The First 90 Days After a Stroke at Home: What Nobody Prepares You For

You go home with a folder.

Inside the folder are exercises, a medication list, appointment cards, and a phone number to call with questions. The folder is accurate. Everything in it is worth doing. And within about ten days, most families discover that the folder answers almost none of the questions they actually have.

It does not explain why he sleeps nine hours and is still too exhausted to get dressed. It does not explain why she eats only the right half of her plate and gets annoyed when you mention it. It does not explain why he burst into tears during a television commercial and then seemed completely fine four minutes later. It does not explain why, three months in, when the therapists start talking about discharge, everyone in the house quietly falls apart.

Here is the part that makes this worth your attention. The Stroke Roundtable Consortium divides recovery into defined phases, and the first three months after a stroke are the early subacute phase: the window in which the brain is most capable of reorganizing itself. The 90 days that matter most neurologically are also the 90 days when families are most alone with it.

This article is about what is in that gap. It is not medical advice, and nothing here replaces what your clinical team tells you. It is a map of the terrain the folder leaves out.

1. The fatigue that sleep does not fix

Start here, because families misread this one more than anything else on the list.

Post-stroke fatigue affects roughly 47 percent of stroke survivors according to pooled prevalence estimates, and the third Stroke Recovery and Rehabilitation Roundtable defined it carefully in 2024: an overwhelming exhaustion or lack of energy, involving physical, emotional, cognitive, and perceptual elements, which is not relieved by rest and which interferes with daily life.

Read that last part again. Not relieved by rest. This is not tiredness, and it is not deconditioning. Some people experience post-stroke fatigue despite high fitness levels. It does not track effort, and a full night of sleep does not clear it.

What families see instead is a person who will not get off the couch, and the interpretations that follow are almost always wrong and almost always hurtful: he has given up, she is depressed, he is not trying. The Roundtable authors are explicit that fatigue can superficially resemble depression and apathy, may co-occur with both, and is distinct from both.

There is a further reason nobody warned you. A systematic review of 200 stroke clinical guidelines found no strong recommendations for preventing or managing fatigue. The guidelines are largely silent, so the discharge folder is too.

What helps at home: Stop treating energy as unlimited and start treating it as a budget. Track which hours are usable, and spend them on the things that matter most, which usually means therapy and therapy practice rather than errands. Schedule rest before the demanding activity rather than after it. Expect that a good day will be followed by a bad one, because overspending on Tuesday is paid for on Wednesday. And say the words out loud to the rest of the family: this is a symptom of the stroke, not a character flaw.

2. The half of the world that disappeared

Hemispatial neglect is a failure to attend to one side of space. It is not blindness. The eyes work. The brain has stopped asking about that side.

It is common after right hemisphere stroke, with reported prevalence around 50 percent acutely and roughly 30 percent after left hemisphere stroke. One study of 166 rehabilitation patients with right hemisphere stroke found neglect in 48 percent of them, and also measured its effect on family burden, which tells you something about who ends up carrying it.

At home it looks like this. Food eaten from only one half of the plate. Shaving one side of the face. Bumping the doorframe with the same shoulder every time. Reading the right half of a page and being confused by the sentence.

Now the part that makes it genuinely hard. Neglect very frequently travels with anosognosia, a lack of awareness of the deficit itself. In one cohort of 85 patients with left hemispatial neglect, over 81 percent had anosognosia for it. So the person does not know the left side is missing, cannot be argued into knowing, and experiences your corrections as nagging about a problem that does not exist. That is not denial in the psychological sense. The machinery that would register the deficit is part of what the stroke damaged.

The instinct to resist: When you notice the untouched left half of the plate, the reflex is to rotate the plate. That solves the meal and teaches nothing. The therapy goal is usually to cue the person to scan toward the neglected side themselves, building the habit of looking. Rotating the plate does the looking for them. Ask the occupational therapist what cueing they want you using at home, and use that, consistently, at every meal. Safety hazards are a different matter and should simply be removed.

3. Speaking louder does not help

Aphasia affects roughly one third of stroke survivors. The single most important thing to understand about it is the thing most visitors get wrong: aphasia is a language impairment, not an intelligence impairment. The person in front of you is entirely themselves. They have lost access to the tool, not the thoughts.

Expressive aphasia means they know exactly what they want to say and cannot retrieve or produce it. Receptive aphasia means the words arriving do not decode into meaning. Many people have elements of both. Volume does nothing for either, because the problem was never the ears.

The statistic that should stop you: while roughly a third of the broader stroke population experiences depression, studies have reported that around 70 percent of stroke survivors with aphasia experience post-stroke depression, with anxiety prevalence estimated near 44 percent. And here is the cruelty stacked on top. Researchers have noted that people with aphasia are frequently excluded from depression screening precisely because they cannot complete the questionnaires. The group at the highest risk is the group least likely to be assessed.

