A new cohort study is drawing attention to a statistical approach that could change how scientists study development, maturation, and mental health across the lifespan. Rather than describing people only by where they stand on a growth curve at a particular age, the research uses conditional-longitudinal models to measure how each individual differs from an expected developmental trajectory. The method may offer a more precise way to identify biological patterns linked to birth weight and mental health symptoms, while also opening new possibilities for psychiatric neuroscience.
Growth curves are commonly used in medicine and neuroscience to summarize how traits change over time. Researchers may compare brain structure, body measurements, cognitive performance, or symptoms with an average trajectory calculated from a large population. A person who falls above or below that average can then be classified as showing unusually rapid, slow, or delayed development. However, a single measurement cannot reveal whether that position reflects a persistent pattern, a temporary deviation, or a different pace of maturation. The new modeling framework is designed to address that limitation by incorporating repeated observations from the same individuals.
In a conditional-longitudinal model, each measurement is interpreted in relation to both age-related expectations and the person’s earlier measurements. The model can estimate a participant’s expected value at a given time and then calculate the deviation between that expectation and the observed value. These deviations, sometimes understood as individual trajectory residuals, can be tracked across multiple assessments. This creates a dynamic profile of development rather than a static label. Two people may occupy the same position on a growth curve at age 15, for example, while arriving there through very different developmental paths.
That distinction is important because maturation is not perfectly synchronized across individuals. Some people may experience earlier changes followed by stabilization, while others may show slower but more prolonged development. Conventional cross-sectional analyses can blur these differences by treating measurements taken at different ages as if they were interchangeable snapshots. Longitudinal modeling, by contrast, can examine within-person change and separate it from differences between people. The approach therefore has the potential to reveal developmental signals that would remain hidden when researchers rely only on average group trends.
The study reports robust associations between individual trajectory deviations, birth weight, and mental health symptoms assessed over time. The findings do not mean that birth weight alone determines later psychiatric outcomes, nor do they establish that deviations from an expected trajectory directly cause symptoms. Instead, the associations suggest that early developmental conditions and the pace or pattern of later maturation may be connected in ways that are not fully captured by conventional analyses. Because the research is observational, unmeasured genetic, environmental, social, and medical factors could contribute to the relationships.
Birth weight is often used as an early indicator of prenatal growth and developmental conditions. Lower or higher birth weight can reflect a range of influences, including gestational age, maternal health, placental function, nutrition, and complications during pregnancy. Its relationship with later brain development and mental health is complex and cannot be reduced to a single biological pathway. The value of the new analysis lies in its attempt to connect this early-life measure with longitudinal patterns, rather than examining birth weight only as an isolated risk factor.
The researchers also emphasize the relevance of the method for psychiatric neuroscience. Mental health symptoms often fluctuate, emerge gradually, or follow different courses in different people. A participant’s symptom score at one assessment may not indicate whether symptoms are increasing, decreasing, or remaining stable. By linking symptom assessments to individual developmental trajectories, conditional-longitudinal models could help investigators study when biological changes coincide with the emergence or persistence of anxiety, depression, psychosis-related experiences, or other psychiatric features. Such models may eventually support more individualized approaches to risk assessment, although they are not clinical diagnostic tools on their own.
The framework may also prove useful beyond adolescence. Developmental trajectories continue to change during adulthood and aging, and the same logic could be applied to cognitive performance, brain imaging measures, neurological function, or physical health. Researchers could ask whether a person is aging along an expected pathway, diverging from it, or changing at an unusually rapid rate. In this sense, the model offers a common mathematical language for studying maturation from childhood through later life. Its broader promise will depend on replication, transparent reporting, diverse cohorts, and evidence that trajectory-based measures improve prediction beyond established clinical and demographic factors.
The study was conducted by investigators at the University of Pennsylvania, including Eren Kafadar, BS, and Aaron F. Alexander-Bloch, MD, PhD. Published in JAMA Network Open, the work highlights a shift in emphasis from static developmental rankings to the timing and direction of individual change. As large longitudinal datasets become increasingly available, approaches that preserve information about personal trajectories could become central to research on the brain, behavior, and psychiatric illness. The immediate message is not that one new model has solved the complexity of human development, but that studying how people change may be more informative than simply measuring where they are.
Subject of Research: Conditional-longitudinal modeling of individual growth trajectories, birth weight, maturation, neuroimaging, and longitudinal mental health symptoms.
Web References: https://doi.org/10.1001/jamanetworkopen.2026.27643
References: Kafadar E, Alexander-Bloch AF, et al. Study published in JAMA Network Open. DOI: 10.1001/jamanetworkopen.2026.27643
Keywords: Neuroimaging, brain, adolescents, cohort studies, trajectories, factorization, modeling, mental health, body weight, symptomatology, aging populations, psychiatry, neuroscience, observational data.
