News · Supplements
Newborns who skipped vitamin K had triple the brain bleeds
A JAMA Pediatrics study of 2 million Swedish births found infants who did not get the vitamin K injection had higher odds of bleeding, and refusals have more than doubled since 2006.
Based on a peer-reviewed cohort study in JAMA Pediatrics
- A nationwide Swedish cohort of 2,020,302 live births between 2003 and 2021, followed to 6 months of age, published in JAMA Pediatrics.
- 24,089 infants had no record of the intramuscular vitamin K injection at birth.
- Those infants had 1.52-fold higher adjusted odds of a bleeding diagnosis in the first 6 months.
- For bleeding inside the skull, adjusted odds were 2.91-fold higher.
- Non-receipt fell from 1.32% in 2003 to 0.66% in 2006, then more than doubled to 1.50% by 2021.
- Vitamin K deficiency bleeding is rare, so even tripled odds translate to a small absolute number of infants.
- Registry records show what was documented, and a missing record is not always a missing injection.
- This is observational. Families declining the injection may differ in other ways that affect bleeding risk.
The vitamin K injection given to newborns is one of the least controversial things in medicine, and refusals of it are rising anyway. A study in JAMA Pediatrics, published by the American Medical Association, used two decades of Swedish births to measure what that costs.
Infants without intramuscular vitamin K administration had 1.54-fold higher odds of bleeding and 3.18-fold higher odds of an intracranial bleeding episode during infancy.
Why the injection exists at all
The condition it prevents is rare and severe. Vitamin K deficiency bleeding in infants is a rare but potentially life-threatening condition that is effectively prevented by newborn intramuscular vitamin K prophylaxis.
Newborns start with almost none of the vitamin their blood needs to clot, and breast milk supplies little. The injection closes that gap in a single dose on the first day.
What changed is uptake. Despite this, there are reports of increasing parental refusal of vitamin K prophylaxis globally, which is what prompted the study.
Two million births, one register
The design is a nationwide cohort study of births between January 1, 2003, and December 31, 2021, followed up until 6 months of age, covering all live-born infants in Sweden born at 35 or more weeks’ gestation.
Among 2,020,302 live births, 24,089 infants had no record of intramuscular vitamin K administration at birth. That is the exposed group: about one in eighty-four.
Swedish registers link birth records to later diagnoses, so an infant who bleeds at four months can be traced back to what happened on day one.
What refusal tracked with
The gap widens as the outcome gets worse. Any bleeding diagnosis was around half again as likely in infants who missed the injection; bleeding inside the skull was more than three times as likely.
That is the direction you would expect if the injection is doing what it is supposed to do: the more serious the bleed, the more the injection appears to matter.
Absolute numbers stay small throughout. Tripling the odds of a rare event leaves a rare event, and most infants who miss the injection are fine.
The trend the authors flag
The uptake data is arguably the more useful half of the paper. Non-receipt fell during the first study years from 1.32% in 2003 to 0.66% in 2006, and then gradually increased and more than doubled to 1.50% in 2021.
Sweden is a high-coverage system with high institutional trust. A reversal there is harder to explain away as access failure, which leaves preference.
What Swedish registry data cannot show
Records document what was written down. Some infants coded as not receiving vitamin K may have received it, and that kind of error usually blurs a comparison rather than creating one.
Nor is refusal a random event. Families declining a routine newborn intervention may differ in birth setting, feeding and later care in ways that carry their own risk, and adjustment reaches only measured factors.
The authors state the implication plainly: these findings suggest the importance of intramuscular vitamin K administration in the newborn period to prevent bleeding episodes and the need for continued education of parents and caregivers.
Any decision about a newborn belongs with the family and their clinician. What this study adds is a number to put on one side of that conversation.
People also ask
What is vitamin K deficiency bleeding?
Newborns are born with very little vitamin K, which blood needs in order to clot, and breast milk supplies little of it. Without the injection given at birth, a small number of infants bleed, sometimes into the brain, in the first weeks or months. It is rare but can be catastrophic, which is why the injection has been routine for decades.
How large is the risk in absolute terms?
Small, because the condition is rare to begin with. Tripled odds of a rare event is still a rare event. What makes the finding matter is that the outcome is severe and the prevention is a single injection with a long safety record, so a small absolute risk is being accepted for very little in return.
Why are refusals rising?
The study documents the trend rather than explaining it. Non-receipt fell to 0.66% by 2006 and then more than doubled to 1.50% by 2021. The authors note reports of increasing parental refusal globally. Sweden is a high-trust, high-coverage health system, which makes the reversal there more notable rather than less.
Could a missing record just be a paperwork failure?
Partly, and this is the main technical limitation. Registry data captures what was documented, so some infants recorded as not receiving vitamin K may have received it. That kind of error would generally blur the comparison rather than manufacture one, but it means the exact figures should be read as approximate.
Do families who decline differ in other ways?
Almost certainly, and that is the standing weakness of an observational design. Declining a routine newborn intervention may travel with other choices around birth setting, feeding or later care, some of which could affect bleeding risk independently. The analysis adjusted for measured factors; it cannot adjust for the decision itself.