You are on page 1of 8

Research Article

Received: 18 March 2010; Revised: 8 June 2010; Accepted: 30 June 2010; Published online in Wiley Online Library

(wileyonlinelibrary.com) DOI 10.1002/ffj.2018

Effects of Flavoured Mouth Rinses on Morning


Breath Odour: a Sensory, Analytical and
Microbial Evaluation
Myriam Troccaz,a* Olivier P. Haefliger,a Isabelle Cayeux,a Sabine Beccucci,a
Nicolas Jeckelmann,a Jérôme Barra,a Anthony J. Clark,a Jacques Schrenzelb
and Pierre Baehnic
ABSTRACT: Transient oral malodour such as morning bad breath is very common and is caused by accumulation of bacteria
and food residues, mainly at the posterior part of the tongue. The use of antimicrobial flavours may help to reduce malodours
with limited side effects. The purpose of this work was to compare the effects of five alcohol-free mouth rinses containing 0.2%
w/w antimicrobial flavour oil on 24 human subjects in a randomized, double-blind clinical study. Two mouth rinses containing
triclosan and peppermint oil were used as positive controls. Oral malodours and flavour intensities, as well as analytical and
microbial analyses of saliva samples, were performed before rinsing and 1 min and 1 h after rinsing the mouth with a 20 ml
mouth rinse for 1 min. Immediately after rinsing, oral malodour scores and salivary bacterial counts were markedly reduced
(85 ⫾ 9% and 54 ⫾ 12% decreases, respectively). At 1 h after rinsing, maximum efficacy (61 ⫾ 18% reduction in oral malodours,
40 ⫾ 11% decrease in salivary bacterial count) was observed with the mouth rinses containing the highest levels of menthol,
limonene, carvone and eugenol, and these ingredients could still be detected in the saliva of most subjects. The finding that
selected antimicrobial flavours have beneficial effects against salivary microorganisms and transient oral malodours may be
important for the design of oral care products that promote fresh breath up to 1 h after rinsing without compromising taste.
Copyright © 2010 John Wiley & Sons, Ltd.

Keywords: oral malodour; sensory analysis; mouth rinse; antimicrobial; flavour

Introduction oral cavity-associated bacteria such as Streptococcus mutans and


various other oral Gram-positive bacteria are more commonly
Transient oral malodour results from the hydrolysis of peptides found in individuals with low or no levels of malodours.
and proteins present in saliva, gingival crevicular fluid and oral Within this context, different approaches have proven to be
soft tissues by bacteria present in the oral cavity and more par- efficient at controlling and preventing oral malodour.[23] The
ticularly on the surface of the tongue.[1–3] We may distinguish biofilm responsible for bad breath may be removed by careful
morning bad breath, which is very common and may be attrib- brushing or scraping (mechanical effect) and possibly by the
uted to reduced salivary flow during sleep and increased anaero- adjunctive use of oral care products containing antimicrobials
bic bacterial counts, from other temporary halitosis resulting and rehydrating agents.[5,6,24–27] Today, salty water mouth rinses
from poor oral hygiene, tobacco use, spicy foods or alcohol, have been replaced by efficient formulations containing antimi-
which may have a drying effect in the mouth.[1–4] Morning breath crobial components such as chlorhexidine, cetylpyridinium
is often used as a model to test the clinical efficacy of different chloride, triclosan (5-chloro-2-(2,4-dichlorophenoxy)phenol),
treatment modalities, as subjects with pathological halitosis essential oils or zinc salts, either alone or in combination.[25,28] The
associated with infection of oral tissues are difficult to recruit and
standardize.[4–7]
Odourous metabolic products include volatile sulfur com-
pounds (VSCs), such as hydrogen sulfide, methanethiol and dim-
ethyl disulfide,[8–12] as well as other volatiles such as cresol, indole, * Correspondence to: Myriam Troccaz, Corporate R&D Division, P.O. Box 239,
skatole, diamines and short chain fatty acids that may act as CH-1211 Geneva 8, Switzerland. E-mail: myriam.troccaz@firmenich.com
odour modifiers.[13–19] Oral Gram-negative anaerobic bacteria a
Firmenich SA, Corporate RandD, Route des Jeunes 1, P.O. Box 239, CH-1211
such as Porphyromonas gingivalis, Prevotella intermedia, Tre- Geneva 8, Switzerland
ponema denticola, Fusobacterium nucleatum, Tannerella forsythia,
b
Bacteroides spp. and Fusobacterium spp., when present in high Geneva University Hospital, Laboratoire Central de Bactériologie (LCB),
number in periodontal pockets, interdental spaces, and dental Service des Maladies Infectieuses, Rue Micheli-du-Crest 24, CH-1211 Geneva
14, Switzerland
and tongue biofilms, may produce those malodorous com-
pounds.[2,20,21] Yet uncultivable bacteria may also play a role and c
School of Dental Medicine, University of Geneva, Rue Barthélemy-Mann 19,
contribute to oral putrefaction and malodours.[11,16,22] In contrast, CH-1205 Geneva, Switzerland

Flavour Fragr. J. 2010 Copyright © 2010 John Wiley & Sons, Ltd.
M. Troccaz et al.

