Efficacy of different mouthrinse formulations in reducing oral malodour: a randomized clinical trial

Download 41.1 Kb.
Date conversion29.01.2017
Size41.1 Kb.
Efficacy of different mouthrinse formulations in reducing oral malodour: a randomized clinical trial

For figures, tables and references we refer the reader to the original paper.

Halitosis is a general term used to define an unpleasant odour emanating from the breath of a person. When the cause of the bad breath can be found in the oral region the condition is defined as oral malodour (Quirynen et al. 2011). Approximately, 90% of all cases of bad breath are caused by oral conditions, especially tongue coating, gingivitis and/or periodontitis (Quirynen et al. 2009). Non-oral causes of halitosis are less frequent and include disturbances of the upper and lower respiratory tract, disorders of the gastrointestinal tract, some systemic diseases, metabolic disorders and carcinomas (Tangerman 2002, Porter & Scully 2006).

Oral malodour is the consequence of degradation of organic substrates by anaerobic bacteria. From the resulting malodorous component, the volatile sulphur compounds (VSCs) are far the most extensively studied and probably also the most important. In particular, hydrogen sulphide (H2S) and methyl mercaptan (CH3SH) have demonstrated to contribute to oral malodour (Tonzetich 1977, Yaegaki & Sanada 1992, Tangerman 2009, Vandekerckhove et al. 2009). In addition to the VSCs, other compounds like indole, skatole, diamines and short-chain fatty acids may also play a role in certain conditions (Goldberg et al. 1994, Loesche & Kazor 2002, Dadamio et al. 2011).

In an effort to prevent or at least alleviate oral malodour, several treatment strategies have been developed. An improvement in oral hygiene and especially tongue cleaning in case of the presence of a tongue coating are a “condition sine qua non” (Coil et al. 2002, Yaegaki et al. 2002). The later can be combined with the use of a mouthrinse. The effects envisaged with mouthrinses can be grouped under (i) masking the odour (e.g. aromas), (ii) reduction of the bacterial load, (iii) decrease in the production of volatile components (e.g. interference with enzymatic reactions) and/or (iv) decrease in volatility of the sulphur compound (Quirynen et al. 2002b).

A recent Cochrane review concluded that besides the numerous articles that have been published on this topic, randomized controlled trials comparing the effectiveness of available mouthrinses were urgently needed (Fedorowicz et al. 2008). At the time the study started, there were no large scale comparative clinical trials available examining the efficacy of different products/active ingredients in patients with obvious forms of bad breath.

The primary objective of this clinical trial was to benchmark the anti-malodour efficacy of different mouthrinse formulations. With that in mind, the efficacy of five different mouthrinses (including a negative and a positive control) was evaluated in patients with obvious oral malodour. A secondary objective of this clinical study was to assess the beneficial effect of the addition of 0.2% zinc lactate to an existing AmF/SnF2 rinse. The affinity of zinc for sulphur and its ability of oxidize sulfhydryl groups is well known (Pauling 1988, Young et al. 2001). This metal exhibit low toxicity and does not cause dental stain and in a form of organic salt presents less metallic taste (Wåler 1997).

Material and Methods

Study design

This study was designed as a single-centre, double-blind, randomized, parallel group clinical trial (five formulations, one per group, including a positive and negative control). The efficacy of each formulation was evaluated by comparing oral malodour indicators before using the rinses, with the same parameters 15 min. after a first rinsing (masking effect) and 12 h after the latest rinse, after 7 days of use (therapeutic effect) respectively.

The organoleptic rating (OLS) of breath and tongue coating were considered primary outcome variables. The secondary ones included the level of VSC determined by a portable gas chromatograph. Microbiological changes were also recorded. A detailed study flow chart is presented in Table 1.

Table 1. Study flow chart, with examinations during the first visit: baseline and masking effect (15 min. after rinsing with the allocated rinse); and during the second visit: therapeutic effect (12 h after the last use of the mouth rinse)

Investigational events

Visit 1 (day 0)

Visit 2 (day 8)


Masking effect

Therapeutic effect

Medical history


Dental status


Inclusion/exclusion criteria



Organoleptic evaluation




Adverse events



VSC reading




Microbiological samples





This study received full approval from the Ethical Committee of the University Hospital of the Catholic University of Leuven (ML4047). Subjects were well informed of the study protocol and objectives, and gave their written consent before participation. The study was conducted according to the European directives and ICH Harmonised Tripartite Guideline E6: Note for Guidance on Good Clinical Practice, CPMP/ICH/135/95 Step5 (http://www.ema.europa.eu).

