pregnancy to HHS visit (continuous), and (7) seen a general doctor who treats a variety of illnesses? (yes/no). Model 3 is a parsimonious model achieved through backward variable selection: (1) Age at visit 5 (continuous) and (2) seen a general doctor who treats a variety of illnesses? (yes/no).

4. DISCUSSION

In a cohort of post‐pregnancy participants who had one or more CRFs 2–7 years after their first pregnancy, we did not identify an association between APO history and CRF awareness after adjusting for potential confounders. Despite most participants indicating that they had been evaluated by a women's health or primary care provider (PCP) in the prior 2 years, only 22% of participants with a CRF reported awareness of their diagnosis. This finding should lead to further investigation about utilization of the postpartum visit to educate patients on their cardiovascular risk, identification knowledge gaps among both providers and patients when it comes to the association between APOs and increased CRFs, and transitions to long‐term preventative care visits to manage these risk factors.

The American Heart Association (AHA) highlights that pregnancy may be the first time a woman formally interacts with the healthcare system and that most women who die of CVD have no formal diagnosis of their CVD [29]. The American College of Obstetricians and Gynecologists (ACOG) recommend that women with APOs and/or CRFs have a comprehensive cardiovascular risk screening within 3 months of delivery [30], and that those with HDP have a blood pressure check no later than 7–10 days after delivery [31]. Ideally, women would have follow‐up at 6 weeks, 12 weeks, 6 months, and 12 months with transition to a PCP around 8–12 weeks [32]. However, obstetric care after delivery is only utilized in 50%–60% of women with a complicated pregnancy [33] and only 18%–25% of postpartum patients with APOs or chronic health conditions are seen by a PCP within 6 months [32]. Considering that the cumulative incidence for premature cardiovascular events is increased even within one year of the index pregnancy [34], connecting women with a healthcare provider after delivery is essential to identify, treat, and educate them on CRFs. 90% of study participants reported access to and utilization of healthcare, likely due to their participation in the nuMoM2b‐HHS study. Despite this access, awareness of CRFs was low.

It is important to recognize the challenges of counseling and treating these women, including successfully transferring them from postpartum obstetric care to general medical care, educating providers about the association between APOs and CRFs, and effectively educating patients. To address the gap between postpartum obstetric care and general medical care, several groups have studied the concept of postpartum transition clinics. A specialty clinic focusing on HDP patient education, medication management, and transition to a PCP was feasible and able to sustain an internists annual salary, but the no show rate was high at 25% [35]. Cusimano et al. established a Maternal Health Clinic and successfully had 92 patients referred; a similar clinic established in Kingston, Ontario had successful referrals, however, only 50% of women scheduled a visit and many were seen >1 year postpartum [36, 37]. Janmohamed et al. established a hospital‐based postpartum clinic focused on lifestyle modification counseling. In 6 months, 21 women had no statistically significant change in weight [38]. Even when seen regularly, female patients who are established with a healthcare provider are less likely to receive guideline‐recommended statin therapy, more likely to decline treatment, and less likely to continue statin therapy when compared to male patients [39]. When 173 internal medicine and obstetrics and gynecology providers were surveyed, internists were more likely to appropriately test and treat identified cardiovascular disease compared to obstetricians [40]. Thus far, provider‐patient interactions seem to fall short of the education needed. We know that increased awareness is associated with improved outcomes. Bakhit et al. completed a systemic review and meta‐analysis demonstrating that communication of CVD risk information reduced overall risk factors and enhanced patients’ perception of their risk [41]. Altman et al. developed a pilot program where 35 patients participated in a 4‐month, community‐based bilingual preventative cardiovascular education program. Postintervention, patients demonstrated improvement in their knowledge of CV symptoms and improved clinical health status [42]. With this in mind, future education efforts could focus on initiatives such as public heath marketing or clinical outreach to disseminate education materials to communities in a tailored or personalized fashion to increase awareness of the association of APOs and increased risk of CVD. Some educational resources are available via ACOG [31] and the Preeclampsia Foundation [43].

