Background
Middle-aged women are at higher risk than men of death after coronary artery revascularization. Maternal placental syndromes (gestational hypertension, preeclampsia, placental abruption, and placental infarction) are associated with premature coronary artery disease, but their influence on survival after coronary artery revascularization is unknown.
Objective
The purpose of this study was to determine whether a history of maternal placental syndromes alters the risk of death after coronary artery revascularization in middle-aged women.
Study Design
We completed a population-based retrospective cohort study among all hospitals in Ontario, Canada, where universal health care includes all aspects of antenatal and delivery care as well as all outpatient and inpatient health care, which includes coronary revascularization. We included 1985 middle-aged women who underwent a first percutaneous coronary intervention or coronary artery bypass grafting between 1993 and 2012 and who had ≥1 previous delivery. We excluded those with cardiovascular disease ≤1 year before or coronary revascularization ≤90 days after any delivery. The main study outcome, determined a priori, was all-cause death. Hazard ratios were adjusted for age, socioeconomic status, parity, revascularization type, time since last delivery, hypertension, diabetes mellitus, obesity, dyslipidemia, tobacco or drug dependence, and kidney disease.
Results
Three hundred sixty-two of 1985 women (18.2%) who underwent coronary artery revascularization had a previous maternal placental syndrome event. The mean age at index coronary revascularization was 45 years; percutaneous coronary intervention comprised approximately 80% of procedures. After a mean follow-up time of approximately 5 years, 41 deaths (2.2 per 100 person-years) occurred in women with previous maternal placental syndromes and 83 deaths (1.1 per 100 person-years) in women without maternal placental syndrome (adjusted hazard ratio, 1.96; 95% confidence interval, 1.29–2.99). Of the maternal placental syndrome subtypes, the risk of death was significant in women with placental abruption (adjusted hazard ratio, 2.79; 95% confidence interval, 1.31–5.96), placental infarction (adjusted hazard ratio, 3.09; 95% confidence interval, 1.23–7.74), and preeclampsia (adjusted hazard ratio, 1.61; 95% confidence interval, 1.00–2.58). Women with maternal placental syndrome in ≥2 pregnancies had the highest adjusted hazard ratio of death (4.31; 95% confidence interval, 1.71–10.89).
Conclusion
In middle-aged women who undergo coronary revascularization, previous maternal placental syndrome doubles the risk of death; recurrent maternal placental syndrome quadruples that risk. Some covariates and secondary measures may not have been well-captured and classified herein, leading to residual confounding.
Complications after coronary artery revascularization include death, acute myocardial infarction (MI), stroke, and the need for repeat revascularization. After percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG), young and middle-aged women are more likely to die and incur major complications than their male counterparts, but this gap narrows with advancing age. The reasons that middle-aged women are especially at higher risk of death are not explained by conventional cardiovascular risk factors.
Women whose pregnancy was affected by a maternal placental syndrome (MPS; gestational hypertension, preeclampsia, placental abruption, and placental infarction ) are especially at higher risk of premature coronary artery disease. This risk is further magnified when MPS is affected concomitantly by preterm delivery, poor fetal growth, and/or stillbirth, each a reflection of worse placental vascular function. In the years that follow, women whose pregnancy was affected by MPS also exhibit high rates of the metabolic syndrome (MetSyn), heart failure, chronic hypertension, and type 2 diabetes mellitus.
Because middle-aged women are at higher risk of adverse outcomes after coronary revascularization and because MPS predisposes to premature onset of coronary artery disease, it is worthwhile to know the degree to which a history of MPS impacts the risk of death and major complications after coronary artery revascularization. We measured death and major adverse cardiovascular events (MACE) in relation to a history of MPS among a large cohort of middle-aged women who underwent coronary artery revascularization.
Methods
We completed a population-based retrospective cohort study in the entire province of Ontario with the use of linked health care administrative databases. All Ontario residents are enrolled in the Ontario Health Insurance Plan, which covers all aspects of antenatal and delivery care and all outpatient and inpatient health care, including coronary revascularization. More than 99% of deliveries are done in hospital.
