Why Very Premature Babies Have Fragile Bones — and How Neonatal Teams Are Getting Better at Finding It

What four recent studies from Saudi Arabia, Türkiye, South Korea, and an international research review tell families about preterm bone health

Babies born very early miss the last three months of pregnancy, when the skeleton is built and the placenta supplies calcium and phosphorus faster than any feeding method can match afterwards. The result — softer, thinner bones, called metabolic bone disease of prematurity — is usually silent and found on a routine blood test. Four studies from 2025 and 2026 clarify how common it is, how to detect it, and what prevents it.

Calcium and phosphorus are the two minerals that make bone hard, and a baby born at 25 weeks simply misses most of that delivery. Although the name sounds alarming, this is not a disease a parent would notice at the cot side, and in most cases it improves over the first year or two of life. What has changed recently is the confidence with which neonatal teams can say who is affected and what helps.

The problem families used to face, and what changed

The condition was first properly described in the early 1980s, when doctors X-rayed premature babies and found thin bones, ragged edges at the growing ends, and — in the worst cases — healed fractures of ribs and arms that nobody had realised had happened [1]. At that time, intravenous nutrition contained far less mineral than it does today, and babies born as early as 25 weeks rarely survived. As survival improved, the bone problem did not vanish; it became quieter. Instead of babies arriving at follow-up clinic with visible rickets, teams began finding abnormal blood results in babies who looked well. That created a new difficulty, which reviews of the field spent the next two decades describing without solving: there is still no single agreed test that says yes or no [2] [3] [4]. The most commonly used blood marker, an enzyme called alkaline phosphatase, is raised in many premature babies whether or not their bones are affected. Blood calcium usually stays normal, because the body pulls calcium out of the bones to keep the blood level steady — so a normal calcium can be falsely reassuring. And an X-ray does not show anything until roughly a quarter or more of the bone mineral has already been lost. Recommendations on how much calcium and phosphorus preterm babies should receive have been reasonably settled for over a decade [5] [6], but a national survey of specialists found no agreement at all on when to test, what number counts as abnormal, or when to X-ray [7]. That is the gap these new studies address.

How common is it? It depends on how you define it

A hospital in Riyadh, Saudi Arabia, studied 487 babies born before 28 weeks and under 1,000 grams — about two pounds three ounces — cared for between 2017 and 2024 [8]. Every one of them had a blood test at four weeks of age as part of a formal bone health programme. Using a hormone called parathyroid hormone as the main marker, 41.5% of these very small babies met the definition. A hospital in Bursa, Türkiye, studied 413 babies born at or before 32 weeks or under 1,500 grams, and used a stricter definition requiring three abnormal results together [9]. On that definition only 19.8% qualified. Both studies are honest and well conducted; they simply drew the line in different places on the same spectrum. This is the single most important thing for families to understand about published numbers: when one hospital reports 4 in 10 and another reports 2 in 10, it usually reflects the test used, not a real difference in how sick the babies were.

The Turkish study also showed what happens at the more severe end. Among the babies who met their stricter definition, about 1 in 4 had a fracture detected — and every single fracture in the whole group of 413 babies occurred in that category. Fractures in premature babies are usually found by accident on an X-ray taken for another reason, often after they have already begun healing, and babies typically show little or no distress at the time. In the Riyadh group, fractures affected 1.8% of all the babies studied. So the honest summary is: fractures are uncommon overall, but not rare among the most affected babies.

What raises the risk — and what does not

Both studies pointed to a similar list. Being born earlier and smaller raises risk, which cannot be changed. But several things can be. The clearest is how long a baby needs intravenous nutrition — a drip that delivers nutrition directly into the bloodstream while the gut is still learning to feed. In the Turkish study, needing this for more than about 19 to 20 days was on its own a useful warning sign, and each extra day added a small amount of risk [9]. The Riyadh study found the same pattern for intravenous nutrition beyond 28 days, and also flagged two medicines: steroids given after birth and water tablets (diuretics) given for more than two weeks, both of which are used mainly for babies with serious lung disease [8]. This does not mean these treatments are wrong — they are often necessary and lifesaving — but it does mean teams now watch bone health more closely in babies who receive them. Interestingly, the Riyadh study found that affected and unaffected babies received similar amounts of calcium and phosphorus, so this is not simply a story about underfeeding. Repeated blood transfusions also emerged as a risk marker in the Turkish study, probably as a sign of overall illness rather than a direct cause.

