How Mei, a Boston vestibular physical therapist, cured her vestibular migraines after two years of specialists.
Two years of ENTs, neurologists, and neuro-otologists. Thirty-five sessions of vestibular rehab at her own clinic. Four preventive medications. Botox. A full migraine elimination diet. Six thousand dollars out of pocket. Then a Curable podcast episode and a 2017 paper she'd printed years earlier and filed away.
Mei is a 34-year-old vestibular physical therapist in Boston. She treats dizziness for a living. When vestibular migraine hit her after a virus, none of her medications, rehab, or diets worked. Two years in, she applied central sensitization research to herself. Ten months later she was running the Charles again. A 2025 case series in Headache documented similar outcomes with brain-based therapy for three patients with chronic migraine.
Fishbein et al., 2025, HeadacheTwo years of ENT workups, medications, and vestibular rehab that barely moved the needle.
Mei was 32 in November 2022. She caught a nasty virus going around her clinic. Three days of fever, sore throat, body aches. She slept it off and went back to work a week later. On her second morning back, she was guiding a 72-year-old post-concussion patient through gaze stabilization in Exam Room 3, and the floor started to tilt. Not dramatically. A slow sustained lean. Like the clinic was on a boat.
She thought it was her blood sugar. She'd skipped breakfast. She went to the break room, ate a granola bar, drank water. The tilt didn't go away. She finished her afternoon sessions with a tight controlled walk and a hand on the wall whenever no patient was watching. She told David she'd picked something up from a patient and went to bed early. The tilt was still there the next morning. And the morning after. Two weeks later she was at her primary care doctor. Three months later she'd been to four specialists. Two years later she was still dizzy.
She moved through a specific ladder every vestibular patient at her clinic knew by heart. An ENT workup at Mount Auburn Hospital that came back clean. A neurologist at a large academic medical center who ran an MRI (normal), diagnosed vestibular migraine, and started her on propranolol. Eight weeks of propranolol, then 80 milligrams for a month. Dizziness dropped maybe 20 percent. Side effects: fatigue heavy enough she had to sit down mid-session, cold hands, eight pounds of weight gain, and an inability to run. Discontinued. Thirty-five sessions of vestibular rehab with her closest colleague, Jen, a specialist at Mei's own clinic. Daily gaze stabilization, habituation drills, balance work on foam. Compliance log at 99 percent. DHI dropped six points over six months and plateaued.
A second neurologist at the Beth Israel Deaconess headache and vestibular clinic. Venlafaxine, 37.5 milligrams titrated to 150. Five months. Dizziness at work dropped maybe 40 percent. The side effects came in quickly: dry mouth, insomnia, emotional flattening, and what she could only describe as feeling one inch behind her own face. Tolerated it until she couldn't, and tapered off. Nortriptyline for four weeks with a worse side effect profile. Botox injections off-label for vestibular migraine, $1,800 out of pocket because insurance denied the claim. Zero benefit at 12 weeks. She did not do a second round.
A neuro-otologist at Mass Eye and Ear on a six-month wait. Detailed videonystagmography, rotary chair testing, VEMPs. All normal or minimally abnormal in ways that didn't explain her symptoms. Diagnosis: vestibular migraine with features of PPPD. Recommended an SSRI next. Mei declined. She was done with medications. She tried the full Heal Your Headache diet for four months. No aged cheese, no cured meat, no alcohol, no chocolate, no nuts, no citrus, no caffeine after noon. Brown rice and steamed broccoli at her sister's wedding. Lost six pounds. Did not get less dizzy. In September 2024 she bought The Steady Coach online program out of reluctance and watched the first three modules. Some of it felt not scientific enough to her, despite having the research right there in the citations. David had been saying for months, "This isn't your identity. It's something that's happening to you." She'd nod. Then she'd open PubMed.
The breaking point was her sister's wedding rehearsal dinner in October 2024. An Italian restaurant in the North End. Forty people. Dim lighting. A slowly rotating Edison-bulb chandelier. Italian waitstaff moving quickly past the table. Forty-five minutes in, Mei had to excuse herself. She sat on a bench in the hallway with her eyes closed. Her mother found her after ten minutes, sat down, and said in Mandarin: you've been sick for a long time now, you can't keep doing this by yourself. Her sister, the pediatric resident, came out to find them. "Jiejie, I love you. But you have to try something different. Whatever you're doing isn't working." Mei went home with David early. That night, lying in bed, she said out loud for the first time, "I think my brain is doing this. I've been pretending it was a puzzle I could solve with more data." David just said, "Okay. Now what?"
"I had become the patient every PT hates to see," Mei says. "The one who intellectualizes her own condition. I could cite the papers. I could draw you a picture of the parieto-insular vestibular cortex on a napkin. I was still dizzy every single day."
