Vertigo is not a diagnosis. It is a symptom with a short list of common causes, and the list matters, because one of them is reversed in a single consultation and another one can be a stroke.
Before anything else: Proco sells magnesium glycinate, omega-3 and three sleep and cognitive products. None of them treats dizziness, vertigo or any vestibular disorder, and we found no supplement with randomised trial evidence in any of the conditions below. The effective treatment for the most common cause is a physical manoeuvre performed by a clinician. It is not something you can buy.
The 2017 American Academy of Otolaryngology–Head and Neck Surgery (AAO-HNS) guideline notes that recent work emphasises timing and triggers over the words a patient uses. It sorts acute dizziness into a triggered episodic syndrome (episodes follow an obligate action such as a head position change, generally under a minute), an acute vestibular syndrome (persistent continuous dizziness lasting days to weeks), and a spontaneous episodic syndrome (untriggered episodes of minutes to hours).
BPPV is the triggered episodic case. The AAO-HNS guideline diagnoses posterior canal BPPV when vertigo with torsional, upbeating nystagmus is provoked by the Dix–Hallpike manoeuvre, a positional test done in clinic. That statement is a strong recommendation at Grade B evidence. If the history fits but the Dix–Hallpike shows horizontal or no nystagmus, the guideline recommends a supine roll test for lateral canal BPPV, which the American Academy of Neurology (AAN) practice parameter — published in 2008 and since formally retired, though not superseded on these points — puts at 10% to 17% of cases.
Vestibular neuritis is an acute vestibular syndrome. The Bárány Society's 2022 criteria require sustained vertigo of moderate to severe intensity lasting at least 24 hours, a spontaneous direction-fixed peripheral nystagmus, documented one-sided reduction of vestibulo-ocular reflex function, and no acute central neurological, otological or audiological signs. The criteria state plainly that there is no definitive test; the diagnosis requires excluding central lesions.
The NHS draws the distinction that matters at the bedside: vestibular neuritis closely resembles labyrinthitis, "but it does not cause hearing loss or tinnitus." It gives recovery of balance as usually 2 to 6 weeks, "although it can take longer."
Ménière's is a spontaneous episodic syndrome. Criteria agreed in 2015 by the Bárány Society with four other international bodies, the AAO-HNS Equilibrium Committee among them, define definite Ménière's as episodic vertigo with low- to medium-frequency sensorineural hearing loss and fluctuating aural symptoms — hearing, tinnitus or fullness — in the affected ear, episodes lasting 20 minutes to 12 hours. Probable Ménière's is broader: 20 minutes to 24 hours, no documented hearing loss required.
Those duration bands are a sorting heuristic, not a diagnosis; the guideline's differential table runs to more than twenty other conditions, vestibular migraine and posterior circulation stroke among them.
The AAO-HNS guideline calls BPPV "the most common vestibular disorder across the life span" and "the most common cause of peripheral vertigo," while noting it "is still often underdiagnosed or misdiagnosed."
The underlying numbers are smaller than that phrasing suggests. A nationally representative German survey of 4,869 people, with validated follow-up interviews, found BPPV accounted for 8% of those reporting moderate or severe dizziness or vertigo. Lifetime prevalence was 2.4%, one-year prevalence 1.6%, one-year incidence 0.6%. BPPV is the most common vestibular cause; most dizziness is not vestibular at all.
In general practice the share is higher. A prospective study of 70 patients presenting to 13 GPs with vertigo found BPPV, vestibular neuronitis and Ménière's together accounted for 93% of symptoms, though the confidence interval ran from 71% to 100%.
Treatability is the real argument. In the German survey, 86% of people with BPPV had a medical consultation, an interruption of daily activities or sick leave because of it. Only 8% received effective treatment. Median episode duration was two weeks.
A canalith repositioning procedure uses a sequence of head and body positions to move displaced otoconia out of the affected semicircular canal. The best known is the Epley manoeuvre. It is performed or taught by a clinician once the affected canal has been identified, and this article deliberately does not describe the steps.
The 2014 Cochrane review included 11 trials and 745 patients. Against a sham manoeuvre or control, complete resolution of vertigo occurred significantly more often with the Epley manoeuvre: odds ratio 4.42 (95% CI 2.62 to 7.44), five studies, 273 participants, with the proportion resolving rising from 21% to 56%. Conversion of a positive Dix–Hallpike test to negative also favoured treatment, OR 9.62 (95% CI 6.0 to 15.42), eight studies, 507 participants.
Three findings in the same review get quoted less often. There was no difference between Epley and Semont (two studies, 117 participants) or Epley and Gans (one study, 58 participants). The claim that a single Epley beats a week of thrice-daily Brandt–Daroff exercises rests on one trial of 81 participants. And the authors recorded "a high recurrence rate of BPPV after treatment (36%)." The AAO-HNS puts recurrence at 5% to 13.5% by six months, 10% to 18% at one year.
On grades the two bodies agree. The AAN's 2008 practice parameter states that the canalith repositioning procedure "is established as an effective and safe therapy that should be offered to patients of all ages with posterior semicircular canal BPPV (Level A recommendation)." The Semont manoeuvre received only Level C. The AAO-HNS makes repositioning a strong recommendation at Grade A, number needed to treat 1 to 3; the AAN's figure across two Class I and three Class II studies was 1.43 to 3.7, for resolution within four weeks.
Here is the part that constrains what you do with this. That grade attaches to the procedure in a clinical setting. On doing it yourself, the same document concludes: "There is insufficient evidence to recommend or refute self-treatment using Semont maneuver or CRP for BPPV (Level U)." Level U means no recommendation either way. A Grade A for an intervention is not a Grade A for every way of delivering it, and an undiagnosed vertigo is the wrong thing to practise on.
