Abstract
Background
The joint effects of individual health behaviors on pregnancy health are rarely considered. We examined associations between combinations of first trimester health behaviors with adverse pregnancy outcomes (APOs) and blood pressure (BP) 2–7 years after delivery, and estimated the proportion of associations with later BP mediated by APOs.
Methods
Participants in the nuMoM2b Heart Health Study prospective cohort were included. Physical activity, diet, sleep duration, and smoking were scored using the Life’s Essential 8 framework. APOs included hypertensive disorders of pregnancy, preterm birth, gestational diabetes, small-for-gestational-age birth, or stillbirth. We constructed latent profiles of health behaviors using the Life’s Essential 8 scores. Risk of APOs and differences in continuous BP 2–7 years after delivery based on behavioral profiles were assessed with Poisson and linear regression. Mediation analysis examined the proportion of associations between behavioral profiles and BP mediated by APOs. Where appropriate, analyses were stratified by first trimester obesity status.
Results
Among 8700 nulliparas, four behavioral profiles were identified: Healthiest Behaviors (36%), Healthy Activity/Sleep with Poor Diet/Smoking (21%), Healthy Sleep Only (32%), and Least Healthy Behaviors (11%). Adjusted risk of APOs was 9–18% higher across the less healthy behavioral profiles compared to the Healthiest Behaviors profile. Among women without first trimester obesity, adjusted systolic and diastolic BP were 0.93 and 0.85 mmHg higher, respectively, 2–7 years after delivery in those with less healthy behavioral patterns. Associations were not significant among women with first trimester obesity. APOs did not mediate a significant proportion of the association with later-life BP.
Conclusions
Less healthy behavioral profiles in early pregnancy were associated with APOs and higher BP 2–7 years after delivery, and early pregnancy obesity was an important effect modifier for later-life BP. Associations with later BP were predominately direct, with no mediation through APOs.
Keywords: Life’s essential 8, Adverse pregnancy outcomes, Cardio-obstetrics, Maternal cardiovascular health, nuMoM2b-HHS
Non-standard Abbreviations and Acronyms
- APO
adverse pregnancy outcome
- BMI
body mass index
- CVH
cardiovascular health
- HEI
Healthy Eating Index
- HELLP
hemolysis, elevated liver enzymes, and low platelets
- MVPA
moderate-to-vigorous physical activity
- nuMoM2b-HHS
Nulliparous Pregnancy Outcomes: Monitoring Mothers-to-be Heart Health Study
1. Introduction
Pregnancy is considered a cardiovascular stress test that can provide an early warning for future risk of cardiovascular disease [1]. Adverse pregnancy outcomes (APOs), including hypertensive disorders of pregnancy, preterm birth, or gestational diabetes are associated with a more than two-fold higher risk of developing hypertension within seven years of delivery [[2], [3], [4], [5], [6], [7]], suggesting that shared antecedents may exist. However, accurately identifying those who will develop APOs remains elusive, and prevention strategies are limited.
The American Heart Association’s Life’s Essential 8 framework for cardiovascular health (CVH) [8] includes health factors (blood glucose, lipids, body mass index [BMI], and blood pressure) and health behaviors (physical activity, diet quality, sleep duration, and nonsmoking) that can be assessed in clinical settings and provides a metric to help identify those at higher risk of APOs and later cardiovascular disease [1]. An early-pregnancy atherogenic profile derived from the health factor components has been associated with both APOs and incident hypertension 2–7 years after delivery in nulliparous individuals [9]. Separate modifiable health behavior components also impact pregnancy health [[10], [11], [12], [13], [14], [15], [16]], but importantly, health behaviors do not exist in isolation and adherence to multiple healthy behaviors can have additive beneficial effects [[17], [18], [19], [20]]. Co-occurring profiles of behavioral patterns during pregnancy have yet to be evaluated as precursors of APOs, and it is unknown whether combined health behaviors in early pregnancy are related to long-term cardiovascular outcomes such as hypertension.
Furthermore, there is a growing body of evidence demonstrating that, while APOs are associated both with poor early-pregnancy CVH and later cardiovascular disease, early pregnancy health factors themselves are related to future CVD risk even among those without APOs [5,9,21,22]. While this has been demonstrated for clinical health factors such as BMI, blood pressure, lipids, and glucose, the contributions of health behaviors to these associations are uncertain. The Nulliparous Pregnancy Outcomes: Monitoring Mothers-to-be prospective cohort and its follow-up Heart Health Study (nuMoM2b-HHS) includes early-pregnancy assessment of health behaviors, prospective collection of rigorously defined APO diagnoses, and postpartum follow-up of cardiovascular risk metrics. Therefore, the objectives of this secondary analysis were to: 1) identify profiles of health behaviors in early pregnancy, 2) examine the associations of these profiles with incident APOs and maternal blood pressure several years after delivery, and 3) determine whether APOs mediated associations between behavioral profiles and later maternal blood pressure.
2. Methods
2.1. Data availability and sharing
Due to the sensitive nature of the data collected in this study, requests to access data used in this analysis from qualified researchers trained in human subject confidentiality protocols will be considered by the data coordinating center, and may be submitted at https://numom2b.org.
2.2. Study design and population
Study protocols for the nuMoM2b-HHS prospective cohort have been described previously [23,24]. Briefly, 10,038 nulliparous pregnant individuals were recruited between 2010–2013. Participants were enrolled in the first trimester and followed through delivery at eight clinical sites across the United States. Participants were eligible if they were 6 weeks+ 0 days to 13 weeks+ 6 days pregnant with a singleton gestation and had no prior pregnancy that lasted past 20 weeks. In-person postpartum follow-up occurred 2–7 years following the index pregnancy (n = 4508). Ethical approval was obtained at each clinical site and participants provided written informed consent. Participants included in this analysis delivered at or after 20 weeks+ 0 days of gestation; those included in the APO analysis had complete pregnancy outcomes data, while those included in the analysis of future blood pressure returned for an in-person follow-up assessment and had complete blood pressure data. See Fig. 1 for the study flow diagram.
Fig. 1.
Study flow diagram. APO: adverse pregnancy outcome; nuMoM2b-HHS1: Nulliparous Pregnancy Outcomes: Monitoring Mothers-to-be Heart Health Study.
2.3. Health behaviors
Health behaviors were self-reported at the first trimester assessment. Leisure-time physical activity type, duration, and frequency were assessed using a questionnaire adapted from the Behavior Risk Factor Surveillance System [25,26], converted to metabolic equivalents of the task using the Compendium of Physical Activities, 2nd edition [27], and quantified as minutes per week of moderate-to-vigorous physical activity (MVPA). Smoking history and exposure to secondhand smoke in the home were self-reported by questionnaire. Nightly hours of sleep duration were self-reported separately for weekdays and weekends; the mean was computed to calculate average nightly sleep duration over a 1-week period. Diet quality was assessed using a modified Block Food Frequency Questionnaire [28,29] and converted to a Healthy Eating Index (HEI) score. Physical activity, smoking, and sleep questions referenced the first trimester, while the diet questionnaire referenced the three months immediately preceding conception.
