Test-Retest Reliability and Reliable Change on the NIH Toolbox Cognition Battery

Area of Science:

• Neuroscience
• Psychology
• Cognitive Science

Background:

• Detecting cognitive improvement or decline in individuals is crucial for research and clinical practice.
• Reliable change methodology provides a framework for assessing meaningful changes in cognitive performance over time.
• The National Institutes of Health Toolbox Cognition Battery (NIHTB-CB) is a widely used tool for assessing cognitive function.

Purpose of the Study:

• To examine the reliability of change in cognitive test scores using test-retest data from the NIHTB-CB normative sample.
• To establish cutoffs for determining reliable change in fluid cognition at the individual level.

Main Methods:

• The study utilized test-retest data from 138 adults (ages 18-85) who completed five fluid cognition tests from the NIHTB-CB.
• Participants were assessed approximately one week apart, providing raw scores, age-adjusted standard scores (SS), and demographic-adjusted T-scores (T).
• Reliable change was assessed using intraclass correlation coefficients (ICCs) and confidence intervals (CIs) around change scores, adjusting for practice effects.

Main Results:

• The Fluid Cognition Composite and individual tests demonstrated good to excellent test-retest reliability (ICC range: 0.64-0.87).
• Confidence intervals (70%, 80%, 90%) were calculated to define thresholds for reliable change.
• Using a 90% CI and adjusted T-scores, a significant proportion of participants showed reliable change: 32.3% declined on at least one test, and 36.6% improved on at least one test.

Conclusions:

• It is common for individuals to exhibit reliable change on at least one fluid cognition test, but less common on two or more.
• Specific cutoffs (e.g., 80% CI) were determined for each test and the composite score to indicate reliable change.
• These reliable change cutoffs can be applied in research and clinical settings to accurately detect within-person changes in fluid cognition.