Better ways of reading the X-ray

The third study came from two hospitals in Seoul, South Korea, and asked a different question: can a computer read these X-rays better than a doctor? Researchers trained a computer program on 2,239 wrist X-rays from 814 babies under 1,500 grams, then tested it on a separate group of 261 babies from a different hospital [10]. The program performed very well. More revealing was what happened when real doctors were tested. Paediatricians reading the X-rays without help got the right answer about 65% of the time; with the computer's suggestion available, that rose to about 79%. In other words, these X-rays are genuinely hard to read, and a normal-looking wrist film reported quickly on a busy unit is not strong reassurance on its own. This is why blood tests, not X-rays, remain the main screening tool — and why teams are working towards tools that combine both.

What actually prevents it

The fourth study pulled together all the research to date: 18 studies involving 1,577 premature babies across ten countries [11]. Its main finding is encouraging and quite firm. Making sure babies receive enough calcium and phosphorus early — through the drip and then through fortified milk — cut the chance of developing the condition by roughly half. Statistically, about five babies need to receive good mineral nutrition for one to be spared the condition. The researchers also ran a formal analysis showing that enough evidence has now accumulated that no further studies are needed to establish that this works.

The second finding is about movement. Several small trials tested gentle physiotherapy — a nurse, therapist, or parent moving the baby's arms and legs through their range of motion with light pressure, for about 5 to 15 minutes a day, five days a week. Bone in all of us responds to being loaded; a baby in an incubator has almost no gravity load and very little spontaneous movement compared with a fetus pushing against the womb. These exercises improved measures of bone strength and, more reliably, improved weight gain by about 4.4 grams per kilogram per day. No harm was reported in any of the trials. The evidence for the bone measurements is less certain than the evidence for mineral nutrition, and no study has yet tested nutrition and movement together — but the two work through completely different mechanisms, so combining them is a reasonable expectation for future research.

What this means for your baby, and what comes next

If your baby was born very early, several things are likely already happening. They will have blood tests around four to six weeks of age looking specifically at bone markers. Their nutrition will be adjusted to maximise calcium and phosphorus, usually by adding a fortifier to breast milk once feeds are established — and it is worth knowing that babies fed human milk have been found to have better bone density at six years of age than those fed formula, so fortified breast milk is the target rather than a compromise [12] [6]. Vitamin D will be given, though vitamin D alone does not solve the problem; it helps the body use calcium and phosphorus but cannot replace them [13]. If results are abnormal, mineral supplements may continue after discharge — about a third of affected babies in the Riyadh study went home on them [8]. Some babies benefit from a repeat check at around their original due date, as bone markers do not always normalise by discharge [14].

Two reassurances are worth holding onto. First, this condition is a nutritional and mechanical consequence of early birth, not a sign that anything was done wrong, and not an inherited bone disease. Second, most babies remineralise their skeletons over the first one to two years and grow normally afterwards [15]. Researchers are now working on three things: agreeing a single definition so that hospitals worldwide can compare results, testing nutrition and gentle movement together in a proper trial, and following children for longer to confirm that early bone problems leave no lasting mark [11]. It is fair to ask your baby's team which bone markers they check, when they check them, and whether gentle range-of-motion exercises are something you could be shown how to do yourself — in several of the trials, parents were the ones performing them.

