Abstract
Background:
Life course patterns of change in risk – trajectories – affect health.
Objectives:
To examine how trajectories of cardiovascular risk factors are associated with pregnancy and birth outcomes
Methods:
Data from two cohort studies participating in the International Childhood Cardiovascular Consortium - The Bogalusa Heart Study (BHS; started in 1973, N=903 for this analysis) and the Cardiovascular Risk in Young Finns Study (YFS; started in 1980, N=499) were used. Both followed children into adulthood and measured cardiovascular risk factors, including body mass index (BMI); systolic and diastolic blood pressure (SBP/DBP); total, low density lipoprotein (LDL)- and high density lipoprotein (HDL)-cholesterol; and serum triglycerides,. Discrete mixture modeling was used to divide each cohort into distinct trajectories according to these risk factors from childhood to early adulthood and these groups were then used to predict pregnancy outcomes including small for gestational age (SGA; <10th study-specific percentile of gestational age by sex), preterm birth (PTB; <37 weeks’ gestation), hypertensive disorders of pregnancy (HDP), and gestational diabetes mellitus (GDM), with control for age at baseline and at first birth, parity, socioeconomic status, BMI, and smoking.
Results:
The models created more trajectories for BMI, SBP, and HDL-cholesterol in the YFS than in BHS, for which 3 classes generally seemed to be sufficient to represent the groups in the population across risk factors. In BHS, the association between the higher and flatter DBP trajectory and PTB was aRR 1.77, 95% CI 1.06–2.96. In BHS the association between consistently total cholesterol and PTB was aRR 2.16, 95% CI 1.22–3.85 and in YFS the association between elevated-high trajectory and PTB was aRR 3.35, 95% CI 1.28–8.79. Elevated-increasing SBP was associated with higher risk of GH in BHS and increasing or persistent-obese BMI trajectories were associated with GDM in both cohorts (BHS: aRR 3.51, 95% CI 1.95, 6.30; YFS: aRR 2.61, 95% CI 0.96, 7.08)
Conclusions:
Trajectories of cardiovascular risk, particularly those that represent consistent or more rapid worsening of cardiovascular health, are associated with higher risk of pregnancy complications.
Keywords: life course, trajectory, blood pressure, lipids, birthweight, pregnancy complications
Background
Increasing interest focuses on life course predictors of health. While a single time point in early life is sometimes critical, there is also interest in how risk changes across time, and how the patterns of change in risk – trajectories – affect health. For instance, the International Cardiovascular Cohort Consortium (i3C) found that childhood risk factors predict adulthood cardiovascular outcomes1 and that individuals with persistently elevated blood pressure from childhood to adulthood had increased risk of atherosclerosis, but risk was reduced if the elevated blood pressure during childhood resolved by adulthood. One analytic method is to examine a priori hypothesized changes over time, for instance, examining people with consistently increasing or decreasing levels of some factor. In some cases, however, groups with common experiences are extracted from the data. In the Dunedin Multidisciplinary Health and Development Study, latent trajectories of systolic blood pressure were identified,2 and blood pressure trajectories leading to prehypertension or hypertension were associated with other cardiovascular risk indicators. The South African Birth to 20 study found 3 distinct trajectories of BMI in boys and 4 in girls. Children in the early onset obesity or overweight BMI trajectories were more likely to have higher blood pressure in late adolescence.3 The Bogalusa Heart Study (BHS) found five trajectories of life course cardiovascular risk. Those who started with relatively high cardiovascular risk but had lower Framingham risk scores as adults had lower risks of diabetes and lower BMIs, better carotid intima-media thickness, and pulse wave velocity than those who were in the high-risk group in both time periods. Children who improved their relative cardiovascular risk over the life course achieved better midlife atherosclerotic health despite maintaining relatively poor metabolic health through adulthood.4
These methods can also be applied to understanding the relationship between early-life and pregnancy health.5 Preconception cardiovascular and metabolic health predict complications of pregnancy and adverse birth outcomes.6–9 While most previous studies address the short-term preconception period, lifetime experience also determines health status on entering pregnancy, and may directly indicate or affect health during pregnancy. A few studies have examined this lifetime experience with respect to pregnancy, particularly as relates to hypertensive disorders of pregnancy (HDP). An analysis of the trajectories in the 5 years prior to pregnancy among Chicago women found those in the overweight/prehypertensive groups were more likely to develop HDP.10 Increase in weight over time was associated with increased risk of HDP in Australian women.11 A Canadian study found that women who developed gestational diabetes mellitus (GDM) during pregnancy had a previously higher annual rate of change in HbA1c, glucose, and triglycerides, and greater decrease in high-density lipoprotein (HDL).12 We are not aware of studies that have associated preconception trajectories of cardiometabolic health with birth outcomes. In the BHS, an overall measure of childhood blood pressure was associated with higher risk of preterm birth,13 but trajectory was not assessed.
In this study, we examine how life-long trajectories of cardiovascular risk factors predict pregnancy outcomes in two independent cohorts. We hypothesized that a steeper worsening of cardiovascular risk factors would be associated with an increased risk of adverse pregnancy and birth outcomes.
Methods
Data from two cohort studies participating in the i3C14 were used. The Bogalusa Heart Study (BHS) began in 1973, examining cardiovascular risk factors in children.15 Cross-sectional and longitudinal studies have been conducted regularly since then; more recent studies have addressed midlife cardiovascular, cognitive, and physical function. The Bogalusa Babies study interviewed women (n=1803) and linked to vital statistics and medical records where available.16 The Cardiovascular Risk in Young Finns study (YFS) began in 1980 with children aged 3–18.17 They have been followed at regular intervals since then, with the most recent follow-up in 2018–2020. Similar to BHS, the primary focus of the YFS has been on cardiovascular risk factors and outcomes since childhood, while new topics such as cognitive function and fatty liver have been introduced in the latest follow-ups.
Exposure
Trajectories and change in BMI, SBP, DBP, serum total cholesterol, serum HDL- and low-density lipoprotein (LDL)-cholesterol, and triglycerides were examined as predictors.
BMI
At each BHS visit (up to 23 total visits), replicate measures of height and weight, using a stadiometer and balance beam metric scale, were obtained. At each YFS visit, unclothed weight was measured to the nearest 0.1 kg using digital scales and standing height was measured to the nearest 0.1 cm using a wall-mounted stadiometer. BMI (kg/m2) was calculated using mean values.
