ADHD in Adults: A Meta-Analytic Review
From MindLinc Wiki
The primary objective of this meta-analysis was to perform a comprehensive review of the entire body of published neuropsychological test results in patients with ADHD.
A total of 33 studies using over 25 different neuropsychological tests were identified and analyzed in this study. The authors grouped these tests into 7 cognitive domains (attention, response inhibition, executive function, memory, processing speed, intelligence and “other”) and calculated effect sizes to measure each test’s ability to discriminate between a group of pts. w/ ADHD and a control group. Weighted mean effect sizes were calculated whenever more than one study employed the same neuropsychological test.
Here is a brief summary of selective results: In the attention, response inhibition and intelligence domains, the effect sizes for the different neuropsychological tests were largely in the medium range, suggesting these domains are significantly impaired in adults w/ ADHD when compared to controls. In the executive function domain, the Wisconsin Card Sorting Test demonstrated a small effect size, an interesting finding given the WCST has been used extensively in the child literature to document ADHD deficits in executive functioning. In the memory domain effect sizes were also largely in the medium range, although ADHD patients appeared to have less deficits on tests which used visual stimuli, e.g. parts of the Wechsler Memory Scale. The processing speed domain yielded mostly small effect sizes.
Study limitations: Confidence intervals for the effect sizes were not included, variability in effect sizes between 2 studies performing the identical test was often large, the study question was broad, there was no assessment of the validity of specific tests, no test of heterogeneity was done before pooling results, control groups were usually unmatched, there was no assessment of unpublished studies.
Commentary: While an in-depth explanation of each neuropsychological test and an exhaustive analysis of the results was beyond the scope of the presentation, Dr. Susan Johnson, our expert discussant, was able to clarify some important points. The neuropsychological tests reviewed in the study were by and large well established, validated tests. The authors’ decision to study each test in a discreet domain was problematic since a single test may actually measure factors in more than one domain (the authors mention this as well). With respect to the large variability in effect sizes observed when the same test was performed in different studies, Dr. Johnson stated that a large degree of variability often occurs when a given neuropsychological test is administered by different persons.
Presented by Cori Burdine, MD, May 11, 2005
Comments
