A Bayesian latent variable mixture model for longitudinal fetal growth

An application of group-based trajectory modeling to define fetal growth phenotypes among small-for-gestational-age births in the LIFECODES Fetal Growth Study

BACKGROUND

Reductions in fetal growth are associated with adverse outcomes at birth and later in life. However, identifying fetuses with pathologically small growth remains challenging. Definitions of small-for-gestational age are often used as a proxy to identify those experiencing pathologic growth (ie, fetal growth restriction). However, this approach is subject to limitation as most newborns labeled small-for-gestational age are constitutionally, not pathologically, small. Incorporating repeated ultrasound measures to examine fetal growth trajectories may help distinguish pathologic deviations in growth from normal variability, beyond a simple definition of small-for-gestational age.

OBJECTIVE

This study aimed to characterize phenotypes of growth using ultrasound trajectories of fetal growth among small-for-gestational-age births.

STUDY DESIGN

This study identified and described trajectories of fetal growth among small-for-gestational-age births (<10th percentile weight for gestational age; n=245) in the LIFECODES Fetal Growth Study using univariate and multivariate trajectory modeling approaches. Available ultrasound measures of fetal growth (estimated fetal weight, head circumference, abdominal circumference, and femur length) from health records were abstracted. First, univariate group-based trajectory modeling was used to define trajectories of estimated fetal weight z scores during gestation. Second, group-based multi-trajectory modeling was used to identify trajectories based on concurrent measures of head circumference, abdominal circumference, and femur length z scores. Last, how these trajectories were related to patient demographics, pregnancy characteristics, and birth outcomes compared with those observed among appropriate-for-gestational-age controls was described.

RESULTS

Of note, 3 univariate trajectories of estimated fetal weight and 4 multivariate trajectories of fetal growth among small-for-gestational-age births were identified. In our univariate approach, infants with the smallest estimated fetal weight trajectory throughout pregnancy had poorer outcomes, including the highest risk of neonatal intensive care unit admission. The remaining univariate trajectory groups did not have an elevated risk of adverse birth outcomes relative to appropriate-for-gestational-age controls. In our multivariate approach, 2 groups at increased or moderately increased risk of neonatal intensive care unit admission were identified, including infants that remained extremely small for all parameters throughout pregnancy and those who had disproportionately smaller femur length and abdominal circumference compared with head circumference. The remaining multivariate trajectory groups did not have an elevated risk of adverse birth outcome relative to appropriate-for-gestational-age controls.

CONCLUSION

Latent class group-based trajectory modeling applied to ultrasound measures of fetal growth may help distinguish pathologic vs constitutional growth profiles among newborns born small-for-gestational age. Although trajectories cannot be fully characterized until delivery, limiting the direct clinical application of these methods, they may still contribute to the development of approaches for separating growth restriction from constitutional smallness.

Fetal growth trajectories of babies born large-for-gestational age in the LIFECODES Fetal Growth Study

BACKGROUND

Babies born large-for-gestational age have an increased risk of adverse health outcomes, including birth injuries, childhood obesity, and cardiometabolic disorders. However, little work has been done to characterize patterns of fetal growth among large-for-gestational age births, which may further elucidate high- and low-risk subgroups.

OBJECTIVE

This study aimed to identify subgroups of large-for-gestational age births based on trajectories of fetal growth derived from prenatal ultrasound measurements and explore differences in sociodemographic, pregnancy, and birth outcome characteristics across subgroups.

STUDY DESIGN

This study identified and described trajectories of fetal growth among large-for-gestational age births (n=235) in the LIFECODES Fetal Growth Study. Ultrasound measurements of fetal growth in middle to late pregnancy were abstracted from health records. Group-based multi-trajectory modeling was applied to measurements of head circumference, abdominal circumference, and femur length z-scores to identify multivariate trajectories of fetal growth. Moreover, sociodemographic variables, pregnancy characteristics, and birth outcomes based on trajectory membership were summarized.

RESULTS

This study identified 4 multivariate trajectories of fetal growth among large-for-gestational age births: catch-up growth (n=28), proportional abdominal circumference-to-femur length growth (n=67), disproportional abdominal circumference-to-femur length growth (n=96), and consistently large (n=44). Fetuses in the "catch-up growth" group exhibited small relative sizes in midpregnancy (ie, below average head circumference, abdominal circumference, and femur length z-scores) and large relative sizes in late pregnancy. Growth among these births was driven by increases in relative abdominal circumference and head circumference sizes. Participants who delivered births assigned to this group were less likely to have normal glucose control (40% vs 65%-75%) and more likely to have pregestational diabetes mellitus (36% vs 10%-17%) than other large-for-gestational age subgroups. In addition, the babies in this trajectory group were more likely to have macrosomia (86% vs 67%-73%) and to be admitted to the neonatal intensive care unit (32% vs 14%-21%) than other large-for-gestational age subgroups. In contrast, babies in the "consistently large" group had the largest relative size for all growth parameters throughout gestation and experienced a lower risk of adverse birth outcomes than other large-for-gestational age subgroups.