Systolic and diastolic blood pressure (SBP/DBP)
In BHS, blood pressure levels were measured on the right arm of subjects in a relaxed, sitting position by 2 trained observers (3 replicates each). Systolic blood pressure and diastolic blood pressure were recorded using a mercury sphygmomanometer. In YFS, blood pressure was measured in a sitting position after 5 minutes’ rest, using a random-zero sphygmomanometer (Hawksley & Sons Ltd, Lancin, UK). The fifth Korotkoff phase was used as the sign of diastolic blood pressure. Readings were performed at least three times on each subject, and the mean used for analysis.
Lipids
In BHS, fasting blood samples were drawn for lipids analysis. Prior to 1987, serum total cholesterol levels were determined by a Technicon AutoAnalyzer II (Technicon Instrument, Tarrytown, NY). From 1987 to 1996, cholesterol levels were determined by enzymatic procedures on the Abbott VP instrument (Abbott Laboratories, Chicago, IL) and on the Hitachi 902 Automatic Analyzer (Roche Diagnostics, Indianapolis, IN) after 1996. Both chemical and enzymatic procedures met the performance requirements of the Lipid Standardization Program of the Centers for Disease Control and Prevention, which has routinely monitored the precision and accuracy of cholesterol and triglyceride measurements since 1973. At each YFS visit, a fasting blood draw was conducted. Cholesterol and other lipids were measured at each visit by the enzymatic cholesterol esterase–cholesterol oxidase method (Cholesterol reagent, Olympus, Ireland). As changes in determination methods had occurred, the values were standardized to be comparable between the study years.18, 19
Outcomes
BHS:
1803 women were interviewed about their reproductive history, which formed the baseline group for this analysis. Of these women, 381 also had medical records data, and 1026 were matched to vital records data on their births (the sample included both parous and nulliparous women). Outcomes assessed were: small for gestational age(SGA; <10th sex-specific percentile of birthweight by gestational week for all births in the study), preterm birth (PTB; <37 weeks’ gestation), gestational hypertension (GH), and gestational diabetes mellitus (GDM), all at any pregnancy. Maternal report of birthweight in Babies compared with vital statistics was excellent; mean and median differences were statistically indistinguishable from 0 and kappa for categorized birthweight was >0.9; mean difference for gestational age was 0.01 week.20 Therefore, self-report data of these outcomes was used. As HDP and GDM are under-reported in vital statistics,21 these complications were compared only to medical records for the women who had them available (n=381). For GH, kappa was 0.65 and 0.73 when clustering was accounted for, depending on how incomplete medical records were treated. For GDM, kappa for agreement was between 0.73 and 0.80 (depending on how incomplete medical records were treated) when clustering was not accounted for, and 0.84 to 0.92 when clustered kappas22 were calculated. This level or higher quality of recall has been shown for GDM in other studies as well (specificity=98%, sensitivity=92%).23 For pregnancy complications, if more than one data source was available, medical records were used unless information was available from BHS that contradicted them (i.e., if BHS exam records indicated pre-pregnancy hypertension or diabetes, participants were not considered at risk for GH or GDM). The number of self-reported pre-eclampsia cases was implausible, so this outcome was not examined.
YFS:
Data from the cardiovascular study cohort were merged with the Finnish national birth registry. The birth registry contains information on every birth in Finland since 1987, including data on the pregnancy. Of the 1832 women originally enrolled in the Cardiovascular Risk in Young Finns study, 1323 had a pregnancy in the birth registry; 1316 had at least one singleton in the registry. The birth registry data include direct reports of some medical information, as well as up to 10 spaces to record diagnoses using ICD-9 and ICD-10 codes.
Outcomes analyzed were:
SGA, PTB, HDP, and GDM. SGA was defined as <10th sex-specific percentile of birthweight by gestational week for all births in the study; preterm birth was defined as birth at <37 weeks’ gestation. Overall HDP were defined as any hypertensive disorder specified as having begun during pregnancy, or an eclamptic disorder (ICD-9 codes 6423–6427, ICD-10 codes O12-O15). Gestational diabetes was defined as failing a glucose tolerance test, ICD-9 code 6488, or ICD-10 code O24.
Statistical analysis
BHS:
The exposures (BMI, SBP, DBP, cholesterol, LDL-cholesterol, HDL-cholesterol, and triglycerides) were first used to build a discrete mixture model to separate the population into subpopulations. Each participant was then identified by their respective class. We selected a class structure and number that minimized the Bayesian Information Criterion (BIC) and the Likelihood test statistic and ensured that each class comprised at least 5% of the sample. We applied group-based trajectory modeling (using SAS proc traj [13]) to estimate distinct trajectories of individual risk factors from childhood to adulthood. We tested models using 3–5 distinct groups and selected the model with the lowest absolute BIC value. This analysis was performed limited to visits before the pregnancy. The classes were modeled as predictors, with reproductive outcomes as the dependent variable. The number of trajectories that was optimal varied by risk factor and study (Table 1, Table S1, Figure S1 and S2).
Table 1.
Trajectories of pre-pregnancy risk factors
| The Bogalusa Heart Study (n=903) | The Young Finns Study (n=499) | ||
|---|---|---|---|
| N (%) | N (%) | ||
| BMI | BMI group | ||
| slow increase | 563 (62.4) | stable slim | 208 (41.7) |
| moderate increase | 255 (28.2) | stable normal | 207 (41.5) |
| steepest increase | 85 (9.4) | progressively overweight | 66 (13.2) |
| persistent increasing obese | 18 (3.6) | ||
| systolic blood pressure group | systolic blood pressure group | ||
| slow increase | 304 (33.7) | low-stable | 203 (43.4) |
| moderate increase | 476 (52.7) | moderate-increasing | 241 (51.5) |
| steepest increase | 123 (13.6) | elevated-increasing | 24 (5.1) |
| diastolic blood pressure group | diastolic blood pressure group | ||
| rising | 232 (25.7) | low-stable | 241 (51.8) |
| starts higher, rising | 564 (62.5) | normal-stable | 224 (48.2) |
| higher and flatter | 107 (11.9) | ||
| cholesterol | cholesterol group | ||
| consistently low | 436 (48.3) | low-stable | 106 (21.8) |
| middle | 419 (46.4) | elevated-stable | 233 (47.8) |
| consistently high | 48 (5.3) | high-stable | 120 (24.6) |
| elevated-high | 28 (5.8) | ||
| LDL cholesterol | LDL cholesterol group | ||
| consistently low | 389 (43.1) | low-stable | 78 (16.2) |
| higher but flat | 429 (47.5) | normal-stable | 212 (43.9) |
| increasing | 85 (9.4) | elevated-stable | 149 (30.9) |
| high-stable | 44 (9.1) | ||
| HDL cholesterol | HDL cholesterol group | ||
| consistently low | 138 (15.3) | low-stable | 150 (30.7) |
| medium/flat | 629 (69.7) | normal-stable | 272 (55.6) |
| declining | 136 (15.1) | elevated-stable | 67 (13.7) |
| triglycerides | triglycerides group | ||
| consistently low | 173 (19.2) | low-stable | 441 (90.4) |
| moderate and flat | 705 (78.1) | normal-increasing | 47 (9.6) |
| steep rise | 25 (2.8) | ||
Cardiovascular Risk in Young Finns Study
A detailed description of the creation of the cardiovascular risk factor trajectories have been reported elsewhere.24 Heterogeneity in the longitudinal development of cardiovascular risk factors was investigated using group-based trajectory modeling performed with SAS proc traj procedure9 to identify subgroups of YFS participants who shared similar underlying trajectories between ages 9 and 49 years. Participants who used antihypertensive (N=145 women) or dyslipidemia medication (N=32 women) were excluded from the cardiovascular risk factor-specific trajectory modeling analyses. All other participants were included in trajectory analyses but, for reliability, a minimum of three measurements was required with at least one being from childhood and adolescence (ages 9–18 years) and at least one from adulthood (ages 21–49 years). BICs were checked and the mean posterior probability for each group was >0.70. Frequency of >5% was preferred for the size of trajectory groups. Both overall and sex-specific models were applied; as this analysis is limited to women, the sex-specific models are used. This analysis was performed limited to visits before the pregnancy.