Mount Auburn Hospital · ENT Clinic
12 / 07 / 2022 · Pt: M. Zhang, DPT
I could draw you the parieto-insular vestibular cortex on a napkin. I was still dizzy every day.Mei, Year 2
The Curable podcast and a 2017 papershe'd printed and filed away.
After the rehearsal dinner, Mei went back to The Steady Coach program she'd half-engaged with the month before and actually did the work. A week later a Curable podcast episode popped into her queue. The guest was Dr. Yonit Arthur, the physical therapist who ran The Steady Coach and had herself recovered from vestibular migraine and PPPD. Arthur described a 2017 vestibular neuritis study Mei had printed years earlier for a case conference and then filed away.
The study was Cousins 2017 in Annals of Clinical and Translational Neurology. Forty patients tracked for a year after inner ear infections. Peripheral vestibular data explained 12 percent of who developed chronic dizziness. Visual dependence and psychological factors explained 59 percent. Mei paused the podcast, opened the paper on her laptop, and read it twice. Then she ordered The Way Out by Alan Gordon and read it over a weekend.
Three things clicked. Her ear had healed 22 months ago. What remained was a brain that had learned to predict dizziness and was generating the signal from the top down. She knew this in theory for her patients. She had not applied it to herself.
Her vestibular rehab hadn't worked because she'd been doing it wrong for herself. She'd been grinding through exercises with hypervigilance, tracking compliance, anxiously monitoring whether she felt worse afterward, turning each session into a referendum on her recovery. The research Arthur cited showed that exact pattern slowed vestibular compensation. Mei had been the reason her own VRT wasn't working.
The message of The Way Out, The Steady Coach, and the central sensitization literature was the same. Safety signals are what allow an over-sensitive system to calm down. Mei had treated every moment of dizziness for 22 months as danger to be eliminated. She'd never let the dizziness just be present without responding to it.
One sentence from The Way Out, Chapter 4, did most of the work: "The brain that learned to generate the pain is the same brain that will unlearn it. You don't need new tissue. You need a new message." Mei treated nervous systems all day. This wasn't a metaphor to her. This was mechanism.
Functional
Her Mount Auburn ENT workup was clean. Her MRI was normal. Her VNG, rotary chair, and VEMPs at Mass Eye and Ear showed nothing that explained her symptoms. Her viral illness had cleared 22 months earlier. In Cousins 2017, peripheral vestibular tests explained only 12 percent of chronic dizziness outcomes. Her 35 sessions of VRT, delivered by one of the best specialists in the region, dropped her DHI six points. A peripheral problem would have responded.
Strong matchInconsistent
Dizziness was worse in visually busy environments (supermarkets, the T, bright restaurants). Better on forest walks. Worse on Monday mornings before clinic. Better on Saturday mornings. Shape-shifted in quality: some days a tilt, some days a fog, some days a floaty disconnection. Two places were nearly dizziness-free: swimming laps at the BU pool and playing the piano. Peripheral damage doesn't read the day of the week.
Strong matchTriggered
Flared before insurance audits at the clinic. Flared the day she misread an email from her father's cardiologist. Flared through her sister's wedding planning. Did not flare from the known peripheral triggers in her own rehab manual: head turns, rising from sitting, rapid scans. Her dizziness responded to emails. The inner ear doesn't read email.
Strong matchWhat recovery actually looked like.
It was not a clean slope down. She had an extinction burst in month two, during Thanksgiving week at her parents' house in Newton, and for three days she was sure the whole approach was wrong. It wasn't.
DHI 58. Dizziness at 7 most days. Came home from the rehearsal dinner and told David out loud that her brain was doing this. Re-enrolled in The Steady Coach and actually did the work this time. Stopped the Heal Your Headache diet. Ate a slice of pizza for the first time in five months and didn't get dizzier. Stopped tracking her DHI daily. Threw out the printed log. Kept going to work, including to her vestibular patients. Drove on I-93. Let the dizziness be present without rescuing herself from it.
DHI 48. Dizziness at 5. Noticed something she'd have noticed in a patient. The dizziness was dropping fastest in places where she'd stopped fighting it. It was still high in places where she'd been most afraid, the Star Market on Mass Ave and the Red Line T. Textbook central sensitization. She'd been reading the textbook for six years and ignoring it in her own chart.
DHI 64. Dizziness at 8. Thanksgiving week at her parents' house in Newton. The dizziness came back worse than it had been in six months. Her mother fed her bone broth and herbs. Mei called into the Steady Coach group coaching call on Thursday and said she was regressing, the whole approach was wrong. A man in his 50s who'd recovered from four years of vestibular migraine said quietly, "This is the bit where it feels wrong. Keep going." Five days later the dizziness dropped to a 4. Lower than it had been before the flare.
DHI 28. Dizziness at 2. Ran a 10K along the Charles for the first time in two years. Had been avoiding running because it had been a trigger. It wasn't a trigger. It had been a fear. Back to treating her own vestibular patients and introducing them to central sensitization research in the first session. Two of them had the aha moment faster than she had.
