archives of oral biology 53, Suppl.

1 (2008) S8–S12

j o u r n a l h o m e p a g e : w w w. i n t l . e l s e v i e r h e a l t h . c o m / j o u r n a l s / a r o b

Topographic distribution of bacteria associated with oral

malodour on the tongue

Robert P. Allakera, *, Richard D. Waiteb , Jenneth Hicklinga , Mairead Northc , Rod McNabc ,
MaryLynn P. Bosmac , Francis J. Hughesa
a
Institute of Dentistry, b Institute of Cell and Molecular Sciences, Barts & The London School of Medicine and Dentistry, Queen Mary
University of London, London, UK
c
GlaxoSmithKline Consumer Healthcare, Weybridge, UK

article info abstract

Keywords: Objective: To investigate the topographic distribution of bacterial types and loads
Tongue associated with mid-morning oral malodour on the tongue surface.
Oral malodour Design: Fifty subjects with good oral health and at least 20 natural uncrowned
Lingual tonsil teeth were included. Samples were taken with sterile brushes from the dorsal
Volatile sulphur compounds anterior (DA), dorsal middle (DM), dorsal posterior (DP), dorsal posterior to the
circumvallate papillae (DPCP), lateral posterior (LP) and ventral posterior (VP) tongue
surfaces. Samples were cultured on appropriate media for anaerobic bacteria,
aerobic bacteria, Gram-negative anaerobic bacteria, volatile sulphur compound (VSC)-
producing bacteria and Streptococcus salivarius. Malodour was assessed by trained
judges on an intensity basis.
Results: The counts of all bacterial groups were consistently highest at the DPCP
surface. Mean VSC-producing bacterial counts (colony forming units/brush ×105 )
were 1.45, 5.67, 32.52, 88.94, 6.46 and 0.33 at DA, DM, DP, DPCP, LP and VP surfaces,
respectively. Anaerobic, Gram-negative and VSC counts at DPCP surfaces increased
with malodour intensity, whereas aerobic and S. salivarius counts decreased; however
these differences were not statistically significant.
Conclusion: It is concluded that the DPCP area consistently carries the highest load
of bacteria capable of contributing to oral malodour. The study demonstrates that
tongue surfaces not accessible to routine oral hygiene procedures can significantly
contribute to oral malodour.
© 2008 Elsevier Ltd. All rights reserved.

time. Approximately 90% of the cases of oral malodour

1. Introduction
are associated with the production of malodorous
Oral malodour, or halitosis, is a common complaint compounds by oral bacteria. Other causes include
that can affect up to half of the population at any one gastrointestinal disorders, hepatic diseases, ingestion of
certain foods and smoking. 1
* Corresponding author. Dr R.P. Allaker. Oral Microbiology, Bacterial products that contribute to oral malodour
Institute of Dentistry, Barts & The London School of result from the breakdown of proteins, peptides and
Medicine and Dentistry, Queen Mary University of London,
mucins found in saliva, gingival crevicular fluid, host
4, Newark Street, London E1 2AT, UK. Tel.: +44 20 7882 2388;
fax: +44 20 7882 2191. cells and residual food. 2 The sulphur-containing amino
E-mail address: R.P.Allaker@qmul.ac.uk (R.P. Allaker). acids namely cysteine, cystine and methionine, found

0003-9969/ $ – see front matter © 2008 Elsevier Ltd. All rights reserved.
 archives of oral biology 53, Suppl. 1 (2008) S8–S12 S9

either free in gingival crevicular fluid, in saliva or undertaken. Samples were taken from the dorsal anterior
produced from proteolytic breakdown are the key (DA), dorsal middle (DM), dorsal posterior (DP) and dorsal
precursor molecules for malodorous volatile sulphur posterior to the circumvallate papillae (DPCP), lateral
compounds (VSCs). VSCs are considered to be the posterior (LP) and ventral posterior (VP) tongue surfaces.
most significant products as regards oral malodour Samples were placed into a reduced transport fluid, 7
and include hydrogen sulphide, methyl mercaptan and stored at 4ºC for no more than 2h, and then vortexed
dimethyl sulphide as the main contributors. 3 Many for 30s to disaggregate bacteria from the brush. Samples
Gram-negative anaerobic species of bacteria, found in the were cultured on appropriate media using a spiral plater
oral cavity, produce malodorous compounds including (Don Whitley Instruments, Shipley, UK). Blood agar base
VSCs, short-chain organic acids (butyrate, propionate, No.2. (Oxoid CM0271) with defibrinated horse blood (5%
valerate), diamines (cadaverine, putrescine) and phenyl v/v) was used to culture aerobic bacteria. Anaerobe
compounds (indole, skatole, pyridine). These species basal agar (Oxoid CM0972) with defibrinated horse blood
include Treponema denticola, Porphyromonas gingivalis, (5% v/v) was used to culture anaerobic bacteria. This
Prevotella intermedia, Tannerella forsythensis, Porphyromonas medium was supplemented with vancomycin (5 mg/ml)
endodontalis and Eubacterium species. 4 In contrast, to enumerate Gram-negative anaerobic bacteria. A dif-
Streptococcus salivarius and other Gram-positive bacteria ferential medium was used to enumerate VSC-producing
are more commonly found in individuals with low/no bacteria. 8 This medium contained Columbia Agar Base
malodour. 5 (CM0331) supplemented with haemin (0.0001% w/v),
The dorsum of the tongue, with a papillary structure glutathione (0.12% w/v) and lead acetate (0.02% w/v).
that allows the retention of considerable quantities VSC-producing colonies exhibit a grey or black colour
of food and other debris, supports a large population following incubation. This medium was validated with
of bacteria including Gram-negative anaerobes. 1 Most known VSC-producing species including F. nucleatum,
published studies on the microbiota of the tongue have P. gingivalis and Prev. intermedia. TYC medium (Lab M)
treated the dorsal surface as a single habitat for sampling was used to enumerate S. salivarius. Biochemical tests
purposes and have reported an association between were used on representative colonies to ensure correct
bacterial numbers and oral malodour. 5,6 The objective of identification to species level. Aerobic and anaerobic
this study was to more closely examine the topographic cultures were incubated in air with 5% CO2 at 37ºC for
distribution and loads of those bacterial types associated 24 h, and 80% N2 , 10% CO2 , 10% H2 at 37ºC for 48 h,
with oral malodour on the tongue surface, including respectively.
the dorsal area posterior to the circumvallate papillae Malodour was assessed by trained judges on an
intensity basis and scored as 0 (no malodour), 1 (very
which has received little attention to date. Possible
slight), 2 (slight), 3 (moderate), 4 (strong) and 5 (very
associations between bacterial counts and mid-morning
strong).
oral malodour intensity scores were also investigated.
Such information has particular relevance to the
2.1. Statistical methods
successful control of oral malodour.
Raw data were converted to colony forming units/sample
(CFU/sample). Duplicate counts were averaged and
2. Materials and methods then log10 transformed for statistical analysis. Counts
recorded as <102 were artificially transformed to 50 prior
The study was carried out at a single centre (Institute to log10 transformation to account for the limit of detec-
of Dentistry, London) and did not involve any treatment tion of the system. For the comparison between tongue
intervention. Consent was sought and ethical clearance sites, a mixed model was fitted, with tongue site and
for the study was obtained from the East London and malodour group as fixed effects and subject as random
the City Health Authority, UK. Eligible subjects, with good effect. Results were adjusted for multiple testing using
general oral health, periodontally healthy and at least 20 the Tukey-Kramer method. For the comparison between
natural uncrowned teeth, attended two visits. At the first malodour groups, a linear model was fitted for each
visit subjects underwent an oral soft tissue examination. tongue site with malodour group as the fixed effect.
After a further phase of between 7 and 14 days, subjects
attended a second time (mid-morning) for oral soft tissue
3. Results
examination, oral malodour intensity assessment and an
oral biofilm sampling from various areas of the tongue. 19 men and 32 women were recruited to the study.
The tongue was sampled by gently pressing a The ethnicity mix comprised 21 Caucasian, 10 Asian, 15
Wisdom® baby toothbrush (surface area 84 mm2 ) onto Afro/Caribbean and 5 Others with a mean age of 34.72
the surface and oscillating it slightly without actual years (minimum 18; maximum 59; median 31.95 years).
lateral movement. Randomized sampling of subjects on Final data was available for 50 subjects. 22 subjects fell
either the left or right hand side of the tongue was within the strong malodour category (scores of 4 or 5)
S10 archives of oral biology 53, Suppl. 1 (2008) S8–S12

