Abstract

AbstractBackgroundAccumulating evidence suggests that fluctuations in reaction time (RT) performance across assessments, usually defined by intraindividual variability in RT (IIV‐RT), are associated with cognitive impairment and decreased neurobiological integrity. Computerized cognitive testing facilitates the study of IIV‐RT, allowing for precise and frequent RT measurements in remote settings. We investigated whether quantifying IIV‐RT over months on various remotely administered computerized RT tasks could aid in the detection of emerging cognitively decline in initially clinically unimpaired (CU) older adults with elevated Alzheimer’s disease (AD) biomarkers.MethodN = 114 CU individuals (age 77.6±5.1, 61% female, MMSE 29±1.2, n = 33 amyloid‐positive as measured using PiB‐PET imaging) from the Harvard Aging Brain Study (HABS) completed the self‐administered Computerized Cognitive Composite (C3) monthly at‐home on an iPad for up to one year (11.7±3.2 C3 assessments). Baseline C3 assessment coincided with participants’ HABS in‐clinic visit including Preclinical Alzheimer’s Cognitive Composite (PACC) testing, and participants had 1.8±1.2 subsequent years follow‐up PACC testing available. The C3 includes two basic RT tasks (Detection [DET]; Identification [IDN]), and two complex RT tasks (One Card Learning [OCL]; One Back [ONB]) that also require working memory skills. IIV‐RT measures were derived for each task by computing the intraindividual standard deviation between mean RTs across all monthly visits for each participant. Subsequently, IDN and DET IIV‐RT were combined into a basic IIV‐RT composite, and OCL and ONB IIV‐RT into a complex IIV‐RT composite. Linear mixed models were used to investigate the association between both IIV‐RT composites and PACC decline adjusting for covariates (age, sex, education, amyloid status (+/‐)) and their interaction with time.ResultGreater IIV‐RT on complex RT tasks was associated with steeper decline on the PACC (Time*IIV‐RT ß = ‐1.3, 95%CI[‐2.57 – ‐0.03],p = .04) (Figure 1), which was driven by amyloid‐positive individuals (Time*Amyloid[+]*IIV‐RT ß = ‐3.52, 95%CI[‐6.43 – ‐0.61],p = .02) (Figure 2). In contrast, greater IIV‐RT on basic RT tasks was only associated with baseline PACC performance (IIV‐RT ß = ‐4.29, 95%CI[‐7.32 – ‐1.26],p = .007) but not with PACC change over time.ConclusionOur findings suggest that fluctuations in RT performance over monthly‐repeated complex RT tasks may reflect evidence of early amyloid‐related neurodegeneration and provide a marker of incipient cognitive decline in preclinical AD.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call