A study of more than 10 million siblings found different patterns of diagnosed illness among first-born and second-born children. The associations may offer clues about early immune development, pregnancy biology, and family life, but they do not mean that a child’s place in the family determines future health.
A Disease Atlas from Family Data
Previous research has usually examined birth order one disease at a time. Later-born children, for example, have often been found to have lower rates of hay fever and asthma. Autism studies have been complicated by parental age and by the possibility that parents may decide not to have another child after an older child receives a diagnosis. The new analysis widened the lens to the human “phenome,” the full range of observable diseases and traits recorded across the body.
Health insurance claims collected in the United States between 2003 and 2024 were used to screen 569 groups of diagnoses. Of the 418 conditions common enough for the primary analysis, 150 were associated with birth order after a strict statistical correction was applied to reduce the chance of false-positive findings.
The question was approached in two ways. First, 1.6 million matched pairs, representing 3.2 million people from different families, were compared. First-born and second-born participants were matched on characteristics including sex, year of birth, geographic region, urbanization, parental age, length of insurance follow-up, and the spacing between siblings.
The second analysis compared siblings within the same family. This method helped control for much of the genetic background, socioeconomic environment, geography, and approach to health care shared by two children. No observational design can eliminate every difference between siblings, but agreement between these two methods made the broad pattern less likely to be a statistical accident.
First-Born Children Showed More Neurodevelopmental and Allergic Diagnoses
The strongest concentration of first-born excess was found among neurodevelopmental conditions. Compared with first-born children, second-born children had 26 percent lower odds of an autism diagnosis in the between-family analysis. That difference narrowed to about 20 percent when siblings within the same family were compared. Lower odds among second-born children were also reported for tics and Tourette syndrome, attention-deficit/hyperactivity disorder, and a broader diagnostic category called other or unspecified pervasive developmental disorder.
Second-Born Children Showed a Different Pattern
The diagnoses found more often among second-born children were concentrated in other parts of the health map. Substance abuse diagnoses were associated with 19 percent higher odds in the between-family analysis and 14 percent higher odds in the comparison of siblings from the same family. Gastritis and duodenitis were associated with approximately 14 percent higher odds, and biliary tract disease with about 18 percent higher odds.
Herpes zoster, commonly called shingles, produced one of the strongest second-born associations, with about 35 percent higher odds in the primary analysis. Several joint, connective tissue, and other musculoskeletal conditions also appeared more often among second-born participants. Claims records cannot reveal why these patterns emerged. Biological exposures, differences in behavior, diagnostic timing, and unmeasured family or social factors could all have contributed.
Older Siblings May Help Shape the Developing Immune System
The allergy findings fit a pattern that has been reported for decades. Younger siblings are often exposed earlier to microorganisms carried home by an older child from school, day care, friends, and the wider environment. Greater microbial diversity during infancy may help educate the immune system, strengthen regulatory pathways, and reduce the likelihood of reacting to harmless substances such as pollen or food proteins.
This idea is sometimes described as the hygiene hypothesis, but the “old friends” framework offers a more precise interpretation. Protection is not thought to come simply from catching more infections or living in an unclean home. It may instead depend on contact with a diverse community of largely harmless microorganisms with which human immune systems evolved.
Sibling spacing provided an additional clue. The second-born advantage for allergic rhinitis gradually weakened as the age gap between siblings increased. That pattern would be expected if children close in age share more microorganisms during the younger child’s early immune development. It does not prove the mechanism, however, and the results should not be taken as a reason to expose infants deliberately to infectious illnesses.
The Microbial Wake of an Older Child
The allergy findings fit an idea that has circulated in immunology for decades. In 1989 epidemiologist David Strachan reported that hay fever appeared less often in children with more older siblings. It was proposed that infections passed among children might train the developing immune system and make allergic reactions less likely.
That original “hygiene hypothesis” has since been refined. Simply becoming sick more often is not considered beneficial, and excessive cleanliness is not regarded as the singular cause of modern allergies. Attention has shifted toward the “old friends” hypothesis, which proposes that immune regulation is supported by early contact with a diverse community of harmless or beneficial microorganisms that accompanied humans throughout evolution.
An older sibling may expand that microbial world. Organisms encountered at school, day care, playgrounds, and friends’ homes can be carried back into the household, where a younger child encounters them during a formative period of immune development. Research in farming environments has similarly linked greater microbial diversity with lower rates of childhood asthma, although the exact organisms and immune pathways responsible remain under investigation.
Sibling spacing added an intriguing detail. The lower rate of allergic rhinitis among second-born children gradually weakened as the age gap between the siblings widened. Children close in age may share more space, respiratory secretions, toys, food, and microorganisms while the younger child’s immune system is developing. This pattern is compatible with microbial sharing, but it does not prove the explanation and should not be interpreted as a reason to expose infants deliberately to infectious diseases.