Each behavioral component was assigned a score from 0–100 using the American Heart Association’s Life’s Essential 8 algorithm (Supplemental Table 1). Scores ≥80 indicate “ideal” health, while scores of 50–79 indicate “moderate” health and scores <50 indicate “poor health [30]. Component scores for physical activity ranged from 0 (no weekly MVPA) to 100 (≥150 min/week MVPA) with intermediate scores for values between 0–150 min/week. HEI score percentiles were used to calculate the diet score. For example, 0 points corresponded to HEI <25th percentile, while 100 points corresponded to HEI ≥95th percentile. Sleep scores accounted for the U-shaped association between sleep duration and cardiovascular health, in which both low and high sleep durations are adverse [31]. A sleep score of 100 corresponded to 7–9 h of sleep, with lower scores for fewer or greater hours of nightly sleep. Finally, the smoking score was adapted for pregnancy from the standard scoring algorithm. A score of 0 indicated current smoking during pregnancy, while a score of 100 indicated never smoking tobacco; intermediate scores were assigned for former smokers. In accordance with the standard scoring algorithm, twenty points were subtracted in cases of exposure to secondhand smoke in the home for all except current smokers [8].
2.4. Adverse pregnancy outcomes
Pregnancy outcomes were determined through medical record abstraction. We used a binary composite APO variable, defined as presence of any of the following: hypertensive disorders of pregnancy (>140 mmHg systolic or >90 mmHg diastolic at or after 20 gestational weeks with or without proteinuria, end-organ involvement, and/or HELLP syndrome), gestational diabetes, preterm birth (<37 weeks of gestation at delivery), small-for-gestational-age birth (<5th percentile using Alexander curves), or stillbirth.
2.5. Blood pressure 2–7 years after delivery
Resting blood pressure was measured at the in-person nuMoM2b-HHS1 visit, occurring 2–7 years following the index pregnancy, using a standardized protocol [24]. Following a seated rest, blood pressure was measured three times using an automated device (Omron Healthcare Incorporated, Lake Forest, IL); the last two measurements were averaged and used for analysis. Incident hypertension was defined as new onset of stage 1 hypertension (≥130 mmHg systolic or ≥80 mmHg diastolic) or the use of blood pressure lowering medication at follow-up. Those with measured first trimester stage 1 hypertension were excluded from incident hypertension analyses. Finally, we examined associations with continuous values of systolic and diastolic blood pressure at follow-up.
2.6. Covariates and demographic characteristics
Questionnaires and interviews assessed participant age at study enrollment, income, education, insurance, perceived racial discrimination (Krieger Discrimination Scale [32], categorized as none, moderate, or high), and the social construct of self-reported racial and ethnic identity. To estimate untreated resting blood pressure for those self-reporting anti-hypertensive medication usage at the follow-up visit (n = 117), we added 10/5 mmHg to measured systolic/diastolic blood pressures, respectively [33,34]. Height and weight were measured at the first trimester study visit using standardized protocols and were converted into body mass index (BMI) in kg/m2. Self-reported pre-pregnancy diabetes mellitus was assessed at the first trimester visit by interview. Resting blood pressure was measured at the first trimester study visit using the same protocol described above.
2.7. Statistical analyses
Latent (unobserved) classes of combinations of the four health behaviors were constructed empirically using the “gsem” command in Stata, which can compute the models using continuous values. Models were built using first trimester CVH scores on a scale of 0–100 for each of the four health behaviors (physical activity, diet, sleep duration, and smoking) based on the full nuMoM2b sample (N = 10,038), and the best-fitting model was selected based on values of Akaike/Bayesian Information Criterion, sample size in each class, and model interpretability; posterior probabilities were also assessed.
We assessed the risk of having an APO among behavioral latent profiles using Poisson regression with robust error variance. Crude models were first adjusted for participant characteristics including age, insurance status, income, education, and perceived racial discrimination; a final model also included first trimester blood pressure and BMI, which are health factors that may both mediate and confound the association between health behaviors and our outcomes [35,36]. Informed by evidence that pre-pregnancy obesity is an important predictor of APO risk and later cardiovascular health [21,22], we tested a multiplicative interaction between behavioral profiles and a binary BMI variable (<30 kg/m2 versus ≥ 30 kg/m2) and presented stratified models as appropriate.
Next, we estimated the risk of incident stage 1 hypertension 2–7 years following delivery associated with early pregnancy behavioral profiles using Poisson regression with robust error variance. In addition, we tested the association between behavioral profiles and continuous blood pressure 2–7 years after delivery; separate models were tested for systolic and diastolic blood pressure using linear regression. Individuals with early-pregnancy stage 1 hypertension (n = 598, 13 %) were excluded from the incident hypertension models but were included in analyses of APOs and continuous blood pressure values. We used the same model adjustment and interaction/stratification framework as described in the APO analysis.
Next, we conducted a mediation analysis to estimate the proportion of any significant associations between early-pregnancy health behaviors and blood pressure outcomes 2–7 years after delivery that were mediated by APOs. We used a multiplicative model with an interaction term between behavioral profile and APO status (modeled as a count outcome using Poisson regression) using the mediate command in Stata. Assumptions include the absence of unobserved confounding in the associations between 1) the exposure and outcome, 2) the mediator and the outcome, or 3) the exposure and the mediator; an additional assumption is the absence of confounding of the association between the mediator and the outcome that may be caused by the exposure. Decomposition of the total effects included estimations of how much early pregnancy health behaviors directly affected later blood pressure (natural direct effect) and how much of the effect was explained by APOs (natural indirect effect); we also estimated the proportions of the total effect mediated by APOs. Crude models were adjusted using the framework depicted in Supplemental Figure 1.
Finally, we conducted the following sensitivity analyses: first, we evaluated associations between early-pregnancy health behaviors and the APO and continuous blood pressure outcomes when excluding those with early-pregnancy stage 1 hypertension or diabetes. Second, the clinical cutoff for hypertension in pregnancy is ≥140 mmHg systolic and/or ≥90 mmHg diastolic and the pregnancy measurements were taken prior to the implementation of revised blood pressure categorizations [37], so we examined whether associations with incident hypertension were different using incidence of stage 2 hypertension. All analyses were conducted using Stata version 18 (StataCorp, College Station, TX) and significance was accepted at p < 0.05.