References

  1. Koo WW, Gupta JM, Nayanar VV, Wilkinson M, Posen S. Skeletal changes in preterm infants. Archives of Disease in Childhood. 1982;57(6):447–452. doi:10.1136/adc.57.6.447 ↩
  2. Harrison CM, Gibson AT. Osteopenia in preterm infants. Archives of Disease in Childhood — Fetal and Neonatal Edition. 2013;98(3):F272–F275. doi:10.1136/archdischild-2011-301025 ↩
  3. Rustico SE, Calabria AC, Garber SJ. Metabolic bone disease of prematurity. Journal of Clinical & Translational Endocrinology. 2014;1(3):85–91. doi:10.1016/j.jcte.2014.06.004 ↩
  4. Chinoy A, Mughal MZ, Padidela R. Metabolic bone disease of prematurity: causes, recognition, prevention, treatment and long-term consequences. Archives of Disease in Childhood — Fetal and Neonatal Edition. 2019;104(5):F560–F566. doi:10.1136/archdischild-2018-316330 ↩
  5. Abrams SA; Committee on Nutrition. Calcium and vitamin D requirements of enterally fed preterm infants. Pediatrics. 2013;131(5):e1676–e1683. doi:10.1542/peds.2013-0420 ↩
  6. Embleton ND, Moltu SJ, Lapillonne A, van den Akker CHP, Carnielli V, Fusch C, Gerasimidis K, van Goudoever JB, Haiden N, Iacobelli S, Johnson MJ, Meyer S, Mihatsch W, de Pipaon MS, Rigo J, Zachariassen G, Bronsky J, Indrio F, Köglmeier J, de Koning B, Norsa L, Verduci E, Domellöf M. Enteral nutrition in preterm infants (2022): a position paper from the ESPGHAN Committee on Nutrition and invited experts. Journal of Pediatric Gastroenterology and Nutrition. 2023;76(2):248–268. doi:10.1097/MPG.0000000000003642 ↩
  7. Chinoy A, Mughal MZ, Padidela R. Metabolic bone disease of prematurity — national survey of current neonatal and paediatric endocrine approaches. Acta Paediatrica. 2021;110(6):1855–1862. doi:10.1111/apa.15654 ↩
  8. Alsaif S, Alrahili M, Aljarbou T, Alsherbini L, Maghoula M, Alluwaymi A, Alsenani M, Altuwaym A, Alamer F, Mandurah A, Baylon B, Ali I, Ali K. Metabolic bone disease in extremely preterm infants: incidence, risk factors, and outcomes from a structured bone health program. Frontiers in Pediatrics. 2025;13:1676540. doi:10.3389/fped.2025.1676540 ↩
  9. Tunç G, Güney Varal İ, Ören A, Ülker D, Denkboy Öngen Y, Bağcı O. Biochemical metabolic bone disease in preterm infants: clinical determinants, markers and risk of fracture in a cohort study. BMC Pediatrics. 2026;26(1):487. doi:10.1186/s12887-026-06824-4 ↩
  10. Park SG, Jeong S, Cho M, Kim MJ, Choi CW, Kim EK, Cheon JE, Kong HJ. Deep learning model for identification of metabolic bone disease of prematurity using wrist radiographs. Scientific Reports. 2026;16(1):7885. doi:10.1038/s41598-026-37116-7 ↩
  11. Alalwani YJ, Alatta MA, Almana KA, Alomari A, Alshammari RH, Alsharif GN, Soweedi AM, Abualmeza SS, Alharbi SA, Maghrabi MA, Alshehri KK, Alburaey A, Azzam AY. Interventions for metabolic bone disease of prematurity: a systematic review and meta-analysis. Metabolism Open. 2026;29:100445. doi:10.1016/j.metop.2026.100445 ↩
  12. Toftlund LH, Beck-Nielsen SS, Agertoft L, Halken S, Zachariassen G. Higher bone mineral density at six years of age in very preterm-born infants fed human milk compared to formula feeding: a secondary analysis of an RCT. European Journal of Pediatrics. 2025;184(2):160. doi:10.1007/s00431-024-05935-3 ↩
  13. Radu IA, Ognean ML, Ștef L, Giurgiu DI, Cucerea M, Gheonea C. Vitamin D: what we know and what we still do not know about vitamin D in preterm infants — a literature review. Children (Basel). 2025;12(3):392. doi:10.3390/children12030392 ↩
  14. Asfa AS, Surender K, B S CL, Jaya Chandra T. A cross-sectional study on corrected-term screening for osteopenia/metabolic bone disease of prematurity in preterm and low-birth-weight neonates: prevalence, risk factors, and feeding-related predictors. Cureus. 2026;18(6):e111595. doi:10.7759/cureus.111595 ↩
  15. Faienza MF, D'Amato E, Natale MP, Grano M, Chiarito M, Brunetti G, D'Amato G. Metabolic bone disease of prematurity: diagnosis and management. Frontiers in Pediatrics. 2019;7:143. doi:10.3389/fped.2019.00143 ↩