The 'unified hypothesis' of Geddes et al. is not supported by the data.

Inflicted head injury to the developing brain frequently results in serious disability. The pathogenesis of the neuraxial and ocular findings in infants believed to have suffered inflicted head injury remains the subject of considerable debate. Recent neuropathology studies of fatal cases of inflicted head injury and of a foetal/perinatal non-traumatic model have led to the proposal that there is a 'unified hypothesis', the essential feature of which is hypoxic brain swelling secondary to cervicomedullary injury. It has been suggested that less than violent forces may be involved and even that some cases may not be due to trauma at all. The purpose of this paper is to provide a critical review of the data upon which these suppositions are based on a background of what is already known. It is submitted that there are serious flaws in the methodology; the conclusions reached cannot logically be drawn from the data; and the 'unified hypothesis' is not supported by the evidence. On the basis of the data presented, it is also difficult to sustain the secondary hypothesis purporting to describe a minority cohort with 'infantile encephalopathy with subdural and retinal bleeding' of non-traumatic causation.