3. Results
In total, 8700 participants (mean ± SD baseline age 26.9 ± 5.6 years, 26.2 ± 6.3 kg/m2 BMI, 40 % underrepresented race or ethnicity) with complete pregnancy outcomes data were included in the APO analysis (Table 1). Model fit criteria for latent classes of early-pregnancy health behaviors found a four-class solution to be the most appropriate (AIC = 343,338.1, BIC = 343,504.0, Entropy = 0.85; Supplemental Table 2). Participants were categorized as engaging in Healthiest Behaviors (n = 3158, 36.3 %), Healthy Activity/Sleep with Poor Diet/Smoking (n = 1807, 20.8 %), Healthy Sleep Only (n = 2804, 32.2 %), or Least Healthy Behaviors (n = 931, 10.7 %) (Fig. 2). Participants in the Healthiest Behaviors profile were more likely to be older, have a lower BMI, have higher education and income, self-identify as non-Hispanic White, have different systolic and diastolic blood pressure in the first trimester of pregnancy, and have a lower prevalence of stage 1 hypertension at baseline compared to those in the other behavioral profiles.
Table 1.
Participant characteristics by health behavior latent profiles.
| Healthiest Behaviors (n = 3158; 36 %) | Healthy Activity/Sleep with Poor Diet/Smoking (n = 1807; 21 %) | Healthy Sleep Only (n = 2804; 32 %) | Least Healthy Behaviors (n = 931; 11 %) | P | |
|---|---|---|---|---|---|
| Maternal age (years) | 29.0 ± 4.9 | 26.2 ± 5.7 | 25.8 ± 5.6 | 24.4 ± 5.5 | <0.001 |
| Body mass index (kg/m2) | 25.1 ± 5.1 | 27.1 ± 6.9 | 26.8 ± 6.5 | 27.1 ± 7.3 | <0.001 |
| Body mass index ≥ 30kg/m2 | 465 (15 %) | 470 (26 %) | 687 (25 %) | 236 (25 %) | <0.001 |
| Maternal Race | <0.001 | ||||
| American Indian | 2 (<1 %) | 2 (<1 %) | 0 (0 %) | 3 (<1 %) | |
| Asian | 167 (5 %) | 51 (3 %) | 107 (4 %) | 28 (3 %) | |
| Black non-Hispanic | 183 (6 %) | 298 (16 %) | 481 (17 %) | 203 (22 %) | |
| Hispanic | 338 (11 %) | 275 (15 %) | 659 (24 %) | 226 (24 %) | |
| Multiracial | 103 (3 %) | 87 (5 %) | 111 (4 %) | 48 (5 %) | |
| Native Hawaiian | 9 (<1 %) | 11 (<1 %) | 8 (<1 %) | 5 (<1 %) | |
| White non-Hispanic | 2334 (74 %) | 1074 (59 %) | 1422 (51 %) | 415 (45 %) | |
| Not available | 22 (<1 %) | 9 (<1 %) | 16 (<1 %) | 3 (<1 %) | |
| Educational Attainment | <0.001 | ||||
| Less than high school | 58 (2 %) | 186 (10 %) | 314 (11 %) | 142 (15 %) | |
| Completed high school or GED | 98 (3 %) | 254 (14 %) | 455 (16 %) | 209 (22 %) | |
| Some college | 362 (11 %) | 429 (24 %) | 670 (24 %) | 243 (26 %) | |
| Associate or technical degree | 261 (8 %) | 198 (11 %) | 324 (12 %) | 99 (11 %) | |
| Completed college | 1156 (37 %) | 448 (25 %) | 641 (23 %) | 146 (16 %) | |
| Degree work beyond college | 1219 (39 %) | 292 (16 %) | 394 (14 %) | 92 (10 %) | |
| Percentage of Federal Poverty Level | <0.001 | ||||
| <100 % | 177 (6 %) | 284 (20 %) | 452 (22 %) | 194 (31 %) | |
| 100–200 % | 263 (9 %) | 250 (18 %) | 387 (18 %) | 115 (18 %) | |
| >200 % | 2459 (85 %) | 887 (62 %) | 1253 (60 %) | 318 (51 %) | |
| Insurance Payor | <0.001 | ||||
| Commercial | 2000 (64 %) | 892 (50 %) | 1371 (49 %) | 334 (36 %) | |
| Medicare/Medicaid | 296 (9 %) | 583 (32 %) | 990 (36 %) | 448 (48 %) | |
| Military/Other | 39 (1 %) | 44 (2 %) | 42 (2 %) | 24 (3 %) | |
| Two or more sources | 730 (23 %) | 219 (12 %) | 296 (11 %) | 91 (10 %) | |
| None (personal income) | 80 (3 %) | 56 (3 %) | 75 (3 %) | 27 (3 %) | |
| 1st Trimester Systolic BP (mmHg) | 108.6 ± 10.4 | 110.0 ± 11.0 | 109.2 ± 11.4 | 108.8 ± 11.5 | <0.001 |
| 1st Trimester Diastolic BP (mmHg) | 67.0 ± 8.1 | 67.5 ± 8.5 | 67.0 ± 8.5 | 66.4 ± 8.7 | 0.014 |
|
1st Trimester Stage 1 Hypertension |
332 (10.7 %) | 257 (14.5 %) | 369 (13.5 %) | 122 (13.4 %) | <0.001 |
|
1st Trimester Stage 2 Hypertension |
35 (1.1 %) | 36 (2.0 %) | 47 (1.7 %) | 18 (2.0 %) | 0.058 |
| Composite APO | 882 (28 %) | 635 (35 %) | 948 (34 %) | 349 (37 %) | <0.001 |
| Attended follow-up visit | 1659 (52 %) | 886 (49 %) | 1486 (53 %) | 466 (50 %) | 0.053 |
| Incident Stage 1 HTN at follow-up | 247 (17 %) | 140 (19 %) | 234 (19 %) | 67 (18 %) | 0.680 |
| Incident Stage 2 HTN at follow-up | 67 (4 %) | 41 (5 %) | 79 (6 %) | 21 (5 %) | 0.339 |
| Systolic BP at follow-up (mmHg) | 110.7 ± 11.0 | 111.9 ± 11.0 | 112.1 ± 11.3 | 111.8 ± 11.2 | 0.003 |
| Diastolic BP at follow-up (mmHg) | 71.3 ± 10.1 | 72.5 ± 9.5 | 72.8 ± 10.0 | 72.7 ± 10.1 | <0.001 |
Reported as mean ± standard deviation or frequency and percentage, as appropriate. Differences among behavioral latent classes tested using one-way analysis of variance for continuous variables or Chi-squared tests for categorical variables. Incident HTN at follow-up calculated from those participants without baseline HTN. APO: adverse pregnancy outcomes; BP: blood pressure; GED: general education diploma; HTN: hypertension.
Fig. 2.
Health behavior cardiovascular health component scores by behavioral profiles. Mean (SD) cardiovascular health (CVH) scores for each health behavior component within each behavioral profile. Higher score indicates more ideal CVH; scores <50 are “poor” CVH; 50–79 are “moderate” CVH, and scores ≥80 are “high” CVH as indicated by markers on the y-axis.