Once the trajectories were calculated for the two cohorts, similar analytic strategies were followed. Demographic and health characteristics of the final trajectory groupings were tabulated and compared using Pearson’s chi-square tests (categorical variables) or analysis of variance (ANOVA) (continuous variables). Next, multivariable generalized linear modeling were performed (log-binomial or robust log-Poisson models)25. Trajectory assignment probabilities served as the independent variables in this model, while pregnancy outcomes were used as the dependent variables. The covariates controlled for were slightly different due to differences in data collected across studies. For BHS, models controlled for age at first pregnancy, age at last pregnancy, total births, education, smoking, and lifetime BMI (for non-BMI exposures), estimated with area under the curve.26 For YFS, models controlled for age at baseline, age at first birth, total births, socioeconomic status, smoking in 2001, and BMI in 2001. These covariates were chosen as likely to be associated with the outcome and exposure, and not on the causal pathway. Interactions with parity were examined to determine whether the associations with exposure trajectories differed in those who had more births.
Missing data
Multiple imputation was used to address missing covariate data (321 observations from BHS and 127 observations in YFS with missing data on at least one covariate; 50 imputations with Markov chain Monte Carlo methods; more details in the supplementary material). The adjusted models use the multiply-imputed data.
Ethics approval
The Bogalusa Heart Study and Bogalusa Babies studies were approved by the Tulane University Institutional Review Board. The Young Finns study was approved by the local ethical committees (1st Ethical Committee of the Hospital District of Southwest Finland, Regional Ethics Committee of the Expert Responsibility area of Tampere University Hospital, Helsinki University Hospital Ethical Committee of Medicine, The Research Ethics Committee of the Northern Savo Hospital District and Ethics Committee of the Northern Ostrobothnia Hospital District) of the participating sites (ETMK: 88/180/2010).
Results
The analysis was limited to women with pregnancies and descriptions of samples for the two cohorts are provided in Table 2. The age at last included visit was higher for YFS participants (mean 37.7) than BHS (mean 31.2), and the age range was narrower in YFS than BHS. The models created more trajectories for BMI, SBP, and HDL-cholesterol in the YFS (and fewer for triglycerides) than in BHS, for which 3 classes generally seemed to be sufficient to represent the groups in the population across risk factors. For BHS, 3 classes generally seemed to be sufficient across risk factors. In both cohorts, BMI and SBP were likely to increase substantially, at least in some groups, while cholesterol was more stable (Table 1, Figures S1 and S2).
Table 2.
Description of the analytic cohorts (women with pregnancies)
| Bogalusa Heart Study | Young Finns Study | ||||||
|---|---|---|---|---|---|---|---|
| Mean (SD) | min, max | Mean (SD) | min, max | ||||
| age at first visit | 8.5 (2.9) | 3.4, 15.6 | 12.0 (5.1) | 3.0, 18.0 | |||
| age at last included visit | 30.4 (11.5) | 7.7, 51.0 | 39.0 (5.1) | 30.0, 45.0 | |||
| age at last visit prior to first pregnancy | 19.2 (5.3) | 7.7, 39.0 | 27.2 (4.4) | 15.0, 39.0 | |||
| BMI at visit closest to age 18 | 22.2 (4.6) | 12.3, 44.7 | 21.1 (2.3) | 14.8, 29.3 | |||
| BMI at last included visit | 23.0 (5.2) | 12.6, 53.9 | 23.2 (3.8) | 17.1, 41.3 | |||
| SBP at visit closest to age 18 | 109.0 (8.8) | 81.7, 145.3 | 116.6 (10.7) | 90.0, 155.3 | |||
| SBP at last included visit | 108.2 (8.7) | 81.7, 145.3 | 113.8 (12.5) | 84.7, 199.3 | |||
| total cholesterol at visit closest to age 18 | 4.2 (0.8) | 2.2, 9.0 | 5.1 (1.0) | 3.0, 10.4 | |||
| total cholesterol at last included visit | 4.4 (0.9) | 2.5, 8.4 | 5.0 (0.8) | 3.0, 8.0 | |||
| Median (IQR) | min, max | median | min, max | ||||
| total number of visits | 6 (4, 9) | 3, 14 | 5 (4, 5) | 2, 7 | |||
| total number of visits < 18 years | 4 (3, 4) | 1,8 | 3 (1, 4) | 1,5 | |||
| education | N | % | N | % | |||
| < high school | 58 | 6.4 | comprehensive | 97 | 21.3 | ||
| high school | 250 | 27.7 | high school | 150 | 33.0 | ||
| associates or some college | 295 | 32.7 | trade | 57 | 12.5 | ||
| college+ | 299 | 33.2 | college+ | 151 | 33.2 | ||
| race | |||||||
| Black | 327 | 36.2 | a | ||||
| White | 576 | 63.8 | |||||
| pregnancy outcomes (at any pregnancy) | |||||||
| Small-for-gestational-age | 155 | 17.2 | 82 | 16.3 | |||
| Preterm birth | 154 | 17.1 | 50 | 9.9 | |||
| Gestational hypertension | 81 | 9.9 | 20 | 4.0 | |||
| Gestational diabetes | 96 | 10.6 | 63 | 12.5 | |||
Not collected in YFS
Higher and more strongly worsening risk factors were associated with birth outcomes (Tables 3 and 4). The elevated-increasing SBP group was associated with an increased likelihood of SGA (aRR 2.12, 95% CI 0.90, 5.00) in YFS, while the steepest increase in SBP (aRR 1.58, 95% CI 0.98, 2.57) and DBP (aRR 1.77, 95% CI 1.06, 2.96) were associated with PTB in BHS. The elevated high total cholesterol group was also associated with PTB in YFS (aRR 3.38, 95% CI 1.28, 8.79), while the high-stable LDL group (aRR, 3.13, 95% CI 1.04, 9.39 ) in YFS and the LDL group with the highest increase in BHS had higher risks of PTB (aRR 1.59, 95% CI 0.95, 2.67).
