Get access to over 1,300 articles and reader Q&As.
It is difficult to miss the recent attention to trends in child health. The Make America Healthy Again movement has put a laser focus on chronic diseases in children, with extensive claims about their possible causes. As I’ve talked about before, many of these purported causes are not well supported by data (food dyes, vaccines).
Amidst this discussion of what might be at fault here, we may also ask a basic question: how bad is the problem? Have chronic diseases really gone up? Are children in the U.S. really much worse off than children in Europe? And what can those basic patterns in the data tell us about what is going on?
A new paper in JAMA, released July 8, 2025, tackles some of these questions with an impressive collection of data. Combining mortality data from the U.S. and Europe and health records and survey data from within the U.S. over time, the authors establish several core facts. First, infants under the age of 1 in the United States have a higher mortality rate than infants in other wealthy countries. Second, the same is true of children aged 1 to 19. Finally, the rate of chronic disease diagnoses has increased over time within the U.S.
The facts in this paper have been established in broad strokes before and are not surprising in the literature, but the paper provides a comprehensive window into the data and an opportunity to further unpack these three issues.
Why are infant mortality rates higher in the United States?
The infant mortality rate is defined as the number of deaths in the first year of life, and this rate in the United States is higher than in most of our peer countries. In 2024, our estimated infant mortality rate was 5.1 deaths per 1,000 live births, comparable to Slovakia and Romania. By contrast, the figure in Sweden was 2.3 deaths per 1,000 live births and in Norway 1.8.
The recent JAMA paper compares the U.S. to an average of 18 other high-income countries on this dimension and, looking into cause of death, shows that by far the most important cause of death is “prematurity,” which is a commonly used cause for infants whose deaths are a direct result of preterm birth and its associated complications.
This paper isn’t able to dive into many details, given the data limitations, but this problem isn’t new, and in a paper I co-authored almost 10 years ago, we were able to do a more detailed analysis of the U.S. disadvantage. The analysis was made possible by accessing individual-level data from the U.S. and several European countries.
That analysis reveals a few important factors in explaining the higher infant mortality in the U.S. One is reporting: In many countries, the mortality statistics are based only on births after a later gestational age. However, in the U.S., more babies born extremely preterm (22 or 23 weeks) are reported as live births in the data. Because survival is much less common in these early weeks, this increases the reported mortality rate. This issue explains about a third of the overall differences.
Most of the rest of the differences across countries — about 75% of the remaining — are due to higher preterm birth rates in the U.S., even putting aside births at very young gestational ages. The remaining issue in the U.S. is higher postneonatal mortality — deaths after one month of life.
Our analysis also revealed that these differences, especially the deaths later in the first year, are driven by differences across socioeconomic groups. To somewhat simplify, wealthier women in the U.S. experience very similar infant mortality rates to wealthier women in Austria, for example. Where the U.S. differs is in a much higher mortality rate for infants from families with fewer resources; in Austria, the mortality rates in the two groups are extremely similar.
The picture our research paints, consistent with this current work in JAMA, is a need for better interventions to decrease preterm birth rates and better support, especially for lower-income families, after birth. (For example, we know that paid parental leave has been shown to decrease infant mortality. That’s a good place to start.)
What about higher child mortality rates?
Child mortality is much lower overall than infant mortality; for children aged 1-14, the death rate is in the range of 0.2 per 1,000 children per year, versus 5.1 for infants under 1. However, death rates for children aged 1-19 in the U.S. are considerably higher than in comparable countries, almost twice as high.
This JAMA paper unpacks the primary excess causes, which have not changed over time: motor vehicles, firearms, and substance abuse. Both firearm death and substance abuse have increased significantly in the post-pandemic period (2020 to 2022).
By including all children from 1 to 19 in this data, there are some subtleties across age groups that are masked. In particular, firearm and substance abuse deaths are much higher among teens than among younger children. In those younger age groups, car accidents are more prevalent.
Death rates in the U.S. are higher than those in comparison countries, largely due to gun deaths and car accidents. Cars because we drive more, and our cars are larger. Guns because we have more guns. It’s important to say that although many parents may associate child gun deaths with mass shootings in school, the vast majority of gun deaths occur in isolated incidents. The U.S. also has more school shootings, but it is a smaller share of gun deaths. (This interview with Megan Raney may provide some context.)
Before moving on to the discussion of chronic disease, I think it is worth pausing here to say that these differences in mortality — for both children and infants — do not closely reflect the issues or causes raised in the MAHA report. The policies that would affect them are ones that improve maternal health or support, reduce access to guns, or make cars safer, none of which are a core component of MAHA efforts.
What’s behind the rise in chronic disease?