Early pregnancy behavioral profiles were significantly associated with risk of APOs (Table 2), with those in less healthy profiles having a 21–34 % higher unadjusted relative risk compared to those in the Healthiest Behaviors profile [overall Chi-squared (3) = 33.28, p < 0.001]. These significant associations remained following adjustment for sociodemographic characteristics, ranging from a 12–23 % higher risk. These associations persisted in the Healthy Activity/Sleep with Poor Diet/Smoking (15 % higher, 95 % confidence interval 2–30 %) and Least Healthy Behaviors (18 % higher, 95 % confidence interval 1–38 %) profiles upon further adjustment for early pregnancy blood pressure and BMI; no significant effect modification was found for first trimester obesity status (p for interaction = 0.580). A full summary of individual APO diagnoses by behavioral profiles is shown in Supplemental Table 3.
Table 2.
Incident rate ratios (IRR) of adverse pregnancy outcomes by behavioral profiles.
| Group | Model 1 IRR (95 % CI) | P | Model 2 IRR (95 % CI) | P | Model 3 IRR (95 % CI) | P |
|---|---|---|---|---|---|---|
| Healthiest Behaviors | 1.00 (ref) | - | 1.00 (ref) | - | 1.00 (ref) | - |
| Healthy Activity/Sleep with Poor Diet/Smoking | 1.26 (1.16, 1.37) | <0.001 | 1.20 (1.09, 1.32) | <0.001 | 1.15 (1.04, 1.27) | 0.004 |
| Healthy Sleep Only | 1.21 (1.12, 1.31) | <0.001 | 1.12 (1.02, 1.22) | 0.016 | 1.09 (0.99, 1.19) | 0.068 |
| Least Healthy Behaviors | 1.34 (1.21, 1.48) | <0.001 | 1.23 (1.09, 1.39) | 0.001 | 1.18 (1.04, 1.34) | 0.010 |
Model 1: unadjusted; Model 2: adjusted for maternal age, education, insurance, income, and perceived racial discrimination; Model 3: additionally adjusted for first trimester blood pressure and body mass index. Estimates modeled with Poisson regression with robust error variance. Bold font indicates significant difference (p < 0.05).
Of the original cohort, 4497 returned for post-delivery follow-up at 3.2 ± 0.9 years (median [IQR] 3.1 [2.5–3.7] years, range 2–7 years) following the index pregnancy and had complete blood pressure data. Given the substantial overlap in APO risk among the three less favorable behavioral profiles, we dichotomized the groups as Healthiest Behaviors and Less Healthy Behaviors (including the Healthy Activity/Sleep with Poor Diet/Smoking, Healthy Sleep Only, and Least Healthy Behaviors profiles) for analyses relating to maternal blood pressure 2–7 years following delivery. Incident stage 1 hypertension was found in 18.1 % of the sample who participated in the follow-up visit and did not have early-pregnancy stage 1 hypertension (688 of 3793 participants), with no significant differences in incidence between early-pregnancy behavioral profiles (Healthiest Behaviors profile: 247 of 1429, 17.3 %; other three Less Healthy Behaviors profiles combined: 441 of 1923, 18.7 %, IRR = 1.08 [95 % confidence interval 0.94 – 1.24, p = 0.290], Supplemental Table 4). We found no effect modification by first trimester obesity status (p for interaction = 0.295)
A significant interaction between behavioral profiles and first trimester obesity status was found in analyses of continuous systolic (p for interaction = 0.019) and diastolic (p for interaction = 0.017) blood pressures. Among women without first trimester obesity, crude linear regression results for continuous blood pressure found differences that were retained after adjustment for visit date, sociodemographic factors, and first trimester blood pressure (Table 3), wherein participants in the combined less favorable behavior profiles had 0.93 mmHg higher systolic blood pressure (p = 0.032) and 0.85 mmHg higher diastolic blood pressure (p = 0.025) compared to the Healthiest Behaviors profile. Among women with first trimester obesity, behavioral profiles were not associated with continuous blood pressure 2–7 years following delivery (Table 3).
Table 3.
Association of continuous blood pressure 2–7 years after delivery by behavioral profiles, stratified by first trimester obesity status.
| Group | Model 1 | P | Model 2 | P | Model 3 | P |
|---|---|---|---|---|---|---|
| Women without first trimester obesity | ||||||
| Systolic Blood Pressure, mmHg [ß coefficient (95% CI)] | ||||||
| Healthiest Behaviors | Reference | - | Reference | - | Reference | - |
| Less Healthy Behaviors | 1.40 (0.65, 2.14) | <0.001 | 1.05 (0.18, 1.92) | 0.018 | 0.93 (0.08, 1.77) | 0.032 |
| Diastolic Blood Pressure, mmHg [ß coefficient (95% CI)] | ||||||
| Healthiest Behaviors | Reference | - | Reference | - | Reference | - |
| Less Healthy Behaviors | 1.10 (0.44, 1.75) | 0.001 | 0.92 (0.15, 1.68) | 0.019 | 0.85 (0.11, 1.60) | 0.025 |
| Women with first trimester obesity | ||||||
| Systolic Blood Pressure, mmHg [ß coefficient (95% CI)] | ||||||
| Healthiest Behaviors | Reference | - | Reference | - | Reference | - |
| Less Healthy Behaviors | −0.33 (−2.20, 1.53) | 0.725 | −0.72 (−2.68, 1.24) | 0.469 | −0.80 (−2.69, 1.09) | 0.406 |
| Diastolic Blood Pressure, mmHg [ß coefficient (95% CI)] | ||||||
| Healthiest Behaviors | Reference | - | Reference | - | Reference | - |
| Less Healthy Behaviors | −0.13 (−1.63, 1.37) | 0.867 | −0.42 (−2.08, 1.24) | 0.619 | −0.63 (−2.26, 1.00) | 0.447 |
Model 1: unadjusted; Model 2: adjusted for time elapsed between visits, maternal age, education, insurance, income, and perceived racial discrimination; Model 3: additionally adjusted for first trimester blood pressure. Estimates modeled with linear regression. Bold font indicates significant difference (p < 0.05).
Given that a significant association between early-pregnancy behavioral profiles and long-term blood pressure was only found among women without first trimester obesity, we only examined mediation by APOs in this subgroup. We found significant total and natural direct effects in crude models for both systolic and diastolic blood pressure (Table 4) which were retained upon adjustment for visit date, sociodemographic factors, and first trimester blood pressure. Natural indirect effects and proportions mediated were not significant across models.
Table 4.