We considered women aged ≥20 years at the time of an index PCI or CABG that was performed between July 1, 1993, and February 28, 2012. Each participant was eligible if she had at least 1 recorded obstetric delivery of a liveborn or stillborn child a minimum of 90 days before the index coronary artery revascularization, in which she was ≤50 years old at the time of delivery. We excluded women who had been hospitalized with any form of cardiovascular disease, cardiac anomaly, dysrhythmia, heart failure, or coronary revascularization within 365 days before or 90 days after any obstetric delivery ( Supplemental Table 1 ). This was to minimize the likelihood of including those with an MI from a spontaneous coronary artery dissection in pregnancy, heart failure from a peripartum cardiomyopathy, or a stroke caused by severe preeclampsia.
Exposures and outcomes
A history of MPS was our main exposure of interest. MPS was defined as any diagnosis of gestational hypertension, preeclampsia, placental abruption, and/or placental infarction during a previous delivery hospitalization ( Supplemental Table 1 ). All placentas are required to be examined by the delivery care provider; a diagnosis of placental abruption or placental infarction was based on a clinical description of the placenta at delivery. The index coronary revascularization (PCI or CABG) was captured as an inpatient or same-day procedure.
The main study outcome of interest, defined a priori, was all-cause death after the index coronary revascularization. Secondary study outcomes included (1) a MACE composite of death, any MI, or any stroke ( Supplemental Table 1 ), (2) nonfatal MI, (3) stroke, and (4) the MACE composite outcome that was expanded to include any repeat coronary artery revascularization. Although there is debate about the use of MACE as a composite outcome, we applied both a basic and expanded definition of MACE.
Database sources
We used linked administrative health databases housed at the Institute for Clinical Evaluative Sciences to identify all study variables. These datasets were linked with the use of unique encoded identifiers and analyzed at the Institute for Clinical Evaluative Sciences, as described elsewhere and in Supplemental Table 1 .
Statistical analysis
Time-to-event analyses were conducted with Cox proportional hazards regression models to derive an incidence rate, hazard ratio, and 95% confidence interval (CI) for each study outcome that compared women with a history of MPS to those without a history of MPS. Time-to-event curves that were created with the use of the Kaplan-Meier estimate of the survival function and tested by the log-rank test were presented with only up to 14 years of follow-up evaluation to avoid required suppression of small numbers thereafter; 1- and 5-year cumulative probability estimates of death also were determined for both exposure groups.
For each endpoint analysis, a participant was censored (ie, determined to have not had a study outcome event) at the point in time in which he/she either reached the end of the period of March 31, 2013, without having experienced that study endpoint event or died during the period of study.
In the main model, the hazard ratios were adjusted (aHRs) for participant age, income quintile, rural residence, parity, and type of index coronary artery revascularization (PCI or CABG), each at the time of the index coronary artery revascularization, the time between the last obstetric delivery and index coronary artery revascularization, and chronic hypertension, diabetes mellitus, obesity, dyslipidemia, tobacco or drug dependence, or kidney disease and each at ≤24 months preceding the index coronary artery revascularization ( Supplemental Table 1 ). Testing of the proportional hazards assumption was done by including the interaction between MPS and a function of survival time.
We performed several analyses that are specific to MPS and its relation to death, which is the primary study outcome. The first analysis explored the association by each MPS component. The second analysis examined MPS with preterm delivery at <37 weeks gestation, which is a more pathologic representation of MPS. The third analysis explored MPS in 1 or ≥2 pregnancies. The fourth analysis expanded the exposure of interest to MPS plus a diagnosis of poor fetal growth, stillbirth 20 weeks gestation, and/or preterm birth <37 weeks gestation. The fifth analysis expanded the exposure of interest to MPS plus 0, 1, or ≥2 features of the MetSyn in the last pregnancy. Rather than use the formal diagnostic criteria for MetSyn, we considered each individual feature separately and additively ( Supplemental Table 1 ), without adjusting for chronic hypertension, diabetes mellitus, obesity, or dyslipidemia in this specific analysis. We reran the main model for the outcome of death and adjusted for an inpatient diagnosis of acute MI or heart failure at ≤7 days before the revascularization procedure, including the day of the procedure.
We calculated the population attributable fraction (PAF) and 95% CI for death in association with previous MPS ( Supplemental Table 2 ).
All probability values were 2-sided, at a significance level of .05. All statistical analyses were performed with SAS software (version 9.3, SAS System for Unix; SAS Institute Inc, Cary, NC).
