Shaken baby syndrome (SBS), in its many guises (abusive head trauma, non-accidental injury, etc.) has been widely accepted and taught among paediatricians for more than 50 years. The central tenet of the hypothesis is that shaking can cause any or all of subdural haemorrhage (SDH), retinal haemorrhage (RH), and encephalopathy. These same pathologies are seen in normal newborn babies and infants after a range of insults, including trauma, and reflect the immature anatomy and pathophysiology of the infant brain and its covering membranes. Spinal damage is increasingly invoked to support the shaking diagnosis. This chapter examines the various brain, eye, and spinal pathologies claimed to be due to shaking, setting them in the context of the anatomy and specific vulnerabilities of the infant. We evaluate the empirical evidence that neuropathology can provide to support or refute these claims.

A systematic review by the SBU identified evidence gaps in diagnosing shaken baby syndrome. Population epidemiological studies, and clinical epidemiology, case-series and case-control studies, from Sweden, based on health registers (ICD-codes) and records for infants born 1997 to 2014, and forensic investigation, may add information to improve the diagnostic process of infant abuse. Our findings to date can be summarised as: perinatal exposure; small-for-gestational age, preterm, multiple birth, or male sex, increase the risk for SDH (subdural haemorrhage). Infants with chronic SDH more often had an abnormal increase in head circumference before or at the time of diagnosis. Intra- and inter-country differences in abuse diagnosis, and findings attributed to SBS/AHT indicate different prevailing practices and different interpretation of current understanding of injuries caused by abuse. A false-positive diagnosis of abuse is detrimental to the family. Further research on infant abuse, its circumstances and the specific findings indicative of abuse, is urgently needed to support evidence-based child protection, and to keep false positives and false negatives to a minimum.