Patient population

Patients on a first visit to the bad breath clinic (UZ Leuven) were invited to participate if they fulfilled the following inclusion criteria: age between 18 and 70 years, organoleptic scores (OLS) ≥2, Halimeter® readings of at least 150 p.p.b., an intra-oral cause of bad breath, and non-smoker. Exclusion criteria included: an ongoing dental or medical treatment, allergy to previously used oral hygiene products or any known allergy to any of the ingredients of the study products, pathological change of the oral mucosa, use of prohibited treatments or therapies and/or abuse of drugs or alcohol, pregnancy or breastfeeding, participation in a clinical study within the previous 30 days, active caries, acute sinusitis, severe oro-pharyngeal infection, medication which can cause malodour, reduced salivary flow due to pathological reasons (e.g. Sjögren syndrome), conditions not compatible with the study according to the investigator's opinion (e.g. patients eating very spicy food, persons under homeopathic therapy, patients who used antibiotics during the 2 months before the study, patients frequently using chewing gum, patients under corticosteroids or other serious medications), non-Caucasians, patients unwilling to abstain from additional oral hygiene, particularly mouthrinses, chewing gums, breath strips.

At least one week before their appointment at the clinic the patients received written instructions with the precautions to be taken beforehand including: avoidance of spicy foods, garlic and onions the 2 days prior to consultation and abstaining from alcohol the last 12 h before all measurements. The morning of the readings the use of scented cosmetics was forbidden. Habitual oral hygiene (tooth brushing without tooth paste) and breakfast were allowed.

All readings were taken the earliest 1 h and the latest 3 h after breakfast and oral hygiene. When the subject volunteered for the study and before they received a bottle containing the mouthrinse and the instructions of use, microbiological samples were collected. The volunteers were asked to rinse at home 2 times per day (morning and evening) with 15 ml (special dispenser cups) for 1 min. (timed with stopwatch) during the next 7 days. In the meantime, they were instructed to continue their habitual oral hygiene procedures, but were no longer allowed to use any other mouthrinse. The volunteers were also asked to rinse for the first time at the consultation after the baseline readings had been taken. Fifteen min after, the odour judge repeated the organoleptic rating and the VSC level of a new sample of mouth air was recorded to evaluate the masking effect. At day 8, a control visit was scheduled for which the subjects had to follow the same guidelines as at the first visit regarding food and drinks consumption, cosmetic use and oral hygiene procedures. That morning, the subjects were not allowed to rinse, so that the measurements could be taken 12 h after their last rinse. At that moment, the therapeutic effect was scored (Table 1). Dropouts and withdraws from the study were replaced by new volunteers.

Test formulations and randomization

Five groups of 18 patients each were formed, via a blind allocation of the rinses (see below). Each group used one of the following formulations: a fluoride rinse [considered negative control (NC)], a 0.12% CHX-based rinse [considered positive control (PC)], HalitaTM (H), meridol® (M), and a meridol® formulation with the addition of zinc lactate (M + Zn). A detailed composition of the rinses is given in Table 2. The formulations were provided by the sponsor in undistinguishable bottles.

Table 2. Composition of rinses compared in the study

The randomization schedule was generated for a parallel group trial design with 23 blocks of 5 subjects using software employing a pseudo-random number generator according to Algorithm AS 183 (Wichmann & Hill 1982). The mouthrinse bottles were consecutively numbered with subject numbers 1 to 115. The random allocation sequence was generated by the GMP manager of the sponsor. Patients were enrolled and assigned to random treatment by the clinical investigator who assigned consecutive subject numbers up to 90. The extra subject numbers 91 to 115 served for potential replacements of dropouts. Patients, odour judge and investigator were blinded regarding the product allocation. In case of dropouts, the corresponding subject number was communicated to the GMP manager who provided a new subject number to replace the dropout by a new patient receiving the same treatment as the one the dropout had. Knowledge of the randomization list was limited to the one person, responsible for creation of the randomization list, for the preparation of the random code envelopes, and the preparation of the study products until treatment of the last subject (i.e. final examination of the last subject), quality control and verification of the CRFs was completed. Sealed emergency envelopes with product allocation were kept at the study site to be used in case of severe adverse event (SAE).

Malodour indicators

At each study point: day 1, baseline and 15 min. after first rinse; day 8, 12 h after last rinse the following parameters were recorded:

Organoleptic assessment (OLS)

The organoleptic score was determined by a trained and calibrated judge who tested its ability to distinguish odours using the Smell Identification Test® (Sensonics Inc., Haddon Heights, NJ, USA) and to detect odours at low concentrations (Doty et al. 1984). The judge scored an average of 200 cases every year and was initially calibrated towards other experienced breath examiners. Breath was scored as described by Rosenberg at rest (open mouth without breathing) and when the patients counted from 1 to 11 (Rosenberg, 1996). A 0–5 score was given (Greenman et al. 2004). At baseline the judge also smelled the nasal breath to exclude extra-oral causes. For the evaluation of the tongue coating smell, the same scores were used. The organoleptic score preceded all other measurements to avoid any bias.