Table 3.
Trajectories of risk factors and birth outcomes, women in the Bogalusa Heart Study (n=903).
| small for gestational age | preterm birth | |||
|---|---|---|---|---|
| adjusted for age at baseline | adjusted for age at baseline, BMI, parity, SES, smoking, age at first birth | adjusted for age at baseline | adjusted for age at baseline, BMI, parity, SES, smoking, age at first birth | |
|
|
||||
| RR (95% CI) | RR (95% CI) | RR (95% CI) | RR (95% CI) | |
| BMI | ||||
| slow increase | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| moderate increase | 0.77 (0.54, 1.08) | 0.72 (0.49, 1.05) | 0.97 (0.69, 1.35) | 0.96 (0.67, 1.38) |
| steepest increase | 0.81 (0.48, 1.38) | 0.72 (0.40, 1.30) | 1.21 (0.77, 1.91) | 1.23 (0.74, 2.05) |
| systolic blood pressure group | ||||
| slow increase | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| moderate increase | 1.00 (0.73, 1.38) | 1.16 (0.81, 1.66) | 1.00 (0.72, 1.39) | 1.05 (0.72, 1.52) |
| steepest increase | 1.15 (0.74, 1.79) | 1.47 (0.88, 2.44) | 1.45 (0.96, 2.18) | 1.58 (0.98, 2.57) |
| diastolic blood pressure group | ||||
| rising | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| starts higher, rising | 0.82 (0.59, 1.12) | 0.96 (0.67, 1.38) | 0.98 (0.70, 1.39) | 1.04 (0.71, 1.53) |
| higher and flatter | 0.86 (0.52, 1.41) | 1.18 (0.67, 2.08) | 1.61 (1.04, 2.50) | 1.77 (1.06, 2.96) |
| cholesterol | ||||
| consistently low | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| middle | 0.92 (0.69, 1.25) | 1.00 (0.72, 1.40) | 1.28 (0.94, 1.73) | 1.32 (0.94, 1.85) |
| consistently high | 1.30 (0.74, 2.27) | 1.46 (0.77, 2.76) | 2.15 (1.33, 3.47) | 2.16 (1.22, 3.85) |
| LDL cholesterol | ||||
| consistently low | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| higher but flat | 0.88 (0.65, 1.20) | 0.99 (0.70, 1.39) | 1.20 (0.88, 1.65) | 1.24 (0.88, 1.76) |
| increasing | 1.17 (0.74, 1.87) | 1.34 (0.79, 2.27) | 1.55 (0.98, 2.44) | 1.59 (0.95, 2.67) |
| HDL cholesterol | ||||
| consistently low | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| medium/flat | 1.05 (0.70, 1.58) | 0.97 (0.61, 1.53) | 0.87 (0.59, 1.27) | 0.81 (0.52, 1.25) |
| declining | 0.98 (0.57, 1.69) | 0.94 (0.51, 1.71) | 0.87 (0.52, 1.46) | 0.85 (0.48, 1.52) |
| triglycerides | ||||
| consistently low | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | |
| moderate and flat | 1.05 (0.72, 1.53) | 1.00 (0.65, 1.52) | 0.61 (0.42, 0.89) | |
| steep rise | 1.46 (0.67, 3.17) | 1.62 (0.66, 3.94) | 0.89 (0.35, 2.27) | |
BMI group not adjusted for BMI in multivariable models
Number of trajectories analyzed limited due to small sample size
Table 4.
Trajectories of risk factors and birth outcomes, women in the Young Finns Study (n=499)
| small for gestational age |
preterm birth |
|||
|---|---|---|---|---|
| adjusted for age at baseline | adjusted for age at baseline, BMI, parity, SES, smoking, age at first birth | adjusted for age at baseline | adjusted for age at baseline, BMI, parity, SES, smoking, age at first birth | |
|
|
||||
| RR (95% CI) | RR (95% CI) | RR (95% CI) | RR (95% CI) | |
| BMI group | ||||
| stable slim | 1.15 (0.75, 1.77) | 1.15 (0.72, 1.85) | 1.24 (0.72, 2.13) | 1.20 (0.67, 2.13) |
| stable normal | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| progressively overweight | 0.98 (0.51, 1.88) | 0.93 (0.46, 1.90) | 0.15 (0.02, 1.09) | 0.14 (0.02, 1.02) |
| persistent increasing obese | 0.71 (0.18, 2.80) | 0.73 (0.17, 3.08) | 1.04 (0.27, 4.10) | 1.16 (0.27, 5.02) |
| Systolic blood pressure group | ||||
| low-stable | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| moderate-increasing | 1.03 (0.67, 1.59) | 1.02 (0.63, 1.66) | 1.39 (0.76, 2.57) | 1.52 (0.79, 2.91) |
| elevated-increasing | 1.90 (0.95, 3.80) | 2.12 (0.90, 5.00) | 1.73 (0.54, 5.57) | 2.45 (0.68, 8.85) |
| Diastolic blood pressure group | ||||
| low-stable | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| normal-stable | 0.80 (0.53, 1.21) | 0.82 (0.51, 1.32) | 1.02 (0.58, 1.81) | 1.15 (0.62, 2.13) |
| cholesterol group | ||||
| low-stable | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| elevated-stable | 0.88 (0.52, 1.49) | 0.87 (0.49, 1.56) | 1.13 (0.54, 2.40) | 1.12 (0.51, 2.45) |
| high-stable | 0.93 (0.51, 1.70) | 0.93 (0.48, 1.78) | 0.90 (0.37, 2.22) | 0.88 (0.35, 2.23) |
| elevated-high | 1.89 (0.96, 3.75) | 1.79 (0.80, 4.02) | 3.39 (1.44, 8.00) | 3.35 (1.28, 8.79) |
| LDL cholesterol group | ||||
| low-stable | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| normal-stable | 0.58 (0.34, 1.01) | 0.56 (0.30, 1.04) | 1.63 (0.64, 4.17) | 1.51 (0.57, 4.02) |
| elevated-stable | 0.77 (0.44, 1.34) | 0.76 (0.41, 1.42) | 1.29 (0.47, 3.55) | 1.23 (0.43, 3.53) |
| high-stable | 1.04 (0.53, 2.07) | 1.00 (0.45, 2.19) | 3.24 (1.16, 9.12) | 3.13 (1.04, 9.39) |
| HDL cholesterol group | ||||
| low-stable | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| normal-stable | 1.40 (0.85, 2.30) | 1.33 (0.78, 2.29) | 0.97 (0.54, 1.74) | 0.81 (0.43, 1.51) |
| elevated-stable | 1.55 (0.80, 3.01) | 1.48 (0.71, 3.06) | 0.57 (0.20, 1.63) | 0.47 (0.15, 1.45) |
| triglycerides group | ||||
| low-stable | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| normal-increasing | 1.20 (0.63, 2.27) | 1.30 (0.63, 2.66) | 0.83 (0.31, 2.22) | 1.00 (0.35, 2.86) |
For pregnancy complications (Tables 5 and 6), trajectories of blood pressure and BMI had the strongest associations. The group with the steepest increase in SBP had the highest risk of GH for BHS (aRR 2.69, 95% CI 1.31, 5.56); in YFS, the point estimate for the groups with increasing trajectories were similar but very imprecise. For both BHS and YFS, groups with higher and increasing BMI were at higher risk for GDM (BHS: aRR 3.51, 95% CI 1.95, 6.30; YFS: aRR 2.61, 95% CI 0.96, 7.08).
