Population-based studies on the epidemiology of migraine and Parkinson's disease

With epidemiologic analyses the effect of certain exposures on diseases can be studied in large population samples. Pharmacoepidemiology is a speciality which focuses on beneficial or harmful drug effects on the development of diseases. In my thesis I carried out different epidemiological studies in order to increase the knowledge on the natural history of migraine and Parkinson's disease. Another focus was to evaluate the effect of certain drug therapies on teh risk of developing migraine or Parkinson's disease (PD) or complications of the diseases.
I used data from the General Practice Research Database (GPRD), which contains electronic records from primary-care of several million people in the United Kingdom (UK). Additionally, information on patient demographics (e.g. age, gender, body mass index, smoking status) is available for a large portion of the patiens as well as data on hospital and specialist diagnoses. The GPRD has been the source for many important studies in epidemiology as well as in drug safety.
In my first project I identified 51'688 individuals of the FPRD with a firyst-time migraine diagnosis between 1994 and 2001 and an equal amount of control subjects without such a diagnosis. The incidence rates (IR) of first-time diagnoses of migraine by the general practitioners (GPs) were 2.5 times higher in women than in men and highest in puberty. The comorbid disorders of the migraineurs were also quantified in migraineurs and controls. By means of a case-control study design which included matching on several important confounders such as gender, age, general practice and index date, the odds ratios (ORs) for the comorbidities in migraineurs compared to non-migraineurs were investigated. This resulted in an increased OR for the migraineurs for most chronic diseases. Determination of the health resource utilisation (HRU) revealed that migraineurs with triptan prescriptions needed more health care, defined as visits to their GP or neurological specialists as well as prescriptions for headeche related drugs.
In a second part of the migraine project I followed a cohort of migraineurs and their matched controls until they developed a stroke, a transient ischaemic attack (TIA), they died of until they were diagnosed with asthma for the first time. Again IRs were calculated and a nested case-control analysis performed. A previous history of migraine was associated with an approximately twofold increased risk for stroke or TIA, however, residual confounding by migraine recency or severity could not totally be ruled out. Furthermore it is challenging to determine the stroke risk in association with prior triptan use because in the GPRD the actual timing of the drug intake is not recorded. The mortality of migraineurs was slightly decreased and no increased asthma risk was seen in migraineurs with or without triptan use.
In my second project I investigated the impact of prior drug use on the risk of being diagnosed with PD. During the study period from 1994 to 2005 3'637 individuals with idiopathic PD were identified from the GPRD. The majority of the cases with a first-time PD diagnosis were men older than 60 years of age. In two separate case-control studies, in which I used the same matching criteria as in the migraine project, I found a decreased risk of PD in patients with current use of calcium channel blockers. This finding is in accordance with a recent hypothesis regarding the involvement of calcium channels in the PD pathophysiology. After the assumption of an increased risk for PD associated with the use of statins, the results of the other case control study gave reassurance that in a large population sample from the GPRD the risk for a PD diagnosis was not increased for current or past use of statins.
To conclude, the GPRD data is very useful for the description of the natural history of diseases as well as for the investigation of particular drug safety questions. The potentials of the database could be further increased if genetic information was also available in future. Certainly, special diligence has to be exercised regarding the issue of data protection.