Background
The incidence of pregnancy in advanced age among women is increasing because of the availability of assisted reproduction, although the long-term health consequences are not known.
Objective
The purpose of this study was to determine the effect of pregnancy in advanced age on the occurrence of cardiovascular events in a large cohort of postmenopausal women.
Study Design
We analyzed the data for 72,221 women aged 50-79 years who were enrolled in the observational arm of the Women’s Health Initiative study. We determined the effect of pregnancy in advanced age (last pregnancy at age ≥40 year) on the risk of ischemic stroke, hemorrhagic stroke, myocardial infarction, and cardiovascular death over a mean period (±standard deviation) of 12±1 years using Cox Proportional Hazards analysis after adjusting for potential confounders.
Results
A total of 3306 of the 72,221 participants (4.6%) reported pregnancy in advanced age. Compared with pregnancy in normal age, the rates of ischemic stroke (3.8% vs 2.4%), hemorrhagic stroke (1.0% vs 0.5%), and cardiovascular death (3.9% vs 2.3%) were significantly higher among women with pregnancy in advanced age. In multivariate analysis, women with pregnancy in advanced age were 50% more likely to experience a hemorrhagic stroke (hazard ratio, 1.5; 95% confidence interval, 1.0–2.1) after adjustment for age, race/ethnicity, congestive heart failure, systolic blood pressure, atrial fibrillation, alcohol use, and cigarette smoking. There was no significant difference in the risk of ischemic stroke, myocardial infarction, and cardiovascular death among women with pregnancy in advanced age after adjustment for potential confounders.
Conclusion
Women with pregnancy at an advanced age have a higher risk for hemorrhagic stroke in the postmenopausal period.
There is an increasing number of women with pregnancy and childbirth in the fourth decade of life for different reasons. In United States between 2007 and 2011, birth rates have decreased for all women aged <35 years, unchanged for women aged 35–39 years, and increased for women aged 40–44 years. Several studies have found increased rates of stillbirth, perinatal death, preterm birth, low birthweight, and maternal complications such as gestational diabetes mellitus and hypertension, preeclampsia, and need for interventions such as cesarean delivery among women ≥40 years. The rate of stroke and myocardial infarction during pregnancy and puerperium is higher among women with pregnancy in advanced age. Preconception counselling regarding the short-term risks of pregnancy that are associated with advanced maternal ages is recommended for women >40 years. However, the long-term health consequences in women with pregnancy in advanced age are not known.
We performed this study to determine the effect of pregnancy in advanced age on long-term occurrence of cardiovascular events in a large cohort of postmenopausal women.
Materials and Methods
We analyzed the data for 93,676 women aged 50–79 years who were enrolled in the observational arm of the Women’s Health Initiative Study. Study participants were enrolled at 40 centers throughout the United States between October 1, 1993, and December 31, 1998. All participants provided informed consent using materials approved by Institutional Review Boards at each center. Women who had never been pregnant were excluded from the analysis.
Baseline evaluation
Standardized questionnaires were used at baseline assessment to assess demographic and clinical data and data regarding family and medical history. Data were acquired based on self-report with the use of standard questionnaires. For other data, certified staff members took physical measurements that included blood pressure, height, and weight and collected blood samples at the clinic visit. The participants were asked how old they were at the end of the last pregnancy; response was sought in following age groups: <20, 20–24, 25–29, 30–34, 35–39, 40–44, and ≥45 years. Use of oral anticoagulant was ascertained from the Medication and Supplement Inventory, which was collected once at baseline and subsequently at year 3. We categorized the participants into 2 groups based on the response: those aged <40 years and those aged ≥40 years at the time of the last pregnancy.
Cardiovascular events ascertainment
The participants were followed for 8–12 years. Annual mailed follow-up forms updated information on hormone treatment and other selected risk factor information and facilitated structured initial reporting of clinical events. Specific details of myocardial infarction (MI) and stroke-related hospitalizations were obtained if applicable via a standardized questionnaire that was administered by phone, in-person interview, or self-completed form. Portions of the medical record (discharge summary and results of relevant diagnostic and laboratory tests) were reviewed by a designated local adjudicator for conformation of event.
