UVM Theses and Dissertations
Format:
Print
Author:
Gillett, Sarah R.
Dept./Program:
Center for Clinical and Translational Science
Year:
2013
Degree:
Ph. D.
Abstract:
Cognitive dysfunction and dementia are devastating and costly conditions prevalent in the U.S. and worldwide. Incidence of dementia increases with age and the number of Americans suffering from the disease is expected to increase as the population ages. Treatments and preventive strategies are desperately needed. Despite decades of research, the underlying mechanisms of this disease remain unclear. Biological processes that contribute to cognitive decline (CD) probably begin long before clinical signs are apparent. Biomarkers related to these processes could be useful in identifying those at risk and eventually designing treatments. Vascular risk factors and inflammation have been implicated in CD. Therefore, the overall goal of this research was to study the risk factors for CD and determine if hemostasis and inflammation biomarkers are associated with CD. Racial and regional disparities in CD were also studied. All original research was done in the Reasons for Geographic and Racial Differences in Stroke (REGARDS) study, a nationally representative cohort study of over 30,000 black and white Americans age 45 or older. Multivariable-adjusted logistic regression models were used to study associations.
The first chapter is a comprehensive review paper that examines the body of literature surrounding inflammation and CD. Studies examined CD in both quantitative and qualitative form. Although the reviewed studies did not support a role for clinical assessment of inflammation biomarkers for CD, heterogeneity ofthe studies and lack of prospective studies with long-term follow up warrant future investigation.
Chapter 2 analyzes the association of the inflammation biomarker C-reactive protein (CRP) with CD. We found that CRP at high levels (above the 90th percentile) was associated with significant decline in learning and memory, but not with decline in executive function. CRP did not mediate the association of race with CD.
Chapter 3 describes a novel definition of clinically relevant incident CD using brief cognitive tests administered via telephone. A regression-based normative approach was used to identify around 500 subjects who scored> 1.5 SD below an age, race, sex, and education adjusted mean score. Demographic and vascular risk factors were associated with CD. Transient ischemic attack (TIA), income, region, diabetes, coronary artery disease and incident stroke were independent predictors of cognitive decline.
Chapter 4 assesses the internal validity of assay results in REGARDS by simulating REGARDS processing and comparing results obtained from REGARDS and ideal-processed samples. Most analytes produced valid results, but proteins released by platelets were highly variable and should not be measured in REGARDS.
Chapter 5 uses a nested case-control study design to investigate the association of hemostasis biomarkers with CD as defined in Chapter 3. D-dimer>0.50 [mu]g/mL, fibrinogen>90th percentile, and factor VIII> 90th percentile were associated with CD in univariate models, but the association was attenuated when demographic and vascular risk factors were accounted for. Participants with CD had higher odds of having at least 2 elevated hemostasis biomarkers in fully adjusted models, suggesting these markers may identify high-risk individuals when used in combination.
In conclusion, this dissertation work showed that high levels CRP, fibrinogen, D-dimer and factor VIII are associated with CD. This supports the hypothesis that both inflammation and procoagulant biologies play a role in CD.
The first chapter is a comprehensive review paper that examines the body of literature surrounding inflammation and CD. Studies examined CD in both quantitative and qualitative form. Although the reviewed studies did not support a role for clinical assessment of inflammation biomarkers for CD, heterogeneity ofthe studies and lack of prospective studies with long-term follow up warrant future investigation.
Chapter 2 analyzes the association of the inflammation biomarker C-reactive protein (CRP) with CD. We found that CRP at high levels (above the 90th percentile) was associated with significant decline in learning and memory, but not with decline in executive function. CRP did not mediate the association of race with CD.
Chapter 3 describes a novel definition of clinically relevant incident CD using brief cognitive tests administered via telephone. A regression-based normative approach was used to identify around 500 subjects who scored> 1.5 SD below an age, race, sex, and education adjusted mean score. Demographic and vascular risk factors were associated with CD. Transient ischemic attack (TIA), income, region, diabetes, coronary artery disease and incident stroke were independent predictors of cognitive decline.
Chapter 4 assesses the internal validity of assay results in REGARDS by simulating REGARDS processing and comparing results obtained from REGARDS and ideal-processed samples. Most analytes produced valid results, but proteins released by platelets were highly variable and should not be measured in REGARDS.
Chapter 5 uses a nested case-control study design to investigate the association of hemostasis biomarkers with CD as defined in Chapter 3. D-dimer>0.50 [mu]g/mL, fibrinogen>90th percentile, and factor VIII> 90th percentile were associated with CD in univariate models, but the association was attenuated when demographic and vascular risk factors were accounted for. Participants with CD had higher odds of having at least 2 elevated hemostasis biomarkers in fully adjusted models, suggesting these markers may identify high-risk individuals when used in combination.
In conclusion, this dissertation work showed that high levels CRP, fibrinogen, D-dimer and factor VIII are associated with CD. This supports the hypothesis that both inflammation and procoagulant biologies play a role in CD.