What actually works: One idea per sentence. Then stop talking. Give ten or fifteen seconds of silence, which will feel unbearable to you and is exactly what the person needs. Ask questions that can be answered yes or no. Keep a pad and pen within reach, because a written key word often unlocks a sentence. Use gesture and point at objects, since context does real work. Never finish their sentence, however kindly you mean it. Never quiz, and never test them in front of visitors. And talk to them, not about them, when other people are in the room.

4. Crying that is not sadness

He cries at a car advertisement. He cries when the neighbor says hello. Then it stops, and he seems bewildered that you are upset.

This may be pseudobulbar affect, also called emotionalism or emotional lability: involuntary outbursts of crying or laughing that are disconnected from, or wildly out of proportion to, what the person actually feels. The American Stroke Association is direct that it is not depression, and notes something that should change how you respond to it. Often the person understands perfectly well that the response does not match the feeling. They are watching it happen to them.

The most useful distinction is duration. Pseudobulbar affect episodes last seconds to minutes, with normal mood in between. Depression is a sustained state lasting weeks to months. If the crying stops as abruptly as it started and the person seems fine underneath it, you are probably not looking at sadness. It is frequently mistaken for depression, and the two are treated differently, so it is worth raising with the physician by name.

In the moment: Do not leave the room, and do not make a thing of it. Stay, keep your face calm, wait, and carry on with what you were doing. Then handle the social cost in advance, because the anticipation of an outburst in public is what drives people into isolation. A short sentence to visitors ahead of time works: sometimes the stroke makes him cry when he is not sad, it passes in a minute, just keep talking to him.

5. Depression arrives on a schedule, and it arrives now

An updated meta-analysis of 79 studies put post-stroke depression at roughly 24 percent when measured by clinical interview and around 29 percent by rating scales, with other reviews landing near one third.

The timing is the part relevant to a 90-day article. That same analysis found that about two thirds of the depressive episodes occurring during the first year after stroke begin within the first three months. And of those early cases, over half went on to become persistent. The window this article covers is the window in which it starts.

It is also routinely missed. One review noted that non-psychiatric physicians fail to identify somewhere between 50 and 80 percent of post-stroke depression cases, because the symptoms are read as reasonable sadness about a stroke, or as fatigue, or as the stroke itself.

What families can do: Ask for screening explicitly, and ask again at six weeks and twelve weeks rather than assuming it happened. Bring specific observations rather than impressions, because “he seems down” is easy to wave off and “he has stopped calling his brother, he skipped therapy twice, and he no longer watches the games” is not. If the person has aphasia, say so and ask how the team plans to screen them anyway.

6. The swallowing rules are not suggestions

This section stays deliberately narrow, because swallowing is squarely clinical territory and belongs to the speech-language pathologist. But families need to understand why the rules exist, because the rules are what get broken first.

Dysphagia is present initially in at least half of stroke patients, and more than half of those who have it aspirate. Aspiration means food or liquid entering the airway. The reason this matters more than it sounds: in a widely cited review, the relative risk of pneumonia was roughly 3 times higher in patients with dysphagia and around 11 times higher in those with confirmed aspiration.

Now the detail that undoes families. Much of it is silent. Silent aspiration means material enters the airway and the person does not cough, does not choke, and shows no distress at all. Research indicates silent aspiration is a primary driver of aspiration pneumonia. So the sentence “he is not coughing, he is fine” is not reassurance. It is the exact scenario the precautions were built for.

The family role, precisely: Follow the plan the speech-language pathologist wrote, exactly, every time, including for snacks, including for visitors bringing treats, including when the person begs for a normal glass of water and it feels cruel to refuse. Do not upgrade the diet because things seem better. That decision belongs to the clinician who can actually assess the swallow. Do report what you observe: a wet or gurgly voice after drinking, coughing at meals, pocketing food in the cheek, a meal that now takes three times as long, or an unexplained low fever. Reporting is enormously valuable. Improvising is not.

7. The most dangerous thirty seconds of the day

Nobody tells you that the transfers are the risk. Bed to chair, chair to toilet, toilet to standing. Those few seconds, repeated a dozen times a day, are where the injuries come from.

Estimates of falls among stroke survivors in the first six months after discharge range from about 37 percent to 73 percent, with the variation driven largely by how the studies were designed and how mild or severe the population was. In one study following more than 1,100 stroke survivors, 37 percent reported at least one fall within six months. Of those who fell, 37 percent were injured badly enough to need medical treatment and 8 percent sustained a fracture. More than three quarters of the falls happened at home.

