Age-invariant approaches to cognitive reserve.

Abstract

Cognitive reserve (CR) and its measurement by proxies have gained interest in cognitive aging research. While CR proxies seem valuable for predicting cognitive function, their measures are often conflated with age effects. The current study aims to address this by introducing an age-invariant approach of CR. We included 380 participants (age = 18-79) from the Advanced Brain Imaging on aging and Memory (ABRIM) study who completed the Cognitive Reserve Index questionnaire (CRIq), a measure to estimate verbal IQ, and several neuropsychological tasks in the domains of memory, executive function and attention/speed. With various regression models and structural equation modeling, we assessed age effects on the CRIq subscales and their predictive value on cognitive function. Results showed a significant non-linear age effect on the Education and Occupation subscale of the CRIq and a linear age effect on the Leisure subscale. New age-corrections derived from these effects were more accurate than age-corrections from the original norm scores. Moreover, the three cognitive domains were significantly predicted in the expected direction by the new age-corrected CRIq scores, and not by the raw scores or original age-corrected scores. However, compared to verbal IQ, the predictive value of these CRIq scores was still low. Associations between the CRIq and cognitive function seemed to vary across the lifespan, but were not consistently stronger for older adults. These findings illustrate the importance of age adjustments in CR research. Most importantly, appropriate age-adjustments may be sample specific and non-linear effects to properly correct for age must be considered.