Diabetes and dementia; the bitter taste of a sweet disease

Type 2 diabetes (T2DM) is associated with an roughly doubled risk of dementia. Although this association is well established, it is less clear which factors account for this increased risk. Moreover, it is unknown which individuals are at increased risk, what are vulnerable periods in life, and what are windows of opportunity for treatment. This thesis focuses on investigating the link between T2DM and dementia. The main aims were to further elucidate who are at risk of dementia and to examine the role of dysglycemia and vascular factors in patients attending a memory clinic. In the first part of this thesis risk factors for dementia were evaluated in large cohorts from an Integrated Health Care Delivery System in North California. Based on these factors we aimed to develop prediction models for dementia. We first addressed a previously published dementia risk score (Kivipelto et al. Lancet Neurol 2006) that uses readily collectible sociodemographic and cardiovascular risk factors in midlife to predict dementia in late life. Our external validation showed a similar predictive accuracy of the likelihood of dementia four decades later in a large representative population. Next we developed the diabetes specific dementia risk score. In patients with T2DM a combination of diabetes associated comorbidities, depression, education, and age is highly predictive of the likelihood of dementia within the next decade. In the second part of the thesis the phenotype of dementia is studied in relation to disturbance in glucose metabolism (dysglycemia) and vascular risk factors in a memory clinic population. Dysglycemia, short of diabetes, is associated with dementia and cognitive impairment. However, it is uncertain if the relation between dysglycemia and dementia is primarily reflected in neurodegenerative changes, vascular pathology, or both. Our results show that once patients report memory complaints and visit a memory clinic there is no association between dysglycemia and a specific phenotype of dementia. Furthermore, a cluster of cardiovascular risk factors, metabolic syndrome (MetS), is not associated with a specific phenotype of dementia or faster progression of cognitive complaints at a memory clinic. In conclusion, the studies presented in this thesis indicate that dysglycemia and a combination of vascular risk factors do not influence the clinical profile at a first presentation at a memory clinic. A metabolically risky phenotype, represented by MetS, is also not associated with a higher conversion rate to dementia at a memory clinic. However, a combination of these same vascular risk factors at midlife is predictive of an increased dementia risk 40 years later. Furthermore, a combination of – mostly vascular - diabetes associated comorbidities, education, and age is highly predictive of the likelihood of dementia within the next decade for an individual with T2DM.Both of the dementia risk scores presented in this thesis can be used to select patients at high risk of dementia for future research. The challenge for future studies is to develop effective treatments strategies that can delay or prevent the onset of dementia.