Perinatal and Early-Childhood Exposures and Likelihood of Autism: A Cosmos Cohort Study of 1.9 Million U.S. Children
This large-scale retrospective cohort study of nearly 2 million U.S. children identified 26 perinatal and early-childhood factors, including family history, male sex, prematurity, and maternal obesity, that show clinically meaningful associations with autism diagnosis, while clarifying that these findings indicate risk characteristics rather than causality.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Autism is a common way the human brain develops, characterized by differences in how people communicate and interact with the world, alongside a tendency toward focused or repetitive behaviors. While scientists have long known that genetics play a major role in who develops autism, the full picture of why it happens remains complex. Many researchers suspect that factors occurring before birth or in the earliest years of life might also influence the likelihood of a child receiving an autism diagnosis. These factors, known as exposures, could range from the health of the mother during pregnancy to the circumstances of the birth itself. Understanding these early-life connections is not about finding a single cause, but rather about identifying patterns that might help doctors recognize children who would benefit from earlier support and intervention.
To explore these connections on a massive scale, a team of researchers turned to a vast digital archive of medical records. They analyzed data from nearly two million children across the United States, drawing from a network of hospitals and clinics that spans the country. This approach allowed them to look at a wide array of potential factors—eighty-eight in total—ranging from the mother's weight and health conditions to the baby's birth weight and early medical history. Because the dataset was so large, the researchers had to be careful not to get lost in tiny statistical differences that might look important on a computer screen but mean little in the real world. Instead, they focused on finding associations that were strong enough to be considered genuinely meaningful for a child's health.
The study confirmed what many experts have suspected for some time: certain early-life characteristics are linked to a higher likelihood of an autism diagnosis. The strongest connections were found with factors related to the child's own biology and family history. For instance, boys were significantly more likely to be diagnosed than girls. Children with a family history of autism, or those whose birth parent had an autism diagnosis, showed a much higher likelihood of receiving a diagnosis themselves. Other powerful links included being born prematurely, having a very low birth weight, or being diagnosed with cerebral palsy or early developmental delays. These findings suggest that the physical stresses of a difficult pregnancy or birth, as well as specific genetic and neurological conditions, are closely tied to the development of autism.
The researchers also looked at the health of the mother during pregnancy. They found that severe obesity in the mother was associated with an increased likelihood of autism in the child. Similarly, conditions like diabetes and hypertension during pregnancy showed connections to higher rates of autism diagnosis. The study also examined medications taken during pregnancy. It found that exposure to certain neuroactive drugs, such as antipsychotics and a specific medication called topiramate, was linked to a higher likelihood of autism. However, the researchers were careful to note that this does not necessarily mean the drugs caused the autism. It is possible that the underlying health condition requiring the medication is the true factor, rather than the medicine itself.
Not every factor the team investigated showed a clear link. Many common concerns, such as the use of acetaminophen for pain or fever during pregnancy, did not show a strong, meaningful connection to autism in this large group of children. The study also found that some factors previously thought to be important, like low Apgar scores (a quick check of a newborn's health) or the length of time a baby stayed in the hospital after birth, were likely just signs of other underlying issues, such as prematurity, rather than independent causes. This distinction is crucial: it helps separate the root causes from the symptoms that appear alongside them.
One of the most valuable aspects of this research is its sheer size and the way it was conducted. By looking at almost two million children, the study could detect patterns that smaller studies might miss, while using a rigorous method to ensure the results were not just random noise. The researchers used a specific framework to decide which findings were important enough to matter in a clinical setting, filtering out results that were statistically significant but too small to be useful. They also accounted for the fact that families who visit doctors more often might have their children diagnosed earlier, which could skew the data. By adjusting for these factors, the study provides a clearer view of the real-world associations between early life and autism.
Ultimately, this work does not prove that any single factor causes autism. Instead, it maps out a landscape of risk, showing which early-life characteristics tend to appear together with an autism diagnosis. The findings suggest that autism is often the result of a complex interplay between genetics, the mother's health, and the conditions surrounding birth. For families and doctors, these insights offer a way to identify children who might need closer attention and earlier support. While the study cannot tell us exactly why autism happens, it provides a solid foundation for future research to understand the mechanisms behind these connections and to develop better ways to help children thrive.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.