CONCLUSION

This study characterized several trajectories of fetal growth among large-for-gestational age births, which were related to different pregnancy characteristics and the distribution of adverse birth outcomes. Although the number of individuals within some trajectories was small, a subgroup that exhibited a catch-up growth phenotype during gestation was identified, which may be uniquely associated with exposure to pregestational diabetes mellitus and a higher risk of admission to the neonatal intensive care unit. These results have highlighted that the risk of adverse outcomes may not be evenly distributed across all large-for-gestational age births.

A Bayesian latent class model for predicting gestational age in health administrative data

Health administrative data are oftentimes of limited use in epidemiological study on drug safety in pregnancy, due to lacking information on gestational age at birth (GAB). Although several studies have proposed algorithms to estimate GAB using claims database, failing to incorporate the unique distributional shape of GAB, can introduce bias in estimates and subsequent modeling. Hence, we develop a Bayesian latent class model to predict GAB. The model employs a mixture of Gaussian distributions with linear covariates within each class. This approach allows modeling heterogeneity in the population by identifying latent subgroups and estimating class-specific regression coefficients. We fit this model in a Bayesian framework conducting posterior computation with Markov Chain Monte Carlo methods. The method is illustrated with a dataset of 10,043 Rhode Island Medicaid mother-child pairs. We found that the three-class and six-class mixture specifications maximized prediction accuracy. Based on our results, Medicaid women were partitioned into three classes, featured by extreme preterm or preterm birth, preterm or" early" term birth, and" late" term birth. Obstetrical complications appeared to pose a significant influence on class-membership. Altogether, compared to traditional linear models our approach shows an advantage in predictive accuracy, because of superior flexibility in modeling a skewed response and population heterogeneity.

Bayesian consensus clustering for multivariate longitudinal data

In clinical and epidemiological studies, there is a growing interest in studying the heterogeneity among patients based on longitudinal characteristics to identify subtypes of the study population. Compared to clustering a single longitudinal marker, simultaneously clustering multiple longitudinal markers allow additional information to be incorporated into the clustering process, which reveals co-existing longitudinal patterns and generates deeper biological insight. In the current study, we propose a Bayesian consensus clustering (BCC) model for multivariate longitudinal data. Instead of arriving at a single overall clustering, the proposed model allows each marker to follow marker-specific local clustering and these local clusterings are aggregated to find a global (consensus) clustering. To estimate the posterior distribution of model parameters, a Gibbs sampling algorithm is proposed. We apply our proposed model to the primary biliary cirrhosis study to identify patient subtypes that may be associated with their prognosis. We also perform simulation studies to compare the clustering performance between the proposed model and existing models under several scenarios. The results demonstrate that the proposed BCC model serves as a useful tool for clustering multivariate longitudinal data.

Fetal Growth Trajectories Among Small for Gestational Age Babies and Child Neurodevelopment

BACKGROUND

Being born small for gestational age (SGA, <10th percentile) is a risk factor for worse neurodevelopmental outcomes. However, this group is a heterogeneous mix of healthy and growth-restricted babies, and not all will experience poor outcomes. We sought to determine whether fetal growth trajectories can distinguish who will have the worst neurodevelopmental outcomes in childhood among babies born SGA.

METHODS

The present analysis was conducted in Generation R, a population-based cohort in Rotterdam, the Netherlands (N = 5,487). Using group-based trajectory modeling, we identified fetal growth trajectories for weight among babies born SGA. These were based on standard deviation scores of ultrasound measures from mid-pregnancy and late pregnancy in combination with birth weight. We compared child nonverbal intelligence quotient (IQ) and attention deficit hyperactivity disorder (ADHD) symptoms at age 6 between SGA babies within each growth trajectory to babies born non-SGA.

RESULTS

Among SGA individuals (n = 656), we identified three distinct fetal growth trajectories for weight. Children who were consistently small from mid-pregnancy (n = 64) had the lowest IQ (7 points lower compared to non-SGA babies, 95% confidence interval [CI] = -11.0, -3.5) and slightly more ADHD symptoms. Children from the trajectory that started larger but were smaller at birth showed no differences in outcomes compared to children born non-SGA.

CONCLUSIONS

Among SGA children, those who were smaller beginning in mid-pregnancy exhibited the worst neurodevelopmental outcomes at age 6. Fetal growth trajectories may help identify SGA babies who go on to have poor neurodevelopmental outcomes.

