Patients. The study included 19 adult subjects extracted from a pool of 186 unilateral definite MD patients recruited at the outpatient clinic of the Vestibular Disorders Unit in a tertiary referral university Hospital during the last 3 years. They were12 females and 7 males (mean age: 57.0 ± 10.8 years). They fulfilled all criteria for definite MD according to the AAO-HNS guidelines  and were negative for retrocochlear lesions at magnetic resonance imaging.
MD patients were included if:
1) altered intestinal permeability was previously confirmed by the validated “double sugar test” and faecal calprotectin > 50 μg/g;
2) they were "symptomatic” from a vestibular point of view, i.e. they suffered at least two major episodes of vertigo per month with sensorineural hearing loss and aural fullness during the 3 months prior to the admission, and
3) had a functional level (FL) ≥ 4, according to the consensus paper guidelines. 
Exclusion criteria were: 1) a positive history for gastrointestinal disease, hypertension, celiac or bowel disease; 2) abnormal thyroid hormones; 3) history of malignant tumours or autoimmune diseases; 4) specific diets (including low-salt [Na/K] intake)
Those patients who had previously received systemic steroids or intratympanic injections (either gentamicin or steroids) were also excluded from the study, as well as those being treated with diuretics, protonic pump inhibitors and/or antihistamines.
Controls. Fourteen healthy volunteers (10 females, 4 males, mean age: 42.0 ± 17.0 years) served as a “control group”, in order to check the reference values. All healthy subjects were normally hearing, reported a negative familiar and personal history of vertigo or dizziness, and never suffered from otological diseases. In the inclusion criteria, they should have never suffered from gastrointestinal diseases, they should have a negative familiar history for intestinal bowel disease or celiac disease. They had a negative “double sugar test” and faecal calprotectin was <50 μg/g.
Test. The standardized procedure described as the “double sugar test”  was modified by testing only mannitol. When awakening in the morning, each patient was instructed to collect and evaluate the volume of a pre-test urine sample. Then, they were asked to drink a solution containing 1g mannitol dissolved in 200 ml water. Urines were collected during the next 4 h and the total volume was measured. Patients were instructed to avoid eating (not even a chewing-gum), drinking or smoking during the test, but were allowed to drink a fixed dose of water (200 ml), only half an hour after the test start.
Statistical analysis. Statistical analysis was performed using the SPSS statistical package version 24.00 (SPSS Inc., Chicago, Illinois). The significance of difference between the two groups compared with each other was evaluated by U-Mann Whitney test for independent samples. Results are expressed as means and standard deviations (SD). A p < 0.05 was considered statistically significant.