The Association Between Early-life Written Language Skills and Late-life Cognitive Resilience to Alzheimer's Disease

Abstract

As the population ages, projections suggest that the number of individuals living with age-related diseases such as Alzheimer’s disease will increase. Prevention of Alzheimer’s disease is a major priority since there is currently no cure for the disease. Cognitive resilience is a hypothetical construct designed to explain why some individuals manage to avoid cognitive changes despite the presence of Alzheimer neuropathology. Educational attainment is one of the well-documented examples of building cognitive resilience since high levels of educational attainment have been associated with delayed onset of cognitive impairment. Written language skills developed in early life may reflect the development of early intellect and are essential to educational attainment. Weak early-life written language skills (i.e., low idea density and low grammatical complexity) have been associated with poor cognitive function in later life. However, there is limited understanding of the influence of written language skills and their potential contribution to cognitive resilience. This research aimed to assess the association between written language skills and cognitive resilience using data from the Nun Study. The Nun Study is a longitudinal study of aging in religious sisters who were a minimum of 75 years of age at baseline. Idea density and grammatical complexity were determined using coded autobiographies. Autobiographies were obtained from archival records and were written at a mean age of 22 years. Cognitive resilience was operationalized based on whether individuals met the clinical diagnosis of dementia at last assessment prior to death according to DSM-IV criteria while fulfilling Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) neuropathologic criteria (“definite” or “probable”) or National Institute on Aging and Reagan Institute (NIA-RI) neuropathologic criteria (“definite”, “intermediate” or “high” likelihood) for Alzheimer’s disease. Analyses included descriptive analyses (univariate and bivariate) as well as logistic regression models. The purpose of this project was to strengthen current knowledge on the potential association between early-life written language skills and late-life resilience to cognitive impairment. This study also aimed to better understand the implications of indicators of cognitive and brain reserve on this potential relationship. Based on descriptive and multivariable analyses, a relationship between written language skills (idea density and grammatical complexity) was found particularly in the CERAD sample where cognitive resilience was defined using CERAD criteria for Alzheimer neuropathology. In logistic regression models adjusting for standard covariates (age and APOE), low idea density was associated with decreased likelihood of cognitive resilience (Odds Ratio (OR): 0.15, 95% Confidence Interval (CI): 0.02-0.72). These findings meant that higher idea density (vs. low) was associated with six times greater odds of cognitive resilience. Similarly, low grammatical complexity was significantly associated with cognitive resilience in adjusted models for age and APOE (OR: 0.13, 95% CI: 0.03-0.50). That is, the odds of cognitive resilience in later life increased seven-fold among those with higher grammatical complexity compared to those with low grammatical complexity. Further analyses also suggested that grammatical complexity remained a significant predictor of cognitive resilience in the presence of indicators of cognitive (education) and brain (cerebral infarcts and cortical atrophy) reserve. In comparison, idea density was significant when separately adjusted for presence or number of infarcts along with standard covariates. However, idea density was not significant in a few full models (e.g., including adjustments for standard covariates (age and APOE), cortical atrophy and presence of infarcts, or standard covariates and education). These findings suggested the strong influences of both education and structural brain changes on the relationship between idea density and cognitive resilience. Future studies should aim to assess whether other forms of writing from early life (e.g., written language in social media) can also be associated with cognition in later life. Findings from this research contribute to the understanding of cognitive resilience and provide the foundation for further exploration into the influence of written language on the prevention of Alzheimer’s disease.