addition of oral flavours provides taste benefits and refreshes the flavouring compositions were created with those materials to maximize
breath mainly by sensorial masking.[6,29] Flavour materials include, the content of antimicrobial actives (MIC values under 0.25% w/w).[31–33]
however, a wide range of compounds that may have significant Those compositions were screened in vitro to confirm their antimicrobial
antimicrobial effects and/or may inhibit the production of VSCs activity with an automated bacterial contact time test.[34] This method is
based on standardized procedures CEN/TC 216 N59 and the EN-1276
by bacteria.[29,30]
norm. Briefly, the method consists of calculating the log reduction in
The present study was designed to evaluate the efficacy of viability of oral bacteria after a contact time of 80 s between flavours and
flavours to control morning bad breath by either masking the bacterial solution (108 cfu/ml). Three flavouring compositions with the
malodorous gases or by killing bacteria responsible for the pro- highest antimicrobial activity were selected for the clinical trial. The com-
duction of those volatiles.[10,29] For this purpose, a collection of position of the three flavours (flavours 1, 2, 3) is described in Table 1. They
several hundred flavour ingredients were first screened in vitro for contain well-known antimicrobial compounds such as eucalyptol,
their antimicrobial properties. Three flavours with the highest carvone and aromatic alcohols (menthol, methyl salicylate), as well as
antimicrobial activity were then included in mouth rinse formu- other ingredients such as limonene and anethole to mask harsh or
lations, at a 0.2% w/w concentration, and selected for a clinical medicine-like tastes and to increase flavour pleasantness.[35,36] Their anti-
trial. Oral malodours and flavour intensities were performed bacterial activity was shown to achieve up to a log 3 reduction (⫾ 0.1)
against F. nucleatum, P. gingivalis and P. melaninogenica at a concentration
before rinsing and 1 min and 1 h after rinsing. Results were com-
of 0.01% and up to 3.5 log (⫾ 0.1) against S. mutans at a concentration of
pared with two controls containing the widely used active ingre- 0.1%.
dients triclosan or peppermint oil. Saliva concentrations of
individual flavour compounds were also determined up to 1 h
after the rinse to evaluate flavour substantivity. This information Antimicrobial Mouth Rinse Composition
is particularly relevant when creating pleasant and active flavours
Five mouth rinses (M1, M2, M3, M4, M5) were tested for malodour coun-
for the formulation of oral care products that promote fresh
teraction and salivary bacterial reduction in the in vivo clinical study. The
breath with long-lasting properties without compromising taste. three antimicrobial flavours (flavours 1, 2, 3) were incorporated at 0.2%
w/w in a non-alcoholic mouth rinse base (M1, M2, M3) formulated with
demineralized water (82.88% w/w), sodium monofluorophosphate
Experimental (0.05%), sodium saccharin (0.03%), sorbitol 70% (8.02%), glycerin (8.02%)
and Cremophor® CO40 (1.00%) 24 h before clinical testing. No preserva-
tives or antimicrobial substances other than the flavours were added to
Flavour Creation
the mouth rinses. The effects of products containing antimicrobial fla-
Five hundred flavour ingredients from the Firmenich SA (Geneva, Swit- vours were compared with two positive controls, M4 and M5, formulated
zerland) raw materials palette were screened for their minimum inhibi- with 0.18% triclosan (5-chloro-2-(2,4-dichlorophenoxy)phenol) (non-
tory concentration (MIC) against the oral bacterial strains F. nucleatum flavoured control 4) and 0.2% regular peppermint oil-type flavour (flavour
DSMZ 20482, P. gingivalis DSMZ 20709, Prevotella melaninogenica ATCC 5), respectively. Compositions of non-flavoured control 4 and flavour 5
25845 (Bacteroides melaninogenicus) and S. mutans DSMZ 6178. Sixteen are described in Table 1.

Table 1. Antimicrobial flavour composition measured by GC/MS (percentage of mass)

Analyte Retention m/z Flavour composition (% mass)*


time (min) Flavour 1 Flavour 2 Flavour 3 Non-flavoured Flavour 5
control 4
Eucalyptol 1.86 154 3.2 — 1.3 — 1.5
(⫾)-2-Methylhexanoic acid 1.86 74 9.1 9.1 9.1 —
Limonene 1.87 136 5.1 — 2.3 — 1.7
Menthone 2.36 112 6.8 — 6.9 — 6.4
Menthol 2.40 95 44.0 47.0 41.3 — 50.4
3,4-Dimethylphenol 2.44 107 0.8 0.8 0.8 —
Methyl salicylate 2.47
120 — 10.0 — — 0.1
Carvone 2.66 82 2.5 — 14.0 — 9.2
Anethol 2.77 148 10.8 6.8 — — 11.4
Cinnamic aldehyde 2.78 131 0.8 13.4 — — —
Eugenol 2.94 164 1.0 11.5 5.7 — 0.6
Heliotropin 2.96 150 — — 3.7 — —
Ethyl vanillin 3.28 137 — — 6.0 — —
Eugenyl acetate 3.39 164 — 1.2 — — —
Triclosan 4.76 288 — — — 100.0 —
% Quantifiable 84.1 99.8 91.1 100 81.3
* Flavours are a mixture of essential oils and pure compounds. The level of the analysed compounds (GC/MS analysis of 14 flavour
compounds) was investigated on a scale of 100. Therefore, 84.1, 99.8, 91.1, 100 and 81.3% of the total formulation was quantifiable
in the five products, respectively.
m/z, mass-to-charge ratio on mass spectrum.

www.wileyonlinelibrary.com/journal/ffj Copyright © 2010 John Wiley & Sons, Ltd. Flavour Fragr. J. 2010
Effects of flavoured mouth rinses on breath odour

