Even though fractures of the shafts of long bones may well occur from accidental trauma, such fractures in the first year of life, and to some extent in the second year, are often due to abuse, especially in the nonmobile baby. Many statistical reviews, for example [1], have shown a majority of such fractures, notably in femur or humerus, have been found to be abusive. Finding fractures of different ages , such as one without callus (injury less than 10 days old) and another with callus (greater than 10 days old), should increase the suspicion (Figs. 3.1 and 3.2). The orientation of fracture, such as spiral vs. transverse, should not be used as evidence for or against abusive trauma.

Fractures of certain bones are particularly suspicious for abuse, including the clavicle (Fig. 3.3) in the first year of life but beyond birth trauma, the acromion (Fig. 3.3), the sternum (look carefully on lateral chest and thoracic spine images), and the first rib (Fig. 3.4). With regard to first rib fractures, do not misdiagnose as a fracture a pseudarthrosis of a cervical rib to a process extending upward from the subjacent rib (Fig. 3.5). This is a hox (homeobox) gene abnormality seen occasionally (perhaps one in 200 persons) [2]. With regard to acromion fractures (Fig. 3.6), the finding of bilateral acromion fractures of the same age may be a consequence of neonatal tetanus [3] and Currarino has described normal acromion lucencies in infancy that simulate fractures [4]. Healing acromion fracture in infancy may show alterations in radiodensity (callus) without frank periosteal reaction.

Fractures of the small tubular bones of the hands and feet in the first year of life should be considered a consequence of abuse (Fig. 3.7), absent metabolic bone disease or, rarely, self-mutulating conditions such as Lesch-Nyhan or Riley-Day syndromes.

With questionable fractures, the follow up after more than 10 days may reveal callus/periosteal reaction that confirms the fracture (Fig. 3.8a, b). A delayed finding of deformity, including of a small tubular bone (Fig. 3.9), may signify prior abuse.

Beware of the manubrium centers simulating rib callus on oblique infant chest images, when they overlap (see Fig. 2.​13).

One fracture which, in the absence of a compelling other explanation, is strongly suspicious for abuse is widely known as the classic metaphyseal lesion. Corner fractures and bucket handle fractures are both manifestations of the classic metaphyseal lesion, each appearing depending on the angle of the x-ray beam to the fracture (Figs. 3.10 and 3.11). See also Figs. 3.1 and 3.2 for examples of the corner fracture appearance. A CML-like fracture occasionally is a manifestation of childbirth (Fig. 3.12). The use of the follow up at least 10 days later may confirm or rule out a questionable corner fracture (Fig. 3.13).

Corner and bucket handle fractures (CMLs) have occasionally been seen in patients with osteogenesis imperfecta [5], acquired copper deficiency [6], in conjunction with difficult IV line placement [7]. Nonetheless, CMLs have a high specificity for abusive trauma.

In general it takes 10 days (well, perhaps 9 days) for periosteal reaction to calcify sufficiently to be visible on conventional radiographs (Fig. 3.14). It takes that long for the matrix vesicles involved in calcification to achieve a proper calcium phosphorus ratio. I think that those investigators who state calcification can occur days earlier err either because of the coincident presence of physiologic periosteal reaction along an involved long bone (which occurs generally between 1 and 6 months of child’s age) or because of improper dating of the traumatic event. A classic description of physiologic periosteal reaction at an equivalent age in puppies was published by Volberg and colleagues [8]. The high incidence of physiologic periosteal reaction in both healthy and sick children was classically commented upon by Shopfner [9].