Results
We identified 3496 women who underwent coronary revascularization and who had at least 1 preceding obstetric delivery during the period of study. Of these, we excluded 1511 women ( Figure 1 ).
In the final cohort of 1985 women, 362 (18.2%) had a history of MPS in a previous pregnancy, of which 317 (87.6%) had MPS once and 45 (12.4%) had MPS in ≥2 pregnancies ( Table 1 ). The mean age at index coronary revascularization was 44.7 and 46.5 years, respectively, in women with and without previous MPS ( Table 1 ). The mean (± SD) time from the last pregnancy to the index revascularization was 11.3 ± 4.7 and 14.2 ± 5.0 years, respectively. In the 7 days up to and including the index revascularization, 55% women had an acute MI, and approximately 3% had heart failure. PCI comprised 80.1% and 82.9%, respectively, of all index revascularization procedures. Women with and without MPS were observed for a total of 1837 and 7842 person-years after their index coronary revascularization, for a mean duration of 5.1 and 4.8 years, respectively. The assumption of proportional hazards was met for all models.
| Characteristic | Measure | Maternal placental syndrome (n = 362) | No maternal placental syndrome (n = 1623) |
|---|---|---|---|
| In any preceding obstetric delivery, n (%) a | |||
| Maternal placental syndrome b | In 1 pregnancy | 317 (87.6) | — |
| In ≥ 2 pregnancies | 45 (12.4) | — | |
| By maternal placental syndrome subtype, n (%) | |||
| Gestational hypertension | 120 (33.1) | — | |
| Preeclampsia or eclampsia | 206 (56.9) | — | |
| Placental abruption | 48 (13.3) | — | |
| Placental infarction | 33 (9.1) | — | |
| Poor fetal growth, n (%) | 37 (10.2) | 65 (4.0) | |
| Stillbirth, n (%) | 31 (8.6) | 23 (1.4) | |
| Preterm birth, n (%) | 104 (28.7) | 135 (8.3) | |
| Median parity, n (interquartile range) | 1 (1–2) | 1 (1–2) | |
| ≥1 Features of the metabolic syndrome, n (%) c | 285 (78.7) | 953 (58.7) | |
| Within 24 months preceding the index coronary artery revascularization, n (%) | |||
| Chronic hypertension | 203 (56.1) | 606 (37.3) | |
| Diabetes mellitus | 153 (42.3) | 375 (23.1) | |
| Obesity | 36 (9.9) | 125 (7.7) | |
| Dyslipidemia | 82 (22.7) | 354 (21.8) | |
| Kidney disease | 49 (13.5) | 117 (7.2) | |
| Tobacco or drug dependence | 89 (24.6) | 546 (33.6) | |
| At the index coronary artery revascularization | |||
| Mean age, y ± SD | 44.7 ± 6.6 | 46.5 ± 6.3 | |
| Mean years since the most recent obstetric delivery, n ± SD | 11.3 ± 4.7 | 14.2 ± 5.0 | |
| Residential income quintile, n (%) | 1 | 86 (23.8) | 388 (23.9) |
| 2 | 73 (20.2) | 364 (22.4) | |
| 3 | 74 (20.4) | 300 (18.5) | |
| 4 | 77 (21.3) | 301 (18.6) | |
| 5 | 48 (13.3) | 258 (15.9) | |
| Missing | ≤5 | 12 (0.74) | |
| Rural residence, n (%) | 46 (12.7) | 242 (14.9) | |
| Acute myocardial infarction ≤7 days preceding the coronary artery revascularization, n (%) | 194 (53.6) | 902 (55.6) | |
| Heart failure ≤7 days preceding the coronary artery revascularization, n (%) | 11 (3.0) | 42 (2.6) | |
| Type of coronary artery revascularization, n (%) | |||
| Percutaneous coronary revascularization | 290 (80.1) | 1,346 (82.9) | |
| Coronary artery bypass grafting | 72 (19.9) | 277 (17.1) |
a An obstetric delivery that resulted in a livebirth or stillbirth after 20 weeks gestation, in which the most recent delivery was at least 90 days preceding the index coronary artery revascularization date
b Comprises gestational hypertension, preeclampsia/eclampsia, placental abruption, and/or placental infraction
c Limited to the most recent obstetric delivery, comprising chronic hypertension, prepregnancy or gestational diabetes mellitus, obesity, or dyslipidemia.
