This side bar can be used for any related additional links, buttons to recommended shortcuts. or can even be a good place for images

It can be useful to put email links and other contact info here.


There are many unanswered questions about the causes and consequences of strokes and transient ischaemic attacks (TIAs). The Edinburgh Stroke Study is a hospital-based register of patients with stroke and TIA that is aiming to address several of these questions.

Our main objectives are to address the following research questions:

  1. How do the risk factor profiles of different ischaemic strokes subtypes differ, and what can this tell us about differences in the underlying arterial pathologies?
  2. How does the prognosis, in terms of rates of vascular events and patterns of recurrent stroke subtypes, differ among subtypes of ischaemic stroke, and what does it tell us about differences in underlying arterial pathologies?
  3. Do 'novel' risk factors (i.e biomarkers and genes) for vascular disease predict the risk of recurrent vascular events among patients who have had a stroke or TIA?
  4. How well can we predict the risks of death and dependency, and of recurrent vascular events, after a stroke or TIA?
Stroke is important: it is the third commonest cause of death worldwide, is an important cause of disability, and is a major drain on health-care funding. Approximately 80% of strokes are ischaemic, while 15% are due to primary intracerebral haemorrhage and 5% to subarachnoid haemorrhage. Of the ischaemic strokes, around one quarter are caused by lacunar infarcts, thought to be caused by occlusion of small perforating arteries supplying the deep subcortical areas of the brain. Many potentially important differences between lacunar and other subtypes of stroke remain uncertain, and a better understanding of these will inform current and new treatment strategies. Existing epidemiological studies that have investigated differences in risk factor patterns and prognosis between lacunar infarcts and other stroke subtypes are limited by a number of shortcomings. In the Edinburgh Stroke Study, we aimed to address and overcome many of these limitations.

Many recent studies have highlighted the potentially important association between various 'novel' risk factors (such as blood haemostatic factors and various genetic polymorphisms) and cardiovascular disease. To date, results from such studies have been inconsistent. Further studies are also needed to examine whether novel risk factors independently predict the risk of recurrent stroke or other vascular events among patients who have already had a stroke or TIA.

Following a stroke, around one fifth of patients die within one month. By six months, about half have died or are dependent on others. The average risk of a recurrent stroke after a minor stroke or TIA is about 10% in the first year and about 5% per year thereafter. Various prognostic models have been developed to help predict the likely outcome for individual stroke or TIA patients. It is important that these models are validated, particularly in datasets other than those in which they were developed. Our cohort of unselected stroke and TIA patients is an ideal setting in which to evaluate the performance of existing models, and to develop new prognostic models.

The Edinburgh Stroke Study is a hospital-based prospective cohort of stroke and TIA patients. All stroke patients were followed for a minimum of one and a maximum of 4 years for recurrent stroke, myocardial infarction and death, with follow-up ending in May 2006.

We prospectively recruited all patients with a definite or probable stroke or TIA (within the previous six months) admitted to or seen in outpatient clinics at the Western General Hospital, Edinburgh, between April 2002 and May 2005, and who had given their informed consent to take part.

Patients (or their relatives) were given information leaflets about the study and could consent to all or any of the four components:

  • use of their data for research
  • contact with their GP and access to their medical records
  • follow-up in the future
  • storage of a blood sample for future biological and genetic analyses
We collected baseline data on: basic demographic; risk factors; clinical examination findings; clinical stroke subtype (using the Oxfordshire Community Stroke Project Classification) and the results of various routine investigations, including brain imaging. The clinical stroke subtype syndrome was refined according to findings on brain imaging.

Biological/genetic analyses
Where patients gave their consent, we obtained a blood sample following their baseline stroke or TIA. The Wellcome Trust Clinical Research Facility (WTCRF) within the Western general Hospital processed and stored all blood samples. Trained research nurses from the WTCRF also assisted in the sampling of blood from inpatients, and staff within the WTCRF genetics laboratory extracted DNA from blood samples.

All stroke patients living in the Edinburgh area (i.e with an EH postcode – the vast majority of patients) and who had given their consent to be included in follow-up were followed for recurrent stroke, myocardial infarction (MI), death and disability using multiple overlapping follow-up methods. Patients (or their relative or carer) were given an ESS contact card and actively encouraged to let us know as soon as possible about any suspected recurrent stroke or MI. They also received a postal questionnaire at six months from stroke onset, one year and annually thereafter. Patients were asked about further possible vascular events and also asked to complete a modified Rankin score to ascertain level of disability. General Practitioners were informed of patients’ participation in the study and asked to inform us of further vascular events or death. Patients were flagged at the General Register Office for Scotland, from whom we received death notifications and certifications of cause of death. We also added ESS stickers to the front of hospital notes to alert clinicians to their involvement in the study. We aimed to assess patients who suffered a recurrent stroke as soon as possible after the event, to determine the recurrent stroke subtype and to provide appropriate management and clinical advice to the patient, their carer and their GP.

Outcome events

  1. Recurrent stroke

    Patients with a suspected recurrent stroke who had not been admitted to hospital were assessed in outpatient clinics. The medical notes and brain imaging of patients admitted to hospital were reviewed, and where necessary, the patients were re-assessed in outpatient clinics, or underwent further advanced brain imaging. Where a patient had a recurrent stroke, data was collected on: the clinical features of the stroke; treatment at time of stroke; the clinical stroke syndrome; details of any clinical investigations performed, including type and results of brain imaging; and final stroke syndrome. As with the baseline stroke, the recurrent stroke syndrome was assigned using a combination of clinical and brain imaging findings (where brain imaging was performed). Where possible, if the CT brain imaging showed no new lesion, or if the patient had a lacunar syndrome, advanced magnetic resonance imaging (MRI) was performed (including diffusion weighted imaging).

  2. Myocardial infarction

    When a patient had an MI, we recorded the date of MI, as well as the clinical symptoms and (where performed), the results of clinical investigations. Two of the following three indicated a definite MI: ECG changes, chest pain or enzymatic changes. A definite MI was also defined where the patient died suddenly and an autopsy was performed which revealed pathological evidence of an acute MI. A sudden death, in the absence of any indications of death by another cause, was considered a probable MI.

  3. Death

    Each patient was flagged at the General Register Office for Scotland, who provided the date and place of death, whether or not an autopsy was performed, the name of the certifying doctor, and the cause of death. All relevant medical notes, including hospital, GP and nursing home records, as well as autopsy reports of patients whose death certificate gave some suggestion that they appeared to have died of a further stroke or MI were reviewed to verify the occurrence of a further vascular event.

Data management / analyses
All electronic data is stored in a secure password-protected database on a secure server in the Division of Clinical Neurosciences. All hard copies of data (data collection forms, follow-up forms etc) are kept in secure storage in the Division of Clinical Neurosciences. We perform all our analyses using fully annonymised data extracts. We do not share any identifiable data with any external organisation.