The local adjudicator classified incident stroke events into ischemic stroke and hemorrhagic stroke (included both intracerebral and subarachnoid hemorrhages) based on the results of neuroimaging tests. Central oversight was provided by review of a fraction of the stroke events in the observational study by central neurologists to ensure quality of stroke event classification. An incident cardiovascular heart disease event during follow-up period was defined as the first-time occurrence of either (1) acute MI that required overnight hospitalization, (2) coronary revascularization procedures, which included percutaneous transluminal coronary angioplasty, stent placement, and coronary artery bypass graft surgery, or (3) coronary death. MI was identified if the level of any cardiac enzyme (creatine kinase, lactate dehydrogenase, troponin, or myoglobin) was ≥2 times the upper limit when participant had ischemic symptoms or if the level were 1–2 times the upper limit of normal with Q waves or ST segment or T wave abnormalities that were suggestive of an MI or if enzymes were normal or absent; however, evolving Q-wave and evolving St segment or T wave abnormalities were documented. For any death event, information was obtained on any outcomes that occurred between the participant’s last routine contact and her date of death. Data linkage with the National Death Index of the National Center for Health Statistics was used to ascertain survival and cause of death for all Women’s Health Initiative study participants. Women’s Health Initiative study participants who are lost to follow up or who were known to be dead were matched to the National Death Index to search for otherwise unreported deaths and to ascertain causes of death.
Statistical analysis
Cox proportional hazards analysis was used to estimate the hazard ratio (HR) for ischemic stroke, hemorrhagic stroke, MI, and cardiovascular death using the IBM SPSS statistical software (IBM Corp, Armonk, NY). The time-to-event variable was calculated from time enrolled in the study to time of endpoint occurrence. For those subjects without the event, the follow-up period was censored at last known follow-up or death. We calculated the HR for last pregnancy at maternal age of ≥40 years using last pregnancy at age <40 years as reference. The potential confounders were identified from univariate analysis of demographic and clinical variables between women with last pregnancy at age of ≥40 years compared with those at age <40 years. We used chi-square and analysis of variance tests for categoric and continuous variable comparisons, respectively. The Cox proportional hazards model were adjusted for differences in age, ethnicity (American Indian/Alaskan native, Asian/Pacific islander, black/African American, Hispanic/Latino, white [not Hispanic Origin], others), systolic blood pressure (<140 or ≥140 mm Hg), cigarette smoking (never, past, <1 pack per day, ≥1 pack per day), congestive heart failure, atrial fibrillation, alcohol use (nondrinker, past drinker, <1 drink per day, ≥1 drink per day), anticoagulant use recorded at baseline or 2 subsequent ascertainments (mentioned earlier), and other variables that were significantly different in univariate analysis. We adjusted for age as a time-dependent covariate in the model. We calculated 95% confidence intervals (CI) using a Taylor series approximation for the standard error of the HR. In an exploratory analysis, we entered oral anticoagulant use (at baseline or at follow-up ascertainment as mentioned earlier) to the model to adjust for any confounding effect on the relationship between last pregnancy at age of ≥40 years and hemorrhagic stroke. In another exploratory analysis, we also performed another Cox proportional hazards analysis to evaluate the effect of last pregnancy at maternal age of ≥40 years on a combined endpoint of cardiovascular death with ischemic stroke, hemorrhagic stroke, or MI.
Results
Of the 93,676 participants in the Women’s Health Initiative observational study, 72,221 reported ≥1 pregnancy; 21,455 women were excluded because they either did not report age at last pregnancy or were never pregnant. A total of 3306 of the 72,221 participants (4.6%) reported last pregnancy at age ≥40 years. The mean age of women at baseline evaluation with last pregnancy at age ≥40 years was greater than those with last pregnancy at <40 years (68.1±7.1 vs 63.3±7.2). The proportion of white women was lower among women with last pregnancy at age ≥40 years. The mean systolic blood pressure (mm Hg±SD) of women at baseline evaluation with last pregnancy at age ≥40 years was greater than those with last pregnancy at <40 years (130.3±19 vs 126.7±17.8). The proportion of women with diabetes mellitus, congestive heart failure, atrial fibrillation, and any alcohol use was higher among those with last pregnancy at age ≥40 years ( Table 1 ). The proportion of women in all categories of cigarette smoking was lower (those who never smoked cigarettes were used as reference) among participants with last pregnancy at age ≥40 years. The proportion of women in all categories of alcohol use was lower (those who never used alcohol were used as reference) among participants with last pregnancy at age ≥40 years, except for past drinkers, which was higher in women with last pregnancy age ≥40 years. The use of oral anticoagulants was higher among women with last pregnancy at age ≥40 years.