VSC measurements

Besides the initial screening with the Halimeter®, a portable gas chromatograph (OralChroma™, Abilit Corporation, Kanagawa, Japan) was used to measure the concentration of H2S, CH3SH and (CH3)2S separately. Sample collection and analysis were performed according to the manufacturer instructions.

Microbial samples and processing

Saliva and tongue coating samples were taken during both visits for the analysis of the microbiota. Tongue coating was collected by wiping a sterile swab over the dorsum of the tongue, in the area of the foramen caecum. The tip of the cotton swab was placed in a vial containing 2 ml of reduced transport fluid (RTF). An independent sample of non-stimulated whole saliva was collected into a sterile container. All samples were kept at 4°C. Each sample was processed within 6 h. The samples were cultured under anaerobic conditions on non-selective agar plates to quantify the number of black-pigmented colony forming units (BpB in CFU/ml). Details concerning the growth conditions, colony selection and final identification have been summarized previously (Quirynen et al. 1999).

Patient compliance & adverse events

At the beginning of the study, each participant received a bottle of the assigned product. The bottle weight was assessed at visit 1, with the bottle still closed, and at visit 2. The difference between both visits was considered the amount of product used during the study. For safety, evaluation self-reported adverse events were documented during the second visit.


A linear mixed model was applied with time of measurement and mouthrinse as fixed factors and patient as random factor. Contrasts between treatment effects were set up to assess the following null hypotheses: (i) none of the treatments demonstrated any masking or therapeutic effect; (ii) there was no difference in masking or therapeutic effect among the three rinses under evaluation; and (iii) there was no difference in masking or therapeutic effect between the latter and the negative or positive controls. A normal quantile plot and residual dot plot were used as a test of the basic assumptions of the model. A correction for simultaneous hypothesis testing was made for each outcome parameter according to Sidak.

The measurements of H2S, BpB and the ratios between anaerobic and aerobic CFU counts were log-transformed before analysis. Specific considerations were made for BpB measurements, zero values were considered as ‘<20000’ in saliva and ‘<2000’ in tongue and these measurements were treated as left-censored values; thus a frailty model was applied with time and mouthrinse as fixed, and patient as random factor.


Between April 2007 and July 2011, ninety-eight (98) healthy Caucasian volunteers with obvious oral malodour were recruited (Fig. 1). There were four dropouts and four other volunteers were withdrawn from the study by the principle investigator after notification of antibiotic intake during the study period or participation in parallel in another trial. All of them were replaced with new volunteers to maintain the goal of 18 subjects per group.

Flow diagram of patient enrolment.

The final 90 participants included 56 males and 34 females with a mean age of 48.2 years. All of them were non-smokers. No one presented active caries or had taken antibiotics the month before the study; neither undergone any dental procedure. Chronic periodontitis or gingivitis was observed in 58 (65%) patients, with no significant differences between subgroups. Demographic and baseline characteristics of each subgroup are presented in Table 3. No statistically significant differences were found (all p-values > 0.05) between groups for any of the baseline parameters. Patients, odour judge and investigator were blinded regarding the product allocation. The copies of the randomization lists were returned to the sponsor unopened.

Table 3. Baseline characteristics of patient population

At the second visit, participants were asked to bring the bottle with the mouthrinse back to assess the compliance to the rinsing schedule. Two people claimed to have used the entire product and returned an empty bottle. Six volunteers did not bring the bottle back. The average of product used was 212.3 ml (SD 47.3 ml), with an average used per rinse of 14.2 ml. When considering each group separately, some differences could be observed (lowest amount for PC, highest amount for M + Zn); however, they did not reach statistical significance (Table 3).

Four adverse events were reported during the study; three regarding unpleasant feelings after the use of the product (one each for NC, PC and H) and one other involving tooth staining (H). There were no severe adverse events.

Masking effect

The masking effect was evaluated by comparing oral malodour indicators at baseline and after 15 min. of rinsing with 15 ml of the assigned product during the first visit. All rinses showed a clear masking effect as illustrated by a statistically significant decrease in the OLS of the breath (0.8 up to 1.9 score OLS reduction) and the tongue coating (0.8 up to 1.5 score OLS reduction), as well as by a reduction of the level of H2S (129 to 346 p.p.b. reduction) (Table 4a).

