Working memory outcomes following traumatic brain injury in children: a systematic review with meta-analysis

The aim of this review is to systematically examine the literature concerning multicomponent working memory (WM)—comprising a central executive (CE), two storage components (phonological loop, PL and visuo-spatial sketchpad, VSSP), and episodic buffer (EB)—in pediatric traumatic brain injury (TBI). Electronic searches were conducted of MEDLINE, PsychINFO and EMBASE up to October 2014 with the inclusion criteria of children and adolescents with TBI, and quantitative methods to assess at least one component of WM. Meta-analytic procedures calculated pooled effect sizes for WM outcomes. Of the studies examined, 27 met the inclusion criteria. Children with TBI exhibited deficits in the CE and PL, but not in the VSSP, and no study could be found which examined the EB. Qualitative analysis found that greater TBI severity was associated with poorer CE functioning in five out of nine studies. Differences in patterns of brain activation were evident in four out of five fMRI studies that examined WM in TBI children and controls. Deficits in CE were associated with poorer mathematical skills in the only study that examined relations between WM and academic deficits. Notwithstanding the heterogeneity of the studies reviewed, TBI places children at risk of WM deficits. Moreover, this meta-analysis suggests that various components of WM have differential vulnerability to pediatric TBI, with significant deficits found in the CE and PL, but not in the VSSP (although the VSSP has rarely been examined to date). Future studies should be theoretically driven, employ tasks assessing all components of the WM model and examine the functional ramifications (including academic outcomes) of WM deficits in this population.