Prediction models that include longitudinal patterns of blood pressure (BP) changes in early pregnancy with standard clinical risk factors can accurately determine the risk for hypertensive disorders of pregnancy (HDP) in patients considered to have low-to-moderate risk pregnancies. These findings were published in the Journal of the American Heart Association.
The retrospective cohort study sought to create and validate predictive models using early pregnancy BP trajectory patterns before 20 weeks’ gestation.
Data were obtained for BP measurements; prepregnancy information; sociodemographic, clinical, social, and lifestyle risk factors; and perinatal outcomes for pregnancies longer than 20 weeks delivered from January 1, 2009, to December 31, 2019, from the Kaiser Permanente Northern California (KPNC) system.
The first index singleton live or still birth with BP measurements was selected for each participant. Of 249,892 eligible pregnant women, 30% were in the internal validation data set and 70% were included in the development data set .
Early-onset preeclampsia was classified as a diagnosis at 20 to less than 34 weeks’ gestation, later-onset preeclampsia was classified as a diagnosis at 34 weeks’ gestation or longer, and gestational hypertension was classified as a diagnosis after 20 weeks’ gestation. Patients without a HDP were the reference group.
The investigators identified 6 early pregnancy systolic BP trajectory groups, which were ordered based on increasing risk of HDP (ultra-low declining to elevated stable). The predictive models were fit with different combinations of variables included sequentially to quantify and assess the predictive value of the BP trajectory groups and other risk factors.
The mean (SD) maternal age at delivery was 30.9 years (5.3 years), and the mean gestational age at the start of prenatal care was 8.2 (2.0) weeks (range, 0-14 weeks).
For the development data set, 700 (0.4%), 7571 (4.3%), and 7981 (4.6%) of pregnancies had early-onset preeclampsia, later-onset preeclampsia, or gestational hypertension after 20 weeks’ gestation, respectively. The validation set had similar corresponding outcomes, which were 308 (0.4%), 3195 (4.3%), and 3533 (4.7%), respectively.
Multivariable, multinominal logistic regression models in the development dataset showed that the BP trajectory groups were independently associated with an increasing gradient of adjusted odds ratios (aORs) for early-onset preeclampsia, later-onset preeclampsia, and gestational hypertension.
The best prediction model included the 6 BP trajectory groups and all standard clinical risk factors (C-statistics) for early-onset preeclampsia (0.747; 95% CI, 0.719-0.775), later-onset preeclampsia (0.731; 95% CI, 0.723-0.740), and gestational hypertension (0.770; 95% CI, 0.762-0.778; all P <.001).
The models with the BP trajectory groups plus all risk factors had excellent calibration performance, and no evidence of poor fit based on Hosmer-Lemeshow was found (P =.99, .99, and .74 for early-onset preeclampsia, later-onset preeclampsia, and gestational hypertension, respectively).
A limitation of this study is that the researchers could not evaluate individual-level social determinants of health that may affect risk differences or identify individuals with previous gestational hypertension or those treated with artificial reproductive technology. Also, preeclampsia may have been underestimated for births before KPNC health plan membership, and the models were not assessed in an external validation population.
“Classification of early pregnancy systolic BP patterns based on BP changes from 0 through 16 to 20 weeks’ gestation in combination with other standard risk factors (clinical, social, and behavioral) can significantly improve individual risk stratification for early-onset and later-onset preeclampsia and gestational hypertension, allowing more targeted surveillance and potentially interventions to ameliorate hypertensive disorders of pregnancy and adverse outcomes, as well as avoidance of additional monitoring, or unnecessary interventions (ie, low-dose aspirin administration) in low-risk pregnancies,” wrote the study authors.
Gunderson EP, Greenberg M, Sun B, et al. Early pregnancy systolic blood pressure patterns predict early- and later-onset preeclampsia and gestational hypertension among ostensibly low-to-moderate risk groups. J Am Heart Assoc. Published online July 12, 2023.doi: 10.1161/JAHA.123.029617