Date of Award


Document Type

Open Access Thesis


Epidemiology and Biostatistics


The Norman J. Arnold School of Public Health

First Advisor

Anwar Merchant


Background: A decline in cognitive functioning is one of the greatest burdens that older adults face. Studies indicate that periodontal disease is positively associated with cognitive decline. The link between stroke and cognitive decline is well-established, and literature supports that tooth loss and stroke are associated as well, but the role that stroke plays in the relationship between tooth loss and decline in cognitive functioning is, as of yet, unclear. This study uses data from the REGARDS cohort to examine the effect of stroke on the relationship between periodontal disease and cognitive function.

Methods: The REGARDS cohort is comprised of 30,000 African-Americans and white individuals, aged 45 and older, from the United States. While the primary objective of the REGARDS study was to determine the reasons for excess stroke mortality in African Americans and in the Southeastern United States, a large number of variables were collected from participants, among them, tooth loss, history of stroke, incident stroke, and a cognitive function score, which was collected annually. We used a Cox survival analysis approach to assess the impact of tooth loss on risk of cognitive decline. To examine stroke’s impact on the relationship between tooth loss and cognitive function, we analyzed the interaction between tooth loss and stroke (incident and prevalent). As African-Americans tend to experience tooth loss, stroke, and cognitive decline differently than white Americans, we stratified our models by race. For each race, we constructed a crude model and a model adjusted for a variety of demographic, lifestyle, and health characteristics.

Results: Interaction between tooth loss and stroke was not found to be significant in any model, on the additive or multiplicative scale. After adjustment for confounders, there was not sufficient evidence to suggest a positive relationship between tooth loss and cognitive decline among African Americans or white Americans, although there was

Conclusions: We did not find evidence that stroke is an effect modifier between tooth loss and cognitive decline. Our findings indicated that there is likely an increased risk of cognitive decline among those who have lost more teeth in white Americans, but not African Americans. We recommend that stroke be examined as an effect modifier to tooth loss and cognitive decline in high-risk populations, where a significant relationship between tooth loss and cognitive decline has already been observed.

Included in

Epidemiology Commons