Subject Population subjects were asked to perform a self-assessment of flavour intensity and
mouth rinse appeal (overall perceived sensation) 1 h after the mouth
Twenty-four healthy and non-periodontally diseased adult subjects (10
rinse by using a linear scale of 0 to 10 for flavour intensity and a multiple
males, 14 females), aged 19–29 years (mean age 24 years), participated in
choice answer (not at all, a little, moderately, a lot) for rating flavour
the study. This number of subjects is considered to be sufficient to dis-
pleasantness. Recording of scores was performed with Fizz Software (Bio-
criminate between products in studies that use aromas or malodour
systèmes, Couternon, France).
intensities as clinical variables.[37] Statistical analysis of variance (ANOVA)
confirmed that the product effect is stronger than the subject effect for all
descriptors. Recruitment was performed by the School of Dental Medi- Determination of Flavour Material Concentrations in Saliva
cine. To be included in the panel, a subject had to be a non-smoker in The flavour materials present in the 500 mg saliva samples were extracted
good general health with no visible tongue coating and should not have by using 500 ml volumes of ethyl acetate containing 1 mg/l of methyl
taken antibiotics for 2 months before the start of the trial. Subjects were octanoate (Fluka, Buchs, Switzerland) as the internal standard. Homog-
required not to use perfumes or cosmetics for the duration of the study. enization and extraction were achieved with an automatic shaker, and
At each visit, subjects received tubes of toothpaste (in exchange for their phase separation was induced by centrifugation. The organic phases
old tubes) that were free of flavours, preservatives or antimicrobial sub- were then transferred to 400 ml microvials to be analysed by gas chroma-
stances such as triclosan. These tubes were to be used the week preced- tography with mass spectrometric detection. Chromatographic separa-
ing the study, as well as during the entire study period (5 weeks). Strict tion was achieved on a 180 mm ¥ 10 m (0.18 mm film) DB-XLB column
rules were reiterated to the subjects every week. On the day before the mounted in an Agilent 6890 gas chromatograph equipped with a Combi-
test, volunteers were asked to refrain from consuming food or drink after PAL autosampler (CTC Analytics, Zwingen, Switzerland). One microlitre
midnight. Neither breakfast nor drinks were allowed before or during the aliquots were injected in the pulsed splitless mode (1.2 bar pulse for
morning of the test. One week before the beginning of the study, a 0.6 min) by using a 4 mm tapered liner (FocusLiner tapered, SGE, Ring-
training session was organized to validate the protocol. The protocol was wood, Australia). Elution, with He as the carrier gas, started at 50°C for
approved by the Ethical Committee of the Geneva University Hospital. 0.8 min. The temperature was then ramped to 250°C at 50°C/min, and the
Experiments were undertaken with the understanding and written final temperature was held for 0.4 min. The interval between two injec-
consent of each subject and according to ethical principles, including the tions was about 11.5 min. The mass spectrometer used for detection
World Medical Association Declaration of Helsinki. (Agilent 5973 inert; Agilent, Santa Clara CA, USA) was operated in the
single ion monitoring mode. Quantitative data was obtained by external
Clinical Study Design calibration, with three concentration levels per order of magnitude after
correction to account for the internal standard.
The clinical trial was performed at the School of Dental Medicine, Univer- The quantified compounds, selected to account for more than 80% of
sity of Geneva. Subjects were asked to come once a week at fixed hours all investigated flavour compositions were in order of elution: eucalyptol
during five consecutive weeks to test one of the mouth rinses. All tests (1,8-cineole); (⫾)-2-methylhexanoic acid; limonene ((⫾)-8-methoxy-1-p-
were performed between 7:30 and 9:30 a.m. All products were coded, menthene); menthone; menthol; 3,4-dimethylphenol; methyl salicylate;
randomized and tested in a double-blind study (i.e. neither assessors nor L-carvone ((-)-(R)-1(6),8-p-menthadien-2-one); anethole ((E)-1-methoxy-
subjects knew which mouth rinses were tested). Products were coded 4-(1-propenyl)benzene); cinnamic aldehyde (3-phenyl-2-propenal);
with three-digit codes generated and randomized by Fizz Software (Bio- eugenol (4-allyl-2-methoxyphenol); heliotropin (1,3-benzodioxole-5-
systèmes, Couternon, France). The presentation order of the products was carbaldehyde); ethyl vanillin (3-ethoxy-4-hydroxybenzaldehyde); eugenyl
generated on a Latin square order design. Each panelist tested the five acetate (4-allyl-2-methoxyphenyl acetate) and triclosan (2,4,4′-trichloro-
different mouth rinses described previously. On the day of the test, sub- 2′-hydroxydiphenyl ether). Limits of quantification were typically about
jects were first examined for oral odour. A 1 ml sample of saliva was then 10 ng of individual flavour raw material per gram of saliva.
collected for analytical measurements and bacterial counts.[15] Volunteers
were then asked to rinse their mouth with 20 ml of mouth rinse for 1 min.
Assessments of oral malodour and collected saliva were repeated 1 min
Bacteriological Analysis of Saliva Samples
and 1 h after the rinse. To collect the saliva, we asked subjects to mix their Saliva samples were serially diluted (10-3 and 10-4) in saline solution,
saliva with a closed mouth for 1 min and then to spit (at least 4 ml of plated in duplicate on Schaedler agar plates containing 5% sheep blood
saliva) into a polypropylene sterile 50 ml BD Falcon® tube (VWR Interna- (BioMérieux) by using a Spiral Plater DS Plus (Interscience, Paris, France)
tional, Fontenay Sous Bois, France). To avoid flavour absorption, we did and incubated for 72 h at 37°C under anaerobic conditions (Scholzen
not use saliva stimulation with paraffin. After homogenization, aliquots of Microbiology Systems, Wittenbach, Switzerland). After preliminary
500 mg of saliva (precise weight) were transferred into 1 ml sterile tubes experiments, no significant differences were shown regarding bacterial
and frozen at -20°C for flavour ingredient analysis. The remaining saliva counts after 72 h and 7 days of incubation (data not shown). Total micro-
was transferred in Portagerm® transport bottles (BioMérieux, Geneva, bial counts, including facultative anaerobes and anaerobes, were deter-
Switzerland) at room temperature for no longer than 4 h before homog- mined and the results were expressed as numbers of cfu/ml.
enization, dilution and culturing for bacterial counting.
Statistical Analysis
Assessment of Oral Odours
Statistical treatment of the sensory data was performed with XLStat soft-
Assessment of oral malodours and flavour intensities was performed by ware (Addinsoft, Paris, France). First, a three-way ANOVA with interaction
three trained assessors on an intensity basis. Assessors were trained at two factors was performed on the intensity ratings to determine if
during preliminary sessions under the same conditions as the tests, mouth rinses were discriminated by the assessors. Subjects, assessors and
where they evaluated flavour and oral malodour of three different sub- mouth rinses were considered as individual factors for ANOVA to deter-
jects and also of pure chemical compounds. They were asked to replicate mine which factors had a significant influence on the oral malodour and
their evaluations on the intensity scales, and results were discussed after flavour intensities at different times after rinsing the mouth (after 1 min
each session. Subjects were asked to close their mouth and not to and 1 h). The two factors assessor and subject were set to fixed. Second,
swallow for 1 min before exhaling air. Subjects and assessors were sepa- the Duncan test was then computed to determine if the differences
rated by a wall with an aperture to allow the subject to breathe directly between each pair of products were significant; and third, a Pearson’s
into the assessor’s nose. This method was used to reduce any discomfort correlation coefficient was calculated between the malodour intensity
for both the examiner and the subject. We recorded intensities by using and flavour intensity. The ANOVA and Duncan tests were also performed
an unstructured linear sensorial analytical scale with scores of 0 (no with salivary bacterial counts to evaluate differences between mouth
malodour/flavour) to 10 (very intense malodour/flavour).[38] In addition, rinses in reducing the bacterial count after 1 min and 1 h. Percentage

