During the 1980′s a sudden life-lessness attack in a very young infant was described as “near-SIDS”, and a caregiver who observed and tried to revive the infant was considered to be reliable and honourable. If, after the later acceptance of the traditional shaken baby theory, bilateral thin subdural hemorrhages (SDH) and retinal hemorrhages (RH) are detected after a lifelessness attack such a reliable caregiver has by default been transformed into an unreliable perpetrator of violent shaking. In 2016, the term “near-SIDS” was – via “apparent life-threatening event” (ALTE) − replaced by “brief resolved unexplained event” (BRUE). We suggest that such a life-threatening event is not always brief, and not always resolved. Based on a neglected hypoxia theory from 2004 we suggest as a hypothesis that a prolonged episode of apnea can result in brain hypoxia, brain edema, SDH, RH and increased intracranial pressure. The new hypothesis includes two new concepts: Prolonged Non-fatal Unexplained Event (PNUE) and Prolonged Fatal Unexplained Event (PFUE) which together with the hypoxia theory might explain why these two separate diagnoses have been conflated into one, viz. shaken baby syndrome (SBS). We suggest a study design that might corroborate or falsify the hypothesis, and if our hypothesis is corroborated, current “unreliable perpetrators” could be re-established as reliable resuscitators.