Table 1 – Summary of mean bacterial counts by area of tongue a

Bacterial type Bacterial counts by tongue area (CFU/sample ×105 )
DA DM DP DPCP LP VP
Total anaerobes 10.57±1.08 27.38±2.64 175.93±20.00 327.97±27.49 32.02±4.55 5.83±0.88
Total aerobes 9.80±1.17 17.72±1.53 77.51±9.15 165.52±16.95 22.54±2.44 5.03±0.67
Gram-negative anaerobes 2.80±0.41 11.20±1.68 87.78±12.56 175.92±20.14 10.70±1.91 0.85±0.22
VSC-producing bacteria 1.45±0.19 5.67±0.90 32.52±4.68 88.94±14.03 6.46±1.56 0.33±0.13
Streptococcus salivarius 0.50±0.09 2.21±0.33 24.01±5.81 54.62±10.23 1.79±0.38 0.09±0.06
a Data are presented as mean±SE (n = 50).
DA = Dorsal anterior; DM = Dorsal middle; DP = Dorsal posterior; DPCP = Dorsal posterior to the circumvallate papillae;
LP = Lateral posterior; VP = Ventral posterior.

and 28 subjects fell within the slight/moderate malodour 400

CFU/sample X 105
category (scores of 1, 2 or 3). A
300
The mean bacterial counts for each tongue sampling
location are shown in Table 1. Considering the dorsum 200
of the tongue, there was a general trend for increasing
100
bacterial load moving from the anterior to the posterior
of the tongue. The most heavily colonized tongue site 0
DA DM DP DPCP LP VP
was DPCP, with approximately twice as many of each of 250
CFU/sample X 105

the five bacterial types assessed than the DP site, the

200 B
next most heavily colonized. A large increase in numbers,
observed across all five bacterial types, was observed 150
between DM and DP (4–12 fold increase). The site with 100
fewest recovered bacteria was the VP surface with 33 50
(total aerobes) to 270 (VSC-producing bacteria) fold fewer
0
bacteria than DPCP. At each tongue site sampled, the DA DM DP DPCP LP VP
250
relative proportions of each of the five bacterial types
CFU/sample X 105

remained broadly similar. Thus, total anaerobes were 200 C

consistently the most numerous bacterial group at every 150
site, with S. salivarius numbers consistently the lowest. 100
In common with total anaerobes, the number of VSC-
50
producing bacteria was highest at posterior sites. In
addition, recovered aerobe numbers were higher than 0
DA DM DP DPCP LP VP
Gram-negative anaerobe numbers at DA, DM, LP and 120
CFU/sample X 105

VP, Gram-negative anaerobes outnumbered aerobes at 100

D
the posterior of the tongue dorsum (DP, DPCP). The 80
ratio of total aerobes to total anaerobes declined from 60
0.93 at the anterior of the tongue dorsum to 0.44 and 40
0.5 at the dorsal posterior (DP and DPCP, respectively). 20
Comparison between tongue sites within subjects 0
indicated significant differences for each bacterial type DA DM DP DPCP LP VP
100
(data not shown). Exceptions were LP versus DM for
CFU/sample X 105

80 E
all 5 bacterial types and additionally for S. salivarius
comparisons DP versus DPCP and LP versus DA. 60
Subjects were subdivided into those with slight/ 40
moderate (intensity score of 1, 2 or 3) or strong 20
(intensity score of 4 or 5) malodour. Those with
0
strong malodour had higher levels of total anaerobes, DA DM DP DPCP LP VP
Gram-negative anaerobes and VSC-producing bacteria Fig. 1 – Mean bacterial counts (±SE) of: (A) total
recovered from DPCP compared with those with slight/ anaerobes, (B) total aerobes, (C) Gram-negative anaerobes,
moderate malodour (Figures 1A, 1C and 1D). Conversely, (D) VSC-producing bacteria, (E) Streptococcus salivarius, at
total aerobe and S. salivarius numbers were higher at tongue sites sampled for subjects with slight/moderate
this site in the slight/moderate malodour group (Figures (, n = 28) or strong (, n = 22) oral malodour.
1B and 1E). However, the mean differences in bacterial
 archives of oral biology 53, Suppl. 1 (2008) S8–S12 S11