Flavour Fragr. J. 2010 Copyright © 2010 John Wiley & Sons, Ltd. www.wileyonlinelibrary.com/journal/ffj
M. Troccaz et al.

Figure 1. Sensory ratings of mouth malodour and oral flavour intensity before treatment, 1 min and
1 h after treatment with various mouth rinses (M). M1, M2 and M3 contain antimicrobial flavours 1, 2
and 3 respectively; M4, positive non-flavoured control containing triclosan; M5, positive control
containing peppermint oil (flavour 5). *Significant differences between mouth rinses with a confi-
dence interval of 95% (Duncan test). The error bars represent 95% confidence intervals

reductions in bacterial counts were calculated for each participant indi- malodour intensity may not be solely explained by the decrease
vidually and as an average. For flavour intensity self-assessment scores, of one class of chemicals, but rather to a decrease in a variety of
subjects and mouth rinses were considered as individual factors for bacteria-generated volatiles that contributes to halitosis.
ANOVA and Duncan tests. Finally, a chi-squared test was performed with Statistical results of ANOVA (performed with the mouth rinse,
self-assessment pleasantness ratings to evaluate the possible presence of
the assessors and the subjects as individual factors) and Duncan
significant differences in the distribution of the ratings between mouth
rinses. The confidence level was set to 5% (p < 0.05) for ANOVA tests,
tests on sensory data showed that mouth rinse M3, containing
Duncan tests and correlation coefficients. antimicrobial flavour 3, and control mouth rinse M5, containing
peppermint oil-type flavour 5, significantly reduced oral mal-
odour more than the other flavours at both times. Subjects and
Results and Discussion assessors were also found to have some significant effects on the
results, but at lower levels than the mouth rinse effects (data not
All subjects (n = 24) completed the study and tested the five shown). The assessor effect is significant for all evaluations
mouth rinse compositions. Three compositions contained antimi- (ANOVA by evaluation, with assessor effect). Nevertheless, similar
crobial flavours (M1, M2, M3) and two contained controls (M4, conclusions were obtained for each assessor independently or
M5). No adverse effect was reported during the study. for the three assessors taken together (ANOVA by assessor and by
evaluation, with mouth rinse and subject effects).
Sensory Evaluation of Oral Malodours
Assessment of Mouth Rinse Intensity and Appeal
The average oral malodour of the 24 subjects measured before
rinsing (mean ⫾ confidence interval at 95%) over the 5-week Flavour intensities decreased dramatically from a value between
period was judged to be 3.60 ⫾ 0.45 on a scale of 0 to 10 5.80 ⫾ 0.29 and 6.11 ⫾ 0.38 immediately after the rinse to a value
(Figure 1). Two subjects were found to be outliers (maximum at between 0.42 ⫾ 0.27 and 0.93 ⫾ 0.27 at 1 h. As expected, mouth
5.2 and minimum at 1.5) but did not impair the final results. This rinse M4, containing no flavour except triclosan, was perceived as
relatively low malodour intensity (in average) may be explained significantly weaker when compared with the other mouth rinses
by non-pathological halitosis among the subjects; however, the (Duncan test at 95%) at 1 min and 1 h after the rinse. In contrast,
evaluated morning malodour was high enough to be repulsive in M1, which contains high levels of limonene, menthol and aneth-
a social context and to be considered in this study. Immediately ole (flavour 1), was perceived to be significantly stronger by the
after the rinse, all formulations containing the antimicrobial fla- three assessors 1 h after rinsing (Duncan test at 95%). Those
vours and the two controls containing triclosan and peppermint results were confirmed by the subjects themselves: M1 intensity
oil-type flavour performed equally well for oral malodour reduc- was scored at an average of 4.1 on an intensity scale of 0 to 10,
tion. An average decrease of 72% was observed in oral malodour whereas M2, M3, M4 and M5 were rated at 2.7, 2.3, 1.8 and 2.2,
scores (Figure 1). One minute after the rinse, the malodour inten- respectively. No correlation could be established between mal-
sity decreased to a value between 0.56 ⫾ 0.20 with mouth rinse 3 odour reduction (or malodour intensity) and oral flavour intensity
(M3) and 1.41 ⫾ 0.57 with M4. One hour after the rinse, the at 1 min and 1 h after the rinse (Pearson’s correlation coefficient
intensities increased to a value between 1.12 ⫾ 0.27 (M5) and was not significant, data not shown).
1.97 ⫾ 0.36 (M1), but were still lower compared with baseline In parallel, the assessments of flavour pleasantness by the 24
(56% oral odour intensity reduction). The intensity rating scale is subjects themselves were intended to confirm that malodour
considered to be the gold standard for the measurement of oral control and anti-microbial properties did not impair flavour
