Concussion treatment targets what the examination finds. The problem could be vestibular (balance), oculomotor (eye movement), sleep, headache, metabolic or psychological. We treat that, rather than prescribing rest for the label. Care combines vestibular and oculomotor rehabilitation, a graded return to activity, and early sleep and headache treatment. It also includes metabolic care and Acceptance and Commitment Therapy delivered in-house.
The old advice was a dark room and complete rest until symptoms stopped. That is no longer the standard. Prolonged inactivity is now understood to slow recovery in many patients. What replaced it is targeted: identify which systems are affected, and treat those.
How is a concussion treated?
Picture four patients. One has headaches driven by an oculomotor problem. One has dizziness that is vestibular. One has a picture dominated by disrupted sleep. One has a headache that is actually coming from the upper neck. These are four different clinical problems sharing a diagnosis label. Sorting them is the point of the assessment.
- Vestibular and oculomotor rehabilitation — the best-supported intervention where the exam finds dysfunction. It is the reason we test for it on purpose.
- Graded return to activity — a step-by-step, symptom-guided return of physical and cognitive load, rather than open-ended rest.
- Sleep and headache management — usually the highest-yield early targets, because both amplify every other symptom.
- Interventional treatment of neck-origin headache where a diagnostic block identifies it — see occipital neuralgia vs concussion.
Does metabolic health affect concussion recovery?
Recovery from a brain injury is metabolically expensive. The injured brain has increased energy demand at exactly the moment its ability to meet that demand is disrupted. A patient whose baseline metabolic function is already impaired starts that process at a disadvantage.
This is not a side issue in the population we actually see. Insulin resistance, type 2 diabetes, obesity, obstructive sleep apnea and vascular disease are common in adults injured in crashes and falls. Yet they are routinely excluded from the concussion trials the entire field is built on. Those trials recruit young, healthy athletes.
So picture a 54-year-old with poorly controlled diabetes and untreated sleep apnea. When that patient does not recover on the timeline drawn from college football players, the honest reading is not that the patient is exaggerating. It is that the evidence base never included them. We assess and treat the metabolic terrain alongside the injury. We do not treat it as somebody else’s problem.
Why does sleep matter after a concussion?
Sleep is where a great deal of concussion recovery either happens or fails to. Disrupted sleep is a symptom of the injury. It is also an independent driver of headache, cognitive difficulty and mood. And insomnia is one of the features that independently predicts persistent symptoms. Treating it early is not symptomatic tinkering; it removes an amplifier sitting underneath everything else.
Where sleep-disordered breathing is present, it is investigated. That is because no amount of vestibular rehabilitation makes up for a brain that is intermittently hypoxic (short of oxygen) all night.
Can therapy help after a concussion?
Persistent post-concussive symptoms reliably produce a second problem on top of the first: fear of the activities that provoke symptoms. Avoidance is a rational short-term response. But it is a poor long-term strategy, because the deconditioning and withdrawal it causes make the underlying problem harder to treat.
This is measurable. A prospective cohort followed adults who were working at the time of their mild TBI. In that cohort, persistent or increasing fear-avoidance behavior was associated with significantly greater odds of still being off work 6–9 months later. That held independent of the injury itself.
Acceptance and Commitment Therapy (ACT) is aimed at exactly this. It does not try to argue someone out of their symptoms. Instead, it works on psychological flexibility. That means reducing avoidance, clarifying what the person actually wants their life to contain, and rebuilding activity in that direction while symptoms are still present. It is delivered in-house here rather than referred out. That keeps it integrated with the physical rehabilitation instead of running on a separate track.
To be clear about what this is not: offering ACT is not a statement that the symptoms are psychological. The vestibular injury is real and gets treated as such. ACT addresses the avoidance spiral that a real injury sets off.
Snell DL, et al. Fear avoidance and return to work after mild traumatic brain injury. Brain Inj. 2023;37(6):541–550. doi:10.1080/02699052.2023.2180663
What should you eat after a concussion?
If the core problem in the first days is an energy supply-and-demand mismatch, then fueling matters more than it usually gets credit for. The practical points are unglamorous and worth stating plainly:
- Do not under-eat during recovery. The injured brain’s glucose demand is raised. Skipping meals because you feel nauseated or foggy takes away fuel exactly when demand is highest.
- Hydration — dehydration worsens headache and cognitive symptoms on its own. It is common in people who have stopped their normal routine.
- Alcohol — disrupts sleep architecture (the normal pattern of sleep stages) and lowers seizure threshold. The case for avoiding it during recovery is straightforward.
- Omega-3 fatty acids, magnesium and creatine appear repeatedly in the concussion nutrition literature. They have plausible mechanisms. Their human support is mostly preclinical or small-trial. Reasonable, low-risk, not established treatment.