numbers on the DM, DP or DPCP tongue dorsum sites including uncultivable bacteria, may be important in
between the two malodour groups (slight/moderate contributing to malodour. Another study, based upon 16S
malodour and strong malodour) were not statistically rRNA gene sequencing, demonstrated that certain oral
different. bacterial species are more prevalent in subjects with no
malodour, whereas other species tended to predominate
in subjects with halitosis. 12 Streptococcus salivarius was
4. Discussion among those found to be more common in individuals
with low/no malodour. The potential of this species
It is well recognised that individual species of bacteria to control oral malodour has been explored using the
differ in their ability to adhere to different attachment administration of a bacteriocin-producing strain; which
sites within the oral cavity and also to other species. 9 was shown to reduce oral VSC levels. 13
Although an overall load difference between the different The filiform, circumvalate and foliate papillae and the
groups of bacteria at each tongue site was demonstrated crevices associated with the lingual tonsils and mucous
with this study, no indication of site-specificity can be glands all act to increase the accumulation of bacteria
ascertained. However, within each of the four broad on the dorsal surface of the tongue. As shown with
categories (aerobes, anaerobes, Gram-negative anaerobes anaerobic VSC-producing bacteria, the posterior area of
and VSC-producing bacteria) specificities may well exist. the dorsal surface of the tongue supports the highest
Paster et al. 9 attempted to define the bacterial diversity loads of anaerobic nitrate reducing bacteria, including
of the oral cavity by examining nine different oral sites, Veillonella spp. and Actinomyces spp. 14 The dorsal surface
including the lateral side of the tongue and the tongue of the tongue offers a large surface area to support a
dorsum. Some species were found to be common to all higher bacterial density (estimated at 100 bacteria per
sites, while many bacteria were site-specific. For example tongue epithelial cell) than any other mucosal surface in
Neisseria spp. were found not to substantially colonise the the oral cavity, for example the ventral and lateral tongue
teeth or subgingival plaque but were commonly found surfaces. The papillary structure of the tongue represents
on soft tissues. S. salivarius was detected generally on an ecological niche that is unique in nature, and acts
the surface of the tongue dorsum, as was shown in this as a reservoir for both oral debris and microorganisms,
study. most notably the anaerobic species of bacteria, including
Several investigators have identified the dorsal pos- Veillonella spp. and Prevotella spp. 11,14 The tongue clefts
terior surface of the tongue as making the primary themselves provide refuge from the forces of mastication
contribution to oral malodour, however little attention and salivary flow, and provide a low redox potential
has been paid to the area posterior to the circumvallate which encourages the growth of such bacteria. Saliva on
papillae (DPCP). In this study, this area consistently the tongue surface acts to bathe the bacteria providing
showed the highest count for all bacterial groups; while buffering and a supply of nutrients. The area beyond
the ventral posterior area consistently showed the lowest the circumvallate papillae; the lingual tonsil is an
count for all bacterial groups examined. It was observed aggregation of lymphoid tissue comprised of 35–100
that counts of total anaerobic, total Gram-negative separate tonsillar units occupying this posterior region at
anaerobic and VSC-producing bacteria were higher at the the root of the tongue. Further studies to determine the
DPCP surface of subjects with strong malodour whereas significance of the micro-anatomy at this site in relation
counts of total aerobes and S. salivarius were higher in to the accumulation of microorganisms are required.
the slight/moderate malodour subjects. Further studies, This study describes the first detailed mapping of the
with a greater number of subjects in each malodour tongue surface in relation to the bacteria associated with
category, and including subjects with no oral malodour, oral malodour and has demonstrated that the dorsal area
may help to confirm the relationship between bacterial posterior to the circumvallate papillae consistently car-
counts and oral malodour at the DPCP site. In support of ries the highest load of bacteria capable of contributing
these findings, DeBover and Loesche 5 showed that both to malodour through the production of volatile sulphur
total and anaerobic bacterial loads on the dorsal surface compounds. This study also demonstrates that surfaces
of the tongue correlated with organoleptic score. More at the very rear of the tongue and thus not accessible
specifically, the findings of Washio et al. 10 suggest that to routine oral hygiene procedures can significantly
an increase in the load of H2 S-producing bacteria found contribute to oral malodour.
within the tongue biofilm is largely responsible for the
intensity of oral malodour.
Donaldson et al. 11 showed no association between 5. Acknowledgements
halitosis and any specific genus of anaerobic bacteria.
However, they did demonstrate an increase in species This study was financially supported by GSK. GSK were
diversity in samples from individuals with halitosis. involved in the study design, interpretation of data and
They suggest that interactions between several species, writing of the manuscript.
S12 archives of oral biology 53, Suppl. 1 (2008) S8–S12

8. Turng BF, Minah GE, Falker WA. Development of an

6. Conflict of interest statement agar medium for the detection of oral H2 S-producing
organisms. J Dent Res 1997;76:226.
Co-authors MN, RM and MPB were employed by GSK at 9. Paster BJ, Olsen I, Aas JA, Dewhirst FE. The breadth
the time of the study. of bacterial diversity in the human periodontal
pocket and other oral sites. Periodontology 2000 2000
2006;42:80−7.
10. Washio J, Sato T, Koseki T, Takahashi N. Hydrogen
References
sulfide-producing bacteria in tongue biofilm and
their relationship with oral malodour. J Med Microbiol
2005;54:889−95.
1. Scully C, El-Maaytah M, Porter SR, Greenman J. Breath 11. Donaldson AC, McKenzie D, Riggio MP, Hodge PJ, Rolph
odor: etiopathogenesis, assessment and management. H, Flanagan A, Bagg J. Microbiological culture analysis
Eur J Oral Sci 1997;105:287−93. of the tongue anaerobic microflora in subjects with and
2. Loesche WJ, Kazor CE. Microbiology and treatment of without halitosis. Oral Dis 2005;11:61−3.
halitosis. Periodontology 2000 2002;28:256−79. 12. Kazor CE, Mitchell PM, Lee AM, Stokes LN, Loesche WJ,
3. Tonzetich J. Direct gas chromatographic analysis of Dewhirst FE, Paster BJ. Diversity of bacterial
sulphur compounds in mouth air in man. Arch Oral Biol populations on the tongue dorsa of patients with
1971;16:587−97. halitosis and healthy patients. J Clin Microbiol
4. Persson S, Edlund MB, Claesson R, Carlsson J. The 2003;41:558−63.
formation of hydrogen sulfide and methylmercaptan 13. Burton JP, Chilcott CN, Moore CJ, Speiser G, Tagg JR. A
by oral bacteria. Oral Microbiol Immunol 1990;5:195–201. preliminary study of the effect of probiotic Streptococcus
5. DeBover EH, Loesche WJ. Assessing the contribution of salivarius K12 on oral malodour parameters. J Appl
anaerobic microflora of the tongue to oral malodour. J Microbiol 2006;100:754−64.
Am Dent Assoc 1995;126:1384−93. 14. Doel JJ, Benjamin N, Hector MP, Rogers M, Allaker RP.
6. Hartley MG, El-Maaytah MA, McKenzie C, Greenman J. Evaluation of bacterial nitrate reduction in the human
The tongue microbiota of low odour and malodorous oral cavity. Eur J Oral Sci 2005;113:14−9.
individuals. Microb Ecol Health Dis 1996;9:215−23.
7. Allaker RP, Langlois T, Hardie JM. Prevalence of Eikenella
corrodens and Actinobacillus actinomycetemcomitans in the
dental plaque of dogs. Vet Rec 1994;134:519−20.
J Periodont Res 2010; 45: 31–37  2009 John Wiley & Sons A/S
All rights reserved
JOURNAL OF PERIODONTAL RESEARCH
doi:10.1111/j.1600-0765.2008.01199.x

