Editors' Note: Cognitive Activity and Onset Age of Incident Alzheimer Disease Dementia

Abstract

In their prospective observational cohort study, Dr. Wilson et al. explored the relationship between cognitive activity and incident dementia among consecutive elderly patients from the Rush Memory and Aging Project. The investigators hypothesized that self-reported higher levels of cognitive activity such as reading would be associated with older age of dementia onset in patients with clinical Alzheimer disease (AD). Of the 1903 included patients, 457 (24%) developed possible or probable AD over a mean of 6.8 years (±4.6) of follow-up. In the primary analysis, higher cognitive activity was independently associated with later dementia onset in the accelerated failure time model (estimate 0.026, 95% CI 0.013–0.039). This relationship persisted in multiple other exploratory models, including a model that accounted for sex and educational level, a model which adjusted for early-life cognitive activity (which bore no independent association with dementia), and a model excluding patients with possible AD from outcome events. When the top 10th percentile of patients with high cognitive activity were compared with the bottom 10th percentile, patients in the top 10th percentile developed dementia at a mean of 5 years later (93.6 vs 88.6 years). Dr. Krauss suggests that the hypothesized direction of the association (more cognitive activity → delay in dementia diagnosis) may possibly be reversed and that the low level of cognitive activity in late life may be an early sign of dementia. In an attempt to disprove this alternative hypothesis, the investigators analyzed the relationship between cognitive activity with markers of AD at enrollment and found no association. The authors agree with Dr. Krauss that their study cannot fully disprove the reverse causality hypothesis; however, these data complement previous research that continues to build the argument for an important relationship between cognitive activity and the subsequent symptoms of dementia. In their prospective observational cohort study, Dr. Wilson et al. explored the relationship between cognitive activity and incident dementia among consecutive elderly patients from the Rush Memory and Aging Project. The investigators hypothesized that self-reported higher levels of cognitive activity such as reading would be associated with older age of dementia onset in patients with clinical Alzheimer disease (AD). Of the 1903 included patients, 457 (24%) developed possible or probable AD over a mean of 6.8 years (±4.6) of follow-up. In the primary analysis, higher cognitive activity was independently associated with later dementia onset in the accelerated failure time model (estimate 0.026, 95% CI 0.013–0.039). This relationship persisted in multiple other exploratory models, including a model that accounted for sex and educational level, a model which adjusted for early-life cognitive activity (which bore no independent association with dementia), and a model excluding patients with possible AD from outcome events. When the top 10th percentile of patients with high cognitive activity were compared with the bottom 10th percentile, patients in the top 10th percentile developed dementia at a mean of 5 years later (93.6 vs 88.6 years). Dr. Krauss suggests that the hypothesized direction of the association (more cognitive activity → delay in dementia diagnosis) may possibly be reversed and that the low level of cognitive activity in late life may be an early sign of dementia. In an attempt to disprove this alternative hypothesis, the investigators analyzed the relationship between cognitive activity with markers of AD at enrollment and found no association. The authors agree with Dr. Krauss that their study cannot fully disprove the reverse causality hypothesis; however, these data complement previous research that continues to build the argument for an important relationship between cognitive activity and the subsequent symptoms of dementia.