Childhood maltreatment is associated with later traumatic stress and dissociative symptoms, but outcomes and neurobiological correlates are heterogeneous.

To synthesize structural, white-matter, functional-connectivity, and stress-biological findings, distinguish empirical associations from mechanistic hypotheses, and examine the limits of their relevance to clinical formulation.

Structured narrative review drawing on PubMed/MEDLINE, Google Scholar, and citation tracking, updated through six targeted PubMed queries on 11 October 2026. Purposive selection and qualitative appraisal considered study design, exposure ascertainment, population, confounding, and consistency.

Three structural findings recur across syntheses: smaller hippocampal volume in adults with maltreatment-related posttraumatic stress disorder (PTSD), with small estimates in population samples; lower midcingulate thickness and volume; and lower fractional anisotropy in anterior callosal and fornix pathways. Resting-state syntheses in youth report lower insula–cingulate and higher amygdala connectivity, and longitudinal and randomized caregiving studies link neural change to later symptoms and to enhanced care. Dissociation research includes resting-state associations, task-dependent differences, and experimental null findings; the reported hippocampal CA1–amnesia association remains exploratory. Comparative PTSD trials do not establish a general advantage of preparatory phases, although some multicomponent treatments improve outcomes in selected populations. Evidence for dissociative disorders remains less developed.

The central distinction is between an exposure-related association, a symptom mechanism, and a useful treatment decision. Developmental and network models organize testable hypotheses but provide no validated rule for diagnosing dissociation or selecting psychotherapy. Clinical translation should combine direct assessment of symptoms and safety with comparative treatment evidence, patient preferences, and observed response.