Development and Validation of Response Time Inconsistency in the Midlife in the United States (MIDUS) Study

TL;DR

This study validates response time inconsistency as a cognitive health indicator using telephone-based assessments in a midlife U.S. cohort.

Contribution

Demonstrates validity of telephone-derived response time inconsistency for assessing CNS integrity and dementia risk in midlife populations.

Findings

RTI increases with age and is lower in more educated individuals and women.

Higher RTI correlates with worse cognitive performance and poorer self-rated health.

RTI shows no significant link to self-reported positive or negative affect.

Abstract

Response time inconsistency (RTI), the trial-to-trial variability on a RT-based cognitive task, is an important behavioral indicator of cognitive health, cognitive aging, and an early indicator of cognitive pathology, and central nervous system (CNS) dysfunction. Although RTI is well-studied through computer-based speeded tasks, little research has examined the validity of RTI obtained from: (1) telephone-based cognitive assessments employing voice-triggered response time protocols; or (2) midlife cohorts who are at risk for dementia. Using data from the second wave of the MIDUS study (N = 4,285; Mage=55.6, SD = 12.2, Range=28-84; 55%=women; 75% having at least some college education), participants completed the Brief Test of Adult Cognition via Telephone (BTACT), which includes tests of executive function and episodic memory abilities, as well as a RT-based stop-go switching (SGS) task. RTI was derived from 20 trials of the SGS non-switch condition and quantified as an intraindividual standard deviation (ISD) residualized for trial number. RTI was greater with age (r=.26, p<.0001), being less educated (r=-.16, p<.0001), and among women (Cohen’s d=.15, p<.0001). Importantly, RTI was associated with slower task switching performance, lower scores on the BTACT, episodic memory, and executive function cognitive composites (rs=-.19 to -.41, ps<.0001), as well as poorer self-rated mental (r=-.13) and physical (-.18, ps<.0001) health. RTI was not significantly related to self-rated positive affect (r=-.02, p=.24) or negative affect (r=.02, p=.17). Discussion will focus on the validity and promise of RTI obtained from telephone-based cognitive assessments for characterizing CNS integrity and dementia risk in large-scale surveys.