Effect decomposition of the association between behavioral profiles and future blood pressure mediated by adverse pregnancy outcomes among women without first trimester obesity.
| Effect Decomposition |
Proportion mediated, P | |||
|---|---|---|---|---|
| Natural Indirect Effect, P | Natural Direct Effect, P | Total Effect, P | ||
| Systolic Blood Pressure (mmHg) | ||||
| Model 1 | 0.10, p = 0.082 | 1.54, p<0.001 | 1.64, p<0.001 | 6 %, p = 0.090 |
| Model 2 | 0.06, p = 0.303 | 1.38, p=0.001 | 1.44, p=0.001 | 4 %, p = 0.308 |
| Model 3 | 0.05, p = 0.344 | 1.28, p=0.002 | 1.34, p=0.001 | 4 %, p = 0.351 |
| Diastolic Blood Pressure (mmHg) | ||||
| Model 1 | 0.08, p = 0.086 | 1.23, p<0.001 | 1.35, p<0.001 | 6 %, p = 0.098 |
| Model 2 | 0.05, p = 0.305 | 1.25, p=0.001 | 1.30, p=0.001 | 4 %, p = 0.314 |
| Model 3 | 0.04, p = 0.346 | 1.20, p=0.001 | 1.24, p=0.001 | 3 %, p = 0.356 |
Model 1: unadjusted; Model 2: adjusted for maternal age, education, insurance, income, and perceived racial discrimination; Model 3: additionally adjusted for first trimester blood pressure. Bold font indicates significant difference (p < 0.05).
Sensitivity analyses excluding participants with pre-pregnancy stage 1 hypertension and/or diabetes from the models estimating APO risk and associations with continuous blood pressure 2–7 years after delivery did not change the direction of our results, though some adjusted effect estimates lost precision (Supplemental Tables 5 and 6). Similarly, rates of incident stage 2 hypertension were not different between those with healthy versus less healthy behaviors (Supplemental Table 4).
4. Discussion
In this diverse prospective cohort of individuals enrolled during a first pregnancy we found four distinct first-trimester health behavior profiles that were associated with risk of subsequent APOs. Those with less healthy behavioral profiles were 9–18 % more likely to experience an APO than those with the healthiest behavioral profile. In women without obesity in the first trimester, less healthy early-pregnancy behavioral profiles were associated with higher blood pressure 2–7 years following delivery; this association was not seen in those with obesity in the first trimester. Mediation analysis suggested that these effects were not mediated through APOs. Our findings indicate that healthy behavior combinations in early pregnancy exhibit statistically significant associations with APOs and that behaviors in early pregnancy may be important for long-term blood pressure whether or not an APO occurs, but early pregnancy obesity is an important effect modifier for associations with long-term cardiovascular risk (Central Figure).
By utilizing the Life’s Essential 8 framework, we were able to construct latent profiles of healthy behaviors (physical activity, diet quality, sleep duration, and nonsmoking) in early pregnancy. This allowed us to leverage naturally occurring covariance in healthy behaviors during pregnancy, rather than forcing disaggregation. We found that lower adherence to healthy behaviors was associated with greater risk of APOs. This was particularly evident in the Least Healthy Behaviors profile (comprised of individuals with “poor” adherence to physical activity, diet quality, and sleep duration behaviors and “moderate” adherence to nonsmoking behavior), in which the adjusted risk ratio was highest. These results suggest that healthy behaviors in early pregnancy may be additive, and that “poor” adherence to at least two healthy behaviors in early pregnancy is related to elevated risk of APOs. Our findings are in contrast to those from a small cohort (n = 202) [15] in which individual health behaviors, but not a composite score, were associated with features of placental health. This discrepancy may be attributable to cohort differences, distinct scoring algorithms, or study outcomes.
In addition, associations between behavioral profiles and APOs were independent of first trimester blood pressure and BMI. These findings suggest that engaging in healthy behaviors is an important contributor to pregnancy outcomes for all pregnant individuals, regardless of these clinical health factors in early pregnancy. Our results support existing evidence that healthy behaviors are associated with more favorable pregnancy outcomes [[10], [11], [12], [13], [14], [15], [16],38], and add important knowledge about the combined effects of multiple healthy behaviors. Furthermore, these results emphasize the importance of healthy behaviors even for those with favorable early-pregnancy blood pressure and BMI. In an analysis applying a hypothetical intervention to improve first trimester CVH in the nuMoM2b cohort, 12–40 % of APOs could be prevented [39]; our analysis suggests that healthy behavioral patterns may play a key role in these outcomes.
It is also important to note sociodemographic differences among participants in each behavioral profile in our study. Those in profiles with the lowest adherence to healthy behaviors and the highest risk of APOs were likely impacted by structural factors that may have influenced the ability to adhere to and the interplay between the healthy behaviors (e.g., neighborhood factors that may influence access to both physical activity and healthy diet options). While we adjusted for sociodemographic factors, future research should investigate the drivers of these behavioral profiles and identify areas for intervention, such as enhancement of the built environment.
We found no significant association between behavioral profiles and incident hypertension 2–7 years after delivery. However, when examining associations among behavioral profiles and continuous blood pressure over this time, we found a significant interaction between behavioral profiles and first trimester obesity status. Specifically, among women without early pregnancy obesity, those with less healthy behavioral profiles had 0.93 mmHg higher systolic blood pressure and 0.85 mmHg higher diastolic blood pressure 2–7 years following delivery than those in the Healthiest Behaviors profile; there was no association between behavioral profiles and later-life blood pressure among women with obesity in early pregnancy. These findings indicate that BMI in early pregnancy is an important effect modifier of the relationship between behavioral profiles and later cardiovascular risk. For women without obesity in early pregnancy, engaging in healthier behavioral patterns may be an important strategy to impact blood pressure in the years following delivery. While 0.85–0.93 mmHg higher blood pressure over a mean follow-up of three years is not a clinically large difference, small increases in blood pressure in the reproductive years can compound cardiovascular risk over time [40,41]. Future research should evaluate these associations over a longer follow-up period.
In the setting of early pregnancy obesity, we did not note associations between behavioral profiles and continuous blood pressure 2–7 years after delivery. Our findings support prior evidence indicating the importance of maximizing CVH prior to pregnancy [1,22]. Notably, significantly fewer women in our cohort with obesity were in the Healthiest Behaviors profile (15 % versus 25 % in the less healthy profiles). It is possible that BMI was in the causal pathway between early pregnancy health behaviors and blood pressure 2–7 years after delivery (i.e., those with healthier behaviors had more favorable early-pregnancy BMI due to the engagement in those healthy behaviors) [35,36]. Future research should evaluate thresholds of healthy behaviors in pregnant women with obesity and examine whether health behavior change interventions during pregnancy effectively impact blood pressure years after delivery [42].
Our mediation analysis suggests that health behaviors may play a direct role in longer-term continuous blood pressure levels in the years following delivery for women without obesity in early pregnancy, with no evidence of mediation through APOs. This is in contrast to two prior studies in the nuMoM2b-HHS cohort. In the first, the association of an early-pregnancy atherogenic profile of poor lipids, C-reactive protein, and insulin with postpartum incident hypertension was partially mediated by APOs [9]. In the second, hypertensive disorders of pregnancy mediated a small but statistically significant proportion of the association between early pregnancy obesity and incident hypertension [21]. Furthermore, hypertensive disorders of pregnancy mediated 6–12 % of the association between overweight or obesity and blood pressure 10–14 years after delivery in the Hyperglycemia and Adverse Pregnancy Outcomes Follow-Up Study cohort [22]. Importantly, our study evaluated health behavior components of CVH, while these prior studies examined health factor components that are, themselves, likely influenced by health behaviors. Taken together, the results of our study and this prior work emphasize the importance of maximizing pre-pregnancy CVH to lower the risk of APOs and subsequent cardiovascular risk, whether or not an APO occurs. Future interventional research is needed to determine whether improving engagement in healthy behaviors in the prenatal period is an effective means to improve long-term maternal health.
