A Broken Collarbone or an Injured Arm Nerve at Birth: How Often It Happens, Who Is Most at Risk, and Why It Is Becoming Less Common

What a Swedish study of more than 1.2 million births, published in 2026 by Voss, Ahlsson and Ekéus, tells parents about two shoulder injuries of birth

A large Swedish study of more than 1.2 million babies born at full term found that a broken collarbone or an injury to the nerves of the arm during birth is uncommon, affecting roughly 2 and 1 in every 1000 babies, and that both were recorded less often in 2022 than in 2010. The same few things raised the chance of both injuries: a shoulder getting stuck during birth, a baby weighing more than 4 kg, and a birth assisted by a suction cup.

Two injuries that can happen as a baby is born

Most babies arrive without any injury at all. Occasionally, though, the journey through the birth canal puts strain on a baby's shoulder. Two injuries can result. The first is a broken collarbone, which doctors call a clavicle fracture. The collarbone is the thin bone running across the top of the chest to the shoulder. The second is a brachial plexus birth injury, often shortened to BPBI. The brachial plexus is a bundle of nerves running from the neck into the arm; when these nerves are stretched during birth, the baby's arm may be weak or move less than the other one. A broken collarbone usually heals by itself. A nerve injury often gets better on its own too, but not always, which is why doctors take it particularly seriously.

What doctors used to think, and why the question needed a fresh look

The new study from Sweden points out that earlier research gave very different estimates of how often collarbones break at birth, anywhere from 2 to 35 in every 1000 births [1]. Part of the reason is that these fractures are easy to miss. In a hospital in Seoul, South Korea, more than half of the broken collarbones were found only by chance, on an X-ray taken for another reason, and about 1 in 7 were not found until after the baby had gone home [2]. For many years, doctors regarded a broken collarbone as a minor, unavoidable event. A study from Hong Kong concluded that it was of little medical importance and did not reflect the quality of care [3], and an Israeli study found that most fractures happened in healthy babies after normal labours, with no clear way to prevent them [4]. An earlier study covering all babies born in Sweden between 1997 and 2014 found that collarbone breaks made up nearly 9 in 10 of all broken bones found in newborn babies, and that a difficult birth was the main cause of breaks that happened during birth [5].

The nerve injury has always worried doctors more. Across many studies it affects between about 3 in 10 000 and 3 in 1000 births, and about two-thirds to three-quarters of babies recover fully without treatment [6]. The Swedish authors note that 20% to 30% of affected children are left with some lasting weakness [1]. Swedish researchers had studied this injury before: in births from 1987 to 1997, the strongest warning signs were a shoulder getting stuck and a baby weighing 4.5 kg or more [7]. A later Swedish study found that the nerve injury became less common between 1997 and 2019, even though stuck shoulders were being recorded more often [8]. In neighbouring Finland, broken collarbones also became much less common between 2004 and 2017 [9]. The Swedish authors point out that no earlier study had looked at how changes in mothers, pregnancies and births affect how often these two injuries happen, so they set out to study both injuries together, across a whole country, during recent years in which the way babies are delivered has been changing [1].

How the Swedish study worked

The researchers, from Uppsala University, used Sweden's national Medical Birth Register. This is a nationwide record that has existed since 1973 and captures around 98% of all births in the country [10]. They looked at every baby born alive at full term (37 weeks of pregnancy or later), one baby per pregnancy, between 2010 and 2022: 1 224 709 babies in total [1]. Babies born by a planned caesarean section were left out. The team then compared babies who had each injury with babies who did not, looking at features of the mother, the pregnancy and the birth. Because many of these features go together (for example, mothers with diabetes are more likely to have larger babies), they used statistical methods to tease out which factors still mattered once the others were taken into account.

What the researchers found

Among all these babies, 2796 had a recorded broken collarbone (about 2 in every 1000) and 1630 had a recorded nerve injury (a little more than 1 in every 1000) [1]. Only 150 babies had both, so the two injuries usually happened separately.

Both injuries became less common over the study period. Broken collarbones fell from about 36 in every 10 000 babies in 2010 to about 10 in every 10 000 in 2022. Nerve injuries fell from about 16 to about 9 in every 10 000 babies [1].