A More Complicated Neurodevelopmental Puzzle
The concentration of autism, ADHD, tic disorders, and related diagnoses among first-born children is harder to interpret. Pregnancy order is accompanied by biological changes in the uterus, placenta, immune system, and maternal metabolism. In a separate meta-analysis of 16 birth cohorts, birth order was associated with differences in DNA methylation measured in newborn blood. DNA methylation helps regulate gene activity, but these observations do not show that the identified changes cause a developmental condition.
Parental behavior may also shape the medical record. The development of a first child is watched without a previous child for comparison, and parents may seek evaluation differently when milestones are being encountered for the first time. A second child develops in a household where parents and clinicians may have different expectations and experience.
Parental age is another source of uncertainty. Parents are necessarily older during a later pregnancy, and parental age has independent associations with several developmental outcomes. Statistical matching reduced this influence but could not remove it entirely.
Family planning further complicates autism research. If parents are less likely to have another child after an autism diagnosis, families with an affected first child will be overrepresented among completed one-child families and some sibling datasets. In the new study, parents whose first child had autism were about 13 percent less likely to have a second child. That reproductive “stoppage” was too small to explain the full first-born excess, however. Autism remained more common among first-born children in the within-family comparison and after stricter matching for parental age. Stoppage was also absent or reversed for several other neurodevelopmental diagnoses.
No single explanation is likely to account for the entire pattern. Pregnancy biology, parental surveillance, diagnostic timing, family planning, and unmeasured environmental factors may all contribute, in different proportions, to different conditions.
Statistical Footprints, Not Medical Forecasts
Large databases can turn tiny differences into statistically significant results, so the consistency of the findings mattered as much as the number of people studied. Across diseases evaluated by both methods, about 74 percent of the estimates pointed in the same direction. Among the 150 conditions that crossed the strict statistical threshold in the primary analysis, nearly 85 percent showed the same directional pattern in the sibling comparison. Similar results were obtained after accounting for state of residence, restricting the analysis to full siblings, and applying more stringent clinical matching.
Differences in health care use were also considered. First-born participants averaged 24 days with insurance claims, compared with 22.9 for second-born participants. Adjusting for this difference did not substantially alter the overall results. Diseases expected to have little connection to birth order, including several genetic or structural conditions, generally remained close to neutral. Significant findings were also divided almost evenly, with 79 diagnoses more common among first-born children and 71 more common among second-born children. A simple tendency to take first-born children to the doctor more often would not readily explain that balance.
Important limitations remain. Insurance claims capture diagnoses submitted for payment, not the true onset of disease. A condition can be missed when care is not sought, insurance coverage is interrupted, or a different diagnostic code is chosen. Participants were largely covered by employer-sponsored insurance and lived predominantly in highly urban areas, which limits how confidently the findings can be extended to uninsured families, rural communities, and populations outside the United States.
The main cohort was also restricted to families with exactly two eligible children. The findings may not apply in the same way to only children or to children in larger families. Even comparisons within a family cannot control completely for differences between pregnancies, changing finances, moves, child care arrangements, or the many ways in which a household evolves over time.
Birth Order Is Context, Not Destiny
Nothing in the study suggests that medical screening or treatment should be changed according to whether a child was born first or second. Birth order is not a diagnosis, and it is far too blunt to predict individual health. Genetics, prenatal development, diet, infections, environmental exposures, access to care, and many other influences matter more for a particular child.
The value of the research lies elsewhere. Birth order packages together a set of biological and social transitions that touch nearly every family but are rarely measured as a whole. A first pregnancy differs from a later one. A home with no children differs from one occupied by an older sibling. Parents change with experience, and children alter the environments into which their siblings are born.
By revealing where these differences leave statistical footprints, the study provides a map rather than a verdict. More focused research can now be used to investigate which paths involve immune education, which begin during pregnancy, which are shaped by sibling behavior, and which reflect the way health care is sought and recorded. A child’s place in the family may help illuminate those paths, but it does not determine where they lead.
Reference
1. Kramer, B., Kushner, S.A. & Rzhetsky, A. Birth order and disease risk across the human phenome. Nat. Health (2026). doi.org/10.1038/s44360-026-00177-z
2. Strachan DP. Hay fever, hygiene, and household size. BMJ. 1989;299(6710):1259-1260. doi:10.1136/bmj.299.6710.1259
3. Ege MJ, Mayer M, Normand AC, et al. Exposure to environmental microorganisms and childhood asthma. N Engl J Med. 2011;364(8):701-709. doi:10.1056/NEJMoa1007302
4. Hoffmann TJ, Windham GC, Anderson M, Croen LA, Grether JK, Risch N. Evidence of reproductive stoppage in families with autism spectrum disorder: a large, population-based cohort study. JAMA Psychiatry. 2014;71(8):943-951. doi:10.1001/jamapsychiatry.2014.420
5. Li S, Spitz N, Ghantous A, et al. A Pregnancy and Childhood Epigenetics Consortium (PACE) meta-analysis highlights potential relationships between birth order and neonatal blood DNA methylation. Commun Biol. 2024;7(1):66. Published 2024 Jan 9. doi:10.1038/s42003-023-05698-x
