Post-Stroke Depression: Prophylactic vs. Treat-if-Emerges
One trial suggested prophylactic fluoxetine after stroke improves motor recovery; a much larger one found no benefit and a real fracture-risk signal. Which reading should govern depends heavily on who is actually at risk for depression in the first place.
R.G., a 71-year-old retired postal worker, suffered a severe left MCA territory stroke six days ago, leaving him with expressive aphasia and dense right-sided hemiparesis that will require inpatient rehabilitation for weeks. He has a documented history of two depressive episodes in his forties and fifties, both treated successfully with sertraline at the time, and his family — his wife and adult daughter, both closely involved in his care — have asked directly whether starting an antidepressant now, before any depressive symptoms appear, could help protect him given how hard this stroke has hit.
His aphasia makes standard depression screening difficult to administer reliably, which is itself part of what makes his risk profile concerning — PSD is both more common and harder to detect early in patients with significant language impairment, and severe stroke with major functional loss is one of the most consistently replicated risk factors for it. He has no other chronic illness beyond well-controlled hypertension, and no prior adverse reaction to sertraline during either past episode. His wife, a retired occupational therapist herself, has spent the past six days at his bedside nearly around the clock and has already begun adapting simple communication boards to help him express basic needs despite the aphasia.
The trial evidence on prophylactic SSRI use after stroke is genuinely mixed rather than settled in either direction. The FLAME trial, published over a decade ago, reported that fluoxetine improved motor recovery after stroke independent of any antidepressant effect, a finding that helped popularize prophylactic SSRI use broadly. The larger, more recent FOCUS trial found no such functional benefit. It did, however, find that fluoxetine reduced new depression at six months — 13.4% versus 17.2% — while increasing bone fractures and hyponatremia, which is why its authors concluded against routine use for either purpose rather than against prophylaxis alone. Whether that broader null result should still govern a patient whose personal risk profile is this concentrated, rather than the general stroke population FOCUS mostly enrolled, is the actual question in front of the team.
At the stroke-unit family meeting
His risk profile is about as concentrated as PSD risk factors get — prior depression history, a severe stroke, major functional loss, and a screening barrier that means we're likely to catch symptoms later than we would in a patient who can reliably answer a PHQ-9. I'd weigh prophylactic sertraline seriously here specifically because of that combination, not as a default for every stroke admission.
I'd add that his own history removes one of the biggest uncertainties in a prophylaxis decision — we already know he tolerates and responds to sertraline, which isn't true for most patients being considered for prophylactic treatment after a first-ever psychiatric exposure. That's a real point in favor of starting now rather than waiting for symptoms in a patient whose ability to report them is already compromised.
Sertraline 50mg was started prophylactically, at the dose that treated his prior episodes successfully, with structured caregiver-reported behavioral monitoring supplementing formal screening given his aphasia. Two things were documented alongside it: that the FLAME and FOCUS evidence is fluoxetine-specific and is being generalized to sertraline on his own response history, and that adding an SSRI to his secondary-prevention antiplatelet raises bleeding risk enough to warrant explicit mention, along with a sodium check given the hyponatremia signal in an older patient.
The team's reasoning here turns entirely on his concentrated risk profile — the same prophylactic decision does not follow automatically for every stroke patient, which is the exact question Patient B's case below puts to the test.
L.P., a 58-year-old high-school counselor, suffered her first stroke five days ago — a small right-hemisphere infarct causing mild left-hand weakness with no aphasia and no cognitive impairment on formal testing. She has no history of depression at any point in her life, no other chronic illness, and is already walking with minimal assistance and expected to return home within the week rather than requiring extended inpatient rehabilitation. Her husband and two adult children have been present throughout her admission, and she has engaged actively and articulately in every conversation about her recovery plan, including asking detailed questions about her own discharge timeline and outpatient physical therapy options.
Her stroke team raised the same question that came up for Patient A — should she start a prophylactic antidepressant now — and she asked directly, and reasonably, why that would be necessary when she isn't depressed and never has been. The same FLAME-versus-FOCUS evidence tension applies to her as it did to R.G., but her risk profile sits at nearly the opposite end of the spectrum: no depression history, a mild deficit, intact language and cognition, and a strong, present support system, none of which resemble the severe, high-risk presentation that made prophylaxis a genuinely close call for Patient A. The larger and more rigorous FOCUS trial's null functional-benefit finding, combined with its documented fracture-risk signal, weighs more heavily against treating her specifically, since she carries little of the underlying risk that a prophylactic prescription would need to be justifying in the first place.
At the stroke-unit follow-up visit
She's asking a fair question, and the honest answer is that the evidence for routine prophylaxis in a patient with her risk profile is weak. FOCUS was a far larger and more rigorously conducted trial than FLAME and found no functional benefit. I want to be careful not to overstate it though: it did reduce new depression, by about four percentage points. What it also found was more fractures and more hyponatremia, and for someone at her risk level a four-point absolute reduction on a low baseline isn't worth those harms.
I'd frame it the same way — FLAME's motor-recovery finding was provocative but came from a much smaller trial with real methodological limitations, and FOCUS is the study that should carry more weight precisely because it was larger, better powered, and specifically designed to test what FLAME suggested. Extending FLAME's logic to a low-risk patient asks her to accept documented fracture and hyponatremia signals for a functional benefit the stronger trial didn't replicate — and for a depression-prevention benefit that is real but small, and smallest in exactly the low-risk group she belongs to.
Structured screening at scheduled follow-up visits, rather than a prophylactic prescription, matches the actual evidence for a patient with her risk profile far better than starting a drug against a risk she's unlikely to carry in the first place.
No antidepressant was started. Structured PHQ-9 screening was scheduled at each stroke follow-up visit, with an explicit plan to treat promptly if symptoms emerge rather than waiting for a crisis to prompt reassessment.
L.P. left the visit with a clear understanding of why the recommendation differed from what a family member had mentioned hearing about for other stroke patients — the team's explanation centered on her own risk profile and the newer, larger trial evidence, not a blanket rule about stroke and antidepressants either way.