Victims of AHT may also have cervical spinal and ligamentous injuries. Infants have a proportionally larger head and weaker neck muscles, making them vulnerable to cervical injuries that result from high-energy rotational forces as the head moves violently back and forth. Cervical injuries observed with AHT include cervicomedullary junction injury, vertebral body subluxations or fractures, traumatic axonal damage of the cervical spine, and primary cervical cord injury (Kadom et al. 2014; Narang and Clarke 2014).

Another common injury related to AHT are retinal hemorrhages, which results from the violent rotational movements of the eyes during shaking that causes the vitreous humor to exert extreme traction on the retina (Yamazaki et al. 2014). Retinal hemorrhages occur in 50–95% of children who are victims of AHT (Binenbaum and Forbes 2014) and have a 71% positive predictive value (Maguire et al. 2009). These hemorrhages tend to be bilateral, widespread to the outer margins of the retina, too numerous to count, and present in all layers of the retina (Maguire et al. 2013) (Fig. 10.2). Retinal hemorrhages also occur in other pediatric conditions, including unidirectional/blunt head trauma, infections, bleeding disorders, metabolic disorders, vaginal births, and after cardiopulmonary resuscitation (Shaahinfar et al. 2015). However, these hemorrhages are usually unilateral, fewer in number, and localized to the optic disks and posterior pole of the retina (Shaahinfar et al. 2015). The majority of RH associated with vacuum-assisted birth resolve within 1 week and all resolve within 4 weeks of age (Laghmari et al. 2014). Up to 15% of non-abused children in the pediatric intensive care unit may have RH, but these are mild, located only on the posterior pole, and found only in the retinal layer (Agrawal et al. 2012; Longmuir et al. 2014). In the absence of a documented history of major trauma, retinal hemorrhages should be considered highly suspicious for AHT.

Fractures may occur as a result of child maltreatment, and rib fractures are strongly associated with AHT, having a 73% positive predictive value (Maguire et al. 2009). These fractures are thought to occur from squeezing around the infant’s chest during shaking. Healing rib fractures may be seen when the abuse has occurred over time. Other fractures, such as metaphyseal injuries, may occur along with AHT, with the most common type of long bone fractures associated with AHT being those of the humerus and femur, particularly in nonmobile infants. Osteogenesis imperfecta, prematurity, vitamin and mineral deficiencies, and previous injuries can make a child vulnerable to bone fractures, and these should be considered on the list of differential diagnoses (Table 10.1) when evaluating the child for AHT and abuse (Flaherty et al. 2014).

Skull fractures can occur with and without intracranial bleeding, and intracranial bleeding can be present without associated related skull fracture. Skull fractures that result from accidental injuries tend to be unilateral, be linear, and occur in the thinner parietal bones, as the mechanism of injury is generally from lateral movement during accidental falls. Skull fractures in other locations and those that cross suture lines are associated with more severe impact, result in significant brain injury, and thus may be indicative of AHT (Roach et al. 2014) (Fig. 10.3).

Both clinical and subclinical seizures may arise from cellular injury in the brain (Narang and Clarke 2014). Over 50% of victims of AHT develop seizures (Hasbani et al. 2013), and younger age and intraaxial bleeding are risk factors for both status epilepticus and subclinical status epilepticus (Arndt et al. 2013). Seizures are best determined when the child undergoes continuous electroencephalographic (cEEG) monitoring as soon as possible after admission for suspected AHT (Paul and Adamo 2014).

Presenting signs and symptoms of AHT will vary depending on the extent of brain injury and the type and severity of accompanying injuries. Manifestations may include nonspecific clinical findings such as an altered level of consciousness, lethargy, irritability, seizures, poor feeding, vomiting, and respiratory changes, including apnea. Apnea is a critical distinguishing feature for AHT compared to accidental head injury, having a positive predictive value of 93% (Maguire et al. 2009). Other conditions that may be associated with AHT include expanding head circumference, failure to thrive, and developmental delay. Since AHT is the most common cause of brain injury in children less than 2 years old, it should be suspected in all children who present with signs of neurologic trauma, unless that trauma is unquestionably accidental (Narang and Clarke 2014). Bruising may be observed, depending on the mechanism of injury, with patterned bruising being more commonly associated with child maltreatment. However, since bruising occurs <50% of the time in AHT, the absence of bruises does not rule out AHT (Fanconi and Lips 2010). Bruising in nonmobile infants and bruising of the head and neck are concerning for AHT (Pierce et al. 2010). It can be challenging to differentiate inflicted from accidental injuries. Table 10.2 further compares injury patterns common to accidental versus abusive injuries (Maguire et al. 2009; Piteau et al. 2012).