Page 367 - ILAE_Lectures_2015

In a Dutch study of mortality in people with epilepsy followed for over 40 years the SMR was
16 in the first two years decreasing to 2.8 thereafter20. After two years, approximately one-third
of deaths were directly or indirectly attributable to epilepsy. Common non-epilepsy causes of
death cited in mortality studies include pneumonia, cerebrovascular disease, malignancy and
heart disease. SMRs and PMRs are consistently elevated for these causes in population-based
studies and often markedly so in the first few years of follow-up. In a Swedish study looking at
cause-specific mortality in over 9000 adults with epilepsy, the overall SMR was 3.6 (95% CI 3.5,
3.7), with SMRs being increased for specific causes such as cancer (SMR 2.6; 95% CI 2.4, 2.8),
respiratory disease (SMR 4.0; 95% CI 3.6, 4.5), heart and cerebrovascular disease (SMR 3.1;
95% CI 3.0, 3.3) and accidents and poisoning (SMR 5.6; 95% CI 5.0, 6.3)21. The risk of premature
death from heart disease in people with epilepsy was found to be elevated in those aged 25 to 64
but not for those aged 65 years and over in the Rochester cohort22, and also in the NGPSE cohort
during the last five years of follow-up7. Bronchopneumonia is an important cause of mortality in
people with epilepsy of all ages, not just the elderly, and was associated with the highest SMR
(6.6) in the NGPSE7. This may be related to aspiration during seizures but this is unproven, or it
may be the terminal event.

The influence of mental retardation (MR) and epilepsy was investigated in a Swedish study. The
SMR was 1.6 (95% CI 1.3, 2.0) in people with MR only but this increased to 5.0 (95% CI 3.3,
7.5) for those with MR and epilepsy, with the increase in mortality associated with seizure type
and frequency23. In studies from institutions and hospitals, where people have presumably more
severe epilepsy, epilepsy-related deaths are more common. In one study, PMRs were cancer
(26%), bronchopneumonia (25%), circulatory diseases (24%), seizure-related deaths (other than
SUDEP) (12%) and SUDEP (6%)24.

SMRs and PMRs for cancer have been consistently elevated in people with epilepsy even after
excluding CNS neoplasms. Cancer mortality was compared between two cohorts with epilepsy,
one from an institution with more severe epilepsy (SEC) and the other, a community-based
population with milder epilepsy (MEC). The SMR for all cancers was elevated in the SEC (SMR
1.42; 95% CI 1.18, 1.69) but not in the MEC (SMR 0.93; 95% CI 0.84, 1.03). The SMR for brain
and CNS neoplasms was significantly elevated in the group with milder epilepsy25.

Two recent studies from Finland26 and Austria27 have looked at cause-specific mortality in people
with epilepsy, and both demonstrate that the majority of deaths are due to non-epilepsy-related
causes. In the Finnish study26, which was based on a nationwide register-based cohort study of
people aged 10 years or older diagnosed with epilepsy between 1990 and 1994, the predominant
causes of death were CNS cancer (17%), other cancers (15%), ischaemic heart disease (11%) and
cerebrovascular diseases (10%), which may have been related to the probable underlying
aetiology. In contrast the proportion of deaths attributable to epilepsy was small with 3.9% of
deaths attributable to accidents, 3.4% for alcohol-related diseases and 1.6% for suicides. The
Austrian study27 comprised all adults (≥18 years) treated for epilepsy at a single centre
(Innsbruck) between 1970 and 2009. In the overall cohort there were 4295 people, with 822
deaths (overall SMR 1.7; 95% CI 1.6, 1.9). The highest cause-specific SMRs in the overall cohort
were for congenital abnormalities (SMR 7.1; 95% CI 2.3, 16.6), suicide (SMR 4.2; 95% CI 2.0,
8.1), alcohol dependence syndrome (SMR 3.9; 95% CI 1.8, 7.4), malignancy of the oesophagus
(SMR 3.1; 95% CI 1.2, 6.4) and pneumonia (SMR 2.7; 95% CI 1.6, 4.2). The cause-specific
SMRs were broadly similar in those with newly diagnosed epilepsy (1299 individuals with 267
deaths) with an overall SMR of 1.8 (95% CI 1.6, 2.1).