Table 5.
Trajectories of risk factors and pregnancy complications, women in the Bogalusa Heart Study (n=903)
| any gestational hypertension | any gestational diabetes | |||
|---|---|---|---|---|
| adjusted for age at baseline | adjusted for age at baseline, BMI, parity, SES, smoking, age at first birtha | adjusted for age at baseline | adjusted for age at baseline, BMI, parity, SES, smoking, age at first birtha | |
|
|
||||
| BMI | RR (95% CI) | RR (95% CI) | RR (95% CI) | RR (95% CI) |
| slow increase | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| moderate increase | 1.68 (1.05, 2.70) | 1.37 (0.88, 2.14) | 1.35 (0.89, 2.06) | 1.72 (1.04, 2.84) |
| steepest increase | 3.40 (2.03, 5.70) | 3.51 (1.95, 6.30) | 1.61 (0.91, 2.85) | 1.70 (0.91, 3.17) |
| systolic blood pressure group | ||||
| slow increase | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| moderate increase | 2.11 (1.21, 3.70) | 1.80 (0.99, 3.26) | 1.47 (0.94, 2.30) | 1.35 (0.83, 2.18) |
| steepest increase | 3.68 (1.95, 6.93) | 2.69 (1.31, 5.56) | 1.40 (0.75, 2.60) | 1.21 (0.61, 2.40) |
| diastolic blood pressure group | ||||
| rising | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| starts higher, rising | 1.91 (1.07, 3.42) | 1.62 (0.87, 2.99) | 1.34 (0.83, 2.16) | 1.22 (0.73, 2.05) |
| higher and flatter | 2.62 (1.23, 5.57) | 1.91 (0.82, 4.45) | 1.49 (0.77, 2.90) | 1.25 (0.61, 2.59) |
| cholesterol | ||||
| consistently low | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| middle | 0.71 (0.46, 1.10) | 0.76 (0.48, 1.22) | 1.04 (0.71, 1.54) | 1.06 (0.70, 1.61) |
| consistently high | 1.49 (0.72, 3.07) | 1.24 (0.55, 2.79) | 0.96 (0.40, 2.30) | 0.94 (0.37, 2.40) |
| LDL cholesterol | ||||
| consistently low | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| higher but flat | 0.77 (0.50, 1.20) | 0.83 (0.52, 1.32) | 0.96 (0.64, 1.44) | 0.96 (0.62, 1.48) |
| increasing | 1.07 (0.56, 2.06) | 1.04 (0.51, 2.11) | 1.28 (0.70, 2.35) | 1.30 (0.67, 2.49) |
| HDL cholesterol | ||||
| consistently low | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| medium/flat | 0.86 (0.51, 1.44) | 1.07 (0.61, 1.91) | 1.25 (0.71, 2.19) | 1.35 (0.74, 2.47) |
| declining | 0.46 (0.18, 1.13) | 0.58 (0.22, 1.51) | 0.83 (0.37, 1.84) | 0.91 (0.39, 2.11) |
| triglycerides | ||||
| consistently low | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| moderate and flat | 0.87 (0.52, 1.44) | 1.08 (0.61, 1.89) | 0.73 (0.47, 1.14) | 0.78 (0.48, 1.27) |
| steep rise | 1.84 (0.69, 4.90) | 1.72 (0.58, 5.15) | 1.20 (0.45, 3.18) | 1.20 (0.41, 3.54) |
BMI group not adjusted for BMI in multivariable models
Table 6.
Trajectories of risk factors and pregnancy complications, women in the Young Finns Study (n=499).
| any hypertensive disorders | any gestational diabetes | |||
|---|---|---|---|---|
| adjusted for age at baseline | adjusted for age at baseline, BMI, parity, SES, smoking, age at first birth | adjusted for age at baseline | adjusted for age at baseline, BMI, parity, SES, smoking, age at first birth | |
|
|
||||
| RR (95% CI) | RR (95% CI) | RR (95% CI) | RR (95% CI) | |
| BMI group | ||||
| stable slim | 0.51 (0.18, 1.48) | 0.51 (0.17, 1.51) | 1.13 (0.68, 1.91) | 1.01 (0.57, 1.79) |
| stable normal | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| progressively overweight | 1.26 (0.41, 3.89) | 1.27 (0.39, 4.19) | 1.39 (0.70, 2.77) | 1.17 (0.53, 2.54) |
| persistent increasing obese | 0.89 (0.11, 7.58) | 1.10 (0.13, 8.93) | 1.82 (0.85, 3.87) | 2.61 (0.96, 7.08) |
| Systolic blood pressure group | ||||
| low-stable | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| moderate-increasing | 2.40 (0.90, 6.43) | 2.39 (0.80, 7.12) | 1.34 (0.82, 2.19) | 1.31 (0.76, 2.26) |
| elevated-increasing | 2.43 (0.30, 19.58) | 1.77 (0.18, 17.15) | 0.97 (0.25, 3.80) | 0.66 (0.15, 2.95) |
| Diastolic blood pressure group | ||||
| low-stable | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| normal-stable | 1.86 (0.74, 4.66) | 1.60 (0.59, 4.36) | 1.13 (0.71, 1.81) | 1.10 (0.65, 1.89) |
| cholesterol group | ||||
| low-stable | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| elevated-stable | 0.86 (0.29, 2.58) | 0.97 (0.30, 3.08) | 0.81 (0.47, 1.41) | 0.85 (0.46, 1.59) |
| high-stable | 0.87 (0.24, 3.17) | 0.78 (0.21, 2.97) | 0.95 (0.50, 1.80) | 0.84 (0.42, 1.68) |
| elevated-high | 1.59 (0.32, 7.82) | 1.42 (0.27, 7.53) | 0.44 (0.11, 1.78) | 0.42 (0.10, 1.84) |
| LDL cholesterol group | ||||
| low-stable | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| normal-stable | 0.87 (0.28, 2.70) | 0.90 (0.27, 3.01) | 0.91 (0.49, 1.66) | 0.95 (0.48, 1.91) |
| elevated-stable | 0.31 (0.06, 1.65) | 0.32 (0.06, 1.81) | 0.96 (0.50, 1.87) | 1.12 (0.53, 2.36) |
| high-stable | 1.98 (0.53, 7.41) | 1.86 (0.45, 7.59) | 0.51 (0.15, 1.67) | 0.42 (0.12, 1.51) |
| HDL cholesterol group | ||||
| low-stable | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| normal-stable | 0.74 (0.30, 1.83) | 0.91 (0.34, 2.43) | 0.77 (0.48, 1.23) | 0.82 (0.48, 1.42) |
| elevated-stable | 0.35 (0.04, 2.89) | 0.45 (0.06, 3.66) | 0.47 (0.17, 1.29) | 0.59 (0.20, 1.73) |
| triglycerides group | ||||
| low-stable | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| normal-increasing | 0.89 (0.21, 3.77) | 0.43 (0.09, 2.12) | 2.21 (1.34, 3.66) | 1.81 (0.92, 3.57) |
No interactions were found with parity.