Oral malodorous compound

B. Calenic, K. Yaegaki, T. Murata,
T. Imai, I. Aoyama, T. Sato,
H. Ii

triggers mitochondrial- Department of Oral Health, Nippon Dental

University, Tokyo, Japan

dependent apoptosis and

causes genomic DNA
damage in human gingival
epithelial cells
Calenic B, Yaegaki K, Murata T, Imai T, Aoyama I, Sato T, Ii H. Oral malodorous
compound triggers mitochondrial-dependent apoptosis and causes genomic DNA
damage in human gingival epithelial cells. J Periodont Res 2010; 45: 31–37.
 2009 John Wiley & Sons A/S

Background and Objective: Volatile sulfur compounds are the main compounds
causing halitosis. One of these compounds, hydrogen sulfide (H2S), which is
responsible for physiological halitosis, is reported also to have periodontal
pathogenic activities. Hydrogen sulfide has been shown to activate the apoptotic
process in different tissues. Apoptosis plays an important role in the development
of periodontitis. The aim of this study was to determine whether H2S causes
apoptosis in human gingival epithelial cells and to examine the cellular signaling
pathway initiating the process.
Material and Methods: Human gingival epithelial cells were incubated with
50 ng/mL H2S in air contining 5% CO2 for 24, 48 or 72 h. To detect apoptosis, the
cells were stained with annexin V and 7-amino actinomycin D, and analyzed using
flow cytometry. Reactive oxygen species, mitochondrial membrane depolarization
and release of cytochrome C into the cytosol were assessed using flow cytometry
and enzyme-linked immunosorbent assay. Activity levels for the key apoptotic
enzymes caspase-9, -8 and -3 were also determined. Genomic DNA damage was
detected using single-cell gel electrophoresis.
Results: Apoptosis was significantly increased to 24.5 ± 5.7 at 24 h and 41.5
± 8.9% at 48 h (p < 0.01). Reactive oxygen species were enhanced and mito-
chondrial membrane depolarization was collapsed. Cytochrome C release was Ken Yaegaki, DDS, PhD, Department of Oral
Health, Nippon Dental University, 1-9-20 Fujimi,
dramatically increased (0.12 ± 0.02 vs. 0.02 ± 0.01 at 24 h and 0.21 ± 0.02 vs. Chiyoda-ku, Tokyo, 102-8159 Japan
0.02 ± 0.01 ng/mL at 48 h; p < 0.05). Caspase-9 and -3 were strongly activated, Tel: +81 03 3261 8791
while caspase-8 activity remained low. The percentage of DNA strand breaks Fax: +81 03 3261 8796
e-mail: yaegaki-k@tky.ndu.ac.jp
increased, especially at 48 h.
Key words: hydrogen sulphide; periodontitis;
Conclusion: Hydrogen sulfide induces apoptosis in human gingival epithelial cells apoptosis; halitosis
by activating the mitochondrial pathway. Accepted for publication December 5, 2008
32 Calenic et al.

ÔOral malodorÕ, the term used to des- however, the apoptotic effect of H2S on Negative control samples were sub-
cribe offensive odors of the breath human gingival cells has not been jected to an identical procedure, except
originating mostly from the oral cavity, elucidated. that they were incubated in air con-
is an increasing problem in todayÕs Detection of apoptosis in human taining CO2 without H2S.
society. Many compounds can be gingival tissues indicates that it may
found in human breath air, but volatile play an important role in the develop-
Hydrogen sulfide incubation system
sulfur compounds (VSCs) are the sub- ment and control of inflammation and
stances primarily responsible for oral cell destruction in periodontal disease The H2S incubation system allows
malodor (1). The VSCs in the oral (16–19). Thus, programmed cell death incubation in a 37C incubator with a
cavity are mainly hydrogen sulfide might be one of the mechanisms H2S sealed chamber. The 25 cm2 flasks
(H2S) and methyl mercaptan (CH3SH). involved in the pathogenesis of perio- containing the Ca9-22 cells were placed
Of the two, H2S plays a distinct role in dontitis. In a previous study, we showed inside the sealed chamber. A standard
physiological halitosis (2). that H2S, at concentrations lower than H2S permeator (PD-1B-2; Gastec,
Our previous research focused not those found in pathological gingival Kanagawa, Japan), and H2S perme-
only on the esthetic problems of VSCs crevicular fluid, causes genomic DNA ation tubes (Permeacal; Gastec) were
but also on their toxicities. Studies damage and apoptosis in human gingi- used to supply 5% CO2 in H2S at
have shown that increased levels of val fibroblasts (20). In the present study, 50 ng/mL concentration in the cham-
H2S in the oral environment have a we focus on the interrelationship of ber with a constant air flow of more
highly toxic effect on oral tissues and H2S, apoptosis and gingival epithelial than 200 mL/min.
play a role in the etiology and devel- cells, since the gingival epithelial cells
opment of periodontitis (3,4). The serve as the first line of defense against
Detection of apoptosis
VSCs have been shown to alter the periodontal pathogens. We also exam-
permeability of gingival tissues, thus ine the cellular signaling pathways For the detection of apoptosis and for
facilitating the penetration of lipo- implicated in the apoptotic process. distinguishing non-apoptotic from
polysaccharide into gingival epithelium apoptotic cells, a Guava Nexin PCA
and inducing inflammatory responses. (Guava Technologies, Hayward, CA,
Material and methods
Moreover, VSCs can increase the pro- USA) was used. The method is based
duction of collagenase and prosta- on double staining with two fluorescent
Cell culture
glandin E2, both of which are key dyes, allowing direct detection via flow
mediators in the processes of tissue Cells from a human gingival epithelial cytometry. This assay includes annexin
destruction and inflammation. In cell line, Ca9-22 (Health Science V for the detection of early apoptosis
studies focused on their effects on gin- Research Resources Bank, Osaka, and 7-amino actinomycin D (7-AAD)
gival connective tissue, we stress that Japan), were cultured in DulbeccoÕs as an indicator of late apoptosis or
VSCs inhibit collagen synthesis (5) and modified EagleÕs medium (Invitrogen, necrosis. annexin V binds phosphati-
decrease total protein production by Carlsbad, CA, USA) supplemented dylserine translocated to the external
human gingival fibroblasts (6). In with 10% fetal bovine serum (Invitro- membrane surface in early apoptotic
addition, VSCs interfere in the wound- gen) at 37C in an atmosphere of air cells (22), and 7-AAD, a DNA inter-
healing process by inhibiting the pro- containing 5% carbon dioxide (CO2). calator, permeates late-stage apoptotic
liferation of epithelial cells (7) and the For each experiment, the cells were and necrotic cells. After treatment
synthesis of basal membranes (8). plated in 25 cm2 flasks at 4 · 105 cells with H2S for 24 or 48 h, the cells were
Collectively, these findings suggest that well density and allowed to attach washed with PBS, trypsinized, pelleted
VSCs are periodontally pathogenic. overnight. Prior to incubation, the cells and resuspended in cold Guava Nexin
A number of recent studies have were washed twice in prewarmed phos- buffer at a concentration of 1 · 106
demonstrated the involvement of H2S phate-buffered saline (PBS) and then cells mL. Subsequently, the cells were
in the apoptotic process. Under normal placed in fresh medium. The prepared double stained with 2.5 lL of 7-AAD
conditions, apoptosis is essential for samples were placed in an H2S incuba- and 5 lL of Annexin V-PE and imme-
regulating tissues during embryogene- tion system and incubated in air con- diately placed in ice for 20 min. For
sis and for maintaining normal tining 5% CO2 mixed with H2S (50 ng/ each experiment, 2000 cells were ana-
homeostasis (9). Hydrogen sulfide may mL) for 24, 48 or 72 h for apoptosis lyzed using Guava EasyCyte flow
play a bifunctional role in cell survival; detection and 24 and 48 h for apoptotic cytometry. Data acquisition and
that is, while H2S inhibits apoptosis in pathway analysis. However, as a result analysis were performed using Guava
human polymorphonuclear neutroph- of diffusion, the amount of H2S in the CytoSoft software.
ils (10), lymphocytes (11) and cardiac medium was found to be only 18 ng/mL
myocytes (12), it may also stimulate (0.5 lmol/L), a much lower concentra-
Reactive oxygen species (ROS)
apoptosis in human aorta smooth tion than the one found in gingival
muscles cells (13), human lung fibro- crevicular fluids from periodontal Oxidative stress was assessed through
blasts (14) and/or pancreatic acinar gingival tissues, which is 1.9 mmol/L measurement of levels of ROS in the
cells (15). In the oral environment, H2S as reported by Persson (21). mitochondria using a red mitochon-
 H2S triggers apoptsis in gingival cells 33

