Halitosis associated with tongue coating results from the production of volatile sulfur compounds by bacterial action. The use of prebiotics and probiotics might be useful in treating such condition. The aim of this study was to evaluate the effect of the prebiotic inulin combined with the probiotic Streptococcus salivarius (SS) on halitosis by coating.

Participants In this randomized, controlled, double-blind, parallel, phase II clinical trial, adult patients (≥18 years old) who complained of halitosis in the presence of tongue coating were invited to participate. They were identified with invitations in the form of posters distributed at the Dental School clinics, University of Passo Fundo (UPF), from August to November, 2015. The Dental School is located on the main campus of UPF,Passo Fundo, a city with >200.000 habitants. Oral halitosis was initially diagnosed by oral examination for tongue coating identification, followed by organoleptic test to confirm the halitosis. Total sample size was estimated in 45 participants, considering a difference of 80% in Halimeter reduction between treatments, for tests with α = 5% and power = 80%. Halitosis was defined by Halimeter values ≥ 75 ppb. All participants gave written consent before entering the study. The research protocol was approved by the local ethics committee (CAAE: 40334914.0.0000.5342).

Clinical examination Oral examination was performed by a trained dentist (CRM), identifying dental cavities, plaque, gingivitis, periodontal pockets and any dental prosthesis. Coating was described after visual inspection of the tongue and coating index was classified into 4 levels: 0 - no coating; 1 - thin coating on 1/3; 2 - thin coating on more than 1/3 or thick coating on only 1/3; 3 - thick coating on more than 1/3.5

Evaluation of halitosis After clinical examination, halitosis was quantified by organoleptic and Halimeter‡ tests. These were performed by a trained and calibrated examiner (CRM), with levels of agreement and reproducibility greater than 80% for organoleptic test and coating index, using test and retest comparisons with an expert judge (CKR). All patients were examined between 8 and 11 am. Patients were asked to avoid eating garlic, onion and other food condiments for two days before testing and avoid overeating margarine, milk, fry, sardines, salami, bologna, sausage, red meat, cheeses, foods with sulfur in the composition (cabbage, broccoli, cauliflower, egg), drinking alcohol and smoking, following protocols reported elsewhere.5, 13, 15 In the examination morning, patients should avoid using candies and gums, but were allowed to have breakfast and tooth brushing with water only. Once included in the study, patients were asked to avoid gums other than those provided in the trial.

Organoleptic test Patients at rest were instructed to hold air in the mouth while breathing through the nose for 1 min. They blow through mouth with the examiner (CRM) positioned 10 cm from the patient mouth. Patient breath was rated from 0 to 5, as follows: 0 - no smell, 1 - poor barely perceptible odour, 2 - slight stench, 3 - moderate smelly, 4 - strong stench, and 5 - severe stench.21 With the purpose of selecting patients, halitosis was considered present when this score was equal or higher than 1 in the presence of tongue coating.

Halimeter Patients had their breath objectively measured using a Halimeter, capable of measuring the VSC level. Halitosis is identified when the device registers ≥75 ppb.1, 22 The measurements were made by inserting a disposable straw into the patient mouth above the tongue. According to the manufacturer, the final result was obtained as the average of three measurements.

Prebiotic, probiotic and placebo Lactobacillus salivarius G60, inulin and placebo§ (composed of gomagron) were prepared in the form of gum‖, identical in physical appearance, odour, taste and consistency making the following experimental groups: LS+inulin: Lactobacillus salivarius G60 1 billion colony forming units (CFU) + inulin 1 g; LS: Lactobacillus salivarius G60 1 billion CFU; and placebo. Patients were instructed to use 1 gum every 12 h (after breakfast and after dinner) for 10 days.

Quality of life assessment For oral related quality of life (QOL), we applied the questionnaire OHIP-14, translated and validated to Brazilian Portuguese.23 The final score range from 0 (best) to 56 (worst). Participants responded the questionnaires on days 0 and 14.

Study protocol Patients were evaluated in a dental clinic, following the steps illustrated in Figure 2. After signing the informed consent, clinical and quality of life data were recorded, followed by oral examination to assess tongue coating, and organoleptic test to confirm halitosis. Afterwards, the Halimeter was used to objectively measure halitosis.

