Intelligence after traumatic brain injury: meta-analysis of outcomes and prognosis.

Abstract

Worldwide, 54-60 million individuals sustain traumatic brain injury (TBI) each year. This meta-analysis aimed to quantify intelligence impairments after TBI and to determine the value of age and injury severity in the prognosis of TBI. An electronic database search identified 81 relevant peer-reviewed articles encompassing 3890 patients. Full-scale IQ (FSIQ), performance IQ (PIQ) and verbal IQ (VIQ) impairments were quantified (Cohen's d) for patients with mild, moderate and severe TBI in the subacute phase of recovery and the chronic phase. Meta-regressions explored prognostic values of age and injury severity measures for intelligence impairments. The results showed that, in the subacute phase, FSIQ impairments were absent for patients with mild TBI, medium-sized for patients with moderate TBI (d = -0.61, P < 0.001) and large for patients with severe TBI (d = -1.09, P < 0.001). In the chronic phase, FSIQ impairments were small for patients with mild or moderate TBI (d = -0.37 and -0.19, P ≤ 0.008) and large for patients with severe TBI (d = -0.80, P < 0.001). Adults with mild TBI had larger PIQ and VIQ impairments in the chronic phase than children (both Q ≥ 5.21, P ≤ 0.02), whilst children with severe TBI had larger FSIQ and VIQ impairments than adults (both Q ≥ 4.40, P ≤ 0.04). Glasgow Coma Scale score, duration of loss of consciousness and post-traumatic amnesia duration moderately to strongly predicted FSIQ, PIQ and VIQ impairments (0.41 ≤ r ≤ 0.82, P ≤ 0.02), but no differences in predictive value were observed. In conclusion, TBI causes persisting intelligence impairments, where children may have better recovery from mild TBI and poorer recovery from severe TBI than adults. Injury severity measures predict intelligence impairments and do not outperform one another.