drial superoxide indicator (MitoSOX, provided standards were pipetted into chem, San Diego, CA) was used. The
Invitrogen). After tryspinization, the wells precoated with a monoclonal assay uses a caspase-9 inhibitor, Leu-
cells were centrifuged for 5 min at antibody specific for cytochrome C. Glu-His-Asp-methyl-fluoromethylke-
1500 g. The supernatant was dis- After the cells were washed with PBS, tone (LEHD-FMK), which binds
carded, and the cells were resuspended an enzyme-linked monoclonal anti- irreversibly to activated caspase-9, and
in DulbeccoÕs PBS (with Ca2+ and body specific for cytochrome C was the fluorescent marker FITC. After
Mg2+) to a final concentration of added to all wells. Following another tryspinzation and centrifugation, the
106 cells/mL. A 100 lL aliquot sample washing step, a substrate solution was concentration of the cells was adjusted
was then added, along with 1 mM added to wells. The color developed in to 1 · 105 cells mL using the tech-
MitoSOX reagent stock solution and proportion to the amount of cyto- nique described for the caspase-3
900 lL DulbeccoÕs PBS (with Ca2+ chrome C bound in the first step. activity assay. The cells were then
and Mg2+) in a 1.5 mL test tube. The When color development was stopped, resuspended in 300 lL of washing
samples were incubated for 10 min at the optical density was determined buffer and analyzed by the Guava
37C and then analyzed by flow using a microplate reader (Bio-Rad EasyCyte flow cytometer using the
cytometry. Benchmark Plus, Bio-Rad Japan, To- FL-1 standard channel.
kyo, Japan) set to 450 nm with a
wavelength correction set to 540 nm.
Detection of mitochondrial Caspase-8 activity assay
membrane potential
To evaluate caspase-8 activity, a Cas-
Caspase-3 activity assay
Changes in mitochondrial membrane pase-8 Detection Kit (Calbiochem)
potential in H2S-exposed epithelial Caspase-3 activity was determined was used. A synthetic caspase-8
cells were detected by Guava EasyCyte using a Caspase-3 Detection Kit inhibitor, Ile-Glu-Thr-Asp-methyl-flu-
MitoPotential. The depolarization of (Oncogene Research Products, San oromethylketone (IETD-FMK), and
the membrane was evaluated through Diego, CA, USA). The assay uses fluo- FITC were used for this assay,
measurement of the uptake of a cat- rescein isothiocyanate (FITC) as a allowing direct detection with a flow
ionic dye, 5,5¢,6,6¢-tetrachloro-1,1¢,3, fluorescent marker, conjugated with cytometer. After tryspinzation and
3¢-tetraethyl-benzamidazolocarbo- Asp-Glu-Val-Asp-O-methyl-fluorom- centrifugation, the concentration of
cyanin iodide, commonly known as ethylketone (also known as DEVD- cells was adjusted to 1 · 105 cells mL
JC-1, into mitochondria. Collapse of FMK), a cell-permeable caspase peptide using the technique described for the
the membrane potential initiates the inhibitor. The FITC–DEVD-FMK caspase-3 activity assay. The cells were
early stages of the mitochondrial binds irreversibly to activated caspase-3 then resuspended in 300 lL of wash-
pathway-dependent apoptosis. In this in apoptotic cells, and its FITC label ing buffer and analyzed by the Guava
assay, JC-1 fluoresces either green or allows for direct detection by a flow EasyCyte flow cytometer using the
orange, depending on mitochondrial cytometer. After trypsinization and FL-1 standard channel.
membrane potential. The dye 7-AAD centrifugation, the cells were resus-
was used for monitoring cell mem- pended in PBS at a final concentration
Single-cell gel electrophoresis assay
brane permeability associated with late of 1 · 105 cells mL. A quantity of
apoptosis or necrotic cell death. After 300 lL of each sample, and a control Genomic DNA damage was detected
incubation with H2S, the cells were sample was combined with 1 lL FITC– using single-cell gel electrophoresis
trypsinized, pelleted and resuspended DEVD-FMK in microfuge tubes and (Comet Assay; Trevigen, Gaithers-
in 200 lL Dulbecco’s MEM together incubated for 1 h in a 37C incubator burg, MD, USA), based on the prin-
with 2 lL JC-1 and 2 lL 7-AAD. The with air containing 5% CO2. Sub- ciple that DNA has a highly organized
cells were incubated for 30 min in a sequently, the cells were centrifuged for structure which, in damaged DNA, is
37C CO2 incubator and then analyzed 5 min at 2500 g, the supernatant was disrupted. When placed in an electric
for fluorescence by a flow cytometer. removed, the cells were resuspended in field, undamaged DNA strands
500 lL of washing buffer (prepared migrate very slowly and remain con-
immediately before use), and the cells fined within the nucleoid, compared
Detection of cytochrome C
were again centrifuged for 5 min at with the small fragments of damaged
For analysis of cytochrome C release 2500 g. The cells were then resuspended DNA, which move much faster. When
into the cytosol, subcellular fraction- in 300 lL of washing buffer and viewed under the microscope, the cell is
ation was performed to separate the analyzed via a Guava EasyCyte flow shaped like a comet, its head corre-
mitochondria from the cytosol, and cytometer using the FL-1 standard sponding to the nuclear region and its
only the cytosolic fraction was used in channel. tail the damaged DNA fragments. The
the assay. The assay (Cytochrome C cells were suspended in PBS at a con-
ELISA, Calbiochem, San Diego, CA, centration of 1 · 105 cells/mL. Fur-
Caspase-9 activity assay
USA) employed a quantitative sand- ther, the cells were incorporated in
wich enzyme immunoassay technique. For the detection of caspase-9 levels, agarose gel (LMAgarose; Trevigen).
Controls, samples and manufacturer- a Caspase-9 Detection Kit (Calbio- They were lysed using a lysis solution
34 Calenic et al.

