Vision loss is a major and growing concern in the aging population of the United States. Approximately 37 million adults aged 50 and older experience some degree of vision impairment, including one in four adults aged 80 and older. Poor eyesight is not merely an inconvenience; it is closely linked to reduced quality of life, depression, falls, loss of independence, and difficulty performing everyday tasks such as reading medication labels, driving, and managing finances. A substantial body of research shows that social conditions across the life course—from childhood poverty to employment and access to preventive care—shape the risk of both acute vision loss and chronic visual impairment. Yet until now, almost nothing was known about how incarceration, one of the most consequential life events experienced by millions of Americans, shapes vision over time.

The scale of the aging prison population gives the question particular urgency. Correctional systems typically define older incarcerated adults as those over approximately age 50, a threshold set far lower than in the general population because incarcerated individuals display health profiles comparable to non-incarcerated adults 10 to 15 years older. Between 1999 and 2016, the number of adults aged 55 and older in state and federal prisons increased by a staggering 280 percent, while the count of prisoners under 55 grew by just 3 percent over the same period. Outside prison walls, an estimated one in 15 American adults over age 50 has been previously incarcerated. These older adults face elevated rates of chronic disease, sensory and functional impairment, and accelerated biological aging, yet jails, prisons, and reentry programs are largely designed for younger populations and often lack geriatric-trained staff, routine sensory screening, and structured protocols for managing age-related conditions.

To investigate whether incarceration leaves a measurable imprint on visual health, a research team led by Alexander Testa of the University of Texas Health Science Center at Houston, along with Luis Mijares, Mengyao Hu, Joana E. Andoh, Matthew W. Wade, and Dylan B. Jackson, analyzed data from the Health and Retirement Study, a nationally representative biennial survey of older Americans. Their analytic sample included 8,347 respondents aged 55 and older who completed a leave-behind questionnaire in 2012 or 2014 assessing incarceration history and who provided vision data across at least four survey waves between 2012 and 2022. Overall, 6.7 percent of respondents—529 individuals—reported having ever been an inmate in a jail, prison, juvenile detention center, or other correctional facility.

The team employed a statistical technique known as group-based trajectory modeling, a finite-mixture approach that identifies latent subpopulations following similar patterns on a measure over time. Rather than treating vision as a single snapshot, the method uses age as the underlying time scale and polynomial functions to capture developmental change, allowing researchers to classify individuals into distinct long-term pathways. Model selection was guided by fit statistics including the Bayesian Information Criterion, the transformed Bayes factor, and the odds of correct classification, with trajectories specified using a logistic distribution for the binary outcome of poor or fair self-rated vision. After estimation, respondents were assigned to the trajectory group with the highest posterior probability of membership, and multinomial logistic regression was used to test associations between incarceration history and group membership, with full accounting for the survey’s complex design through weighting, clustering, and stratification.

Three distinct vision trajectories emerged from the data. The largest group, accounting for 65.5 percent of the sample, was characterized by stable good vision, with a consistently low probability of reporting poor or fair eyesight. A second group, comprising 25.6 percent of respondents, followed a path of increasing poor vision, beginning midlife with modest rates of visual difficulty that rose sharply with age. The smallest group, 8.9 percent, experienced persistently poor vision throughout the study period. In unadjusted models, prior incarceration roughly doubled the relative risk of belonging to either the increasing poor vision group or the stable poor vision group compared with the stable good vision group.

After adjusting for a comprehensive set of covariates—including race and ethnicity, sex, educational attainment, veteran status, maternal education, marital status, income-to-poverty ratio, household wealth, and chronic conditions such as heart disease, hypertension, and diabetes—the association remained robust. A history of incarceration was linked to a 67 percent higher relative risk of membership in the increasing poor vision group (relative risk ratio 1.67, 95 percent confidence interval 1.20 to 2.33), indicating that formerly incarcerated older adults are disproportionately likely to experience eyesight that deteriorates as they age. The researchers ground these findings in life course theory, which conceptualizes aging as a set of long-term trajectories shaped by the timing and sequencing of pivotal life events. Within this framework, incarceration operates as a turning point that redirects trajectories of attainment and health: it erodes employment prospects, depletes wealth, and disrupts health care access—all resources needed to maintain optimal vision health—and has been linked to markers of accelerated epigenetic aging.

