A-012 Meta-Analysis of Neuropsychological Intra-Individual Variability in Dementia and Mild-Cognitive Impairment
Abstract Objective Neuropsychological intra-individual variability (IIV) refers to within-person cognitive fluctuation. It is often measured within-task (i.e., consistency) and across-tasks (i.e., dispersion). This investigation meta-analyzed studies comparing cognitive IIV of dementia and mild cognitive impairment (MCI) participants to normal controls. Method Eight databases within the EBSCO network as well as ProQuest Dissertations & Theses were searched for original research comparing cognitive IIV between clinical and control samples. Studies were coded using a complex multi-comparison, outcome, and subgroup schema. All data were analyzed under random-effects modeling using Comprehensive Meta-Analysis. Main index of effect size was Hege’s g. The Q and I2 statistics were used to formally assess heterogeneity. Continuous and categorical moderator variables were tested using meta-regression and subgroup analysis, respectively. Results This study was a part of a broader meta-analysis looking at IIV across all clinical samples. The initial search strategy yielded 2,962 results, which were reduced to 87 studies for final inclusion. From which, the current meta-analysis included 28 consistency studies (dementia: k = 13, 88 effect sizes; MCI: k = 15, 110 effect sizes) and 16 dispersion studies (dementia: k = 10, 36 effect sizes; MCI: k = 6, 14 effect sizes). Meta-analysis revealed a greater combined effect size for dementia samples (consistency: g = 0.92; dispersion: g = 1.11) than MCI (consistency: g = 0.33; dispersion: g = 0.56). Significant between-study heterogeneity was noted across analyses. Age and education were the only notable moderators. Conclusions All clinical subgroups demonstrated significantly elevated IIV compared to controls. The extent of IIV was proportionate to neurologic burden. Dispersion-based IIV may be a more sensitive marker of neural integrity than consistency.