Background There have been very few controlled studies of antidepressa
nts in dysthymia, particularly in samples diagnosed reliably and with
an adequate length of follow-up, in this investigation, we measured th
e long-term outcome in a large group of patients meeting DSM-III-R cri
teria for dysthymia. This study was designed to investigate whether fl
uoxetine is effective in the treatment of dysthymia. Method This rando
mised study, including 140 patients, compared fluoxetine (91 patients)
and placebo (49 patients) on a double-blind basis in two distinct pha
ses: a short-term end-point (3 months with 20 mg/day fluoxetine) and a
medium-term end-point (6 months) where the initial responders continu
ed double-blind treatment unchanged and non-responders received an add
itional treatment of 20 mg/day fluoxetine. Results After-three months
of treatment, response was seen more frequently in the fluoxetine grou
p (42/72) than in the placebo group (14/39,P < 0.0001). Improved patie
nts at 3 months were still improved at 6 months. Furthermore 50% of th
e non-responders art 3 months improved and rated as responders at 6 mo
nths, after fluoxetine was increased to 40 mg daily. Conclusions This
study showed the significant and persistent action of fluoxetine on dy
sthymia. The finding that 50% of the non-responders at 3 months were i
mproved at 6 months, after fluoxetine dosage was increased to 40 mg da
ily, argues in favour of treating dysthymic patients for at least 6 mo
nths, and with a higher dosage if the initial doses are ineffective.