The warning follows US research suggesting that a 12-week course of fluoxetine or nortriptyline can improve long-term executive function.
At a 21-month follow-up, patients who were prescribed antidepressants showed significant improvement in executive function, but those given placebo showed deterioration.
The drugs may boost executive function recovery by affecting monoaminergic nuclei in frontal cortical-subcortical circuits that control executive function, say the University of Iowa team.
Alternatively, antidepressants may boost neurogenesis.
But Professor Allan House, an expert in depression after stroke at the University of Leeds, said: ‘It is a very interesting theory but this research doesn't help one way or another.'
Although 92 people were randomly assigned to an antidepressant or placebo six months after stroke, only 36 completed all evaluations during the 21-month follow-up.