The findings reveal a striking correlation between premature death and mental illness in these patients and people with epilepsy were four times more likely to have received a psychiatric diagnosis in their lifetime compared with the general population.

The figures are considerably higher than previously thought and have important implications for epilepsy management.

Researchers at the University of Oxford and Karolinska Institutet studied 69,995 people with epilepsy born in Sweden between 1954 and 2009 and followed up over 41 years, between 1969 and 2009. They compared mortality and cause of death information from these patients with 660,869 age- and sex-matched people from the general population. The study also looked at the unaffected siblings of those with epilepsy, in order to rule out the influence of background factors such as genetic risk factors and upbringing.

Throughout the course of the study, almost nine per cent (6,155) of people with epilepsy died compared with less than one per cent (4,892) of people from the general population.

The most important cause of death in people with epilepsy that was not clearly related to the underlying disease process was death by external causes, such as accident or suicide, accounting for almost 16 per cent of deaths. Three quarters of these deaths were amongst patients who also had a psychiatric diagnosis.

Although suicide and deaths from accidents were still relatively rare, the odds of a person with epilepsy committing suicide during the study were four times higher than the general population and there was a strong correlation with mental illness and substance abuse.

Dr Seena Fazel, a Wellcome Trust Senior Clinical Research Fellow at the University of Oxford and main author of the study, said: "This is the largest report to date to look at psychiatric associations in epilepsy and their contribution to premature mortality. Our finding that three quarters of suicide and accident deaths in epilepsy also had a diagnosis of mental illness strongly identifies this as a high risk population to focus preventative strategies and more intensive treatment.

"Improving the identification, monitoring and treatment of psychiatric problems in epilepsy patients could make an important contribution to reducing the risk of premature death that we're currently seeing in these patients."

The study also reveals that the odds of dying in a non-vehicle accident, such as drug poisoning or drowning, were more than five times higher for people with epilepsy than control populations.

"Our findings also highlight general accidents as a major, preventable cause of death in epilepsy patients and suggest that specific warnings, in addition to those already given around driving, should be provided to patients at the time of diagnosis to ensure they are aware of the risks," added Dr Fazel.

Professor Charles Newton from the Wellcome Trust programme at the Kenyan Medical Research Institute (KEMRI) and the Department of Psychiatry at Oxford University, said: "Although it is well -recognised that psychiatric and addiction disorders occur in epilepsy, in high income (Western) countries, epilepsy is often managed by neurologists only. The findings from this study would suggest that clinical epilepsy services should review their liaison with psychiatric and addiction services as a priority."

This is the first study to look at the odds of premature death in people with epilepsy compared with their unaffected siblings, revealing that they do not differ significantly from odds of death in epilepsy compared with general population controls. This provides further weight to the evidence that epilepsy as a disease is an independent risk factor for death by any cause.