Childhood emotional maltreatment and trauma-related outcomes: a systematic review and multilevel meta-analysis
TL;DR
These findings underscore the importance of CEM in trauma-related symptomatology and support the need for its systematic assessment in clinical and research settings, particularly by highlighting differences between emotional abuse and emotional neglect.
Abstract
ABSTRACT Background: Estimating the association between childhood emotional maltreatment (CEM) and trauma-related outcomes is important for prevention and intervention. Objective: We synthesised the evidence by differentiating between emotional abuse and emotional neglect, examining post-traumatic stress disorder (PTSD) and complex PTSD (CPTSD) and exploring moderators and mediators. Method: Following PRISMA guidelines, seven databases were searched up to 19/03/2026 (PROSPERO-CRD42024598358) for studies on the association between CEM and PTSD/CPTSD. Study quality was assessed using an adapted Newcastle-Ottawa Scale (NOS), and certainty of evidence was evaluated using the GRADE framework. Results: Of 11,479 records, 62 were included in the review, of which 57 studies provided 132 effect sizes for a random-effects multilevel meta-analysis (N = 323,776). Study quality was overall fair (Mean NOS = 5.8). A positive association between CEM and PTSD/CPTSD (k = 132, r = .27, 95% CI [.23, .30]) 95% PI [−.04, .53] with substantial heterogeneity (I² = 86.5%). Higher estimates were identified for emotional abuse (k = 71, r = .29, 95% CI [.25, .33]) compared to emotional neglect (k = 55, r = .24, 95% CI [.19, .28]). Effect sizes were similar for CPTSD (k = 29, r = .29, 95% CI [.24, .35]) and PTSD (k = 103, r = .26, 95% CI [.22, .30]). No differences were observed by sex, age, or population type (clinical vs. non-clinical). However, higher estimates were identified in high-income settings and more developed contexts. Narrative synthesis identified cognitive, affective, physiological and interpersonal mechanisms as potential mediators. The certainty of the evidence was generally low. Limitations include the predominance of cross-sectional designs, reliance on self-report measures of CEM and variability across outcome measures. Conclusions: These findings underscore the importance of CEM in trauma-related symptomatology and support the need for its systematic assessment in clinical and research settings, particularly by highlighting differences between emotional abuse and emotional neglect.