Comment
Principal findings
This analysis of two cohorts of different populations found both similarities and differences for the classes. The shapes of the trajectories were similar enough, however, that some comparisons can be drawn. Trajectories of cardiovascular risk factors were associated with some pregnancy and birth outcomes. In both cohorts, the steepest trajectories of cholesterol and LDL were associated with PTB. Higher DBP was associated with PTB in BHS, and point estimates were raised for SBP and PTB in both cohorts, though very imprecisely in some cases. Steeper increases in SBP were associated with higher risk of GH in BHS and increasing or persistent-obese BMI trajectories with associated with GDM.
Strengths of the study
Strengths of this analysis include the two cohorts in diverse populations with detailed and consistent measurements of cardiovascular risk factors. Because the studies were not limited to tests ordered for clinical reasons, the populations reflect a range of nonclinical and nonsymptomatic values. The trajectories go back to childhood for a fuller understanding of the life course. Some previous studies have been limited to adult measures;12 for pregnancy, which usually occurs in early adulthood, this is a narrow window. Trajectory modeling creates data-driven groupings of participants in longitudinal studies, which may reveal patterns that are not apparent when participants are categorized a priori, and allow for using more than two datapoints per participant, unlike using annual rate of change12, 27 or piecewise linear models28 may impose a linear relationship on the data.
Limitations of the data
Limitations include the variation in number of visits and the lack of information on cardiovascular risk factors during pregnancy. Particularly the YFS was left with a relatively small sample size when limited by the number of visits necessary in pre-pregnancy. Pregnancy complications were self-reported in some cases (BHS) or limited to what is recorded in the Medical Birth Registry (YFS); while both sources of information are broadly valid,29–35 details may be missed, particularly for HDP. The assignment to classes is based on probabilities and may not reflect an underlying biological reality.
Interpretation
The two studies had relatively similar designs but different populations. Both included longitudinal follow-up of participants who were children and young adults primarily in the 1980s. The BHS only recruited from a single site, a small town in Louisiana, and includes Black and white participants. YFS recruited from five sites around Finland, centered in urban centers and recruiting from the surrounding area, and includes people of Finnish ethnicity. For the most part, results are consistent in the two cohorts and population differences in baseline measures were not substantial (Table 2; for a comparison of the full cohorts, see 14) In general, health status of the Finnish population has improved more strongly than that of United States over the last forty years, although current population mean levels of most cardiovascular risk factors are not far apart.36, 37
These results are consistent with some previous analyses, in these cohorts and others, including finding increasing BMI associated with GDM11and that overweight/prehypertensive groups were more likely to develop HDP. 10 GDM has been associated with higher annual rate of change in triglycerides and greater decrease in HDL;12 the analysis in YFS found neither of these patterns and in BHS was consistent with the HDL result, though imprecise. In BHS, when childhood and pre-pregnancy was considered as a whole, mean and area under the curve pre-pregnancy blood pressure were associated with PTB and term LBW, and pre-pregnancy total cholesterol was negatively associated with gestational age.13 While other studies have similarly found higher preconception blood pressure was associated with lower birthweight and higher preterm birth,9, 38 previous research on lipids is more ambiguous. Our previous analysis of preconception cardiovascular risk factors in the YFS indicated increased risk of adverse birth outcomes with higher blood pressure and lipids (effect sizes between 1.2 and 1.4 for a 1-SD increase),7 and SBP and total cholesterol were somewhat higher at baseline in the YF compared to BHS. Romundstad et al. found higher preconception levels of cholesterol and triglycerides and lower HDL were associated with higher birthweight,9 while Catov et al. found that both low and high total cholesterol were associated with increased preterm birth risk,39 but not triglycerides, LDL, or HDL. So, while our results are broadly consistent with these studies, we found more associations with LDL.
Conclusions
In summary, life course trajectories of cardiovascular risk are associated with higher risk of pregnancy complications. This analysis contributes to the greater understanding of pregnancy in the life course: it has become increasingly clear that pregnancy is not an isolated event, but intimately intertwined with health before and after pregnancy, and may require interventions beyond the nine months of pregnancy to improve population levels of pregnancy-related morbidity. Future studies should continue to examine cardiovascular health across the life course and attempt to determine groups and inflection points for intervention.
Supplementary Material
Synopsis.
Study Question: How does pre-pregnancy trajectory of cardiovascular health predict pregnancy complications and birth outcomes?
What’s already known: Single pre-pregnancy measures of cardiovascular health have been associated with pregnancy outcomes.