Table 4a. Masking and therapeutic effect (comparison with baseline) of the different formulations in organoleptic ratings (OLS) and H2S recordings (mean and SD)

The reductions in the OLS of breath caused by H, M and M + Zn were statistically superior to those obtained by the positive and negative control (Fig. 2a). M + Zn was the only rinse superior to the NC, in reducing the OLS of the tongue coating. H and M + Zn caused significantly more reduction in H2S when compared to the NC; and M + Zn even also when compared to the PC (Table 4b).

Table 4b. Statistical analyses for masking and therapeutic effect of the different formulations compared with baseline measurements (p-values) as well as an inter-group comparison (towards negative and positive control)

(a) Organoleptic scores of breath before and 15 min. after a single use of one of the 5 formulations indicated via box- and whisker plots (the rectangles extend from the 25th to the 75th percentile and the whiskers extend to the most distant value that is not considered an outlier). (b) Organoleptic scores of breath at baseline, and 12 h after the last rinsing with one of the 5 formulations (rinsing for 7 days) indicated via box- and whisker plots (the rectangles extend from the 25th to the 75th percentile and the whiskers extend to the most distant value that is not considered as an outlier).

Therapeutic effect

This effect was assessed by evaluating the change in the oral malodour indicators 12 h after the latest rinse, after 7 days of use. Only the rinses with antimicrobial ingredients showed a significant decrease in the OLS of breath (Fig. 2b), tongue and H2S level (Table 4a), but no differences were observed between these rinses and neither towards the PC (Table 4b). The amount of tongue coating did not show significant differences except for the PC (p < 0.05).

No significant differences in the ratio anaerobes/aerobes in saliva or the tongue coating were observed for any of the formulations. The PC, H and M + Zn reduced the amount of black-pigmented bacteria in saliva significantly over time. The changes in the microbial composition on the tongue were negligible (Table 4c).

Table 4c. Effect of the rinses on the microbiota of saliva and tongue (log-transformed values)


Because their great social acceptance, the use of mouthwashes has become a common oral hygiene practice. Besides the increasing number of formulations in the market, the scientific evidence supporting their efficacy in the treatment of oral malodour is not conclusive (Fedorowicz et al. 2008).

In this study, the efficacy of five different formulations was compared in patients with obvious oral malodour. The formulations included four commercially available products (FluorigardTM, Perio-plusTM, HalitaTM and meridol®) and one “experimental rinse” (meridol® supplemented with 0.2% zinc lactate). This experimental solution served as starting point for a commercial version currently in the market. CHX is a benchmark product that has demonstrated the ability to reduce plaque at a 0.12% (Berchier et al. 2010) and has been proposed for the treatment of halitosis because of its antibacterial properties (van Steenberghe et al. 2001). Ideally, a placebo should be as similar as possible to the test product except for the active ingredients. As this study included formulations based on two different active ingredients (CHX and AmF/SnF2), this was not possible. We have decided not to use water as negative control to improve the adherence to the treatment and not to compromise the blindness of the investigators. A fluoride rinse with alcohol was chosen because this anti-cavity rinse does not contain any active components and therefore it is not expected to have any anti-malodour activity. Oral hygiene measures were not standardized neither before nor during the study period to contemplate the broad range of conditions in which the rinses are used in real life. This decision may have introduced some variability in our results without, however, increasing the risk of bias. Moreover, the simplification of the protocol probably contributed to the adherence of the patients. Since only individuals fitting strict inclusion and exclusion criteria were included and no monetary compensation was offer to the participants, a 4-year period was necessary to achieve the desired number of volunteers. This might also underline the unpleasant feeling patients have when suffering from oral malodour, and/or from being examined by others. Fortunately, few dropouts (4/98) were encountered.

The average volume of rinse used (14.2 ml) demonstrated a very good compliance to the schedule. Whether the difference in volume used (15.3 ml for the M + Zn versus 12.7 for the PC) has something to do with the unpleasant feeling chlorhexidine can give (Watts & Addy 2001) is difficult to confirm. The number of AEs was too low to allow any statistical analysis. Three patients (all rinsing with different formulations) complained from unpleasantness and only in one patient (rinsing with HalitaTM) tooth staining was observed.

All the rinses caused an immediate decrease in the organoleptic scores of breath and tongue, and the level of H2S. This reduction can be partially attributed to a further dilution of volatile compounds in the mouth (Tonzetich 1978, Suarez et al. 2000), and/or to a masking of the odoriferous molecules by the aromas in the formulations. All rinses included in this trial, even the negative control, included some aromas. Aroma's present in toothpaste and chewing gums are already known to give a temporary improvement of the breath by simply overruling the odorous components (Reingewirtz et al. 1999, Peruzzo et al. 2008). The same seems applicable to our negative control, without obvious antimicrobial compounds.