malodour because it reflects the real-time presence of an objec- acceptance. The assessment of mouth sensation was judged to
tionable odour as detected by the assessor.[9,21,38–40] In addition, be significantly worse with the non-flavoured mouth rinse M4,
this evaluation method is not restricted to VSCs, contrary to containing triclosan (chi-squared test) (Figure 2). Twelve subjects
evaluation with a halimeter.[9,23,39] Therefore, the decrease of oral of the 24 did not like this mouth rinse at all. M2 and M3, contain-

www.wileyonlinelibrary.com/journal/ffj Copyright © 2010 John Wiley & Sons, Ltd. Flavour Fragr. J. 2010
Effects of flavoured mouth rinses on breath odour

Figure 2. Self-assessment of mouth rinse pleasantness by the subjects. Overall perceived sensa-
tions were recorded 1 h after rinsing the mouth with M1, M2 and M3 containing the antimicrobial
flavours 1, 2 and 3, respectively, and the two positive controls, M4 (containing triclosan) and M5
(containing peppermint oil)

five formulations. Both the mechanical/dilution effect of the


mouth rinse and the antimicrobial effects of the ingredients may
be responsible for those high active percentages. A significant
decrease in bacterial counts was observed in 11, 15, 17, 15 or 16
subjects of the 24 after rinsing with M1, M2, M3, M4 or M5,
respectively. In three subjects, an increase in bacterial counts was
found after rinsing with M1 and M4 and in one subject after
rinsing with M2 and M5.
At 1 h, bacterial counts were significantly lower with M3 than
with the other mouth rinses (Duncan test at 95%) (40% decreases
in bacterial counts on average 1 h after rinsing, Figure 3). With
M3, none of the subjects showed an increase in bacterial count
and three subjects had returned to pre-rinsing values. For the
other mouth rinses (M1, M2, M4 or M5), 9, 19, 16 or 14 subjects of
the 24 had bacterial counts similar to or higher than before the
rinse, respectively.
M3 showed the highest antimicrobial effect after 1 h (40%
Figure 3. Average reduction in salivary bacterial counts (n = 24) 1 min reduction). Flavour 3 contained high levels of the antimicrobial
and 1 h after treatment with various mouth rinses (M). M1, M2 and M3 ingredients menthol, limonene, carvone, eugenol and men-
contain antimicrobial flavours 1, 2 and 3, respectively; M4, positive non- thone. Some of these compounds were reported to have addi-
flavoured control containing triclosan; M5, positive control containing tive effects.[41] In addition, in vitro testing revealed that flavour 3
peppermint oil (flavour 5). * Significant differences between mouth rinses was strongly bactericidal against F. nucleatum, a bacterium that
with a confidence interval of 95% (Duncan test). The error bars represent generates odorous compounds (log reduction of 3.2 or 99.9%
95% confidence intervals reduction of F. nucleatum after 80 s contact time) (data not
shown). In previous studies, similar compositions have been
used to counter plaque, inhibit anaerobic bacterial growth or
ing antimicrobial flavours 2 and 3, respectively, were judged glo-
neutralize VSCs in halitosis treatments. For example, US Patent
bally as the most pleasant (ANOVA and Duncan test).
3164524 describes oral antiseptic compositions for the preven-
tion of bad breath comprising an aqueous alcoholic solution
Salivary Bacterial Counts
containing menthol, methyl salicylate, thymol and eucalyptol.[42]
Before rinsing, the average number of bacteria (mean ⫾ confi- US-5472684 discloses an antibacterial oral composition to
dence interval at 95%) was 8.3 ⫾ 0.2 log cfu/ml of saliva (corre- counter plaque comprising thymol and eugenol.[43] Similarly,
sponding to 3.58 ¥ 108 cfu/ml). Significant differences in bacterial menthol, thymol, methyl salicylate and eucalyptol, present in
counts were found between subjects. Two subjects were outliers, the essential oil-containing mouth rinse Listerine® antiseptic,
with a bacterial number of 7.7 log cfu/ml and 9.0 log cfu/ml, were shown to be active on dental plaque bacteria in situ.[44]
respectively, but these numbers did not influence final mouth Polyphenols such as the epigallocatechin gallate contained in
rinse efficacies. Microbial reduction per milliliter of saliva was first tea extracts were shown to have antimicrobial properties and to
calculated individually for each subject and then averaged. neutralize VSCs.[45]
Results are summarized in Figure 3. By contrast, in our study, antimicrobial compounds such as
One minute after the rinse, no significant differences were triclosan, anethole and cinnamic aldehyde had limited effects on
observed between mouth rinse antimicrobial efficacies: an salivary bacteria and mouth malodour when present in mouth
average bacterial reduction of 49.0 ⫾ 8.0% was calculated for the rinse flavour compositions. One explanation may be that the anti-