Two further positions. Postprocedural postural restrictions — sleeping upright, avoiding head movement afterwards — carry a strong recommendation against at Grade A. Observation alone is an option at Grade B, resolution in 15% to 85% at one month without intervention, though the panel "felt strongly in favor of treatment with CRP rather than observation."
Vestibular suppressants for BPPV. The AAO-HNS recommends against routinely treating BPPV with antihistamines or benzodiazepines, at Grade C. The AAN found a single Class III study showing no difference between lorazepam, diazepam and placebo, and concluded there is no evidence to support any medication in the routine treatment of BPPV (Level U). Worth holding alongside that: what benzodiazepines actually do over a longer horizon.
Corticosteroids for vestibular neuritis. Here one positive number travels and the rest do not. The 2011 Cochrane review pooled four small trials, 149 participants, oral corticosteroids versus placebo. Complete caloric recovery at one month favoured steroids: RR 2.81 (95% CI 1.32 to 6.00, P = 0.007). At 12 months, RR 1.58 (95% CI 0.45 to 5.62, P = 0.48) — interval crossing 1, no demonstrated effect. The extent of caloric recovery was non-significant at both timepoints (one month: mean difference 9.60%, 95% CI −20.66 to 39.86, P = 0.53). So was vertigo at 24 hours (RR 0.39, 95% CI 0.04 to 3.57, P = 0.40), and Dizziness Handicap Inventory scores at every timepoint. The authors: "there is currently insufficient evidence from these trials to support the administration of corticosteroids." The single positive result was a surrogate measure at the earliest timepoint.
Betahistine for Ménière's attacks. The BEMED trial was a nine-month, three-arm, double-blind, placebo-controlled trial across 14 German tertiary centres: 74 on placebo, 73 on 24 mg twice daily, 74 on 48 mg three times daily. The primary outcome was attacks per 30 days during months seven to nine. Incidence did not differ between groups (P = 0.759); attack rate ratios against placebo were 1.036 (95% CI 0.942 to 1.140) and 1.012 (95% CI 0.919 to 1.114). "Results were consistent for all secondary outcomes."
The instructive detail is what did change. Monthly attack rates fell significantly, by a factor of 0.758 (95% CI 0.705 to 0.816, P < 0.001) — in every arm, placebo included. Reported as a before-and-after figure in a treated group, that reads as a large benefit. Against placebo it is nothing.
A 2016 Cochrane review of betahistine across mixed vertigo diagnoses did find more patients reporting reduced symptoms than on placebo, RR 1.30 (95% CI 1.05 to 1.60, 606 participants, 11 studies). But it graded the evidence low quality, judged most studies at high risk of bias, and warned the result "should be interpreted with caution as the test for statistical heterogeneity as measured by the I-squared value was high." The AAO-HNS 2020 Ménière's guideline lists diuretics and betahistine as an option for maintenance therapy — its weakest policy level.
Vestibular rehabilitation is better supported, with one exception. A 2015 Cochrane review of 39 studies and 2,441 participants with unilateral peripheral vestibular dysfunction found reduced dizziness frequency versus control, OR 2.67 (95% CI 1.85 to 3.86, four studies, 565 participants). But in BPPV, repositioning beat movement-based rehabilitation on short-term cure rate (OR 0.19, 95% CI 0.07 to 0.49). The AAO-HNS makes rehabilitation for BPPV an option at Grade B, better as an adjunct than a primary treatment.
Sudden vertigo can be a posterior circulation stroke, and the overlap with a peripheral cause is real. The AAO-HNS guideline cites a series of 240 cerebellar strokes in which 10% presented similarly to a peripheral vestibular process, and notes that examination often discloses other posterior circulation findings: dysarthria, dysmetria, dysphagia, sensory or motor loss, or Horner's syndrome.
The HSE advises calling 112 or 999, or going to the nearest emergency department, if someone has vertigo together with any of:
The NHS adds, on its Ménière's page, that you should call 999 if a spinning sensation comes with a face drooping on one side, an inability to keep both arms raised, difficulty speaking or sudden vision problems in one eye, because these can be signs of a stroke — and that you should not drive yourself to A&E. The HSE separately advises an urgent GP appointment for persistent vertigo with severe headache, vomiting, a temperature of 38°C or above, or feeling hot and shivery.
Imaging is not the safety net people assume. In a prospective study of 101 patients with acute vestibular syndrome and at least one stroke risk factor — 76 of whom had central lesions — a three-step bedside oculomotor examination was 100% sensitive and 96% specific for stroke, while initial MRI diffusion-weighted imaging was falsely negative in 12% of cases, all within 48 hours of onset. That was a deliberately high-risk group, so those figures do not transfer to unselected patients. What does transfer: an early normal scan does not exclude a posterior circulation stroke. Conversely, where BPPV criteria are met and nothing inconsistent is present, the AAO-HNS recommends against imaging, at Grade C.
The commercially inconvenient summary: the most common cause of vertigo has a Grade A treatment involving no product, no prescription and usually one appointment, at a number needed to treat between one and three. The other two conditions have widely used treatments whose best trials were negative or inconclusive. And nothing Proco sells — or, as far as we can establish, any supplement — has trial evidence in these conditions. That is the test we apply before stocking anything. If you have positional vertigo, the useful next step is a clinician who can perform a Dix–Hallpike test.
This article is general information, not medical advice. It does not describe how to perform any repositioning manoeuvre, and it is not a substitute for assessment by a clinician. Written 18 October 2026.