Longitudinal changes in response time mean and inconsistency exhibit predictive dissociations for risk of cognitive...

Although mean response time (RT) is a commonly used indicator of cognitive function, trial-to-trial variability (RT inconsistency [RTI]) represents a dissociable dimension of performance arguably more sensitive for characterizing cognitive status. The present study explores whether (a) RT mean or inconsistency reflects a more dispositional characteristic of an individual, particularly with increased cognitive impairment; (b) RT mean and inconsistency exhibit comparable patterns of change across a 4-year period; and (c) these rates of change differentially predict cognitive status.

A sample of 304 adults (64-92 years) at baseline completed a choice RT task weekly for 4-5 weeks, repeating this protocol and a basic neuropsychological assessment annually for 4 years. Three cognitive status subgroups were identified at baseline and Year 4: healthy controls (HCs), as well as cognitively impaired-not-demented (CIND) status based upon single (CIND-S) and multiple (CIND-M) domains.

RTI was disproportionately stable among cognitively compromised individuals at baseline, suggesting that increased RTI is a dispositional characteristic reflecting underlying changes in psychological process or neurological function. Notably, 4-year trajectories of change for RT mean versus RTI were dissociable; whereas the HC and CIND-S groups exhibited faster mean RTs over time, CIND-M individuals exhibited significant RTI increases. Multinomial logistic regressions demonstrated that greater RTI at baseline (OR = 2.58, p < .01) and steeper 4-year increases (OR = 4.40, p < .01) were linked to increased likelihood of CIND-M classification at Year 4, independent of age, education, chronic health conditions, and mean RT.

RT mean and RTI confer distinct sources of information about cognitive function and status. Overall, RTI holds promise as an early indicator of normal and pathological cognitive aging. (PsycINFO Database Record (c) 2020 APA, all rights reserved).