Researchers analyzed health-insurance records from more than 10 million people in 5.1 million two-child families and asked a simple question: does being the first or second child change which diseases you get? Across 418 conditions with enough cases to test, 150 showed a statistically robust link to birth order. Firstborns carried more neurodevelopmental and allergic disease (autism, tics, ADHD, hay fever, food allergy, acne), while second-borns carried more substance abuse, migraine, gut conditions and shingles. The effects are real and consistent across two independent study designs, but individually small. The value is the map, not any single number.
For a century, the idea that your rank among siblings shapes your life has drifted between folk wisdom and scattered science. Firstborns supposedly get more parental attention and score marginally higher on IQ tests; later-borns supposedly take more risks. Individual studies linked birth order to allergies here, autism there, but nobody had looked across the whole span of human disease at once. A team from the University of Chicago and Columbia University has now done exactly that.
Using two decades of anonymized US commercial insurance claims, they built a cohort of over 10 million people, restricted to clean two-child families so that firstborn versus second-born status is unambiguous. They then ran the comparison two different ways. The first, a between-family design, matched 1.6 million firstborns to 1.6 million second-borns from other families with the same sex, birth year and location. The second, a within-family design, compared the two siblings inside each of 5.1 million families directly, which automatically cancels out anything the siblings share: parents, genes, income, neighborhood.
The headline is that birth order touches almost every corner of the disease map. Firstborns showed higher rates of neurodevelopmental conditions, with the strongest signals in a group of childhood developmental disorder codes, tics and Tourette syndrome, autism, and obsessive-compulsive disorder, tapering down to milder effects for anxiety and depression. Firstborns also had more allergic and skin disease: hay fever, food allergy, asthma and acne. Second-borns, by contrast, led in substance abuse, migraine, gastritis, biliary disease, kidney infection and shingles.
The allergy pattern fits a well-known idea, sometimes called the old-friends or hygiene hypothesis: a second child grows up already exposed to the germs an older sibling drags home, training the immune system toward tolerance. The neurodevelopmental pattern is murkier and the authors are careful not to overclaim a mechanism. They rule out the most obvious artifact, that parents stop having children after an autism diagnosis and so inflate the firstborn count, by showing the effect survives in the within-family comparison. What they cannot fully remove is parental age, which is mathematically tangled up with birth order.
The take-home is deliberately modest. No single association is large enough to change an individual’s life. But because birth order is a universal exposure, even small shifts scale into a meaningful population signal, and the resulting atlas gives clinicians and researchers a benchmark they did not have before.
Actionable Insights
These are population associations, not personal risk scores, and every effect is small.
The most defensible practical read is the allergy and immune finding, because it has a plausible mechanism and validated positive controls. Firstborns carried more allergic rhinitis (odds ratio 0.91 for second-borns, so roughly a 9 to 10 percent relative reduction in second-borns), food allergy (OR 0.80, about a 20 percent relative reduction) and asthma (OR 0.97). In absolute terms, with allergic rhinitis affecting about 22 percent of the cohort, that translates into only about a 2 percentage-point difference between a first and second child. This is consistent with, and gives a large-sample endorsement of, early and diverse microbial exposure as protective. It supports, but does not prove, mainstream guidance against over-sterilizing an only or first child’s early environment.
For neurodevelopmental and behavioral conditions, the only real action is vigilance framing: a firstborn is a marginally higher-prior patient for autism (OR 0.74, about a 26 percent relative excess in firstborns) or tics, and a second-born a marginally higher-prior patient for substance abuse (OR 1.19). This might nudge screening attention. It should never drive a diagnosis or a reproductive decision.
Context and Source
- Open Access Paper: Birth order and disease risk across the human phenome
- Authors: Benjamin Kramer, Steven A. Kushner and Andrey Rzhetsky.
- Institutions: University of Chicago (Department of Medicine; Department of Human Genetics) and Columbia University (Department of Psychiatry), United States.
- Journal: Nature Health (Springer Nature).
- Impact evaluation: Nature Health launched in January 2026. Because a journal must be indexed and published for roughly two years before Clarivate assigns a Journal Impact Factor, this journal has no official JIF or CiteScore yet. This is an Unrated impact journal at present, though its Nature-family stablemates (for example Nature Mental Health) sit in the high to elite band, which is the most reasonable prior for where it will land.