The stroke symptom nobody likes to talk about
When people imagine stroke recovery, they picture weakness on one side, slurred words, maybe a walking cane. Few imagine what happens at the dinner table. But for many survivors, the scariest lingering problem is swallowing — specifically, a condition called cricopharyngeal achalasia.
The name sounds clinical, but the experience is not. The cricopharyngeus is a ring-shaped muscle at the top of the esophagus that normally relaxes when you swallow to let food pass. After some strokes, it forgets how. It stays tight. Food and liquid hit it and either pool above or sneak into the airway. Meals become dangerous. Pneumonia becomes a recurring visitor. Nutrition falls off.
Why now
This problem has had a patchwork of treatments for years, and clinicians have argued about which work best. A Chinese research team just published the first large network meta-analysis on it — the kind of statistical tool that ranks treatments even when they were never tested directly against one another.
They combed 10 databases and trial registries through December 2025, pulled 36 randomized trials, and compared 13 different interventions across four swallowing outcome measures. No prior review had mapped the field this completely.
The treatments in the ring
Most approaches start from the same base: balloon dilation, where a thin balloon is eased past the tight muscle and gently inflated to stretch it open. Used alone, it helps — but most of the trials in this analysis paired dilation with something else. The add-ons included:
- rTMS (repetitive transcranial magnetic stimulation, a non-invasive brain-nudging device)
- Acupuncture
- Botulinum toxin (Botox) injection into the muscle
- EMG biofeedback (seeing your own muscle activity on a screen)
- tDCS (transcranial direct current stimulation)
- Tongue-pressure resistance training
The winners depend on what you are measuring
Here is where this study gets genuinely useful — the best treatment changes based on what problem you are trying to solve.
For an overall "it worked" measure, the top combination was balloon dilation plus rTMS — the brain-stimulation add-on. Ranking probability: 91.8%.
For the VFSS score, a videofluoroscopy-based imaging test that shows how food actually moves through the throat, balloon dilation plus acupuncture took the top spot at 92%.
For FOIS, the functional scale that tracks what real foods and drinks a patient can safely eat, balloon dilation plus Botox led at 87.4%.
For the SSA swallowing safety score, acupuncture again topped the list at 80%.
Why do the winners change?
Each outcome measures a different slice of the swallowing problem. Videofluoroscopy cares about physical movement of the muscle. FOIS cares about daily eating. The effective-rate scale blends everything. Brain stimulation, needles, and Botox all work through different mechanisms, so it makes sense they would each shine in different lanes.
For clinicians, this is actually clarifying rather than confusing — it means the treatment plan can be tailored to the patient's biggest problem.
Re-engaging with what it means
A patient who is aspirating liquids and losing weight cares most about the FOIS outcome — can they eat real food again? The review points toward balloon dilation plus Botox.
A patient who passes the eating test but whose imaging still shows a sluggish muscle may benefit more from the acupuncture pairing.
And for patients whose overall recovery has stalled, rTMS is the pairing with the strongest "it worked" signal.
If you or a loved one is recovering from stroke with persistent swallowing trouble, three things follow from this research:
1. Ask specifically about combination therapy. Balloon dilation alone often does not bring patients back to normal eating. The add-on matters. 2. Know which outcome your team is tracking. If they only watch one score, you may miss improvements in another. 3. Acupuncture shows up twice in the top rankings. It is worth asking whether your rehab center offers it in combination with dilation.
Brain stimulation devices are not yet everywhere, but more rehab centers are adding them each year.
The limitations worth naming
The authors are careful to warn readers. Many included trials had small samples, weak allocation concealment, and blinding challenges that are nearly impossible to solve when one treatment involves needles and another involves a magnetic device. Publication bias may have nudged some outcomes higher than they deserve.
Translation: the rankings are a useful map, not a verdict. A treatment at the top of one list might slide down once better-designed trials arrive.
What this field needs next is direct, head-to-head comparisons — ideally between the top two or three combinations on FOIS, which is the outcome that most matters to patients' daily lives. Larger sample sizes and sham-controlled designs for the brain stimulation arms would also tighten the evidence.
For now, the message is hopeful: post-stroke swallowing problems are not a dead end. There is a menu of add-on therapies, and at least some of them are beating dilation alone.