The study’s authors emphasize that the mechanisms linking imprisonment to failing eyesight remain poorly understood and merit deeper investigation. Inside correctional facilities, vision care is frequently limited to intake screening, with access to optometry services and replacement eyewear constrained by copays, specialty-care backlogs, and security-related barriers. Prior research has documented inaccurate ophthalmology follow-up among incarcerated patients and gaps in glaucoma care. Upon release, formerly incarcerated individuals often struggle to reestablish care, making continuity of vision services a critical but understudied challenge. The researchers point to promising strategies, including routine age-based vision screening consistent with clinical guidelines, expanded tele-ophthalmology, and partnerships with academic eye care programs inside prisons, along with pre-release insurance enrollment, warm handoffs to community providers, and integration of vision care into transition clinics after release.

The stakes extend beyond clinical outcomes. Declining vision can undermine core reentry tasks that determine whether a person successfully rebuilds a life after prison, including reading program materials, completing supervision paperwork, obtaining the driver’s license needed for reliable transportation, and securing employment. Studies have shown that vision loss is associated with job loss and reduced workforce participation, meaning that untreated visual impairment may compound the economic disadvantages that incarceration already imposes. The authors argue that screening for and accommodating visual impairment in correctional programming and community supervision could improve reentry outcomes for older adults.

The researchers acknowledge important limitations. Incarceration was measured as any lifetime history, without capturing the length, timing, or type of facility involved, and self-reported vision was dichotomized rather than measured through objective visual acuity testing—though prior studies show self-reported vision correlates well with clinical impairment. Because the sample was restricted to adults aged 55 and older, individuals in the poorest health may be underrepresented, since formerly incarcerated people experience elevated premature mortality with a median age of death before 50. This survivor bias means the reported associations may actually underestimate the true relationship between incarceration and poor vision. The observational design also precludes definitive causal inference, and covariates such as education and income may function either as confounders or as mediators of the incarceration-vision link.

Nevertheless, the study opens a striking new window on the long shadow cast by imprisonment. As the population of older Americans with incarceration histories continues to grow, the findings underscore the need for coordinated strategies across correctional, public health, and aging systems—improved access to preventive eye care during incarceration, continuity of care at reentry, and integration of vision screening into community-based services—to reduce vision-related disparities and support healthy aging for a population that has long been invisible to mainstream public health efforts.

Subject of Research: Association between history of incarceration and vision trajectories in older U.S. adults

Article Title: History of Incarceration and Vision Trajectories in Older Adults

Article References: Testa, A., Mijares, L., Hu, M., Andoh, J. E., Wade, M. W., & Jackson, D. B. (2026). History of Incarceration and Vision Trajectories in Older Adults. American Journal of Criminal Justice. https://doi.org/10.1007/s12103-026-09948-6

Image Credits: AI Generated

DOI: 10.1007/s12103-026-09948-6

Keywords: incarceration, vision trajectories, older adults, Health and Retirement Study, vision impairment, group-based trajectory modeling, life course theory, mass incarceration, aging prison population, eye care access, reentry health, History

The study conducted by Dai, Yang, Huang, and colleagues employs sophisticated statistical techniques to analyze data from a cohort of preterm infants. By applying group-based trajectory modeling, the researchers can categorize infants into distinct developmental paths. This technique enables them to identify patterns over time, revealing critical periods during which various interventions may be beneficial. It effectively illustrates how different infant characteristics and environmental factors contribute to their developmental trajectories.

Moreover, the integration of interpretable machine learning techniques enhances the transparency of the data analysis process. Unlike traditional machine learning methods that often operate as “black boxes,” interpretable models allow researchers and clinicians to understand which specific features influence neurodevelopmental outcomes. This is particularly vital in pediatric care, where comprehending the nuances of development can lead to more tailored and effective intervention strategies.

The findings of this research indicate that several key factors are associated with neurodevelopmental trajectories in preterm infants. These factors include not only medical and biological variables such as gestational age and birth weight but also psychosocial elements like parental involvement and socioeconomic status. By illuminating these associations, the study provides invaluable insights into how different spheres of influence can shape the developmental paths of preterm infants.