Enrolled patients were randomized by the examiner (CRM) using an envelope code that was computer generated, organized by the gums manufacturer in blocks for 3 patients 1:1:1, each containing one of the 3 treatments (LS+inulin, LS and placebo). Therefore, examiner and participants were blinded to treatments. Patients were instructed to use 1 gum every 12 hours starting on the fourth day after the interview, and interrupting 2 weeks after the first interview (10 days treatment). This delay of 4 days before starting the treatment allowed an interval of 14 days for attenuation of recall bias with OHIP-14. Each gum was placed in the mouth to complete dissolution, allowing gentle mastication. The final examination was performed in the last day of treatment, 12 hours after using the last gum, repeating coating index, organoleptic and Halimeter tests, and quality of life assessment (OHIP-14). Instructions about oral hygiene were provided by the examiner (CRM): patients were asked to avoid any new approach that could alter halitosis, such as tongue cleaning or changes in routine tooth brushing. Treatment adherence was assessed as the number of gums not used in relation to the total number of gums received. Safety was characterized according to side effects described by the patients, including headache, oral complaints and abdominal symptoms (yes/not). Primary outcomes were organoleptic test, Halimeter and coating index, whereas secondary outcomes were quality of life and treatment adherence/safety. Allocation into the three treatment groups was revealed only after termination of data analysis.

Metabolism of inulin by Lactobacillus salivarius G60 The capacity of Lactobacillus salivarius G60 to metabolize inulin as a carbon source was tested by inoculating 50 ml of bacterial suspension LS (1x10⁸-⁹ cells/ml) in buffered physiological saline solution in 5 ml of environment half-defined by the limit of detection of the method (LDM) without glucose and sucrose but supplemented with inulin. The following culture mediums were prepared: PYG (peptone yeast + 1% glucose), PYI (peptone yeast + 1% inulin) and PY (peptone yeast). The latter served as negative control. The LS G60 previously cultivated was included in these three solutions respectively, and incubated at 30°C for 24 hours.

Spectrophotometry (Genesys 10S VIS, made Madison, USA) analyses using 600 mm for optical density (OD) were performed. To measure absorbance, the solution with LS G60 was adjusted to OD=1, according to the formula C1.V1=C2.V2 (C1: initial concentration, V1: initial volume, C2: final concentration and V2: final volume). The measurements were performed at the initial time (zero), and after 2, 4, 6, 8, 10 and 12 hours. All tests were repeated for confirmation.

Inhibition of pathogenic bacteria by Lactobacillus salivarius G60 Well diffusion test was performed. Three drops of 10 ml of bacterial suspension (1x10⁸-⁹ cells/ml of LS G60) in buffered saline were placed in soy triptic agar and incubated at 30°C for 3 hours under optimal anaerobic conditions. The plates were covered with 10 ml of soft soy triptic agar (0.7% agar) and inoculated with 1 ml of bacterial suspension (1x10⁸-⁹ cells/ml) in buffered saline of halitosis-causing pathogens (Porphyromonas gingivalis and Prevotella intermedia) under anaerobic conditions. They were incubated at 30°C and zones of inhibition halo of microbial growth around the droplets were examined at 24, 48 and 72 hours.

Statistical analysis Data was statistically analyzed by one of the authors (SMCJ) blinded to the interventions. Initial comparisons among groups of patients were done using ANOVA (age) and chi-square (gender) tests. Treatment values on days 0 and 14 were first compared in each group by Wilcoxon or t tests for paired data. We employed generalized linear models (GZLM) to compare treatment means adjusting for pre-intervention values. The inclusion of covariable gender or age was decided based on Akaike´s index. Gender entered as covariable in the analysis of Halimeter, age was included in the models for OHIP-14, whereas no covariable was used when analyzing organoleptic score. We used Gama distribution and log link for the analyses of Halimeter and OHIP-14, as well as Poisson distribution and log link for organoleptic scores and coating index. The constant 1 was added to OHIP-14 values to make them amenable to the use of logarithms. Pairwise multiple comparisons among groups were made using Least Significant Difference. Results from metabolization of inulin by Lactobacillus salivarius G60 were analyzed using ANOVA and Tukey post hoc test. The statistical analyses were performed using Graph Prism 4.0 and SPSS v.18 softwares. The limit considered for statistical significance was 5%.

Inclusion Criteria:

• adult patients with halitosis by tongue coating;
• participants must accept to participate in the study;
• tongue coating identified by oral examination;
• halitosis must be confirmed by the organoleptic test.

Exclusion Criteria:

• halitosis for other conditions, including periodontal diseases and non-oral conditions;
• use of antibiotics in the last 30 days;
• active smoking (>10 cigarettes/day);
• alcohol consumption (>2 drinks/day);
• report of pregnancy or breastfeeding;
• report of systemic diseases, including diabetes, kidney failure and hepatic cirrhosis.

10 days

Headache in one patient during 10 days use of Inulin + LS Tongue discomfort in two patients during 10 days use of LS Mild diarrhea in one patient during 10 days use of placebo