and then placed in an electrophoretic A 50 * 70

*
Control A Control

Early apoptotic
field at a constant voltage of 1 V/cm 40 Sample 60
Sample
*

cells (%)
* 50

ROS (%)
for 10 min. The slides holding the 30
40
samples (CometSlide; Trevigen) were 20 30
placed in 70% ethanol for 5 min. After 10 20
0 10
drying, the cells were stained using 24 h 48 h 72 h 0
Time 24 h 48 h
SYBR Green 1 (Trevigen). This was Time
followed by image analysis using a B B
50 60 *
fluorescence microscope. The images Control * *

necrotic cells (%)

Depolarisation (%)
Late apoptotic &
Control
40 Sample 50
were obtained with a digital camera Sample
30 40
and processed using imaging software 20 30
(TriTek CometScore, Sumerduck, VA, 10
20
10
USA). For assessing the amount of 0
24 h 48 h 72 h 0
DNA damage induced by H2S, the Time
24 h
Time
48 h
following parameters were analyzed: C
tail length, which indicates the distance Fig. 1. Hydrogen sulfide triggers apoptosis *

Cytochrome C (ng/mL)
0.25
Control
of damaged DNA migration from the in epithelial cells derived from human gin- 0.2 Sample
nucleoid; percentage of DNA in the giva. Apoptosis levels were detected by flow 0.15
*
tail, which expresses the proportion of cytometry after staining with Annexin V 0.1
total DNA present in the tail; and tail and 7-AAD. (A) Percentage of early apop- 0.05
moment, which represents the product totic cells. After both 24 and 48 h, a sig-
0
of two values, namely the percentage nificant difference was found between 24 h
Time
48 h

of DNA in the comet tail and the tail samples and their corresponding controls.
After 72 h, early apoptosis was decreased. Fig. 2. Mitochondrial changes. (A) Reac-
length.
The data show that after 48 h of incubation, tive oxygen species, a marker of cellular
the main biological event was early apop- stress, were significantly increased com-
Statistical analysis tosis, with very few late apoptotic and pared with control groups at each point of
necrotic cells. (B) Late apoptotic and analysis. (B) Mitochondrial membrane
Results from five independent experi- electrical gradient was disrupted after both
necrotic cells. After 24 and 48 h, the pres-
ments are presented as the ence of these cells was < 5% at each time 24 and 48 h of H2S incubation. (C) Release
means ± SD. Statistical analysis was point, while after 72 h the levels of necrosis of cytochrome C from mitochondria into
performed using one-way analysis of and late apoptosis were greatly increased. cytosol is a key event in the activation of the
variance (ANOVA). Statistical signifi- Each bar represents the mean ± SD of 5 intrinsic apoptotic pathway. After both 24
cance was accepted at p < 0.05. independent experiments (*p < 0.05 vs. and 48 h, release of cytochrome C into the
control). cytosol was significantly increased com-
pared with the corresponding control
Results groups. Each bar represents the
late apoptosis was significantly mean ± SD of 5 independent experiments
Hydrogen sulfide-induced apoptosis (*p < 0.05 vs. control).
increased (40.6 ± 8.4 vs. 3.5 ± 1.9%;
The percentage of apoptotic cells p < 0.01, ANOVA), while early
increased in a time-dependent manner. apoptosis decreased to 10% of the
A significant difference in early apop- total cell number (Fig. 1B). JC-1, a fluorescent dye that can
totic levels was found between test monitor mitochondrial membrane
groups (p < 0.01, ANOVA). In addi- depolarization. The percentage of
Mitochondrial changes
tion, the number of early apoptotic membrane-depolarized cells was
cells was significantly increased after The mitochondrial apoptotic pathway significantly increased at each time
both 24 and 48 h compared with their is marked by profound changes in point compared with control groups
control groups (24.5 ± 5.7 vs. mitochondria. The presence of ROS in (45.8 ± 11 vs. 14.5 ± 5% at 24 h and
5.5 ± 2.3% after 24 h and 41.5 ± 8.9 the mitochondria was detected 38.1 ± 10.5 vs. 19.6 ± 5.5% at 48 h;
vs. 4.1 ± 0.8% at 48 h of incubation; according to the number of MitoSOX- p < 0.01, ANOVA; Fig. 2B). A key
p < 0.01, ANOVA; Fig. 1A). After 24 positive cells in five independent event in the intrinsic apoptotic
and 48 h, late apoptotic and necrotic experiments. A significant difference pathway is the release of cytochrome C
events were found to be < 5%. The was observed after both 24 and 48 h from the mitochondria. A significant
data support the observation that the between H2S-incubated samples and increment of cytochrome C levels into
cells were in different stages of early their respective control groups cytosol was observed after both 24 and
apoptosis after 2 days of incubation (42.9 ± 9.5 vs. 7.6 ± 3.6% at 24 h 48 h (0.12 ± 0.02 vs. 0.02 ± 0.01 ng/
with H2S, and very few cells were in a and 56.6 ± 4.1 vs. 9.1 ± 3.7% at mL at 24 h and 0.21 ± 0.02 vs.
late apoptotic or a necrotic stage. After 48 h; p < 0.01, ANOVA; Fig. 2A). 0.02 ± 0.01 ng/mL at 48 h; p < 0.05,
72 h, the percentage of necrosis and Another line of assessment relied on ANOVA; Fig. 2C).
 H2S triggers apoptsis in gingival cells 35

extrinsic pathway (5.5 ± 2.1 vs. A 25

Caspase activities

Tail length (pixels)