As researchers delve deeper into the data, they highlight the importance of early and ongoing assessments of neurodevelopment. By identifying infants at risk of suboptimal outcomes earlier in their lives, healthcare providers can implement strategies that focus on developmental support. This timely intervention could significantly improve long-term outcomes, thereby enhancing the quality of life for preterm infants and their families.

Additionally, the study advocates for a holistic approach to neonatal care that encompasses not just the medical needs of these infants but also the socio-environmental factors that they encounter. Engaging families in the care process, along with providing access to additional resources, can create a supportive environment conducive to healthy development. This perspective is gaining traction within pediatric healthcare, emphasizing that a multidisciplinary approach is essential for addressing the complex challenges faced by preterm infants.

The implications of this research extend far beyond academia. By equipping healthcare professionals with the knowledge derived from group-based trajectory modeling and interpretable machine learning, they can make informed decisions that directly impact prenatal and neonatal care practices. Consequently, initiatives that promote training and education for healthcare providers in these advanced analytical techniques may prove to be highly beneficial.

Moreover, the significance of this research lies in its potential to inspire future studies. As scientists continue to explore the intricacies of neurodevelopment in preterm infants, the methodologies established by Dai and colleagues can serve as a foundational framework for subsequent investigations. These methods can be adapted and expanded to include variables that may not have been fully explored, further refining our understanding of the neurodevelopmental landscape.

The use of advanced computational techniques also opens doors for building predictive models that can assess the risks of developmental delays based on newborn characteristics. Such models could revolutionize how healthcare systems allocate resources and prioritize interventions, ultimately improving the care provided to vulnerable populations. By identifying at-risk infants with greater accuracy and speed, practitioners can adjust care plans proactively rather than reactively.

In summary, the research led by Dai, Yang, Huang, and their team encapsulates a transformative shift in how we approach neurodevelopment in preterm infants. By harnessing the power of group-based trajectory modeling and interpretable machine learning, they provide a clearer picture of the complexities involved in infant development. Their findings underscore the multifactorial nature of development and advocate for an inclusive, data-driven approach to neonatal care.

As the longitudinal impacts of preterm birth continue to be explored, studies such as these serve as critical stepping stones toward improving the lives of millions of children worldwide. By fostering a collaborative environment between researchers and healthcare providers, we can pave the way for innovative interventions that truly make a difference. The work done in this study not only contributes to our scientific repository but also symbolizes hope for countless families navigating the uncertain journey of prematurity.

Given the complexity of this topic, the research demands extensive collaboration across various fields, including pediatrics, psychology, and data science. Continuous advancements in these areas will be pivotal in shaping best practices and crafting policies that better serve preterm infants and their families. As the conversation around preterm development evolves, the findings from this study will undoubtedly inform future research agendas and clinical practices for years to come.

In conclusion, the intersection of advanced modeling techniques and the urgent need for better outcomes for preterm infants creates an intriguing landscape for future exploration. The fusion of data science and traditional healthcare signifies a progressive step toward a more integrated and effective approach to understanding and fostering neurodevelopment in at-risk children. As we continue to glean insights from such research, we must remain steadfast in our commitment to improving the lives of preterm infants and providing them the best possible start in life.

Subject of Research: Neurodevelopmental trajectories in preterm infants

Article Title: Group-based trajectory modelling and interpretable machine learning to identify factors associated with neurodevelopmental trajectories in preterm infants.

Article References:

Dai, K., Yang, X., Huang, M. et al. Group-based trajectory modelling and interpretable machine learning to identify factors associated with neurodevelopmental trajectories in preterm infants.
BMC Pediatr (2026). https://doi.org/10.1186/s12887-025-06476-w

Image Credits: AI Generated

DOI: 10.1186/s12887-025-06476-w

Keywords: preterm infants, neurodevelopment, group-based trajectory modeling, interpretable machine learning, pediatric care, developmental outcomes.

Fetal growth is traditionally assessed using standard biometric measurements at various gestational stages. However, this study leverages an advanced statistical approach known as group-based trajectory modeling (GBTM), which allows researchers to classify fetuses into distinct growth trajectory groups based on longitudinal data. This methodology permits the detailed mapping of growth velocity and patterns throughout pregnancy rather than relying solely on snapshot measurements. By following these trajectories, investigators can detect early acceleration or deceleration in fetal growth—markers that have profound implications for postnatal health.