Flavour Fragr. J. 2010 Copyright © 2010 John Wiley & Sons, Ltd. www.wileyonlinelibrary.com/journal/ffj
M. Troccaz et al.

Figure 4. Mean concentrations of the flavour ingredients present in the saliva of the subjects before rinsing the mouth and 1 min and 1 h after
treatment (number of subjects = 24). The error bars represent 95% confidence intervals. When the investigated raw material was not detected in the
saliva of all 24 subjects, values indicate the number of subjects in which it was detected; M, mouth rinse. M1, M2 and M3 contain antimicrobial flavours
1, 2 and 3 respectively; M4, positive non-flavoured control containing triclosan; M5, positive control containing peppermint oil (flavour 5). E, eucalyptol;
L, limonene; MA, (⫾)-2-methylhexanoic acid; MN, menthone; ML, menthol; DP, 3,4-dimethylphenol; MS, methyl salicylate; C, L-carvone; A, anethol;
CA, cinnamic aldehyde; EL, eugenol; H, heliotropin; EV, ethyl vanillin; EA, eugenyl acetate; T, triclosan

microbial activity of flavours is due to their lipophilic characteris- tion, such interactions may be modulated by the presence of salt,
tics. The bacteria have lipids on and in their surfaces, which may sweeteners, hydrophobic compounds and other additives in the
take up lipophilic flavours that have an antibacterial effect by mouth rinse.[46] Flavours must be active in a complex environ-
destroying the bacterial membranes. Indeed, the lower hydro- ment, the saliva, which may change flavour retention time and
phobicity of these compounds may explain the results. In addi- antimicrobial properties.

www.wileyonlinelibrary.com/journal/ffj Copyright © 2010 John Wiley & Sons, Ltd. Flavour Fragr. J. 2010
Effects of flavoured mouth rinses on breath odour