What this study adds: In two cohorts with diverse populations, the steepest increase of blood pressure and cholesterol were also associated with preterm birth. Steeper increases in BMI was also associated with gestational diabetes in both cohorts.
ii. funding statement:
The Young Finns Study has been financially supported by the Academy of Finland: grants 322098, 286284, 134309 (Eye), 126925, 121584, 124282, 255381, 256474, 283115, 319060, 320297, 314389, 338395, 330809, and 104821, 129378 (Salve), 117797 (Gendi), and 141071 (Skidi); the Social Insurance Institution of Finland; Competitive State Research Financing of the Expert Responsibility area of Kuopio, Tampere and Turku University Hospitals (grant X51001); Juho Vainio Foundation; Paavo Nurmi Foundation; Finnish Foundation for Cardiovascular Research ; Finnish Cultural Foundation; The Sigrid Juselius Foundation; Tampere Tuberculosis Foundation; Emil Aaltonen Foundation; Yrjö Jahnsson Foundation; Signe and Ane Gyllenberg Foundation; Diabetes Research Foundation of Finnish Diabetes Association; EU Horizon 2020 (grant 755320 for TAXINOMISIS and grant 848146 for To Aition); European Research Council (grant 742927 for MULTIEPIGEN project); Tampere University Hospital Supporting Foundation, Finnish Society of Clinical Chemistry and the Cancer Foundation Finland.
The Bogalusa Heart Study has been financially supported by the National Institutes of Health grant R01HD069587, R01AG016592, R01AG041200, P50HL015103, and R01HD032194. This analysis was also supported by the Fulbright Finland Foundation.
Footnotes
conflict of interest disclosure: none of the authors has a conflict of interest
ethics approval statement: The Bogalusa Heart Study and Bogalusa Babies studies were approved by the Tulane University Institutional Review Board. The Young Finns study was approved by the local ethical committees (1st Ethical Committee of the Hospital District of Southwest Finland, Regional Ethics Committee of the Expert Responsibility area of Tampere University Hospital, Helsinki University Hospital Ethical Committee of Medicine, The Research Ethics Committee of the Northern Savo Hospital District and Ethics Committee of the Northern Ostrobothnia Hospital District) of the participating sites (ETMK: 88/180/2010).
patient consent statement: NA
permission to reproduce material from other sources: NA
i. data availability statement:
The datasets generated and/or analysed during the current study are not publicly available. For BHS, data are confidential to protect participant privacy. However, data are available upon request from BHS Steering Committee. Data requests can be submitted by email and are subject to approval. For YF, due to legal restrictions, the Ethics Committee of the Hospital District of Southwest Finland has in 2016 stated that individual level data cannot be stored in public repositories or otherwise made publicly available. However, data are available upon request from YFS Project Application. Data requests can be submitted by email and are subject to approval by the YFS Board.
References
- 1.Jacobs DR, Jr.,Woo JG, Sinaiko AR, Daniels SR, Ikonen J, Juonala M, et al. Childhood Cardiovascular Risk Factors and Adult Cardiovascular Events. N Engl J Med 2022; 386:1877–1888. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Theodore RF, Broadbent J, Nagin D, Ambler A, Hogan S, Ramrakha S, et al. Childhood to Early-Midlife Systolic Blood Pressure Trajectories: Early-Life Predictors, Effect Modifiers, and Adult Cardiovascular Outcomes. Hypertension 2015; 66:1108–1115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Munthali RJ, Kagura J, Lombard Z, Norris SA. Childhood adiposity trajectories are associated with late adolescent blood pressure: birth to twenty cohort. BMC Public Health 2016; 16:665. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Pollock BD, Stuchlik P, Harville EW, Mills KT, Tang W, Chen W, et al. Life course trajectories of cardiovascular risk: Impact on atherosclerotic and metabolic indicators. Atherosclerosis 2019; 280:21–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Fleming TP, Watkins AJ, Velazquez MA, Mathers JC, Prentice AM, Stephenson J, et al. Origins of lifetime health around the time of conception: causes and consequences. Lancet 2018. [DOI] [PMC free article] [PubMed]
- 6.Catov JM, Ness RB, Wellons MF, Jacobs DR, Roberts JM, Gunderson EP. Prepregnancy Lipids Related to Preterm Birth Risk: the Coronary Artery Risk Development in Young Adults Study. J Clin Endocrinol Metab 2010; 95:3711–3718. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Harville EW, Viikari JS, Raitakari OT. Preconception cardiovascular risk factors and pregnancy outcome. Epidemiology 2011; 22:724–730. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Harville EW, Juonala M, Viikari JS, Raitakari OT. Preconception metabolic indicators predict gestational diabetes and offspring birthweight. Gynecol Endocrinol 2014:1–5. [DOI] [PubMed]
- 9.Romundstad PR, Davey Smith G, Nilsen TI, Vatten LJ. Associations of prepregnancy cardiovascular risk factors with the offspring’s birth weight. Am J Epidemiol 2007; 166:1359–1364. [DOI] [PubMed] [Google Scholar]
- 10.Lane-Cordova AD, Tedla YG, Carnethon MR, Montag SE, Dude AM, Rasmussen-Torvik LJ. Pre-pregnancy blood pressure and body mass index trajectories and incident hypertensive disorders of pregnancy. Pregnancy Hypertens 2018; 13:138–140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Adane AA, Mishra GD, Tooth LR. Adult Pre-pregnancy Weight Change and Risk of Developing Hypertensive Disorders in Pregnancy. Paediatr Perinat Epidemiol 2017; 31:167–175. [DOI] [PubMed] [Google Scholar]
- 12.Retnakaran R, Shah BR. Impact of pregnancy on the trajectories of cardiovascular risk factors in women with and without gestational diabetes. Diabetes Obes Metab 2021. [DOI] [PubMed]
- 13.Harville EW, Wallace ME, He H, Bazzano LA. Lifetime cardiovascular risk factors and maternal and offspring birth outcomes: Bogalusa Babies. PLoS One 2022; 17:e0260703. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sinaiko AR, Jacobs DR, Jr., Woo JG, Bazzano L, Burns T, Hu T, et al. The International Childhood Cardiovascular Cohort (i3C) consortium outcomes study of childhood cardiovascular risk factors and adult cardiovascular morbidity and mortality: Design and recruitment. Contemp Clin Trials 2018. [DOI] [PMC free article] [PubMed]
- 15.Berenson GS. Bogalusa Heart Study: a long-term community study of a rural biracial (Black/White) population. Am J Med Sci 2001; 322:293–300. [PubMed] [Google Scholar]
