Abstract

OBJECTIVES: To describe the incidence of Guillain-Barré syndrome in Sweden during the period 1978-93 and its temporal and geographical variations. METHODS: Stratified and Poisson regression analyses and tests for detection of small epidemics were applied to population based hospital discharge data from 2257 incident cases of Guillain-Barré syndrome in Sweden during the study period. RESULTS: The incidence of Guillain-Barré syndrome was (1) 1.77 per 100000 person-years when age adjusted to the European population; (2) higher in males; and (3) stable across time, although occasional increases of annual incidence rates were found-namely, in 1978 (relative risk (RR) 1.30 (95% CI 1.10-1.54)), and in 1983 (RR 1.24 (95% CI 1.06-1.40)). The incidence increased with age and was bimodal, with peaks at 20-24 and 70-74 years. There was a moderate but significant seasonality with a peak in August, particularly among the young age groups. The age adjusted incidence by county varied from 1.11 to 2.57 per 100000 person-years. Neither temporal nor spatial clustering was significant, except during the period July-September in 1983 at ages below 40 years. CONCLUSIONS: The incidence of Guillain-Barré syndrome in Sweden during the period 1978-93 had a magnitude similar to those described in other surveys, a bimodal distribution by age, and modest geographical and temporal variations with significantly high rates in 1978 and 1983 and in autumn. Minor outbreaks might have passed unnoticed up to the present. Whereas reported drug induced cases of Guillain-Barré syndrome may in part explain the high incidence in 1983, the cause of the aberrant incidence in 1978 remains unknown. Epidemiological surveillance of Guillain-Barré syndrome in Sweden might have been useful.