While data confirm increased CVD risk in women with APOs, integrating this information into risk prediction models to include women of childbearing age has not led to improved risk classification. This leads to difficulty educating people on their cardiovascular risk. There are no evidence‐based guidelines or randomized controlled trials to guide providers on how to treat these patients [32, 44]. Historically, when estimating future risk of atherosclerotic cardiovascular disease (ASCVD), tools such as the Framingham Risk score [45] or ASCVD 10‐year risk calculator [46] have been utilized. Thus far, preeclampsia and preterm delivery have been added to the ACC/AHA prevention guidelines as risk‐enhancing factors to be used in adults who are at borderline or intermediate 10‐year ASCVD risk [47]. Further attempts to add APOs to our current risk‐stratification models have yielded discordant results [48, [49]. While specifics on reclassifying cardiovascular risk are unclear, APOs should prompt aggressive prevention strategies [23]. While our study appears to overcome the barrier of access to care, there is still work to be done educating providers and patients on their CRF and future implications for cardiovascular health.

4.1. Study limitations

Our study has several limitations. Our analyses rely on participant‐reported data collected via standardized interview in the form of a survey. Questions from this survey may be interpreted differently among participants. For example, questions were worded as “have you ever been told by a doctor or other healthcare professional that you have (hypertension, high cholesterol, or diabetes).” Participants may answer this question “no” but still be aware of their diagnosis through other means that we have failed to measure. Further, participants may deny a diagnosis if it is now well‐controlled (i.e., they have a diagnosis of hypertension and are on medical therapy with a subsequent reduction of blood pressure into the normal range). We considered participants who were taking treatment for their CRFs to be aware of the risk factors even if they answered “no” to the survey questions regarding a diagnosis of that risk factor, which should mitigate this potential for misclassification of participant awareness. Participants were not explicitly asked if they were aware of increased cardiovascular risk following an APO, which could have provided further insight into their perceived risk.

The rate of APOs reported in our cohort was high at 35% when compared to other studies (10%–20%) [1]. This is not surprising given that we limited our study population to participants with CRFs at the time of the follow‐up study visit. Our cohort also reported higher interaction with the healthcare system when compared to other studies. Participants in this prospective study were enrolled in the first trimester of their pregnancy and agreed to several study visits over the course of their pregnancy as well as additional follow‐up in the years following. Thus, participants in this cohort study may have resources and access to healthcare that are not necessarily generalizable to the broader U.S. population.

Reasons for non‐participation in the follow‐up study after a stillbirth (70% exclusion rate among stillbirth deliveries versus 55% in live birth deliveries) included declining to participate or having a subsequent pregnancy that precluded completing the follow‐up study visit. Since these participants were largely excluded from our secondary analysis, we are unable to reliably measure an association between this specific APO and the outcome of interest.

5. CONCLUSIONS

Both pregnancy and postpartum are valuable time periods to assess and reduce CVD risk in young adult women, as many have limited or no knowledge of the link between APOs and increased risk of CVD. In our study, only one in five participants recognized their diagnoses of CRFs despite interaction with a healthcare provider. There is need for reliable, consistent, and thorough postpartum care for women, especially those experiencing APOs, as well as a focus on education for both providers and patients regarding cardiovascular disease risk after pregnancy. Further study is needed to appropriately stratify cardiovascular risk in postpartum women and identify ways to improve their engagement with and the effectiveness of healthcare services.

CONFLICT OF INTEREST STATEMENT

Lauren H. Theilen is supported by K23HL159316. The remaining authors declare no conflicts of interest.

ETHICS STATEMENT

The nuMoM2b‐HHS study was approved by the local governing institutional review boards at all study sites, and all participants provided written informed consent.

Supporting information

Supporting Information

Supporting Information