Control
4.5 ± 2.3% at 24 h and 7.0 ± 1.5 vs. 20
*
Sample
Following H2S incubation, caspase-9, 2.8 ± 2.1% at 48 h; Fig. 3C). 15 *
an upstream regulator of the apoptotic 10
process, was significantly activated 5
Genomic DNA damage
compared with control groups (18.7 ± 0
24 h 48 h
3.8 vs. 5.3 ± 2.5% at 24 h and 51.1 ± Using single-cell gel electrophoresis, we Time
2.8 vs. 4.6 ± 1.1% at 48 h; p < 0.01, also observed the genotoxicity of H2S
B 25 *
ANOVA; Fig. 3A). Significant differ- by evaluating the DNA ÔcometÕ tail Control

DNA in tail (%)

20 Sample *
ences were also found between caspase- shape. Tail length, percentage of DNA
15
3 activity levels in test groups compared in tail, and their product, tail moment,
10
with their control groups (25.6 ± 8.4 are parameters correlated with the
5
vs. 3.3 ± 1.1% at 24 h and 46.1 ± 8 number of DNA strand breaks. All
0
vs. 3.0 ± 1.2% at 48 h; p < 0.05, parameters were increased, suggesting 24 h Time 48 h
ANOVA; Fig. 3B). At the same time, significantly more DNA strand breaks
caspase-8 levels remained very low and in test groups than in the correspond- C 5 *
Control

Tail moment (%)

were comparable to control groups, ing control groups: for tail length, 4 Sample
suggesting that the apoptotic process 7.82 ± 2.94 vs. 1.97 ± 0.06 at 24 h 3
*
initiated by H2S is independent of the and 16.71 ± 6.86 vs. 0.97 ± 0.06 at 2

48 h (p < 0.05, ANOVA; Fig. 4A); 1

0
A
for DNA in tail, 12.92 ± 4.12 vs. 24 h 48 h
Time
2.26 ± 0.08% at 24 h and 16.52 ±
Caspase-9 activity (%)

70
Control *
60 4.14 vs. 2.73 ± 0.57% at 48 h (p <
50
Sample Fig. 4. Genomic DNA damage following
40
0.05; Fig. 4B); and for tail moment, hydrogen sulfide incubation. Genomic
30 * 1.51 ± 0.51 vs. 0.09 ± 0.05% at 24 h DNA damage was detected using single-cell
20
and 3.95 ± 0.72 vs. 0.11 ± 0.07% at gel electrophoresis and quantified using
10
0 48 h (p < 0.05, ANOVA; Fig. 4C). three different parameters. (A) Tail length,
24 h 48 h
Time expressing distance of damaged DNA
B migration from the nucleoid, was signifi-
Discussion
Caspase-3 activity (%)

60 * cantly increased after 24 and 48 h compared

Control
50
Sample
* Apoptosis, or programmed cell death, with control cells. (B) Percentage of DNA
40
30 is a normal process in the development in tail, showing the proportion of total
20 of multicellular organisms. Cells initi- DNA in tail, was increased after both 24
10 ate the apoptotic process in response to and 48 h. (C) Tail moment, representing the
0 product of the first two parameters, was
24 h 48 h a wide variety of stimuli; cell death is
Time significantly increased at each time point
thus regulated and homeostasis main-
C compared with control groups. Each bar
tained. Dysfunctions during the apop-
Caspase-8 activity (%)

50
Control represents mean ± SD of 5 independent
40 Sample totic process may lead to either
experiments; 75 nuclei analyzed per experi-
30 prolonged cell survival or premature
ment (*p < 0.05 vs. control).
20 cell death. Therefore, excessive or
10 insufficient apoptosis contributes to a
0 wide range of human diseases, such as Human gingival fibroblasts infected
24 h 48 h
Time cancer, viral infections and degenera- with Porphyromonas gingivalis for
tive disorders (23,24). 24–36 h showed activation of the mito-
Fig. 3. Caspase activities. For caspase
Studies of the relationship between chondrial pathway and nuclear DNA
activity detection, FITC, a fluorescent
apoptosis and the development of perio- degradation (28), while apoptotic
marker allowing direct detection by a flow
dontal disease have shown that apop- inducers, such as butyric acid, caused
cytometer, was conjugated with caspase
tosis plays a critical role in host immune apoptosis in gingival fibroblasts
peptide inhibitors for caspase-9, -3 or -8.
(A,B) Caspase-9 and -3 levels, respectively,
response and inflammation, both of isolated from inflamed periodontal
were significantly increased compared with which are involved in periodontitis. lesions (29). Gingival tissues in chronic
their control groups, suggesting that H2S- Previous studies have reported apopto- periodontitis have increased apoptotic
induced apoptosis is activated via an sis of polymorphonuclear leukocytes in activity (16,28,30), while severe perio-
intrinsic apoptotic pathway. (C) Caspase-8 gingival tissues with periodontitis dontitis is marked by a rise in the
activity remained low, comparable to the (25,26). Periodontal disease leads to number of pro-apoptotic genes (31).
control group, proving the inactivity of the progressive degradation of connective Thus, apoptosis is actively involved in
death-receptor apoptotic pathway. Each bar tissue. Apoptosis among periodontal periodontal pathogenesis.
represents mean ± SD of 5 independent ligament fibroblasts increased in the In healthy gingiva, oral epithelium
experiments (*p < 0.05 vs. control). early stages of periodontitis (27). plays a key role as a barrier against
36 Calenic et al.

