Post Doctoral Fellow McLean Hospital Nashua, New Hampshire, United States
Objective: This meta-analysis summarized and compared the effect sizes of neuroimaging to diagnose Alzheimer’s Disease (AD) to neuropsychological tests to diagnose AD and traumatic brain injury (TBI). Data Selection: EBSCO network was searched for original research examining the efficacy of neuroimaging for AD and common neuropsychological tests for AD and TBI within the last 5 years. Other inclusion/exclusion criteria were: clinical and control group, efficacy statistics, and sufficient descriptive statistical information (e.g., mean, SD) for each group. Data Synthesis: Of 3,488 studies identified, 56 were retained for meta-analyses (502 effect sizes extracted; 3,856 clinical and 4,112 control participants represented). Using comprehensive meta-analysis software and random-effects modeling, results indicated the neuroimaging effect size [Hedges g = -1.670] was not significantly different than the neuropsychological test efficacy [Hedges g = -1.548; Q(1) = 0.213, p = 0.644]. Within neuropsychological tests, the AD neuropsychological test effect size [Hedges g = -2.213] was significantly different than the TBI neuropsychological test efficacy [Hedges g = -0.649; Q(1) = 42.821, p = 0.000]. There was a significant degree of between-study heterogeneity for both neuroimaging [Q(21) = 298.655, p < .001, I2 = 93.303] and neuropsychological [Q(39) = 860.715, p < .001, I2 = 95.469] studies.
Conclusions: The diagnostic efficacy of neuroimaging and neuropsychological tests were both substantial and non-significantly different from one another. These findings are consistent with prior research suggesting neuropsychological tests are as effective as neuroimaging and provides clinicians and consumers convincing evidence that neuropsychological tests are a reliable diagnostic tool among people with acquired and neurodegenerative brain disorders.