The biggest warning sign for both injuries was shoulder dystocia, which is when a baby's shoulder gets caught behind the mother's pelvic bone after the head has been born. To put this in everyday numbers: shoulder dystocia happened in about 2 of every 1000 babies without a nerve injury, but in more than a third of the babies who had one [1]. A large baby was the next most important factor. Nearly a third of babies with a nerve injury weighed more than 4.5 kg, compared with about 1 in 30 babies without the injury. Birth with the help of a vacuum cup (a soft suction cup placed on the baby's head to help guide it out) was also strongly linked to both injuries. Shorter mothers and mothers with obesity had a somewhat higher chance too [1].

A few factors affected only one injury. Babies born bottom-first (breech) had a higher chance of the nerve injury but not of a broken collarbone; in Sweden, more than 92% of breech babies are born by caesarean section [1]. Having labour started with medication or other methods (called induction) was linked to a small increase in the nerve injury, but not to broken collarbones. Diabetes in the mother mattered too: type 1 diabetes was linked to broken collarbones and type 2 diabetes to nerve injuries, once other factors were taken into account [1]. Earlier research shows why diabetes and weight are hard to separate from baby size: type 1 diabetes makes a very large baby far more likely [11], and a mother's obesity raises the chance of a large baby, with a higher chance of a stuck shoulder when obesity is combined with diabetes that develops during pregnancy [12].

Some things were linked to a lower chance of both injuries: being born by an emergency caesarean section, being born facing upwards, and weighing under 3 kg at birth [1]. This does not mean a caesarean is the right choice for everyone. Caesarean birth has its own risks, and an earlier Swedish study, which looked at offering a caesarean when a baby was thought to be very large, estimated that hundreds of caesareans would be needed to prevent a single nerve injury [7].

Why fewer injuries are being recorded, and what we still do not know

It is encouraging that both injuries are being recorded less often, but the authors are careful about what this means. The study could only count injuries that doctors noticed and wrote into the register. The authors note that up to 40% of broken collarbones may not be found before a baby goes home, and they could not tell whether that proportion changed over time [1]. So the fall is a fall in recorded injuries, not necessarily in all injuries.

Some changes in Swedish birth care might help explain it: fewer babies weighing more than 4 kg, fewer vacuum-assisted births and slightly more caesareans [1]. Birth teams can also practise for stuck-shoulder emergencies using realistic simulations, and studies elsewhere, including one at a large hospital in Helsinki, Finland, have linked this kind of training to fewer nerve injuries [13][14]. The Swedish study did not test whether any of these changes caused the decline.

What this means for families

If your baby was large, if their shoulder got stuck during birth, or if a vacuum cup was used, it is reasonable to ask whether your baby's collarbones and arms have been checked. If a broken collarbone is found, the authors suggest that parents be shown gentle ways to handle the baby, for example during feeding and lifting, to reduce pain [1]. If your baby's arm seems weak, doctors will usually examine it repeatedly over time, because this is the best way to tell which babies need specialist help [15]. Many babies with a nerve injury recover well, and in one large American health system, babies who also had a broken collarbone recovered at least as often as those who did not [16]. Families whose child has a lasting nerve injury may find our article on the certainty gap in brachial plexus birth injury helpful, and parents interested in how assisted births compare can read our article on vacuum, forceps or second-stage caesarean.

What researchers are working on next

Researchers want to know how many of these injuries are being missed, which babies end up with lasting arm weakness rather than a short-lived problem, and whether changes such as earlier induction of labour (now much more common in Sweden) or regular shoulder-emergency training make a difference. The authors stress that monitoring should continue as birth care keeps changing [1].

A reassuring overall picture

For the great majority of families, these injuries will never be part of their story: in this study, more than 99 in every 100 babies had neither. When they do happen, a broken collarbone usually heals by itself and many nerve injuries improve. Knowing the main warning signs helps birth teams take extra care and check the babies most likely to need attention.