Deaths
For the primary outcome, there were 41 deaths in women with a history of MPS and 83 deaths in women without MPS, corresponding to 1-year cumulative probability estimates of 3.6% and 1.8%, 5-year estimates of 5.9% and 4.9%, and incidence rates of 2.2 and 1.1 per 100 person-years, respectively ( Figure 2 , A; Table 2 ). In the main model, after adjustment for multiple risk factors, the aHR for death was slightly attenuated (1.96; 95% CI, 1.29–2.99). The PAF for death attributable to MPS was 5.6% (95% CI, 2.5–7.5). Further adjustment for acute MI or heart failure within the previous 7 days did not alter the results (aHR, 1.99; 95% CI, 1.31–3.03). In a post hoc analysis, further adjustment within the main model for systemic lupus erythematosus or rheumatoid arthritis that was diagnosed within 24 months preceding the index coronary artery revascularization did not change the aHR (1.94; 95% CI, 1.28–2.96).
| Outcome | Maternal placental syndrome (n = 362) | No maternal placental syndrome (n = 1623) | Hazard ratio (95% confidence interval) | |||
|---|---|---|---|---|---|---|
| Events, n (% of all women) | Incidence rate per 100 person-years | Events, n (% of all women) | Incidence rate per 100 person-years | Crude | Adjusted b | |
| Primary: all-cause death | 41 (11.3) | 2.2 | 83 (5.1) | 1.1 | 2.17 (1.49–3.16) | 1.96 (1.29–2.99) |
| Secondary | ||||||
| Major adverse cardiovascular event composite c | 102 (28.2) | 6.6 | 331 (20.4) | 4.9 | 1.36 (1.09–1.69) | 1.26 (0.99–1.60) |
| Non-fatal myocardial infarction | 68 (18.8) | 4.4 | 254 (15.7) | 3.7 | 1.18 (0.91–1.55) | 1.12 (0.84–1.49) |
| Stroke | 6 (1.7) | 0.33 | 21 (1.3) | 0.27 | 1.25 (0.50–3.10) | 0.98 (0.36–2.70) |
| Expanded major adverse cardiovascular event composite d | 135 (37.3) | 11.1 | 489 (30.1) | 8.6 | 1.26 (1.04–1.53) | 1.10 (0.90–1.34) |
a Gestational hypertension, preeclampsia, placental abruption, and/or placental infarction
b Adjusted for age, income quintile, rural residence, parity, type of index coronary artery revascularization (percutaneous coronary intervention or coronary artery bypass grafting) (1) each at the time of the index coronary artery revascularization, the time between the last obstetric delivery and index coronary artery revascularization, and chronic hypertension, diabetes mellitus, obesity, dyslipidemia, tobacco or drug dependence, or kidney disease and (2) each within 24 months preceding the index coronary artery revascularization
c All-cause death, myocardial infarction, or stroke
d All-cause death, myocardial infarction, stroke, or subsequent coronary artery revascularization.
Of the MPS subtypes, the risk of death was significant in women with placental abruption (aHR, 2.79; 95% CI, 1.31–5.96) and placental infarction (aHR, 3.09; 95% CI, 1.23–7.74) and women with preeclampsia (aHR, 1.61; 95% CI, 1.00–2.58; Figure 3 ). Women with MPS in ≥2 pregnancies had the highest aHR of death (4.31; 95% CI, 1.71–10.89). The risk of death was also more pronounced in women with MPS and ≥2 features of the MetSyn.
The cause of death was available for 88 of 124 fatalities (71%; Supplemental Table 3 ). More than one-third of deaths were due to heart disease. The number of deaths of unknown cause was higher among women with MPS because cause-of-death records were complete only up to December 2011.
MACE
The MACE composite was higher in women with MPS than without MPS ( Figure 2 , B; Table 2 ). After adjustment for all covariates, the risk was no longer significant (aHR, 1.26; 95% CI, 0.99–1.60). The risk of nonfatal MI ( Figure 2 , C) and stroke ( Figure 2 , D) did not significantly differ between groups ( Table 2 ). The addition of subsequent revascularization to the MACE composite outcome did not appreciably alter the aHR ( Table 2 ).
Results
We identified 3496 women who underwent coronary revascularization and who had at least 1 preceding obstetric delivery during the period of study. Of these, we excluded 1511 women ( Figure 1 ).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
