A subtle injury that lurks invisibly inside the developing elbows of infants and toddlers—and one that doctors frequently miss—is finally getting a clearer diagnostic fingerprint. Distal humeral physeal separation, an uncommon fracture in which the entire cartilaginous end of the humerus shears away from the shaft through the growth plate, has long frustrated radiologists because the affected region is largely unossified in very young children. On a plain X-ray, the injured cartilage is essentially invisible, and the true extent of the damage can be masked by anatomy that has not yet turned to bone. Now, a retrospective comparative study from The Children’s Hospital of Philadelphia, published in Pediatric Radiology, has systematically compared the radiographic features of this injury against normal elbows in the same age group, and the results point to one finding that towers above all others: abnormal medial alignment of the forearm relative to the distal humerus.
The clinical stakes of this research are considerable. Distal humeral physeal separation occurs almost exclusively in immature elbows, typically the result of birth trauma, accidental injury, or—critically—non-accidental trauma. Because the ligaments and joint capsule of a young child’s elbow are stronger than the growth plate cartilage itself, twisting or pulling forces tend to fail through the physis rather than dislocating the joint. True elbow dislocation is exceedingly rare in infants, which means that when a radiograph shows the forearm displaced relative to the humerus, a physeal separation should be the leading consideration. The injury also carries a strong association with child abuse, and its radiographic recognition may be the first clue of unsuspected abuse when the history provided is absent or inconsistent. A missed or delayed diagnosis can lead to permanent growth disturbance and lasting elbow deformity.
The challenge for clinicians is that the diagnostic toolkit is limited at the point of first contact. Ultrasound and magnetic resonance imaging can directly visualize the unossified chondroepiphyses, but these modalities are usually deployed only after the injury is already suspected. Ultrasound is portable and well tolerated by neonates without sedation but is highly operator-dependent, while MRI is constrained by availability and the need for immobilization. Plain radiographs remain the first-line imaging modality for evaluating elbow symptoms in children, yet the abundance of cartilage in the immature elbow makes confident interpretation notoriously difficult. Prior literature on this injury consisted largely of case reports and small retrospective series describing posterior or medial displacement of the radius and ulna relative to the distal humerus, without a controlled comparison against normal elbows.
To fill that gap, the research team searched their institution’s electronic imaging archive for all elbow and forearm radiographs performed in children three years of age or younger over a 16-year period, from January 2008 to December 2024. The initial search yielded 311 examinations, reduced to 259 unique studies after excluding incomplete or unavailable imaging. Reports containing terms such as physeal fracture, physeal separation, or transphyseal fracture of the distal humerus were flagged, and the complete medical records were then reviewed for clinical, imaging, or surgical confirmation of the injury. After rigorous exclusion of normal studies, forearm fractures, supracondylar and metaphyseal fractures, congenital abnormalities, dislocations, condylar fractures, and infections, 19 children with confirmed distal humeral physeal separation formed the study group. A comparison group of 32 age- and sex-matched children without elbow fracture was drawn from the same source population, whose imaging indications ranged from evaluation of congenital or syndromic abnormalities to accidental household trauma and birth-related trauma.
Two board-certified radiologists, one with more than 20 years of experience in pediatric radiology and the other with 11 years spanning pediatric and musculoskeletal imaging, independently reviewed all examinations. They were blinded to group assignment, clinical history, and prior reports, and disagreements were resolved by consensus. For each study, they recorded the presence or absence of soft-tissue swelling on both anteroposterior and lateral views, elbow joint effusion assessed by elevation of the anterior fat pad—the so-called sail sign—or visibility of the posterior fat pad, abnormal ossific fragmentation, and abnormal medial or posterior alignment of the forearm relative to the distal humerus. In children who had developed a capitellar ossification center, the radiocapitellar alignment and the anterior humeral line were also assessed. Inter-rater agreement was substantial to near perfect for most findings, with kappa values ranging from 0.68 to 0.95, although elbow effusion showed only slight agreement at 0.24.
The final cohort comprised 51 children—32 boys and 19 girls—with a median age of just 17 days and a range spanning from birth to 1,186 days. Among the 19 children with physeal separation, birth-related injury was the most common mechanism, accounting for 10 cases, followed by accidental injury in 6 and non-accidental trauma in 3. The median age at diagnosis varied strikingly by mechanism: 3.5 days for birth trauma, 101 days for accidental trauma, and 329 days for non-accidental injury. Most children were treated with closed reduction and percutaneous pinning, while the remainder were managed with immobilization, in one case augmented by external traction. Follow-up data, available for 18 children with a median of 94 days, revealed a sobering reality: more than half had residual limitations in elbow motion, and a minority developed cubitus varus, an inward angulation deformity of the arm.
The imaging results were dramatic. Soft-tissue swelling was present in 94.7 percent of injured elbows on anteroposterior views versus 34.4 percent of controls, and in 78.9 percent versus 12.5 percent on lateral views, differences that were highly statistically significant. Abnormal ossific fragmentation appeared in 78.9 percent of the separation group compared with just 3.1 percent of controls. Most striking of all, abnormal medial forearm alignment was seen in 89.5 percent of children with physeal separation and in exactly zero controls, while abnormal posterior alignment was present in 78.6 percent of the injured group and, again, in no control child. In univariable analysis, medial alignment carried an odds ratio of 455, posterior alignment 213.5, and ossific fragmentation 72.3. But when the researchers applied Firth penalized logistic regression—a method chosen to reduce the small-sample bias that plagues standard logistic regression with sparse data—only abnormal medial forearm alignment remained independently associated with the injury, with an odds ratio of 104.2.