References

  1. Voss S, Ahlsson F, Ekéus C. Risk and protective factors for birth-related clavicle fracture and brachial plexus birth injury: a national cohort study. Arch Dis Child Fetal Neonatal Ed. 2026;111(6):F561-F566. doi:10.1136/archdischild-2026-330774 ↩
  2. Ahn ES, Jung MS, Lee YK, Ko SY, Shin SM, Hahn MH. Neonatal clavicular fracture: recent 10 year study. Pediatr Int. 2015;57(1):60-63. doi:10.1111/ped.12497 ↩
  3. Lam MH, Wong GY, Lao TT. Reappraisal of neonatal clavicular fracture: relationship between infant size and neonatal morbidity. Obstet Gynecol. 2002;100(1):115-119. doi:10.1016/S0029-7844(02)02055-002055-0) ↩
  4. Lurie S, Wand S, Golan A, Sadan O. Risk factors for fractured clavicle in the newborn. J Obstet Gynaecol Res. 2011;37(11):1572-1574. doi:10.1111/j.1447-0756.2011.01576.x ↩
  5. Högberg U, Fellman V, Thiblin I, Karlsson R, Wester K. Difficult birth is the main contributor to birth-related fracture and accidents to other neonatal fractures. Acta Paediatr. 2020;109(10):2040-2048. doi:10.1111/apa.15217 ↩
  6. Koshinski JL, Russo SA, Zlotolow DA. Brachial Plexus Birth Injury: A Review of Neurology Literature Assessing Variability and Current Recommendations. Pediatr Neurol. 2022;136:35-42. doi:10.1016/j.pediatrneurol.2022.07.009 ↩
  7. Mollberg M, Hagberg H, Bager B, Lilja H, Ladfors L. High birthweight and shoulder dystocia: the strongest risk factors for obstetrical brachial plexus palsy in a Swedish population-based study. Acta Obstet Gynecol Scand. 2005;84(7):654-659. doi:10.1111/j.0001-6349.2005.00632.x ↩
  8. Mollberg M, Ladfors LV, Strömbeck C, Elden H, Ladfors L. Increased incidence of shoulder dystocia but a declining incidence of obstetric brachial plexus palsy in vaginally delivered infants. Acta Obstet Gynecol Scand. 2023;102(1):76-81. doi:10.1111/aogs.14481 ↩
  9. Kekki M, Salonen A, Koukkula T, Laivuori H, Tihtonen K, Huttunen TT. Incidence changes in risk factors associated with the decreasing number of birth-related clavicle fractures in Finland: A nationwide retrospective birth cohort from 2004 to 2017. Birth. 2023;50(2):428-437. doi:10.1111/birt.12662 ↩
  10. Cnattingius S, Källén K, Sandström A, et al. The Swedish medical birth register during five decades: documentation of the content and quality of the register. Eur J Epidemiol. 2023;38(1):109-120. doi:10.1007/s10654-022-00947-5 ↩
  11. Persson M, Norman M, Hanson U. Obstetric and perinatal outcomes in type 1 diabetic pregnancies: A large, population-based study. Diabetes Care. 2009;32(11):2005-2009. doi:10.2337/dc09-0656 ↩
  12. Weir TL, Majumder M, Glastras SJ. A systematic review of the effects of maternal obesity on neonatal outcomes in women with gestational diabetes. Obes Rev. 2024;25(7):e13747. doi:10.1111/obr.13747 ↩
  13. Wagner SM, Bell CS, Gupta M, et al. Interventions to decrease complications after shoulder dystocia: a systematic review and Bayesian meta-analysis. Am J Obstet Gynecol. 2021;225(5):484.e1-484.e33. doi:10.1016/j.ajog.2021.05.008 ↩
  14. Kaijomaa M, Gissler M, Äyräs O, Sten A, Grahn P. Impact of simulation training on the management of shoulder dystocia and incidence of permanent brachial plexus birth injury: An observational study. BJOG. 2023;130(1):70-77. doi:10.1111/1471-0528.17278 ↩
  15. Pulos N, Shaughnessy WJ, Spinner RJ, Shin AY. Brachial Plexus Birth Injuries: A Critical Analysis Review. JBJS Rev. 2021;9(6). doi:10.2106/JBJS.RVW.20.00004 ↩
  16. Wall LB, Mills JK, Leveno K, et al. Incidence and prognosis of neonatal brachial plexus palsy with and without clavicle fractures. Obstet Gynecol. 2014;123(6):1288-1293. doi:10.1097/AOG.0000000000000207 ↩