Latent class analysis to characterize neonatal risk for neurodevelopmental differences

Abstract

Background: Neonatal risk factors, such as preterm birth and low birth weight, have been robustly linked to neurodevelopmental deficits, yet it is still unclear why some infants born preterm and/or low birth weight experience neurodevelopmental difficulties while others do not. The current study investigated this heterogeneity in neurodevelopmental abilities by examining additional neonatal morbidities as risk factors, utilizing latent class analysis to classify neonates into groups based on similar neonatal risk factors, and including neonates from the full spectrum of gestational age. Methods: Neonates who received neonatal care at an academic public hospital during an almost 10year period (n = 19,951) were included in the latent class analysis, and 21 neonatal indicators of health were used. Neonatal class, sex, and the interaction between neonatal class and sex were used to examine differences in neurodevelopment at 18 months of age in a typically developing population. Results: The best fitting model included five infant classes: healthy, hypoxic, critically ill, minorly ill, and complicated delivery. Scores on the parent-rated neurodevelopmental measure differed by class such that infants in the critically ill, minorly ill, and complicated delivery classes had lower scores. There was no main effect of sex on the neurodevelopmental measure scores, but the interaction between sex and neonatal class was significant for three out of five neurodevelopmental domains. Conclusions: The current study extends the understanding of risk factors in neurodevelopment by including several neonatal medical conditions that are often overlooked and by using a person-centered, as opposed to variablecentered, approach. Future work should continue to examine risk factors, such as maternal health during pregnancy and medical interventions for newborns, in relation to neonatal risks and neurodevelopment by using a personcentered approach.

Profiles of Autistic Youth with and Without Co-occurring Behavioral Health and Neurodevelopmental Disorders: A Latent Class Analysis

“…Data were analyzed using latent class mixture modeling to investigate population heterogeneity on a latent trait. 14,19 Participants were classified into distinct latent groups (termed classes) that best represent the observed response patterns in the data. 20 Entropy represents a weighted average of individuals' posterior probabilities of membership 21 ; despite their appeal to distinguish classes, entropy should not be part of the class enumeration process.…”

Editorial: 'Safety in numbers'? Big data discovery strategies in neuro‐developmental science – contributions and caveats

“…Indeed, in the current issue two such data-driven approaches to this are described. Momany, Jasper, Markon, Nikolas, and Ryckman (2023) When considering the relative strengths of deductive and inductive approaches we must be careful not to fall into the trap of seeing them as in competition. In fact, their respective value comes from their complementary role in the process of scientific advancement and the generation of new knowledge.…”

“…Indeed, in the current issue two such data‐driven approaches to this are described. Momany, Jasper, Markon, Nikolas, and Ryckman (2023) used a person‐centered approach (latent class analysis) to parse heterogeneity in neonatal risk factors to create a five‐class model (healthy, hypoxic, critically ill, minorly ill, and complicated delivery) that could predict which babies (those in the last three classes) are at risk for poor developmental outcomes at age 18 months. Using a different approach, Lorenzi et al.…”

Utilising behavioural and sensory profiles and associated perinatal factors to identify meaningful subgroups in autism spectrum disorder.

“…Further, a study byMomany et al (2023) based on neonatal latent class analysis with variables including birthweight, gestational age, and the diagnostic status of common neonatal morbidities followed by analysis of covariance to examine eighteen-month neurodevelopmental scores by latent class identi ed 5 subgroups(Momany et al, 2023). They included complicated delivery, minor illness, and critically ill classes which attained lower neurodevelopmental scores compared to the healthy class, analogous to the pattern suggested by Class 2(Momany et al, 2023). A notable difference in methodology is that Momany et al (2023) used eighteen-month neurodevelopmental scores, whereas in the current study, age at the time of behavioural and sensory scoring was not accounted for.…”

“…Montogomery et al (2023) performed a latent pro le analysis of variables re ecting core autism traits, psychiatric and medical comorbidity on a sample of 754 children with autism and identi ed a subgroup of individuals with signi cant social communication and cognitive di culties as well as increased sensory seeking behaviours(Montgomery et al, 2023). Further, a study byMomany et al (2023) based on neonatal latent class analysis with variables including birthweight, gestational age, and the diagnostic status of common neonatal morbidities followed by analysis of covariance to examine eighteen-month neurodevelopmental scores by latent class identi ed 5 subgroups(Momany et al, 2023). They included complicated delivery, minor illness, and critically ill classes which attained lower neurodevelopmental scores compared to the healthy class, analogous to the pattern suggested by Class 2(Momany et al, 2023).…”