Pregnancy is a time of near-universal healthcare access, and motivation may be high to engage in behavior change [43,44]. Our results suggest that modifiable health behaviors group together in distinct ways, and that risks of multiple unhealthy behaviors compound one another and may be related to health beyond pregnancy. Notably, we did not see large differences in risk between our three less healthy behavioral profile groups, but each group with two or more behaviors classified as “poor” were at higher risk. This suggests that behavior change interventions may employ strategies to improve adherence to whichever health behavior is the most easily modifiable for the individual rather than targeting the adoption of a particular behavioral profile. Early pregnancy may be an ideal time to assess behavioral profiles and engage individuals in behavior change.
4.1. Strengths, limitations, and future directions
This study has several strengths. We utilized a longitudinal approach to the reproductive years by examining early-pregnancy behavioral profile exposures and followed participants through pregnancy and several postpartum years. Our large, diverse prospective cohort of people experiencing a first birth captured well-documented APOs. In addition, we were able to identify four latent health behavior profiles, demonstrating for the first time how different behavioral combinations in early pregnancy are related to perinatal and long-term maternal health. Those who returned for a postpartum visit 2–7 years after delivery were demographically similar to the full pregnancy cohort [2], suggesting these results may be generalizable to our sample at large.
This study was limited by the relatively short follow-up of 2–7 years after delivery, which may have been too short to capture meaningful differences in blood pressure among behavioral profiles (though prior analyses have found a higher risk of incident hypertension among those in the cohort who had APOs [2]). This cohort continues to be followed, and future analyses may find more clinically relevant associations of behavioral patterns with longer follow-up time. Another limitation of this study included the self-report of health behaviors. However, given that these behaviors were reported prior to the outcomes occurring, recall bias or other measurement error would likely be random. In addition, the diet questionnaire referenced the 3 months prior to conception, which may not have accurately captured first trimester dietary patterns. Despite these limitations, our strong posterior probabilities of latent class membership suggest that the behavioral profiles are distinct and any bias is likely nondifferential. Furthermore, our activity questionnaire was limited to leisure-time physical activities. Future investigations should examine whether behavioral patterns are different among those with differing levels of occupational activity. Our stratified analyses relied on a BMI cutoff of 30 kg/m2, which may not have captured risk at lower BMI levels among Asian populations; this limitation could bias our findings towards the null. Finally, health behaviors may have shifted between delivery and the postpartum follow-up assessment; future research should examine whether behavioral profiles are consistent after childbirth and whether more recent behavioral profiles explain additional variability in cardiovascular risk measures.
5. Conclusions
Our results suggest that co-occurring combinations of modifiable health behaviors are important and are related to pregnancy outcomes, but that early pregnancy obesity status is an important effect modifier of the association with later blood pressure in that less healthy behavioral patterns were only associated with higher blood pressure among women without early pregnancy obesity. These findings add to the growing body of literature indicating that health in early pregnancy is a strong contributor to both pregnancy health and CVH following pregnancy. Among women without first trimester obesity, health behaviors also had direct associations with blood pressure in the years following delivery and thus are important for those with and without pregnancy complications, though the absolute difference in blood pressure was small in this cohort of young women. Improving health behaviors in the prenatal context shows promise for reducing maternal risk during and possibly beyond pregnancy.
Funding sources
Grant funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD): U10 HD063036; U10 HD063072; U10 HD063047; U10 HD063037; U10 HD063041; U10 HD063020; U10 HD063046; U10 HD063048; and U10 HD063053. In addition, support was provided by Clinical and Translational Science Institutes: UL1TR001108 and UL1TR000153. Cooperative agreement funding from the National Heart, Lung, and Blood Institute and the Eunice Kennedy Shriver National Institute of Child Health and Human Development: U10‐ HL119991; U10‐HL119989; U10‐HL120034; U10‐HL119990; U10‐HL120006; U10‐HL119992; U10‐HL120019; U10‐ HL119993; U10‐HL120018, and U01HL145358; and the National Center for Advancing Translational Sciences through UL‐1‐ TR000124, UL‐1‐TR000153, UL‐1‐TR000439, and UL‐1‐ TR001108; and the Barbra Streisand Women’s Cardiovascular Research and Education Program, and the Erika J. Glazer Women’s Heart Research Initiative, Cedars-Sinai Medical Center, Los Angeles. The Sedentary Behavior and Cardiovascular Health in Young Women ancillary study is supported by NHLBI R01HL158652. ACK is supported under NHLBI T32HL083825.
Disclosures
The authors have no financial disclosures relevant to this work.
Central figure. Profiles of combined health behaviors in the first trimester of pregnancy are associated with APOs and higher systolic and diastolic blood pressure 2–7 years following delivery in a sample of participants experiencing a first birth. Early pregnancy obesity is an effect modifier for the association between behavioral profiles and later-life blood pressure, and this association was not mediated by APOs. APOs: adverse pregnancy outcomes; DBP: diastolic blood pressure; SBP: systolic blood pressure.
CRediT authorship contribution statement
Andrea C Kozai: Writing – original draft, Visualization, Methodology, Funding acquisition, Formal analysis, Data curation. Bethany Barone Gibbs: Writing – review & editing, Methodology, Funding acquisition, Conceptualization. Lisa D Levine: Writing – review & editing, Funding acquisition, Conceptualization. Abbi D Lane: Writing – review & editing, Conceptualization. Mitali Ray: Writing – review & editing. William Grobman: Writing – review & editing, Conceptualization. Philip Greenland: Writing – review & editing, Funding acquisition. Matthew K Hoffman: Writing – review & editing, Funding acquisition. C Noel Bairey Merz: Writing – review & editing, Funding acquisition. Phyllis C Zee: Writing – review & editing. Lisa Mims: Writing – review & editing. Natalie A Cameron: Writing – review & editing. George Saade: Writing – review & editing, Funding acquisition. Robert M Silver: Writing – review & editing, Funding acquisition. Natalie A Bello: Writing – review & editing, Methodology. Eliza C Miller: Writing – review & editing. Uma M Reddy: Writing – review & editing, Funding acquisition. Judith H Chung: Writing – review & editing, Funding acquisition. Kara M Whitaker: Writing – review & editing. Janet M Catov: Writing – review & editing, Supervision, Funding acquisition.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
Grant funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD): U10 HD063036; U10 HD063072; U10 HD063047; U10 HD063037; U10 HD063041; U10 HD063020; U10 HD063046; U10 HD063048; and U10 HD063053. In addition, support was provided by Clinical and Translational Science Institutes: UL1TR001108 and UL1TR000153. Cooperative agreement funding from the National Heart, Lung, and Blood Institute and the Eunice Kennedy Shriver National Institute of Child Health and Human Development: U10‐ HL119991; U10‐HL119989; U10‐HL120034; U10‐HL119990; U10‐HL120006; U10‐HL119992; U10‐HL120019; U10‐ HL119993; U10‐HL120018, and U01HL145358; and the National Center for Advancing Translational Sciences through UL‐1‐ TR000124, UL‐1‐TR000153, UL‐1‐TR000439, and UL‐1‐ TR001108; and the Barbra Streisand Women’s Cardiovascular Research and Education Program, and the Erika J. Glazer Women’s Heart Research Initiative, Cedars-Sinai Medical Center, Los Angeles. The Sedentary Behavior and Cardiovascular Health in Young Women ancillary study is supported by NHLBI R01HL158652. ACK is supported under NHLBI T32HL083825.