The honest framing: nutrition supports the conditions for recovery rather than driving it. Anyone selling a supplement protocol as concussion treatment is ahead of the evidence.
Sleep support without benzodiazepines — glycine and GABA
Sleep is one of the highest-value targets in concussion recovery. And insomnia is one of the features that independently predicts persistent symptoms. So the clinical question is not whether to treat sleep. It is what to treat it with.
The default answer in most settings is a benzodiazepine or a related sedative-hypnotic. In a brain-injured patient, that is a poor trade. Here is the reasoning behind our approach.
Why not a benzodiazepine — they block learning
The usual objections to benzodiazepines are sedation, falls and dependence. All are real. And all are beside the main point in this population. Benzodiazepines impair the formation of new memories. Picture a patient whose central complaint is that their memory and concentration are not what they were. For that patient, losing new memories is not a side effect to be weighed — it is the injury being prescribed for a second time.
This is not a fringe concern or a matter of dose. Across the class, benzodiazepines produce dose- and concentration-dependent amnesia (memory loss). The amnestic effect is characterized as impairment of information acquisition, impairment of consolidation and storage, or both. Anterograde amnesia — the failure to lay down memories going forward — is produced so reliably that research uses it as a model of hippocampal cognitive dysfunction. The drug is a standard way to cause, in experiments, the deficit the patient came in with.
Greenblatt DJ. Pharmacology of benzodiazepine hypnotics. J Clin Psychiatry. 1992;53(Suppl):7–13. · Venault P, et al. Benzodiazepine impairs and beta-carboline enhances performance in learning and memory tasks. Nature. 1986;321(6073):864–866. doi:10.1038/321864a0 · Vandesquille M, et al. J Psychopharmacol. 2012;26(6):845–856. doi:10.1177/0269881111416692
And they remove the window where memory is consolidated
The damage is done twice, by two separate routes. The second is easy to miss.
Memory is not filed at the moment of learning. It is consolidated (locked in) during sleep — specifically during slow-wave sleep (deep sleep). This holds specifically for the hippocampus-dependent declarative memories (facts and events) that concussion patients complain about. In a landmark experiment, newly learned material was cued with an odor during slow-wave sleep. That improved retention of declarative memories. Functional imaging showed hippocampal activation in response. The same cue during REM sleep or during wakefulness did nothing. Procedural memories — which do not depend on the hippocampus — were unaffected.
Now recall what long-term benzodiazepine use does to sleep architecture. It causes a marked depression of slow-wave activity. Delta power (deep-sleep brain waves) is reduced during NREM sleep, and the cyclic alternating pattern is sharply reduced.
So look at the drug given to help a concussed patient sleep. It impairs encoding while they are awake, and then suppresses the sleep stage in which whatever did get encoded would have been consolidated. Both hits land on the same hippocampus-dependent system. That is the system already injured. The patient sleeps more and remembers less. The memory complaint that brought them in gets worse while being treated.
Rasch B, Büchel C, Gais S, Born J. Odor cues during slow-wave sleep prompt declarative memory consolidation. Science. 2007;315(5817):1426–1429. doi:10.1126/science.1138581 · Manconi M, et al. Sleep architecture in insomniacs with severe benzodiazepine abuse. Clin Neurophysiol. 2017;128(6):875–881. doi:10.1016/j.clinph.2017.03.009 — note this studied long-term high-dose use, not ordinary short-term prescribing.
There is a final turn of the screw. Look at the features that independently predict who develops persistent post-concussive symptoms. Of these, retrograde amnesia carries the highest odds ratio of any symptom measured — and difficulty concentrating is close behind. The patients most likely to be handed a sedative-hypnotic for poor sleep are disproportionately the ones whose memory is already the problem. They also have the most to lose from a drug that works on exactly that system. See the long-term effects of concussion.
Sedation, unsteadiness, fall risk and dependence are all still true. In a patient already at risk of a second head injury from a fall, they matter. But the reason we reach for something else first is the learning.
Glycine: a thermoregulatory route into sleep, not a sedative
Glycine taken before bed has been reported to improve subjective sleep quality in people with a tendency toward insomnia. The mechanism is the interesting part, because it is not sedation.
Oral glycine raises plasma and cerebrospinal fluid glycine. It increases cutaneous (skin) blood flow and lowers core body temperature — and a falling core temperature is a normal precondition for sleep onset, which is maintained through the night. In rodents, glycine shortened the time it took to reach non-REM sleep. The effect was localized to NMDA receptors in the suprachiasmatic nucleus (the SCN, the brain’s master clock). Ablating (destroying) the SCN abolished both the sleep-promoting and the hypothermic (cooling) effect entirely.