Shape invariant mixture model for clustering non-linear longitudinal growth trajectories

In longitudinal studies, it is often of great interest to cluster individual trajectories based on repeated measurements taken over time. Non-linear growth trajectories are often seen in practice, and the individual data can also be measured sparsely, and at irregular time points, which may complicate the modeling process. Motivated by a study of pregnant women hormone profiles, we proposed a shape invariant growth mixture model for clustering non-linear growth trajectories. Bayesian inference via Monte Carlo Markov Chain was employed to estimate the parameters of interest. We compared our model to the commonly used growth mixture model and functional clustering approach by simulation studies. Results from analyzing the real data and simulated data were presented and discussed.

A class of joint models for multivariate longitudinal measurements and a binary event

Predicting binary events such as newborns with large birthweight is important for obstetricians in their attempt to reduce both maternal and fetal morbidity and mortality. Such predictions have been a challenge in obstetric practice, where longitudinal ultrasound measurements taken at multiple gestational times during pregnancy may be useful for predicting various poor pregnancy outcomes. The focus of this article is on developing a flexible class of joint models for the multivariate longitudinal ultrasound measurements that can be used for predicting a binary event at birth. A skewed multivariate random effects model is proposed for the ultrasound measurements, and the skewed generalized t-link is assumed for the link function relating the binary event and the underlying longitudinal processes. We consider a shared random effect to link the two processes together. Markov chain Monte Carlo sampling is used to carry out Bayesian posterior computation. Several variations of the proposed model are considered and compared via the deviance information criterion, the logarithm of pseudomarginal likelihood, and with a training-test set prediction paradigm. The proposed methodology is illustrated with data from the NICHD Successive Small-for-Gestational-Age Births study, a large prospective fetal growth cohort conducted in Norway and Sweden.

The effect of preterm birth on infant negative affect and maternal postpartum depressive symptoms: A preliminary examination in an underrepresented minority sample

OBJECTIVE

To examine the effect of preterm birth on maternal postpartum depressive symptoms and infant negative affect in an underrepresented minority sample.

METHOD

Participants were 102 mothers and their 3- to 10-month-old infants. Mothers completed the Edinburgh Postnatal Depression Scale and the Infant Behavior Questionnaire-Revised.

RESULTS

Relative to normative samples, the current underrepresented minority sample of mostly Hispanics and Blacks displayed high rates of preterm birth (30%) and maternal postpartum depressive symptoms (17%). Preterm birth had a significant direct effect on postpartum depressive symptoms and infant negative affect. Additionally, there was an indirect effect of postpartum depressive symptoms on the relation between preterm birth and infant negative affect. Specifically, lower birth weight and gestational age predicted higher levels of depressive symptoms in the mother, and higher levels of depressive symptoms in the mother, in turn, predicted higher levels of infant negative affect.

CONCLUSION

Findings emphasize the importance of screening for postpartum depressive symptoms and infant negative affect among mothers and their preterm infants, especially among families from underrepresented minority backgrounds.

A linear mixed model for predicting a binary event from longitudinal data under random effects misspecification

The use of longitudinal data for predicting a subsequent binary event is often the focus of diagnostic studies. This is particularly important in obstetrics, where ultrasound measurements taken during fetal development may be useful for predicting various poor pregnancy outcomes. We propose a modeling framework for predicting a binary event from longitudinal measurements where a shared random effect links the two processes together. Under a Gaussian random effects assumption, the approach is simple to implement with standard statistical software. Using asymptotic and simulation results, we show that estimates of predictive accuracy under a Gaussian random effects distribution are robust to severe misspecification of this distribution. However, under some circumstances, estimates of individual risk may be sensitive to severe random effects misspecification. We illustrate the methodology with data from a longitudinal fetal growth study.

A Bayesian growth mixture model to examine maternal hypertension and birth outcomes

Maternal hypertension is a major contributor to adverse pregnancy outcomes, including preterm birth (PTB) and low birth weight (LBW). Although several studies have explored the relationship between maternal hypertension and fetal health, few have examined how the longitudinal trajectory of blood pressure, considered over the course of pregnancy, affects birth outcomes. In this paper, we propose a Bayesian growth mixture model to jointly examine the associations between longitudinal blood pressure measurements, PTB, and LBW. The model partitions women into distinct classes characterized by a mean arterial pressure (MAP) curve and joint probabilities of PTB and LBW. Each class contains a unique mixed effects model for MAP with class-specific regression coefficients and random effect covariances. To account for the strong correlation between PTB and LBW, we introduce a bivariate probit model within each class to capture residual within-class dependence between PTB and LBW. The model permits the association between PTB and LBW to vary by class, so that for some classes, PTB and LBW may be positively correlated, whereas for others, they may be uncorrelated or negatively correlated. We also allow maternal covariates to influence the class probabilities via a multinomial logit model. For posterior computation, we propose an efficient MCMC algorithm that combines full-conditional Gibbs and Metropolis steps. We apply our model to a sample of 1027 women enrolled in the Healthy Pregnancy, Healthy Baby Study, a prospective cohort study of host, social, and environmental contributors to disparities in pregnancy outcomes.