Footnotes
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ajpc.2026.101489.
Appendix. Supplementary materials
References
- 1.Khan S.S., Cameron N.A., Lindley K.J. Pregnancy as an early cardiovascular moment: peripartum cardiovascular health. Circ Res. 2023;132(12):1584–1606. doi: 10.1161/CIRCRESAHA.123.322001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Haas D.M., Parker C.B., Marsh D.J., et al. Association of Adverse Pregnancy Outcomes With Hypertension 2 to 7 Years Postpartum. J Am Heart Assoc. 2019;8(19) doi: 10.1161/JAHA.119.013092. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sukmanee J., Liabsuetrakul T. Risk of future cardiovascular diseases in different years postpartum after hypertensive disorders of pregnancy: a systematic review and meta-analysis. Medicine. 2022;101(30) doi: 10.1097/MD.0000000000029646. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tassi A., Sala A., Mazzera I., Restaino S., Vizzielli G., Driul L. Long-term outcomes of patients with preeclampsia, a review of the literature. Hypertens Pregn. 2023;42(1) doi: 10.1080/10641955.2023.2217448. [DOI] [PubMed] [Google Scholar]
- 5.Fraser A., Catov J.M. Placental syndromes and long-term risk of hypertension. J Hum Hypertens. 2023;37:671–674. doi: 10.1038/s41371-023-00802-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Giorgione V., Jansen G., Kitt J., Ghossein-Doha C., Leeson P., Thilaganathan B. Peripartum and Long-Term Maternal Cardiovascular Health After Preeclampsia. Hypertension. 2023;80(2):231–241. doi: 10.1161/HYPERTENSIONAHA.122.18730. [DOI] [PubMed] [Google Scholar]
- 7.Boucheron P., Lailler G., Moutengou E., et al. Hypertensive disorders of pregnancy and onset of chronic hypertension in France: the nationwide CONCEPTION study. Eur Heart J. 2022;43(35):3352–3361. doi: 10.1093/eurheartj/ehab686. [DOI] [PubMed] [Google Scholar]
- 8.Lloyd-Jones D.M., Allen N.B., Anderson C.A.M., et al. Life’s Essential 8: updating and Enhancing the American Heart Association’s Construct of Cardiovascular Health: a Presidential Advisory From the American Heart Association. Circulation. 2022;146(5) doi: 10.1161/cir.0000000000001078. 101161CIR0000000000001078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Catov J.M., McNeil R.B., Marsh D.J., 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) doi: 10.1161/jaha.120.017216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Davenport M.H., Ruchat S.M., Poitras V.J., et al. Prenatal exercise for the prevention of gestational diabetes mellitus and hypertensive disorders of pregnancy: a systematic review and meta-analysis. Br J Sports Med. 2018;52(21):1367–1375. doi: 10.1136/bjsports-2018-099355. [DOI] [PubMed] [Google Scholar]
- 11.Bernardo D., Carvalho C., Mota J., Ferreira M., Santos P., Karger S. The Influence of Pregestational Body Mass Index and Physical Activity Patterns on Maternal, Delivery, and Newborn Outcomes in a Sample of Portuguese Pregnant Women: a Retrospective Cohort Study. Port J Public Health. 2023;41(2):140–150. doi: 10.1159/000531587. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Miller C., Boushey C., Benny P., et al. Diet quality predicts hypertensive disorders of pregnancy in Asian and Pacific Islander Cohort. Nutr Health. June 27, 2022 doi: 10.1177/02601060221109668. Published online. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Makarem N., Chau K., Miller E.C., et al. Association of a Mediterranean Diet Pattern With Adverse Pregnancy Outcomes Among US Women. JAMA Netw Open. 2022;5(12) doi: 10.1001/jamanetworkopen.2022.48165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Lu Q., Zhang X., Wang Y., et al. Sleep disturbances during pregnancy and adverse maternal and fetal outcomes: a systematic review and meta-analysis. Sleep Med Rev. 2021;58 doi: 10.1016/j.smrv.2021.101436. [DOI] [PubMed] [Google Scholar]
- 15.Teo S.M., Murrin C.M., Mehegan J., et al. Associations between the maternal healthy lifestyle score and its individual components during early pregnancy with placental outcomes. Placenta. 2023;139:75–84. doi: 10.1016/j.placenta.2023.06.001. [DOI] [PubMed] [Google Scholar]
- 16.Wang J., Yang W., Xiao W., Cao S. The association between smoking during pregnancy and hypertensive disorders of pregnancy: a systematic review and meta-analysis. Int J Gynecol Obstet. 2022;157(1):31–41. doi: 10.1002/ijgo.13709. [DOI] [PubMed] [Google Scholar]
- 17.Reeves M.J., Rafferty A.P. Healthy Lifestyle Characteristics Among Adults in the United States, 2000. Arch Intern Med. 2005;165(8):854–857. doi: 10.1001/archinte.165.8.854. [DOI] [PubMed] [Google Scholar]
- 18.Loef M., Walach H. The combined effects of healthy lifestyle behaviors on all cause mortality: a systematic review and meta-analysis. Prev Med. 2012;55(3):163–170. doi: 10.1016/j.ypmed.2012.06.017. [DOI] [PubMed] [Google Scholar]
- 19.Loprinzi P.D., Fitzgerald E.M., Woekel E., Cardinal B.J. Association of physical activity and sedentary behavior with biological markers among U.S. pregnant women. J Womens Health. 2013;22(11):953–958. doi: 10.1089/jwh.2013.4394. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.McLean M.K., Kozai A.C., Lane A.D. Activity behaviors: influencers of cardiometabolic risk and adverse pregnancy outcomes. Exerc Sport Sci Rev. 2025;53(4):178. doi: 10.1249/JES.0000000000000369. [DOI] [PubMed] [Google Scholar]
- 21.Khan S.S., Petito L.C., Huang X., et al. Body mass index, adverse pregnancy outcomes, and cardiovascular disease risk. Circ Res. 2023;133(9):725–735. doi: 10.1161/CIRCRESAHA.123.322762. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Borrowman J.D., Huang X., Petito L.C., et al. Prepregnancy Adiposity, Adverse Pregnancy Outcomes, and Cardiovascular Disease Risk in Midlife. JACC. 2025;85(15):1536–1546. doi: 10.1016/j.jacc.2025.02.033. [DOI] [PubMed] [Google Scholar]
- 23.Haas D.M., Parker C.B., Wing D.A., et al. A description of the methods of the Nulliparous Pregnancy Outcomes Study: monitoring mothers-to-be (nuMoM2b) Am J Obstet Gynecol. 2015;212(4):539.e1–539.e24. doi: 10.1016/j.ajog.2015.01.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Haas D.M., Ehrenthal D.B., Koch M.A., et al. Pregnancy as a Window to Future Cardiovascular Health: design and Implementation of the nuMoM2b Heart Health Study. Am J Epidemiol. 2016;183(6):519–530. doi: 10.1093/aje/kwv309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Centers for Disease Control and Prevention. A data users guide to the BRFSS physical activity questions. https://www.cdc.gov/brfss/data_documentation/pdf/pa-rotatingcore_brfssguide_508comp_07252013final.pdf.
