Safety and efficacy of fluoxetine on functional recovery after acute stroke (EFFECTS): a randomised, double-blind, placebo-controlled trial

Studies have suggested that fluoxetine could improve neurological recovery after stroke. The Efficacy oF Fluoxetine—a randomisEd Controlled Trial in Stroke (EFFECTS) trial aimed to assess whether administration of oral fluoxetine for 6 months after acute stroke improves functional outcome. EFFECTS was an investigator-led, multicentre, randomised, placebo-controlled, double-blind, parallel group trial that enrolled patients aged 18 years or older between 2 and 15 days after stroke onset in 35 stroke and rehabilitation centres in Sweden. Eligible patients had a clinical diagnosis of ischaemic or intracerebral haemorrhage, brain imaging that was consistent with intracerebral haemorrhage or ischaemic stroke, and had at least one persisting focal neurological deficit. A web-based randomisation system that incorporated a minimisation algorithm was used to randomly assign (1:1) participants to receive oral fluoxetine 20 mg once daily or matching placebo capsules for 6 months. Patients, care providers, investigators, and outcomes assessors were masked to the allocation. The primary outcome was functional status, measured with the modified Rankin Scale (mRS) at 6 months, analysed in all patients with available mRS data at the 6-month follow-up; we did an ordinal analysis adjusted for the minimisation variables used in the randomisation. This trial is registered with EudraCT, 2011-006130-16; ISRCTN, 13020412; and ClinicalTrials.gov, NCT02683213. Between Oct 20, 2014, and June 28, 2019, 1500 patients were enrolled, of whom 750 were randomly assigned to fluoxetine and 750 were randomly assigned to placebo. At 6 months, mRS data were available for 737 (98%) patients in the fluoxetine group and 742 (99%) patients in the placebo group. There was no effect of fluoxetine on the primary outcome—distribution across mRS score categories—compared with placebo (adjusted common odds ratio 0·94 [95% CI 0·78 to 1·13]; p=0·42). The proportion of patients with a new diagnosis of depression was lower with fluoxetine than with placebo (54 [7%] patients vs 81 [11%] patients; difference −3·60% [–6·49 to −0·71]; p=0·015), but fluoxetine was associated with more bone fractures (28 [4%] vs 11 [2%]; difference 2·27% [0·66 to 3·87]; p=0·0058) and hyponatraemia (11 [1%] vs one [