Chemical Analysis of Saliva breath and initial salivary bacterial counts that were similar to
those of the other subjects and did not modify our conclusions.
Menthol was the major flavour ingredient present in saliva 1 min However, at this time, the exact source of triclosan still remains
after the rinse, with a concentration of 1.1 ng/g of saliva undetermined.
(Figure 4). At 1 min after the rinse, an average flavour dilution of In conclusion, the present clinical trial has shown that morning
500 was calculated between the flavour material concentration in breath odours can be reduced up to 1 h after rinsing with alcohol-
20 ml of the mouth rinse (containing 0.2% flavour) and the free mouth rinses containing in-vitro active antimicrobial flavour
flavour ingredient concentration expectorated in 1 ml of saliva. ingredients. Maximum efficacies were obtained with mouth rinse
The flavour dilution was 500 000 at 1 h after the rinse. The mouth- 3 (M3) and the peppermint control M5, containing high levels of
wash M1 was rated as significantly more intense 1 h after the menthol and carvone. The maximum malodour intensity reduc-
rinse by both the assessors and the subjects (as described previ- tions were 61% and 68% reduction after 1 h, respectively. In addi-
ously). These results may be explained by the chemical ingredi- tion, flavour 3, containing high levels of the antimicrobial
ents of flavour 1, menthol, limonene and anethole that are still ingredients menthol, limonene, carvone and eugenol, which are
present in saliva 1 h after rinsing at a concentration of 83.7, 30.0 still present in saliva 1 h after rinsing, may explain the antimicro-
and 46.3 ng/g saliva, respectively. One hour after the rinse, six of bial effect of M3 (40% reduction in bacterial count 1 h after
the 15 ingredients analysed were still present in the saliva of up to rinsing). Today, formulating effective oral care products that
21 subjects. These ingredients were menthol, anethole, cinnamic promote fresh breath with long-lasting properties and limited
aldehyde, limonene, carvone and triclosan. They are generally side effects represents a challenge. In the present study, selected
highly hydrophobic flavour ingredients and have low solubility. flavours were found to have beneficial effects against salivary
Flavour compositions comprising limonene, menthol and aneth- microorganisms as well as transient oral malodours. These fla-
ole with an average octanol–water partition coefficient (log P) of vours may be considered for the designing of oral care products
at least 3.3 (in logarithm form) have been previously shown to that promote fresh breath without compromising the taste.
have an increased substantivity and thus longevity in the oral
cavity.[47,48]
Acknowledgements
Other compounds, including methylhexanoic acid, menthone,
eugenol and ethyl vanillin, which have a log P less than 2.5, were The authors are grateful to Firmenich SA for funding this study.
detected in only a few subjects 1 h after rinsing. Five ingredients The authors thank Hidemi Tashiro for creating the flavour com-
(eucalyptol, 3,4-dimethylphenol, methyl salicylate, heliotropin positions and Dr Christopher Dean for support and advice. They
and eugenyl acetate) were not detected in saliva after 1 h are also grateful to Dr Markus Seyfried, Pauline Gourdain and José
(Figure 4). Our results are in agreement with reports of human Fernandez for the minimum inhibitory concentration assays and
salivary proteins changing the partitioning behaviour of aroma microbiological technical assistance. They are indebted to
compounds in the mouth by binding hydrophobic molecules or Vanessa Petitpierre, Anne Yu and Anna Filieri for recruiting the
by salting out hydrophilic molecules.[49,50] Mucin-glycoproteins, subjects and organizing the olfactive tests.
which have low solubility and adhesiveness properties, were
identified as the key components that affect flavour release.[51] In
addition, mucins contain cysteine-rich regions that may react References
with carbonyls present in those compounds to modify antimicro-
1. M. Rosenberg. J. Am. Dent. Assoc. 1996, 127, 475.
bial flavour properties, as well as the efficacies, to reduce 2. M. Rosenberg, G. V. Kulkarni, A. Bosy, C. A. McCulloch. J. Dent. Res.
malodour.[51] 1991, 70, 1436.
After the rinse with M3, the antimicrobial compounds 3. M. Rosenberg, C. A. McCulloch. J. Periodontol. 1992, 63, 776.
limonene (32 ng/g saliva), menthol (121 ng/g saliva), carvone 4. F. L. Suarez, J. K. Furne, J. Springfield, M. D. Levitt. J. Dent. Res. 2000, 79,
1773.
(18 ng/g saliva) and eugenol (7.4 ng/g saliva) were still found in 5. D. Van Steenberghe, P. Avontroodt, W. Peeters, M. Pauwels, W. Coucke,
saliva at 1 h in most subjects, whereas menthone was detected in A. Lijnen, M. Quirynen. J. Periodontol. 2001, 72, 1183.
only two subjects in an average concentration of 2.7 ng/g saliva. 6. M. Quirynen, H. Zhao, D. Van Steenberghe. Clin. Oral Invest. 2002, 6, 1.
Indeed, the high antimicrobial activity of M3 at 1 h after the rinse 7. M. Quirynen, P. Avontroodt, C. Soers, H. Zhao, M. Pauwels, W. Coucke,
may be explained by the presence of limonene, menthol, carvone D. Van Steenberghe. J. Clin. Periodontol. 2002, 29, 944.
8. H. Senpuku, A. Tada, T. Yamaga, N. Hanada, H. Miyazaki. Int. Dent. J.
and eugenol at a concentration higher than their MIC values. In 2004, 54, 149.
contrast, triclosan was still present at a concentration of 9. W. J. Loesche, C. Kazor. Periodontology 2000 2002, 28, 256.
1752 ng/g of saliva 1 h after rinsing with M4; however, it had a 10. F. J. Hugues, R. McNab. Arch. Oral Biol. 2008, 53, S1.
limited effect on salivary bacterial counts. This may be due to the 11. C. E. Kazor, P. M. Mitchell, A. M. Lee, L. N. Stokes, W. J. Loesche, F. E.
Dewhirst, B. J. Paster. J. Clin. Microbiol. 2003, 41, 558.
solubilizer used to formulate the mouth rinse. The nature of the 12. R. P. Allaker, R. D. Waite, J. Hickling, M. North, R. McNab, M. P. Bosma,
solubilizing agent has been shown to be critical to the plaque- F. Hugues. Arch. Oral Biol. 2008, 53, 8.
inhibiting activity of triclosan.[24] It is interesting to note that one 13. J. Tonzetich. J. Periodontol. 1977, 48, 13.
third of our subjects were found to have an average of 80 ⫾ 14. I. Kleinberg, M. Codipilly. Quintessence Int. 1999, 30, 357.
51 ng/g of triclosan (mean ⫾ standard error) in their saliva in the 15. S. Roldán, D. Herrera, I. Santa-Cruz, A. O’Connor, I. González, M. Sanz.
J. Clin. Periodontol. 2004, 31, 1128.
morning before the rinse despite the strong recommendations 16. A. C. Donaldson, D. McKenzie, M. P. Riggio, P. J. Hodge, H. Rolph, A.
given to the subjects not to use any oral care products containing Flanagan, J. Bagg. Oral Dis. 2005, 11, 61.
triclosan 1 week before and during the study (see Materials 17. S. Goldberg, A. Kozlovsky, M. Rosenberg. In Bad Breath: Research Per-
and Methods). This was previously stated in an in vivo laboratory spectives, M. Rosenberg (ed.). Ramot Publishing: Tel Aviv, Israel, 1995,
p. 71.
scale analysis (data not shown). A concentration of 1200 ng 18. I. Kleinberg, M. Codipilly. In Bad Breath: Research Perspectives, M.
triclosan/g of saliva was detected in one of the subjects. Despite Rosenberg (ed.). Ramot Publishing: Tel Aviv, Israel, 1995, p. 13.
such triclosan concentrations, those subjects had morning bad 19. N. Sterer, R. B. Greenstein, M. Rosenberg. J. Dent. Res. 2002, 81, 182.