So glycine works by nudging the body’s own sleep-onset machinery. It does not work by suppressing the central nervous system. That difference is the whole point — it does not trade slow-wave sleep away to get unconsciousness.
Bannai M, Kawai N. New therapeutic strategy for amino acid medicine: glycine improves the quality of sleep. J Pharmacol Sci. 2012;118(2):145–148. doi:10.1254/jphs.11r04fm · Kawai N, et al. The sleep-promoting and hypothermic effects of glycine are mediated by NMDA receptors in the suprachiasmatic nucleus. Neuropsychopharmacology. 2015;40(6):1405–1416. doi:10.1038/npp.2014.326
GABA: the inhibitory neurotransmitter, working through the gut
GABA is the brain’s major inhibitory neurotransmitter — the signal that damps excitatory activity down. In a concussion, a flood of the excitatory neurotransmitter glutamate is a defining early event. After a concussion, then, restoring inhibitory tone is a coherent thing to want.
The standard objection to oral GABA is that it crosses the blood–brain barrier poorly. That objection is fair. It is also why the relevant site of action is the gut.
GABA turns out to be a live currency in the gut microbial ecosystem, not an inert passenger. In work published in Nature Microbiology, one gut bacterium could not be grown in culture at all until GABA was supplied — GABA was the only tested nutrient that supported its growth. Bacteroides species were found to produce large quantities of GABA. Genome-scale modeling identified many genera (groups of bacteria) that produce or consume it. And stool from healthy people showed GABA-producing pathways at work. They were actively expressed by Bacteroides, Parabacteroides and Escherichia. Take patients with major depressive disorder. In them, the relative abundance of fecal Bacteroides (how much was in the stool) was negatively correlated with brain signatures associated with depression.
That is the route we are working with. GABA takes part in the gut ecosystem, and the gut–brain axis carries the consequences — not a pill of GABA traveling to the brain and acting there directly. Anyone who tells you oral GABA sedates you the way a benzodiazepine does is describing a pharmacology that does not happen.
Strandwitz P, et al. GABA-modulating bacteria of the human gut microbiota. Nat Microbiol. 2019;4(3):396–403. doi:10.1038/s41564-018-0307-3
The route is gut → vagus nerve → brain
The missing link between “GABA is active in the gut” and “this affects how you sleep” is the vagus nerve — the main neural cable of the parasympathetic nervous system and the principal channel through which the gut talks to the brain.
The cleanest demonstration is a study in mice. Mice fed Lactobacillus rhamnosus showed region-specific changes in brain GABA receptor expression. They also had reduced stress-induced corticosterone (a stress hormone) and less anxiety- and depression-related behavior. The decisive part of the experiment came next: none of those neurochemical or behavioral effects appeared in vagotomized mice. Cutting the vagus abolished them entirely. That identified the nerve as the communication pathway between gut bacteria and the brain.
That is the whole mechanism in one experiment. Bacteria live in the gut, with GABA as the signaling currency between them. The vagus carries the message. At the far end, there is measurable change in the brain’s own inhibitory receptors — with no requirement that swallowed GABA reach the brain itself.
Bravo JA, Forsythe P, Chew MV, Escaravage E, Savignac HM, Dinan TG, Bienenstock J, Cryan JF. Ingestion of Lactobacillus strain regulates emotional behavior and central GABA receptor expression in a mouse via the vagus nerve. Proc Natl Acad Sci USA. 2011;108(38):16050–16055. doi:10.1073/pnas.1102999108
The inhibitory side of the nervous system is substantially a gut story. Take the neurotransmitters that quiet things down, and the parasympathetic “rest and digest” state they support. Neither is confined to the brain — they are produced, consumed and signaled from the gut, and relayed upward. Sleep is a parasympathetic activity. Getting into it is not a matter of overriding the nervous system with a sedative but of letting the system that induces it work.
This matters after a concussion specifically. That is because autonomic disturbance is part of the injury — which is why exercise intolerance, poor temperature control and a racing pulse on standing are such common complaints. We will be careful here, though. Heart rate variability (HRV) has been studied as a marker of that disturbance. The findings are openly conflicting, and the reading of its low-frequency component is still debated. Autonomic involvement after concussion is real. But HRV is not a settled diagnostic for it, and we do not present it as one.
Bishop SA, Dech RT, Guzik P, Neary JP. Heart rate variability and implication for sport concussion. Clin Physiol Funct Imaging. 2018;38(5):733–742. doi:10.1111/cpf.12487
Where this honestly sits
We choose precise over impressive:
- There is no randomized trial of glycine or GABA as a treatment for concussion. The reasoning is indirect. Sleep drives recovery, these support sleep, and so they support recovery. That is a chain of inference, not a trial result.