The third fact that is highlighted in the JAMA paper is the rise in “chronic disease” in children. That’s not a well-defined category, but these authors include a set of 15 conditions: ADHD, anxiety, asthma, autism, cerebral palsy, conduct disorder/behavioral problems, depression, developmental delay, diabetes, hearing problems, learning disability, seizure disorder, speech problems, Tourette syndrome, and vision problems.
They find that the share of children ages 2 to 19 reporting one or more of these conditions has increased from 2011 to 2022. For parent-reported conditions, the increase is from 25.8% to 31% over this period; in electronic medical records, the increase is from 39.9% to 45.7%. There is also an increase in obesity rates in this age group, from 17 to 20%.
The changes in chronic conditions are largely driven by increases in mental health conditions (anxiety in particular) and behavioral diagnoses (autism, ADHD). Some conditions, like diabetes and asthma, are unchanged or declining.
These facts reflect what we know in much of the existing literature (although it is very helpful to see them all in one place). The open question is what explains them and if that helps inform policy.
In the case of autism and other behavioral diagnoses like ADHD, as I’ve written before, it is largely due to changes in diagnostic behavior. It is not clear whether this requires a policy change. Better diagnosis may mean better treatment; the important thing is to make sure people get the support they need.
Increases in mental health diagnoses, among teens in particular, have been hotly debated. Is it screens? Sleep? Something else? This is an open question, and one which deserves more work.
Finally, obesity. Increases in obesity reflect increases in calorie consumption. This is one place where MAHA pushes against how ultraprocessed foods could play an important role. We know that ultraprocessed foods prompt greater calorie consumption, probably by limiting the feeling of fullness. The increase in their availability is likely a contributing factor here. There are policies that might affect this, although the problem is complicated; one good place to start might be healthier school lunches.
Closing thoughts
As this new paper highlights, the health of babies and children in the U.S. is not where it should be. Death rates are higher than those of our peer countries, especially among lower-income children. Chronic disease rates have increased over time, some as a result of diagnosis, but some have changed, like the increase in obesity, which is real and does present health threats.
All of these are possibly amenable to policy solutions, but it is crucial to keep our focus on policy solutions that might actually matter based on the data. These policies are often more challenging to implement than some of the MAHA rhetoric. Addressing teen substance abuse or getting universal paid leave is a lot harder than banning food dyes like Orange B, which are barely used anyway. But those are the changes that might actually improve the health of children.
The bottom line
- Infant mortality rates in the United States are higher than in most of our peer countries. While some of the gap can be explained by reporting differences, much of it is driven by differences across socioeconomic groups. Research points to a clear need for better interventions to decrease preterm birth rates and better support, especially for lower-income families, after birth.
- While overall, child mortality in the United States is much lower than infant mortality, U.S. child mortality rates are nearly twice as high as in peer countries, largely due to car accidents and gun deaths.
- There is also a rise in chronic conditions among U.S. children, driven mainly by increases in mental health and behavioral diagnoses like anxiety, autism, and ADHD.
- If we want to improve child health, we need to focus on policies that will actually make an impact — like addressing teen substance abuse or expanding paid parental leave — which are harder to implement than some of the simpler fixes pushed by MAHA (such as banning food dyes).
Log in
Thanks for the helpful breakdown! May I offer a suggestion to switch the language you use in future writing from “car accident” to “car crash” or “car death”? There is a strange acceptance in the U.S. of car-based violence as “normal” when it shouldn’t be. Our cars are bigger, as you said, and U.S. safety standards are measured differently from other parts of the world (e.g. danger to people outside the car is not counted in safety rating). Small things like word choice can help people re-evaluate this problem.
“Cars because we drive more, and our cars are larger” it’s not the point of this site but isn’t that a little pat? California has 1.4 deaths per 100 million miles driven and Australia has 0.39
Clearly support to low income families after birth is essential! I urge folks to reach out to their representatives and save the Healthy Start Initiative that is one the verge of being cut. Launched in 1991, Healthy Start is a federally funded, community-driven program focused on improving birth outcomes and reducing disparities.
Great article, thank you!
Thanks for the great piece! I wonder if you can summarize whether there is an effect from different cultural and legal regimes around pregnancy termination. Essentially, are the babies in the “denominator” of the infant mortality calculation really all sampled from the same underlying population in terms of health? Do the JAMA authors control for that?
Motivated to comment just because I was fist pumping and shouting “shots fired!” in my head the entire time I read this. As a non-US reader, I also feel the need to quote the Onion: “‘No Way to Prevent This,’ Says Only Nation Where This Regularly Happens.”
She does, in fact, talk about the ways to prevent it though.
I don’t know the stats specifically regarding preterm births, but there is quite a lot of literature (under the rubric of the social determinants of health) establishing that bad outcomes in low income populations are not primarily about their income in a vacuum, but about the inequality itself. People of all income groups experience better outcomes when inequalities are reduced. Nothing against targetting interventions to low income folks, but we have to talk about the role inequality plays if we’re going to truly move the needle.