Flavour Fragr. J. 2010 Copyright © 2010 John Wiley & Sons, Ltd. www.wileyonlinelibrary.com/journal/ffj
M. Troccaz et al.

20. M. Rosenberg, I. Septon, I. Eli, R. Bar-Ness, I. Gelernter, S. Brenner, 36. A. K. Talwar, E. J. Carlin. Taste Masking of Thymol. Patent application US
I. Gelernter, J. Gabbay. J. Periodontol. 1991, 62, 487. 4945087, 1990.
21. M. Rosenberg, I. Gelernter, M. Barki, R. Bar-Ness. J. Periodontol. 1992, 37. M. C. Meilgaard, G. V. Civille, B. T. Carr. In Sensory Evaluation Tech-
63, 39. niques, 3rd edition. CRC Press: Boca Raton, FL, 1999, p. 161.
22. V. I. Haraszthy, J. J. Zambon, P. K. Sreenivasan, M. M. Zambon, D. 38. G. Pitts, R. Pianotti, T. W. Feary, J. McGuiness, T. Masurat. J. Dent. Res.
Gerber, R. Rego, C. Parker. J. Am. Dent. Assoc. 2007, 138, 1113. 1981, 60, 1891.
23. E. G. Winkel. In Clinical Periodontology and Implant Dentistry, 5th 39. M. Sanz, S. Roldan, D. Herrera. J. Contemp. Dent. Pract. 2001, 2, 1.
edition. J. Lindhe, N. Lan, T. Karring (eds). Wiley-Blackwell: Oxford, UK, 40. M. Ishikawa, K. Shibuya, F. Tokita. J. Dent. Health 1984, 24, 54.
2008, p. 1325. 41. S. F. Van Vuuren, A. M. Viljoen. Flavour Fragr. J. 2007, 22, 540.
24. G. Rölla, V. Kjaerheim, S. M. Waaler. In Proceedings of the 2nd European 42. T. I. Fand, F. C. Ninger, M. E. Stolar. Oral Antiseptic. Patent application
Workshop on Periodontology, N. P. Lang, T. Karring, J. Lindhe (eds). US 3164524, 1965.
Quintessence Publishing: Berlin, 1995, p. 120. 43. N. Nabi, A. Gaffar, S. Lucchesi, J. J. Afflitto, S. M. Herles. Oral Composi-
25. S. Roldan, D. Herrera, M. Sanz. Clin. Oral Invest. 2003, 7, 189. tions for Plaque and Gingivitis. Patent application US 5472684, 1995.
26. T. L. Outhouse, Z. Fedorowicz, J. V. Keenan, R. Al-Alawi. Gen. Dent. 44. P. Pan, M. L. Barnett, J. Coelho, C. Brogdon, M. B. Finnegan. J. Clin.
2006, 54, 352. Periodontol. 2001, 27, 256.
27. M. D. Carvalho, C. M. Tabchoury, J. A. Cury, S. Toledo, G. R. Nogueira- 45. J. Smullen, G. A. Koutsou, H. A. Foster, A. Zumbe, D. M. Storey. Caries
Filho. J. Clin. Periodontal. 2004, 31, 85. Res. 2007, 41, 342.
28. L. Netuschil, T. Hoffmann, M. Brecx. Int. J. Dent. Hygiene 2003, 1, 46. J. H. Hong, S. E. Duncan, A. M. Dietrich, S. F. O’Keefe. J. Agric. Food
188. Chem. 2006, 54, 9168.
29. D. J. Bradshaw, K. D. Perring, P. M. Cawkill, A. F. Provan, D. A. McNulty, 47. K. D. Perring, K. M. Tuck, D. A. McNulty, M. Wilson. Flavour Composi-
E. J. Saint, J. Richards, M. J. Munroe, J. M. Behan. Oral Dis. 2005, 11, 75. tions Comprising Menthol, a Mint Oil and other Flavour Materials.
30. M. Greenberg, P. Urnezis, M. Tian. J. Agric. Food Chem. 2007, 55, 9465. Patent application WO2004073669, 2004.
31. J. A. Morris, A. Khettry, E. W. Seitz. J. Am. Chem. Soc. 1979, 56, 595. 48. J. J. Stroz, A. I. Bakal, F. Witzel, D. A. M. MacKay. Chewing Gum Having
32. S. G. Deans, G. Ritchie. Int. J. Food Microbiol. 1987, 5, 165. Improved Flavor Duration and Shelf-life. Patent application US
33. C. M. Mann, J. L. Markham. J. Appl. Microbiol. 1998, 84, 538. 4157401, 1979.
34. J. Barra, M. Seyfried, T. Hidemi, M. Troccaz, S. Beccucci. Antimicrobial 49. E. N. Friel, A. J. Taylor. J. Agric. Food Chem. 2001, 49, 3898.
Flavouring Composition. Patent application WO2008068683, 2008. 50. S. M. Van Ruth, I. Grossmann, M. Geary, C. M. Delahunty. J. Agric. Food
35. P. A. Delli Santi, D. G. A. Nelson. Taste Masking of Phenolics Using Citrus Chem. 2001, 49, 2409.
Flavors. Patent application US 5945088, 1999. 51. L. A. Tabak. Crit. Rev. Oral Biol. M. 1990, 1, 229.

www.wileyonlinelibrary.com/journal/ffj Copyright © 2010 John Wiley & Sons, Ltd. Flavour Fragr. J. 2010

You might also like