- The human glycine data are small and rest on subjective sleep measures. The detailed mechanism comes from rodent work.
- The GABA–microbiome–vagus work is mechanistic — the decisive vagotomy experiment is in mice. The human microbiome findings are associative. Neither is an intervention study in brain injury.
- What is solid is that sleep matters here and that these carry low risk. Also solid: the alternative they displace — sedative-hypnotics in a cognitively impaired, fall-prone patient — carries real and documented harm.
They are adjuncts (add-ons) to sleep hygiene, treatment of headache and mood, and investigation of sleep-disordered breathing — not a substitute for any of those.
Other adjunctive approaches
- Transcranial magnetic stimulation (TMS) — non-invasive brain stimulation. It has been studied for persistent post-concussive symptoms, with encouraging but not definitive results.
- Virtual reality rehabilitation — an effective way to deliver graded vestibular and cognitive work, rather than a distinct treatment in itself.
- Whole body vibration therapy — mechanical stimulation used as an adjunct. It is emerging, and drawn from musculoskeletal and neurological rehabilitation.
- Thermal therapies — controlled cooling and sauna-based heat exposure have both been studied in recovery settings. They are mechanistically plausible, with early evidence.
A word about the trial populations
Most of what is known about concussion recovery comes from studies of young athletes. They are healthy, monitored, and rapidly identified. Those studies systematically exclude older adults, people with prior head injury, and people with diabetes, obesity, sleep apnea or vascular disease. Those exclusions describe a large share of the people who actually arrive at a concussion clinic after a crash or a fall at work.
When the evidence base does not cover a patient, the honest response is to say so and treat what the examination shows. It is not to apply a return-to-play timeline built on college athletes and call it settled.
Related reading: concussion care for Illinois patients and seeing a TBI doctor in St. Louis.
Common questions
Is rest still the treatment?
Not as it used to be prescribed. Brief relative rest in the first day or two, then a graded, symptom-guided return to physical and cognitive activity. Prolonged inactivity is now understood to slow recovery in many patients. It also feeds the avoidance cycle that predicts staying off work.
Why would a concussion clinic care about my diabetes or sleep apnea?
Because recovery is metabolically expensive, and both conditions impair it. They are also excluded from most concussion research. That is a large part of why patients who have them do not follow the published timelines, and get told their symptoms must be something else.
The detail is in Why Two People With the Same Concussion Recover Differently.
Does offering therapy mean you think my symptoms are psychological?
No. The vestibular, oculomotor and headache problems are physical and get treated as such. ACT addresses the avoidance and deconditioning that a real injury sets off. That pattern independently predicts still being off work 6–9 months later.
The return-to-work evidence behind that is in The Long-Term Effects of Concussion.
Do supplements help concussion recovery?
The gut-brain rationale is real. The trial evidence specific to concussion is early and thin. Glycine has better-supported evidence for sleep quality than for brain injury recovery. We use these as low-risk adjuncts, and we do not present them as established treatment.
What should you not do after a concussion?
Do not settle into weeks of complete rest in a dark room. Prolonged inactivity slows recovery in many patients. It also feeds the avoidance that keeps people off work. Do not skip meals, because the injured brain needs more fuel, not less. Stay hydrated. Avoid alcohol, which disrupts sleep and lowers the seizure threshold. After the first day or two, return to activity in graded, symptom-guided steps.
What is best for concussion recovery?
Treatment aimed at what your examination finds. Vestibular and oculomotor rehabilitation is the best-supported option where those problems are present. Sleep and headache are usually treated early, because both amplify every other symptom. A graded, symptom-guided return to physical and mental activity replaces open-ended rest. Conditions such as diabetes or sleep apnea are treated alongside the injury, because they slow recovery.
Will a concussion heal if it is left untreated?
Most do. The energy crisis inside the brain settles over days to weeks, and most people are back to themselves within a month without specific treatment. The ones that do not clear are rarely mysterious. An untreated vestibular or eye-movement deficit, a headache coming from the neck, or sleep that never recovered keeps the symptoms running, and time does not repair a sensory or mechanical problem. Those are worth finding in the first weeks rather than after a year of waiting.
Learn more: what slows recovery in adults who are not athletes.
Do physical therapists treat concussions?
Vestibular physical therapists treat one part of it, the balance and gaze problems, and they do it well once someone has established that those are the problem. A concussion is rarely only that. A neck-origin headache, broken sleep, eye-tracking deficits, medication effects and the metabolic side need a physician’s examination first. We test first, then coordinate with vestibular therapy and optometry where the findings call for it, so the therapist starts with a named target instead of a general referral.
What that testing measures is laid out in concussion diagnosis and testing.