DHI 12. Dizziness at 1. Flew to Shanghai for the first time in two years. Fourteen hours. No anti-nausea prescription in her carry-on. Walked through Pudong Airport, rode the Maglev, went to the Yuyuan Garden market with her grandmother. Her grandmother, 87, noticed nothing. Mei noticed that she noticed nothing.
DHI 4. Dizziness at 0. She caught herself, sometimes, looking for the dizziness like a lost set of keys. Then remembered there were no keys. Mei and David found out they were expecting their first child. She'd started giving a talk at PT conferences about vestibular migraine as a central sensitization phenomenon. The talk is anonymized. She'll write the non-anonymized version someday. Or maybe this is that post.
The three things that changed everything.
She stopped treating her own body as a clinical case.
For 22 months Mei had been charting her own condition like a patient record updated in real time. Symptom diary she'd only have recommended a patient keep short-term. DHI score trend spreadsheet. VRT compliance log. Medication timing dashboard. Diet adherence tracker. She was cross-referencing variables with the kind of rigor she'd apply to an actual clinical case.
In her first week post-turning, she threw out the DHI printout. Stopped the food log. Deleted the compliance spreadsheet. Went to work without asking herself once how the dizziness was reading. Sat in her kitchen and ate pizza. Sat in bright supermarkets. Let her body live without being continuously observed and measured.
The clinical detachment was the exact hypervigilance that was keeping her brain alert to danger. Drop the clinical gaze and the brain stops being cued. Her body wasn't a chart. It was a body. It turned out to be fine without the surveillance.
She stopped being afraid of the wrong things.
Before, Mei avoided supermarkets, bright restaurants, overhead fluorescents, freeway driving, and her mother's kitchen lighting. When she had to enter any of them, she pre-medicated with meclizine, braced herself, and counted the minutes until she could leave. Twenty-two months of this had taught her brain that those environments were genuinely dangerous.
After the turning point, she kept going into every one of them without pre-medicating. Lingered in the Star Market on Mass Ave for ten extra minutes after her shopping was done. Drove I-93 at Friday rush hour. Sat in the middle of a loud dinner in the North End facing the kitchen lights on purpose. Let her brain receive the data: I'm in this environment and nothing bad happens when I'm dizzy.
Brains are Bayesian. They update on new evidence. She gave her brain four months of new evidence and the brain updated.
Remove the middle link and the environment stops firing the alarm.
She stopped waiting for the dizziness to leave before she lived again.
Before, she ran the same line about every life she was putting on hold. When the dizziness is better I'll go back to running. When it's better I'll visit my grandmother. When it's better we'll try for a baby. Life was paused waiting for a health state that wasn't coming on its own.
After, she ran along the Charles at a dizziness level of 5. Booked the Shanghai flight at a 3. Started trying for a baby at a 2. Didn't wait for the all-clear. Lived into her own life around the dizziness, until the dizziness became an afterthought.
Pausing life had signaled to her brain that life was dangerous. Resuming life signaled safety. The dizziness faded faster once she stopped making it the arbiter of what she was allowed to do.
Back in her own body.
Mei's been dizziness-free for almost a year. Not flare-free. Dizziness-free.
She runs along the Charles most mornings. Twenty-five miles a week. Leads her clinic's vestibular team. Treats her own patients all day without once thinking about her own balance. When a new vestibular migraine patient comes in, she hands them a photocopy of the Cousins 2017 abstract in the first session and tells them what she wishes someone had told her.
She flew to Shanghai in spring 2025 and spent two weeks with her grandmother, including a full day at the Yuyuan Garden market. She drove the Mass Pike and the Tobin Bridge at rush hour without her polarized sunglasses. She gave her first talk at a PT conference on brain-based vestibular rehabilitation in January 2026. She and David are expecting their first child this summer.
She still gets a half-day of mild floatiness maybe once every six weeks. Usually around a clinic audit, a bad phone call from her mother, or a week of poor sleep. She doesn't track it. She doesn't medicate. She doesn't tell herself a story about it. Usually it's gone by the next morning.
"A 1 out of 10 floatiness used to make me panic and open PubMed for two hours," she says. "Now it's the weather. That's the piece nobody sold me on. Not the absence of dizziness. The absence of the panic that used to follow it. I was a vestibular PT who couldn't treat herself. That's the story I used to tell. The real story is I was a scientist who trusted data until the data was about me, and then I trusted fear. Fear is not data. That's what I learned. I'm the next clinician who knows."
Does Mei's story sound familiar?
If your vestibular migraine started after a virus, an ear infection, a concussion, or a hard stretch of stress, and your MRI is clean, your VNG is clean, your rotary chair is clean, and your medication list is long and your dizziness list is longer, there's a good chance your condition follows the same pattern Mei's did.
Take the 3-minute assessment→