Population-based studies on the natural history of Alzheimer's disease and vascular dementia

Pharmacoepidemiology is defined as the study of the utilization and effects of drugs in large human populations. Beside its classical role in the evaluation of drug safety after marketing, pharmacoepidemiology is increasingly gaining importance in the pre-marketing phase of the drug development process, where it can provide useful information on the natural history of the disease a drug is being developed to treat.
Alzheimer’s disease (AD) is one of the most disabling and burdensome health conditions worldwide. It is the most common form of dementia with more than 26 million cases worldwide. Vascular dementia (VD) is the second most common dementia form, resulting from intracerebral vascular and circulatory pathology.
The aim of this thesis was to increase knowledge on the natural history of AD and VD, thereby focusing on the effect of certain drug therapies as potential risk or protective factors for these diseases or complications thereof.
The studies in this thesis were carried out using data from the United Kingdom (UK) based General Practice Research Database (GPRD), a large and well established physician-based primary care database. This database contains longitudinal records from several million patients representative of the UK population. The information recorded in the medical files includes patient demographics and characteristics (e.g. age, sex, height, weight, smoking status), symptoms, medical diagnoses, referrals to consultants, and hospitalizations.
In the first study (3.1) we identified patients aged >=65 years with an incident diagnosis of AD or VD between 1998 and 2008 and assessed incidence rates (IRs) of AD and VD, stratified by age and sex. To each demented case patient we matched one dementia-free control patient and analyzed co-morbidities and drug use prior to the time of diagnosis. We identified 7,068 AD and 4,438 VD cases. For AD, IRs were higher for women than for men, but not for VD. Except for orthostatic hypotension, the prevalence of all cardiovascular (CV) co-morbidities and exposure to CV drugs was lower in patients with AD than in the corresponding controls, whereas the opposite was true for VD. We concluded that this may be a true finding or the result of diagnostic bias, i.e. that demented patients with CV diseases may be more likely to be diagnosed with VD than AD.
In the second study (3.2) we studied the influence of metformin or other antidiabetic drugs on the risk of developing AD. We performed a case-control analysis within the population of AD cases and corresponding controls identified in the first study (3.1). We found that long-term users of metformin had a slightly increased risk of developing AD as compared to non-users, but there was no consistent trend with increasing duration of use. Use of other antidiabetic drugs such sulfonylureas, thiazolidinediones, or insulin was not associated with an altered risk of developing AD.
In the third (3.3) and fourth study (3.4) we followed the complete study population of the first study (3.1) forward in time to assess IRs of certain diseases (complications) of interest in patients with AD or VD and compared them to patients without dementia. We then performed a nested case-control analysis to identify potential risk factors for developing such diseases of interest. The diseases of interest in the third study were seizures/epilepsy and in the fourth study ischemic stroke, hemorrhagic stroke or transient ischemic attack (TIA). In the third study we found that seizures or epilepsy were substantially more common in patients with AD and VD than in dementia-free patients. Additionally, patients with longer standing (>=3 years) AD had a slightly higher risk of developing seizures or epilepsy than those with a shorter disease duration, while in patients with VD the contrary was observed. In the fourth study we found that patients with AD did not have a materially different risk of developing an ischemic stroke compared to patients without dementia, whereas patients with VD had an about twofold increased risk. AD patients receiving atypical antipsychotic drugs only had a higher risk of developing a TIA than AD patients not receiving any antipsychotic drug treatment, whereas for patients with VD there was no significant difference between users of atypical or typical antipsychotic drugs and those not receiving antipsychotic treatment.
The GPRD is a very useful tool to conduct pharmacoepidemiological research. Its strengths are the large size, the population-based character of the data, and the opportunity to have access to original medical records. On the other hand, data on important confounders such as dietary or exercise habits is largely missing.