Baseline neuropsychological intra‐individual variability predicts cognitive impairment 8 years later in the Brazilian Longitudinal Study of Adult Health (ELSA‐Brasil)

Abstract

AbstractBackgroundNeuropsychological assessment is still vastly based on a nomothetic approach. However, middle‐aged and older adults might start to reveal higher intra‐individual cognitive variability (IICV) as an early sign of cognitive impairment, even when still cognitively normal. IICV refers to inconsistency in function and has been related to an increased risk for dementia. Therefore, we aimed to verify if baseline IICV predicts the occurrence of cognitive impairment after eight years.MethodWe included 10,783 participants from the ELSA‐Brasil study, a multicenter, prospective cohort study with three waves, approximately four years apart each, starting in 2008. The ELSA‐Brasil baseline exclusion criteria included the presence of clinically observed severe cognitive or communication impairment. Participants' health status and cognitive performance were evaluated at baseline (W1) and Wave 3 (W3). Baseline neuropsychological assessment comprised measures of memory and executive function. We used regression corrected‐based norms, from a robust normative subsample, to generate the presence of baseline cognitive deficit (< ‐1.5 SD) (BCD) and the mean standard deviation across baseline cognitive scores (IICV). Mini‐Mental State Exam at W3 was corrected according to education‐based cutoff to verify the presence of cognitive impairment. Logistic regression, adjusted for sociodemographic (age, sex/gender, education, race/ethnicity) and clinical confounders (depressive symptoms, hypertension, diabetes, and alcohol consumption), was used to predict the odds ratios (OR) of being cognitively impaired after eight years of follow‐up when considering IICV, BCD and both.ResultParticipants had a mean age of 59.0±8.7 years old, 56% were women, 7% had less than high school level, and 42% were Black/Brown (Table 1). IICV within tasks ranged from 0.7 to 7.9 standard deviations, and only 1% of the sample had BCD. Higher IICV is associated with a higher occurrence of cognitive impairment after eight years, even after controlling for other risk factors (Table 2). IICV was associated with cognitive impairment even after controlling for BCD.ConclusionHigher cognitive intra‐individual variability among test scores at baseline was associated with a higher risk for cognitive impairment later in life. Cognitive heterogeneity should be considered during the neuropsychological assessment interpretation and can be a target for the detection of pre‐clinical cognitive impairment.