Two more numbers worth holding onto. As many as 58 percent of the people who will fall do so in the first month after discharge, meaning the risk is front-loaded into exactly the period when the family is least practiced. And because stroke survivors tend to fall toward the weaker side, with reduced protective reactions from the affected arm and bone density loss on that side, the risk of hip fracture after stroke has been estimated at up to four times that of healthy peers.

What to do about it: Ask the physical therapist to teach the transfer to whoever is actually doing it at home, and then to watch that person do it, unassisted, before discharge. Watching a professional do it is not learning it. Clear the path to the bathroom completely, because the night route is where this happens. Get the equipment before you need it rather than after the first fall. And treat any change in transfer quality as information worth reporting, not as a bad day.

8. Therapy is three hours a week. Recovery is the other 165.

This is the most important section in this article, and the least intuitive.

Researchers led by Catherine Lang observed what actually happens inside stroke rehabilitation sessions. Task-specific practice of functional upper limb movement occurred in only about 51 percent of the sessions that were meant to address the upper limb. When it did occur, the average number of repetitions per session was 32. For walking, the average was 357 steps per session.

Compare that to the animal research those rehabilitation principles were derived from, where subjects perform something like 400 to 600 repetitions per session, and where monkeys performed roughly 600 repetitions daily to reverse the effects of a cortical lesion. Other work found that patients spent around 47 minutes a day in occupational therapy in the early phase after stroke, of which only 4 to 11 minutes involved the upper limb. The authors of the observation study drew the obvious conclusion: the amount of practice provided during rehabilitation is small compared with the doses that drive neural reorganization.

This is not a criticism of therapists, who are working inside impossible constraints of time and reimbursement. It is a description of arithmetic. If formal therapy is three hours a week, then 165 hours a week are happening somewhere else, and that somewhere else is your kitchen.

Which is why the highest-leverage question a family can ask a therapist is not “how is he doing?” It is: “What should he be practicing between now and the next session, how many repetitions, and how do we know if we are doing it wrong?”

That question also reframes what daily help at home is for. The value of a caregiver during stroke recovery is not that tasks get done faster. It is that someone is present during the 165 hours, reinforcing the therapist’s plan rather than inventing one, encouraging the weaker hand into the task instead of quietly doing it for the person, and noticing on Tuesday what nobody would otherwise report until the following Monday. Families often frame this kind of nonmedical support during stroke recovery as help with daily living, which undersells it. Between therapy visits, consistency is the intervention.

One warning that belongs here. When the strong hand does everything because it is faster, the brain learns that the weak limb is not needed. Clinicians call this learned nonuse, and it is trained into people by well-meaning families every single day. Help that removes the effort removes the recovery.

9. The month-four crash, and the plateau myth

Around month three or four, two things collide. Therapy visits start winding down, and the pace of visible improvement slows. The conclusion everyone draws is that this is as good as it gets.

The honest version is more nuanced than either the pessimists or the inspirational posters will tell you.

The pessimists are partly right. Spontaneous neurological recovery, the healing that happens largely on its own, is at its most powerful in the first weeks and tends toward a relative plateau after about three months, with spontaneous recovery generally at its limit around six months. That slowdown is real, and pretending otherwise is its own cruelty.

But spontaneous recovery is not the only mechanism. Learning-dependent recovery, driven by deliberate practice, continues well beyond it, and improvements in the chronic phase are documented. The crucial thing to separate: being discharged from therapy at six months is a coverage and funding decision. It is not a neurological verdict. Those are two different things that arrive in the same envelope, and families read them as one.

So the useful framing is neither “you have plateaued” nor “you can recover fully if you just believe.” It is this: progress after the early window is slower, requires deliberate effort rather than time, and depends on someone maintaining the practice when nobody is scheduled to show up. A genuine plateau, where nothing changes despite consistent effort, is a signal to change the approach and ask the team what else is available. It is not a signal to stop.

What to ask before you leave the hospital

If you take one practical thing from this, take this list. Ask these before discharge, in writing:

• Has he been screened for neglect, and if so, what cueing do you want us using at home?
• Exactly what diet and liquid consistency is she on, and what specific signs should we call you about?
• Who is going to watch me do the transfer, and when?
• What should he practice between sessions, how many repetitions, and how will we know if we are doing it wrong?
• When will she be screened for depression, and how will you screen her given the aphasia?
• Is the fatigue we are seeing post-stroke fatigue, and what should we do about scheduling?
• What number do we call at 9 p.m. on a Saturday, and what counts as worth calling about?

None of this makes the first 90 days easy. What it does is stop the hardest parts from arriving as a surprise, and stop a family from misreading a symptom as a personality. Almost everything in this article gets mistaken for something else: fatigue for laziness, neglect for stubbornness, aphasia for confusion, emotionalism for despair, and a funding decision for a prognosis. Knowing which is which does not shorten the road. It just means you are walking the right one.