- 26.Shea S., Stein A.D., Lantigua R., Basch C.E. Reliability of the behavioral risk factor survey in a triethnic population. Am J Epidemiol. 1991;133(5):489–500. doi: 10.1093/oxfordjournals.aje.a115916. [DOI] [PubMed] [Google Scholar]
- 27.Shephard R.J. 2011 Compendium of Physical Activities: a Second Update of Codes and MET Values. Yearb Sports Med. 2012;2012:126–127. doi: 10.1016/j.yspm.2011.08.057. [DOI] [PubMed] [Google Scholar]
- 28.Block G., Coyle L.M., Hartman A.M., Scoppa S.M. Revision of Dietary Analysis Software for the Health Habits and History Questionnaire. Am J Epidemiol. 1994;139(12):1190–1196. doi: 10.1093/oxfordjournals.aje.a116965. [DOI] [PubMed] [Google Scholar]
- 29.Parrott M.S., Bodnar L.M., Simhan H.N., Harger G., Markovic N., Roberts J.M. Maternal cereal consumption and adequacy of micronutrient intake in the periconceptional period. Public Health Nutr. 2009;12(8):1276–1283. doi: 10.1017/S1368980008003881. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.U.S. Department of Health and Human Services. Physical Activity Guidelines for Americans. 2nd Edition.
- 31.Yin J., Jin X., Shan Z., et al. Relationship of Sleep Duration With All-Cause Mortality and Cardiovascular Events: a Systematic Review and Dose-Response Meta-Analysis of Prospective Cohort Studies. J Am Heart Assoc. 2017;6(9) doi: 10.1161/JAHA.117.005947. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Krieger N., Sidney S. Racial discrimination and blood pressure: the CARDIA Study of young black and white adults. Am J Public Health. 1996;86(10):1370–1378. doi: 10.2105/AJPH.86.10.1370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Cheng S., Xanthakis V., Sullivan L.M., Vasan R.S. Blood pressure tracking over the adult life course. Hypertension. 2012;60(6):1393–1399. doi: 10.1161/HYPERTENSIONAHA.112.201780. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Wu J., Kraja A.T., Oberman A., et al. A summary of the effects of antihypertensive medications on measured blood pressure. Am J Hypertens. 2005;18(7):935–942. doi: 10.1016/j.amjhyper.2005.01.011. [DOI] [PubMed] [Google Scholar]
- 35.Stuart J.J., Tanz L.J., Rimm E.B., et al. Cardiovascular Risk Factors Mediate the Long-Term Maternal Risk Associated With Hypertensive Disorders of Pregnancy. J Am Coll Cardiol. 2022;79(19):1901–1913. doi: 10.1016/j.jacc.2022.03.335. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Schoenaker D.A., Soedamah-Muthu S.S., Mishra G.D. Quantifying the mediating effect of body mass index on the relation between a Mediterranean diet and development of maternal pregnancy complications: the Australian Longitudinal Study on Women’s Health. Am J Clin Nutr. 2016;104(3):638–645. doi: 10.3945/ajcn.116.133884. [DOI] [PubMed] [Google Scholar]
- 37.Whelton P.K., Carey R.M., Aronow W.S., 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: a Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Hypertension. 2018;71(6):e13–e115. doi: 10.1161/HYP.0000000000000065. [DOI] [PubMed] [Google Scholar]
- 38.Al-Huda F., Shapiro G.D., Davenport M.H., Bertagnolli M., Dayan N. Association between Cardiorespiratory Fitness and Hypertensive Disorders of Pregnancy: a Systematic Review and Meta-Analysis. J Clin Med. 2022;11(15):4364. doi: 10.3390/jcm11154364. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Borrowman J.D., Huang X., Perak A.M., et al. Estimating the Proportion of Adverse Pregnancy Outcomes Prevented With Improved Early Pregnancy Cardiovascular Health. JACC Adv. 2026;5(2) doi: 10.1016/j.jacadv.2025.102501. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Ji H., Kim A., Ebinger J.E., et al. Sex Differences in Blood Pressure Trajectories Over the Life Course. JAMA Cardiol. 2020;5(3):255–262. doi: 10.1001/jamacardio.2019.5306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Kishi S., Gisela Teixido-Tura, Ning H., et al. Cumulative Blood Pressure in Early Adulthood and Cardiac Dysfunction in Middle Age. J Am Coll Cardiol. 2015;65(25):2679–2687. doi: 10.1016/j.jacc.2015.04.042. [DOI] [PubMed] [Google Scholar]
- 42.Lane A., Wilcox S., Wingard E., McLean M.K., Liu J. Association of a Lifestyle Intervention with Blood Pressure Trajectories During Pregnancy and Postpartum in Women with Pre-pregnancy Overweight and Obesity. Matern Child Health J. 2023;27(8):1407–1415. doi: 10.1007/s10995-023-03709-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Zinsser L.A., Stoll K., Wieber F., Pehlke-Milde J., Gross M.M. Changing behaviour in pregnant women: a scoping review. Midwifery. 2020;85 doi: 10.1016/j.midw.2020.102680. [DOI] [PubMed] [Google Scholar]
- 44.Lindqvist M., Lindkvist M., Eurenius E., Persson M., Mogren I. Change of lifestyle habits – Motivation and ability reported by pregnant women in northern Sweden. Sex Reprod Healthc. 2017;13:83–90. doi: 10.1016/j.srhc.2017.07.001. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Due to the sensitive nature of the data collected in this study, requests to access data used in this analysis from qualified researchers trained in human subject confidentiality protocols will be considered by the data coordinating center, and may be submitted at https://numom2b.org.