Convulsive epilepsy in a rural district of northern Tanzania: risk factors and classification

E Hunter, J Rogathe, A Iqbal, D Birchall, R Whittaker, M Jackson, A Jusabani, E Aris, R Walker
Publication year: 

Background Few community-based studies of epilepsy from sub-Saharan Africa (SSA) have included neuroimaging or neurophysiology. There have been only two case-control studies to define risk factors.

Methods We identified 291 people with epilepsy (PWE) in a rural population in Tanzania. All PWE were offered investigation with EEG and CT head scan. Epilepsy was classified according to ILAE criteria. PWE were compared with age- and sex-matched controls to define risk factors for epilepsy.

Results 216 PWE (74.2%) had EEGs and 200 (68.7%) had CTs. Investigations were abnormal in 38.9% of EEGs and 27% of CTs. All 291 PWE had convulsive epilepsy. Seizure onset was focal in 71.5%, generalised in 2.4%, and undefined in 26.1%. Epilepsy was classified as symptomatic (structural/metabolic) in 23.4%, idiopathic in 2.1%, and remained unclassified in 74.2%. Risk factors associated with epilepsy were a family history of epilepsy (OR 6.2; 95% CI 2.6 to 14.8), a history of childhood febrile convulsions (21.6; 2.9 to 159.5) and a history of perinatal complications (3.8; 1.3 to 11.2). Motor or cognitive impairment was also associated with epilepsy (35.8; 4.9 to 261.4 and 39.0; 9.5 to 160.7). On multivariable logistic regression, the strongest predictors of epilepsy in this population were family history, childhood febrile convulsions, and the presence of motor/ cognitive impairment.

Conclusion The high proportion of focal onset seizures and the association with childhood febrile convulsions and motor or cognitive impairment points to a significant burden of acquired epilepsy in this population. The association with family history warrants further study.