- 16.Harville EW, Jacobs M, Shu T, Breckner D, Wallace M. Feasibility of Linking Long-Term Cardiovascular Cohort Data to Offspring Birth Records: The Bogalusa Heart Study. Maternal and child health journal 2018. [DOI] [PMC free article] [PubMed]
- 17.Raitakari OT, Juonala M, Ronnemaa T, Keltikangas-Jarvinen L, Rasanen L, Pietikainen M, et al. Cohort profile: The Cardiovascular Risk in Young Finns Study. Int J Epidemiol 2008; 37:1220–1226. [DOI] [PubMed] [Google Scholar]
- 18.Porkka KV, Raitakari OT, Leino A, Laitinen S, Räsänen L, Rönnemaa T, et al. Trends in serum lipid levels during 1980–1992 in children and young adults. The Cardiovascular Risk in Young Finns Study. Am J Epidemiol 1997; 146:64–77. [DOI] [PubMed] [Google Scholar]
- 19.Raiko JR, Viikari JS, Ilmanen A, Hutri-Kahonen N, Taittonen L, Jokinen E, et al. Follow-ups of the Cardiovascular Risk in Young Finns Study in 2001 and 2007: levels and 6-year changes in risk factors. J Intern Med 2010; 267:370–384. [DOI] [PubMed] [Google Scholar]
- 20.Harville EW, Jacobs M, Shu T, Breckner D, Wallace M. Comparison of reproductive history gathered by interview and by vital records linkage after 40 years of follow-up: Bogalusa Babies. BMC Med Res Methodol 2019; 19:114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Martin JA, Wilson EC, Osterman MJ, Saadi EW, Sutton SR, Hamilton BE. Assessing the quality of medical and health data from the 2003 birth certificate revision: results from two states. Natl Vital Stat Rep 2013; 62:1–19. [PubMed] [Google Scholar]
- 22.Yang Z, Zhou M. Kappa statistic for clustered matched-pair data. Stat Med 2014; 33:2612–2633. [DOI] [PubMed] [Google Scholar]
- 23.Carter EB, Stuart JJ, Farland LV, Rich-Edwards JW, Zera CA, McElrath TF, et al. Pregnancy Complications as Markers for Subsequent Maternal Cardiovascular Disease: Validation of a Maternal Recall Questionnaire. J Womens Health (Larchmt) 2015; 24:702–712. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hakala JO, Pahkala K, Juonala M, Salo P, Kähönen M, Hutri-Kähönen N, et al. Cardiovascular Risk Factor Trajectories Since Childhood and Cognitive Performance in Midlife: The Cardiovascular Risk in Young Finns Study. Circulation 2021; 143:1949–1961. [DOI] [PubMed] [Google Scholar]
- 25.Spiegelman D, Hertzmark E. Easy SAS calculations for risk or prevalence ratios and differences. Am J Epidemiol 2005; 162:199–200. [DOI] [PubMed] [Google Scholar]
- 26.Cook NR, Rosner BA, Chen W, Srinivasan SR, Berenson GS. Using the area under the curve to reduce measurement error in predicting young adult blood pressure from childhood measures. Stat Med 2004; 23:3421–3435. [DOI] [PubMed] [Google Scholar]
- 27.Retnakaran R, Shah BR. Divergent Trajectories of Cardiovascular Risk Factors in the Years Before Pregnancy in Women With and Without Gestational Diabetes Mellitus: A Population-Based Study. Diabetes Care 2020; 43:2500–2508. [DOI] [PubMed] [Google Scholar]
- 28.Catov JM, Sun B, Bertolet M, Snyder GG, Lewis CE, Allen NB, et al. Changes in Cardiometabolic Risk Factors Before and After Gestational Diabetes: A Prospective Life-Course Analysis in CARDIA Women. Obesity (Silver Spring, Md) 2020; 28:1397–1404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Wise LA, Wang TR, Wesselink AK, Willis SK, Chaiyasarikul A, Levinson JS, et al. Accuracy of self-reported birth outcomes relative to birth certificate data in an Internet-based prospective cohort study. Paediatr Perinat Epidemiol 2021; 35:590–595. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Simard JF, Rossides M, Wikström AK, Falasinnu T, Palmsten K, Arkema EV. Evidence of under-reporting of early-onset preeclampsia using register data. Paediatr Perinat Epidemiol 2021; 35:596–600. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Moth FN, Sebastian TR, Horn J, Rich-Edwards J, Romundstad PR, Asvold BO. Validity of a selection of pregnancy complications in the Medical Birth Registry of Norway. Acta Obstet Gynecol Scand 2016; 95:519–527. [DOI] [PubMed] [Google Scholar]
- 32.Gresham E, Forder P, Chojenta CL, Byles JE, Loxton DJ, Hure AJ. Agreement between self-reported perinatal outcomes and administrative data in New South Wales, Australia. BMC Pregnancy Childbirth 2015; 15:161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Bat-Erdene U, Metcalfe A, McDonald SW, Tough SC. Validation of Canadian mothers’ recall of events in labour and delivery with electronic health records. BMC Pregnancy Childbirth 2013; 13 Suppl 1:S3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Troude P, L’Helias LF, Raison-Boulley AM, Castel C, Pichon C, Bouyer J, et al. Perinatal factors reported by mothers: do they agree with medical records? Eur J Epidemiol 2008; 23:557–564. [DOI] [PubMed] [Google Scholar]
- 35.Falkegard M, Schirmer H, Lochen ML, Oian P, Acharya G. The validity of self-reported information about hypertensive disorders of pregnancy in a population-based survey: the Tromso Study. Acta Obstet Gynecol Scand 2015; 94:28–34. [DOI] [PubMed] [Google Scholar]
- 36.Borodulin K, Vartiainen E, Peltonen M, Jousilahti P, Juolevi A, Laatikainen T, et al. Forty-year trends in cardiovascular risk factors in Finland. Eur J Public Health 2015; 25:539–546. [DOI] [PubMed] [Google Scholar]
- 37.He J, Zhu Z, Bundy JD, Dorans KS, Chen J, Hamm LL. Trends in Cardiovascular Risk Factors in US Adults by Race and Ethnicity and Socioeconomic Status, 1999–2018. JAMA 2021; 326:1286–1298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Yang Y, He Y, Li Q, Wang Y, Peng Z, Xu J, et al. Preconception blood pressure and risk of preterm birth: a large historical cohort study in a Chinese rural population. Fertil Steril 2015; 104:124–130. [DOI] [PubMed] [Google Scholar]
- 39.Catov JM, Ness RB, Wellons MF, Jacobs DR, Roberts JM, Gunderson EP. Prepregnancy Lipids Related to Preterm Birth Risk: the Coronary Artery Risk Development in Young Adults Study. J Clin Endocrinol Metab 2010. [DOI] [PMC free article] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated and/or analysed during the current study are not publicly available. For BHS, data are confidential to protect participant privacy. However, data are available upon request from BHS Steering Committee. Data requests can be submitted by email and are subject to approval. For YF, due to legal restrictions, the Ethics Committee of the Hospital District of Southwest Finland has in 2016 stated that individual level data cannot be stored in public repositories or otherwise made publicly available. However, data are available upon request from YFS Project Application. Data requests can be submitted by email and are subject to approval by the YFS Board.