The investigation draws data from a well-characterized cohort of pregnant women and their offspring, tracking growth parameters measured at multiple time points during gestation. The fetal growth trajectories identified included a normative growth group and an accelerated growth group among others. Crucially, the accelerated growth trajectory group demonstrated a significantly greater propensity towards being overweight or obese at two years of age. This correlation suggests that fetal overnutrition or altered intrauterine metabolic environments may predispose children to adiposity very early in life, setting a potential precedent for metabolic disorders.

Maternal metabolic health emerges as a pivotal influence on fetal growth trajectory classification. Factors such as maternal pre-pregnancy body mass index (BMI), glucose regulation, insulin sensitivity, and lipid profiles appear intricately tied to accelerated fetal growth patterns. These findings hint at a complex interplay between maternal metabolic milieu and fetal development, where dysregulated maternal metabolism may alter nutrient delivery and fetal anabolic signaling pathways, fostering excess fetal growth.

This study’s insights bear significant clinical relevance. Understanding that accelerated fetal growth trajectories confer a fourfold increased risk for early childhood overweight or obesity highlights a narrow window for intervention during pregnancy. Detecting at-risk pregnancies through repeated biometric assessments and maternal metabolic screenings could enable the implementation of tailored nutritional and metabolic interventions before birth, potentially averting the trajectory toward pediatric obesity.

The use of group-based trajectory models in this research exemplifies a broader shift in perinatal epidemiology towards more nuanced analytical frameworks. Unlike traditional linear or cross-sectional analyses, GBTM accommodates heterogeneity in growth patterns and timing, offering a dynamic perspective of fetal development. This approach enhances our grasp of the etiological pathways that underlie the early origins of obesity and may revolutionize how prenatal care is personalized.

Furthermore, this study prompts a reevaluation of how fetal growth guidelines are constructed. Conventional relevance placed on small and large for gestational age extremes may overlook subtler distinctions in growth velocity and patterning that carry substantial lifelong metabolic consequences. Recognizing accelerated growth trajectories as a distinct risk phenotype encourages the refinement of monitoring protocols during pregnancy.

In the context of public health, these revelations advance the developmental origins of health and disease (DOHaD) paradigm. The fetal environment is increasingly acknowledged as a critical determinant of chronic disease susceptibility, with obesity standing prominently among them. By pinpointing fetal accelerated growth as a quantifiable risk factor, this research supports the prioritization of maternal health optimization as a strategy to combat the burgeoning childhood obesity epidemic.

The lack of specific funding for this work speaks to the independent rigor and authenticity of the findings, grounded in scientific inquiry rather than commercial interests. The authors’ declaration of no competing interests further strengthens the credibility and objectivity of the conclusions drawn.

The study also highlights important avenues for future research. Delineating precise metabolic pathways through which maternal factors influence fetal growth acceleration is essential. Moreover, extended follow-up into later childhood and adolescence would clarify the persistence and evolution of obesity risk associated with prenatal growth trajectories. Integrative studies incorporating genetic, epigenetic, and environmental data are anticipated to deepen our mechanistic comprehension.

In summary, this pioneering research conducted in China and published in PLOS One applies advanced trajectory modeling to fetal growth data, conclusively linking accelerated prenatal growth patterns with a markedly elevated risk of overweight and obesity by age two. Maternal metabolic health emerges as a key determinant in this process, underscoring the significance of prenatal metabolic monitoring and intervention. These findings mark a vital step forward in unraveling the complex origins of pediatric obesity and forging new preventative strategies rooted in early life.

As the obesity epidemic continues to challenge global health, such insights offer crucial pathways to mitigate risk from the earliest stages of development. Targeting fetal growth trajectories may ultimately lead to transformative outcomes in childhood health and beyond.

Subject of Research: Association between fetal growth trajectories and childhood overweight and obesity risk, with a focus on maternal metabolic factors.

Article Title: Application of group-based trajectory models to evaluate the association of fetal growth trajectories and childhood overweight and obesity: A longitudinal study with 2-year follow-up

News Publication Date: 17-Sep-2025

Web References: http://dx.doi.org/10.1371/journal.pone.0330715

Keywords: fetal growth trajectory, childhood obesity, accelerated fetal growth, maternal metabolic factors, group-based trajectory modeling, prenatal development, early childhood overweight, developmental origins of health and disease