Emotion recognition and traumatic brain injury : A systematic review and meta-analysis

Traumatic brain injury (TBI) can lead to impairments in emotion recognition that can present considerable challenges to social communication and the maintenance of interpersonal relationships. This review aimed to estimate the magnitude of emotion recognition impairments in TBI patients overall, and at the emotion category level, and to determine if the magnitude of observed impairments were moderated by modality (e.g., face, voice, multi-modal) of emotional expression, and severity of injury. Searches of PubMed, Scopus, PsycINFO, and Medline databases identified 17 studies which satisfied strict inclusion and exclusion criteria for the systematic review (comparing TBI patients to matched controls). Of these studies, 15 were included in the meta-analysis (NTBI = 474; NControl = 461). Moderate/large average deficits emerged for TBI patients relative to controls (Hedges’ g = 0.79, 95% CI: 0.61 – 0.96, p < .001; Q = 22.53, p = .068, τ2 = 0.04, I2 = 37.84; indicating low heterogeneity). TBI patients were impaired across all emotion categories, with moderate/large effect sizes observed for fear and anger, moderate effect sizes for disgust, neutral and sadness, while effect sizes for happiness and surprise were small. The magnitude of impairment for individuals with TBI severity classified as moderate/severe TBI was moderate, whereas severe TBI was large. Moderate/large effect sizes were observed across the different modalities of presentation. This meta-analysis provides evidence for marked global impairments in emotion recognition, with the magnitude of impairment greatest for negative emotions (i.e., anger and fear). This meta-analysis provided no evidence to suggest that the magnitude of impairment is influenced by injury severity or modality of stimulus presentation. Recommendations for future research are discussed.