Concurrent and longitudinal relationships between cognitive activity, cognitive performance, and brain volume in older adult women.

Abstract

We investigated (a) cross-sectional associations between cognitive activity, cognitive performance, and MRI measures and (b) longitudinal associations between cognitive activity and change in cognitive performance, using structural equation modeling (SEM). Women's Health Initiative Memory Study (WHIMS) Extension participants who continued annual neuropsychological assessments by telephone and completed a concurrent questionnaire of cognitive activities and MRI scans were included (mean age = 81.4 years; N = 393). Cognitive performance was measured by tests of attention, working memory, verbal fluency, executive function, and memory. Cognitive activity was measured by self-reported participation in a variety of cognitive activities (e.g., reading books, playing games, computer activities; N = 11 items) during the previous 12 months. MRI measures included gray and white matter normal and white matter lesion volumes. SEM demonstrated a significant association between cognitive activity and baseline cognitive performance but not change over 2-3 years. Gray and white matter was associated with cognitive performance but not cognitive activity. All effects remained significant after modeling covariates (age, education, depressive symptoms, WHIMS intervention assignment, and intracranial volume). Cognitive activity benefits current cognitive performance but is not associated with change over 2-3 years. Cognitive activity and MRI volumes are independently associated with cognitive performance, suggesting distinct cognitive and brain reserve constructs.