| Variable | Last pregnancy at <40 years of age | Last pregnancy at ≥40 years of age | Relative risk (95% confidence interval) |
|---|---|---|---|
| Overall, n | 68,915 | 3306 | |
| Mean age, y a | 63.3±7.2 | 68.1±7.1 | 1.1 (1.1–1.1) |
| Race/ethnicity, n (%) | |||
| American Indian/Alaskan native | 271 (0.4) | 20 (0.6) | 1.5 (0.9–2.4) |
| Asian/Pacific islander | 1,838 (2.7) | 119 (3.6) | 1.4 (1.1–1.6) |
| Black/African American | 4,980 (7.2) | 313 (9.5) | 1.3 (1.2–1.5) |
| Hispanic/Latino | 2,205 (3.2) | 162 (4.9) | 1.6 (1.3–1.8) |
| White (not Hispanic origin) | 58,708 (85.2) | 2638 (79.8) | 0.7 (0.6–0.7) |
| Other | 736 (1.1) | 43 (1.3) | 1.2 (0.9–1.7) |
| Body mass index, kg/m 2 a | 27.3±5.7 | 27.4±5.6 | 1.0 (1.0–1.0) |
| Systolic blood pressure, mm Hg a | 126.7±17.8 | 130.3±19.0 | 1.0 (1.0–1.0) |
| Cardiovascular risk factors, n (%) | |||
| Diabetes mellitus | 3741 (5.4) | 247 (7.5) | 1.4 (1.2–1.6) |
| High cholesterol | 10,020 (14.5) | 483 (14.6) | 1.0 (0.9–1.1) |
| Congestive heart failure | 749 (1.1) | 52 (1.6) | 1.5 (1.1–1.9) |
| Atrial fibrillation | 3,215 (4.7) | 199 (6.0) | 1.3 (1.2–1.5) |
| Anticoagulation use b | 1,490 (2.2) | 108 (3.3) | 1.5 (1.3–1.9) |
| Cigarette smoking status, n (%) | |||
| Never | 34,181 (49.6) | 1853 (56.0) | Reference |
| Past | 29,626 (43.0) | 1220 (36.9) | 0.8 (0.7–0.8) |
| <1 Pack/d | 2,853 (4.1) | 114 (3.4) | 0.7 (0.6–0.9) |
| ≥1 Pack/d | 466 (0.7) | 14 (0.4) | 0.6 (0.3–0.9) |
| Alcohol use, n (%) | |||
| Nondrinker | 7,431 (10.8) | 455 (13.8) | Reference |
| Past drinker | 12,680 (18.4) | 651 (19.7) | 1.1 (1.1–1.3) |
| <1 Drink/d | 39,802 (57.8) | 1828 (55.3) | 0.8 (0.7–0.8) |
| ≥1 Drink/d | 8,632 (12.5) | 347 (10.5) | 0.7 (0.6–0.8) |
a Data are given as mean±standard deviation
b Use of oral anticoagulant was ascertained at baseline and at year 3.
Compared with women with last pregnancy at <40 years, the rate of ischemic stroke (3.8% vs 2.4%), hemorrhagic stroke (1.0% vs 0.5%), and cardiovascular death (3.9% vs 2.3%) was significantly higher among women with last pregnancy at age ≥40 years. There was a trend towards a higher rate of MI among women with last pregnancy at age ≥40 years (3.0% vs 2.5%). In multivariate analysis, women with last pregnancy at age ≥40 years were 50% more likely to experience a hemorrhagic stroke (HR, 1.5; 95% CI, 1.0–2.1) after adjustment for age, race/ethnicity, congestive heart failure, systolic blood pressure, atrial fibrillation, alcohol use, and cigarette smoking ( Table 2 ). Adjustment for anticoagulant use in the multivariate model for hemorrhagic stroke did not have any significant effect on the HR (HR, 1.4; 95% CI, 1.0-2.1). There was no significant difference in the risk of ischemic stroke (HR, 1.0; 95% CI, 0.8–1.2), MI (HR, 0.8; 95% CI, 0.7–1.0), and cardiovascular death (HR, 1.0; 95% CI, 0.8–1.2) among women with last pregnancy at age ≥40 years after adjustment for potential confounders. Compared with women with last pregnancy at <40 years, there was no difference in the rates of combined endpoint of cardiovascular deaths with ischemic stroke, hemorrhagic stroke, or MI among women with last pregnancy at age ≥40 years in age-adjusted or multivariate adjusted analysis.