The study also illuminated several technical pitfalls that complicate diagnosis in this population. Elbow effusion, a classic sign of intra-articular injury in older children, proved unreliable here: it was uncommon even among injured children and showed the poorest agreement between readers. The authors explain that in very young children the anterior coronoid and posterior olecranon fossae are underdeveloped, making it difficult to distinguish physiologically prominent fat pads from the pathologic displacement that signals an effusion. Positioning posed another obstacle. Among the 16 children with capitellar ossification, three had suboptimally positioned lateral views that precluded assessment of the radiocapitellar line and anterior humeral line, and the injured group was particularly prone to poor positioning—possibly because additional rotation across the injured physis makes standard views harder to obtain. Among children with adequate lateral views, an abnormal anterior humeral line was present in all children with separation and none of the controls, though the small evaluable subgroup prevented its inclusion in the regression models.
The pattern of ossific fragmentation added further nuance. Fragments in the injured group were distributed relatively evenly across the medial, posteromedial, and posterior distal humerus, mirroring the predominance of medial and posterior displacement. The authors note that some of these appearances likely represented early accessory ossification of the capitellum—the first secondary ossification center to mineralize in the elbow, typically between one and two years of age—that was overcalled as pathology in the presence of an effusion. Only a minority of children had capitellar ossification at all, underscoring the extreme skeletal immaturity of this cohort. Even when the capitellum is present, its small and often irregular early ossification makes confident assessment of the radiocapitellar line and anterior humeral line challenging, since the line drawn along the radial shaft can miss the capitellum in a substantial fraction of normal young elbows, and the anterior humeral line may transect only the anterior third of the immature ossification center in up to 40 percent of children under four.
The authors are candid about their limitations. The retrospective design restricted the clinical information available and precluded repeat imaging, and despite being the largest single-institution cohort reported to date, the sample of 19 cases remains small—a consequence of the injury’s rarity. That small size likely inflated the apparent discriminative power of medial alignment, and the final model’s area under the curve of 1.00, with 100 percent sensitivity and specificity, almost certainly reflects overfitting rather than true performance, the researchers caution. The comparison group, drawn from children imaged for syndromic workups or injury exclusion, may not be entirely normal, though the authors argue this better reflects routine clinical practice. Even so, the central message stands with unusual clarity: in an infant or toddler whose elbow radiograph shows the forearm displaced medially relative to the distal humerus, distal humeral physeal separation should be presumed until proven otherwise, prompting urgent further imaging with ultrasound or MRI, orthopedic referral, and—in cases where the history does not fit—careful consideration of non-accidental trauma. Future prospective studies incorporating comparison views of the asymptomatic side could refine these criteria and help prevent the treatment delays and growth disturbances that have shadowed this elusive injury for decades.
Subject of Research: Radiographic diagnosis of distal humeral physeal separation in infants and toddlers
Article Title: Distal humeral physeal separation: radiographic findings in infants and toddlers
Article References: Yaya-Quezada, C., Nguyen, J. C., Forero-Millan, J., Hailu, S. S., Lerebo, W. T., & Taragin, B. H. (2026). Distal humeral physeal separation: radiographic findings in infants and toddlers. Pediatric Radiology. https://doi.org/10.1007/s00247-026-06790-0
Image Credits: AI Generated
DOI: 10.1007/s00247-026-06790-0
Keywords: distal humeral physeal separation, pediatric radiology, elbow fracture, growth plate injury, infants, toddlers, radiography, child abuse, birth trauma, forearm alignment, elbow effusion, ossific fragmentation
Cite Scienmag News
APA
MLA
Chicago
Harold Sullivan. (September 26, 2026). Hidden Elbow Fracture in Babies: One X-ray Clue Stands Out, Study Finds. Scienmag. https://scienmag.com/hidden-elbow-fracture-in-babies-one-x-ray-clue-stands-out-study-finds/
Harold Sullivan. “Hidden Elbow Fracture in Babies: One X-ray Clue Stands Out, Study Finds.” Scienmag, 26 September 2026, https://scienmag.com/hidden-elbow-fracture-in-babies-one-x-ray-clue-stands-out-study-finds/. Accessed 26 September 2026.
Harold Sullivan. “Hidden Elbow Fracture in Babies: One X-ray Clue Stands Out, Study Finds.” Scienmag. September 26, 2026. https://scienmag.com/hidden-elbow-fracture-in-babies-one-x-ray-clue-stands-out-study-finds/
Copy citation
Download RIS
Tags: birth traumachallenges in pediatric fracture detectionchild abusedistal humeral physeal separationdistal humeral physeal separation in childrenearly detection of infant elbow injurieselbow effusionelbow fractureforearm alignmentgrowth plate injurygrowth plate injury in toddlersimportance of medial forearm alignment in diagnosisInfant elbow fracture diagnosisinfantsnon-accidental trauma in infantsossific fragmentationpediatric elbow injury imagingpediatric elbow X-ray interpretationpediatric radiologypediatric radiology fracture cluesradiographic signs of elbow trauma in infantsradiographysubtle signs of unseen elbow injuriestoddlers