The presence of metal ions can explain the differences observed in the masking of the odour between the control rinses and the other formulations. Zinc and stannous ions, like other metals, are able to capture VSC (Pauling 1988, Young et al. 2001, 2003, Rölla et al. 2002). This was first clearly illustrated in oral malodour with a small scale study by Hoshi & van Steenberghe (1996) showing a four hours reduction in the VSC levels when using zinc containing tooth paste. Later, other studies confirmed the short and long-term effect of zinc ions on reducing the level of VSC and OLS in subjects with morning bad breath (Navada et al. 2008, Newby et al. 2008, Young & Jonski 2011). For mouthrinses the same applies. The formulations with these ions (HalitaTM, meridol®, and meridol® with zinc lactate added) scored in this trial clearly superior in their masking capacity. Even though zinc ions have a lower affinity for sulphur compounds than stannous ions, their capacity to inhibit the VSC is bigger (Wåler 1997). The latter might explain why HalitaTM and meridol® with zinc lactate added were best in decreasing the level of H2S. The added value of zinc to formulations with CHX has been demonstrated before (van Steenberghe et al. 2001, Roldán et al. 2004). The results of this study are in agreement with two recent papers evaluating the effect of a commercial formulation (meridol® HALITOSIS: AmF/SnF2, 0.2% of zinc lactate and oral malodour counter-actives). This new formulation significantly reduced, on short and long term, oral malodour in volunteers with morning bad breath (Wigger-Alberti et al. 2009, Wilhelm et al. 2010).

meridol® with zinc lactate added was the only rinse with masking effect significantly different from the positive control. The presence of both stannous and zinc ions could explain this observation. A dose–response inhibitory effect has indeed been reported previously (Kleinberg & Codipilly 2002).

Only the rinses with antimicrobial ingredients in sufficient concentration showed a long-term reduction of organoleptic scores of breath and tongue, and the level of H2S. Two possible mechanisms can be considered: (i) a reduction in the number of bacteria able to produce the malodour compounds and (ii) the neutralization of the sulphur compounds by the zinc and stannous ions. Several studies have reported the effectiveness of CHX and AmF/SnF2 on plaque control based on their anti-microbial properties (Brecx et al. 1993, Auschill et al. 2005, Øgaard et al. 2006). However, their anti-microbial effect in patients with halitosis is less well documented. Quirynen et al. (2002a) reported reductions (0.5 log) in bacterial load (aerobic and anaerobic) in saliva but not in tongue coating as results of the use of HalitaTM and meridol® in a morning bad breath model. Roldán et al. (2003), working with patients with moderate to severe malodour without periodontal involvement, were also not able to prove that the difference that they found in anaerobic counts in saliva or tongue after 2 weeks of treatment differed from the changes observed for the placebo. In this study, including patients with obvious bad breath, with and without periodontitis, changes in the ratio anaerobic/aerobic could not be detected, neither in the saliva, nor in the tongue coating. The only significant microbial change was a reduction in black-pigmented species in the saliva for patients rinsing with the positive control, Halita TM or meridol® with zinc lactate. On the basis of our previous experience, we opted to use culturing techniques to assess the microbiological effect of the tested rinses. Black-pigmented bacteria are known for their ability of producing large amounts of VSC and have been previously related with oral malodour (Nakano et al. 2002). The reason why major changes were not detected remains unclear, but might be because of the severity of the tongue coating, the lack of mechanical cleaning, the selection of patients and the impossibility of these antiseptics to penetrate thick biofilms. Moreover, no change in the amount of coating could be established.

A long-term beneficial effect of the zinc by itself could not be discovered. The therapeutic effect of the two meridol® rinses was similar except for the amount of BpB, although the difference was minimal. A long-term reduction of OLS of breath and tongue below 1 and 2, respectively, appears to be impossible without mechanical intervention. Even though, in the context of this study, this is only an hypothesis it is interesting to note that only one third of the population reported interdental cleaning as routine procedure (data not shown), and the use of a tongue scraper was forbidden during the study period.

In conclusion, the results of the present clinical trial support the effectiveness of mouthrinses containing CHX and AmF/SnF2 in the management of halitosis. The addition of zinc ions to the AmF/SnF2 formulation clearly improved the short term effect of the rinse on oral malodour.

The database is protected by copyright ©dentisty.org 2016
send message

    Main page