Depression and depressive symptoms have a relatively poor prognosis. P
rophylactic maintenance treatment with antidepressants has only been r
ecommended recently. Most studies of chronicity and relapses have been
carried out with initially hospitalized patients or with patients att
ending psychiatric treatment facilities. All incident ambulatory antid
epressant-treated patients (n=456) in a Swedish community during 1980-
84 were identified. They were followed up for 8 years after the index
date. Survival analysis was used to study chronic antidepressant use.
Concomitant use of other prescription drugs was studied using generali
zed estimating equations regression analysis. More than half of the pa
tients had one or several periods of antidepressant treatment after th
e index year. The relapse rate was between 23% and 33% for each year a
fter the index year. Eleven per cent had continuous use every year aft
er the index year. Patients with one or more relapses had a significan
tly higher number of other prescription drugs, as compared with those
treated only once. Once antidepressant treatment has been initiated, t
he relapse rate is considerable. The high use of other prescription dr
ugs indicated continued somatic and psychiatric health problems among
recurrent users. There is a need for long-term outcome studies of anti
depressant treatment in the population, particularly in view of the in
creased sales after introduction of the selective serotonin reuptake i
nhibitors.