pathogens or toxic compounds. The of early apoptosis. Apoptotic levels especially after 48 h of incubation. In
maintenance of this epithelial barrier were also detected after 72 h. Necrosis turn, cytochrome C release initiates the
is therefore extremely important for and late apoptosis was dramatically apoptotic caspase cascade by activat-
the preservation of normal gingival increased at 72 h, while early apoptotic ing caspase-9, an initiator caspase
structure and function. However, levels decreased to one-quarter of the responsible for the upstream regulation
during the progression of periodontal 48 h incubation levels. Therefore, to of apoptosis (38,39). In this study,
disease, this barrier can be affected. analyze the apoptotic pathway we caspase-9 levels were markedly
Several studies have focused on the employed 24 and 48 h incubation times, increased. Exposure to H2S also
relationship between oral malodorous since at both these time points the main increased the activity of caspase-3, an
compounds and this epithelial barrier. cell death event was early apoptosis. executioner caspase responsible for the
A study conducted by Tonzetich & Ng The two main mechanisms in the downstream regulation of the apopto-
used a porcine model for gingival apoptotic process involve activation of tic process (40). Meanwhile, caspase-8
crevicular epithelia to show that an intrinsic pathway, in which the levels remained low and were compa-
exposure to concentrations of VSCs mitochondrion plays a central role, rable to control levels. Caspase-8 is an
much lower than those found in and activation of an extrinsic pathway, initiator caspase activated in the
periodontal pockets caused increased involving a receptor–ligand-mediated receptor–ligand-mediated apoptosis
permeability of sublingual nonkerati- mechanism (36). To distinguish (41), and its inactivity suggests that the
nized mucosa (32). Another study, between the two pathways, we ana- extrinsic pathway is not involved in
also conducted by Tonzetich and lyzed mitochondrial changes. Increased H2S-induced apoptosis.
others, shows that VSCs can cause production of ROS in mitochondria Taken together, these results suggest
total disruption of the basement causes disruption of the electrochemi- that H2S induces apoptosis by activat-
membrane (33). Furthermore, VSCs cal gradient across the inner mito- ing the mitochondrial intrinsic path-
induced an important decrease in the chondrial membrane, which then way (Fig. 5). In a previous study, H2S
collagen content of the VSC-exposed activates the apoptotic process (37). was shown to induce genomic DNA
cell cultures (5,6). Other studies We found that H2S increased ROS and damage in human gingival fibroblasts
showed that increased concentrations caused a significant loss of the (20). Using a single-cell gel electro-
of CH3SH have an inhibitory effect on mitochondrial inner transmembrane phoresis, we also observed an incre-
both cell growth and proliferation in potential. Collapse of this potential is ment in the number of DNA strand
human oral epithelial cell lines (34). associated with early stages of apop- breaks at the genomic level, proving
Hydrogen sulfide was also shown to tosis; moreover, it leads to a key event the genotoxic effect of H2S. Genomic
induce cell cycle arrest in oral epithe- in the mitochondrial pathway of apop- DNA damage suggests that other
lial cells, which may contribute to tosis, that is, the release of cyto- molecular pathways, such as the p53
delayed epithelial repair (35). How- chrome C from the mitochondria pathway, might be involved in the
ever, the apoptotic effect of H2S on intermembrane into cytosol. In apoptotic process and that H2S may
oral epithelial cells and the molecular response to H2S, the release of cyto- have pathological effects on human
mechanisms of this process remain chrome C was significantly increased, gingiva at the genomic level.
unexplained. In the healthy epithe-
lium, apoptosis occurs after basal Intrinsic pathway H2S Extrinsic pathway Extracellular
keratinocytes mature and then differ- Receptor
entiate. In the oral cavity, an enhanced Intracellular
apoptotic process may attenuate the Mitochondria
barrier functions of epithelium against ROS Caspase-8
periodontally pathogenic strains of inactive

microorganisms or their products.

In the present study, we demon- m
strated that H2S induces apoptosis in Cytochrome C Caspase-9 Caspase-3 DNA
epithelial cells derived from human active active fragmentation
gingiva, and we identified the mecha-
nisms underlying activation of the Apoptosis
apoptotic signaling pathway. After 24
and 48 h incubation in a physiological Fig. 5. Hydrogen sulfide triggers mitochondrion-dependent apoptosis via the intrinsic
concentration of H2S, more than one- pathway. Hydrogen sulfide increases levels of ROS in the mitochondria and collapses the
third of the cells were found to be in mitochondrial membrane potential (DYm). These events are followed by release of cyto-
different stages of early apoptosis (20), chrome C from the mitochondria intermembrane into the cytosol. Cytochrome C activates
whereas late apoptotic and necrotic caspase-9, an upstream caspase, which in turn activates caspase-3. Caspase-3 activation
levels were lower than 5%. Our data ultimately leads to DNA damage and apoptosis. Caspase-8 inactivity suggests that the
show that after 2 days of incubation, the extrinsic pathway is not involved in the process. For the techniques employed please refer to
main events caused by H2S are features the ÔMaterial and methodsÕ section.
 H2S triggers apoptsis in gingival cells 37

cells via activation of mitogen-activated 27. Ekuni D, Tomofuji T, Yamanaka R,

Acknowledgements protein kinases and caspase-3. FASEB J Tachibana K, Yamamoto T, Watanabe T.
2004;18:75–80. Initial apical migration of junctional epi-
This research was supported by the
14. Baskar R, Li L, Moore PK. H2S induces thelium in rats following application
grant-in-aid no. 203905380001 from the DNA damage and changes in apoptotic of lipopolysaccharide and proteases.
Japanese Ministry of Education, Cul- gene expression in human lung fibroblast J Periodontol 2005;76:43–48.
ture, Sports, Science and Technology. cells. FASEB J 2007;21:247–255. 28. Urnowey S, Ansai T, Bitko V, Nakayama
15. Cao Y, Adhikari S, Ang AD, Moore PK, K, Takehara T, Barik S. Temporal acti-
Bhatia M. Mechanism of induction of vation of anti- and pro-apoptotic factors
References pancreatic acinar cell apoptosis by H2S. in human gingival fibroblasts infected with
1. Oho T, Yoshida Y, Shimazaki Y et al. Am J Physiol Cell Physiol 2006;291:C503– the periodontal pathogen, Porphyromonas
Characteristics of patients complaining of C510. gingivalis: potential role of bacterial pro-
halitosis and the usefulness of gas chro- 16. Leung WK, Wu Q, Hannam PM, teases in host signaling. BMC Microbiol
matography for diagnosing halitosis. Oral McBride BC, Uitto V-J. Treponema den- 2006;8:6.
Surg Oral Med Oral Pathol Oral Radiol ticola may stimulate both epithelial pro- 29. Kurita-Ochiai T, Seta S, Suzuki N et al.
Endod 2001;91:531–534. liferation and apoptosis through MAP Butyric acid induces apoptosis in inflamed
2. Yaegaki K, Sanada K. Volatile sulfur kinase signal pathways. J Periodontal Res fibroblasts. J Dent Res 2008;87:51–55.
compounds in mouth air from clinically 2002;37:445–455. 30. Ekuni D, Tomofuji T, Yamanaka R,
healthy subjects and patients with perio- 17. Jewett A, Hume WR, Le H et al. Induc- Tachibana K, Yamamoto T, Watanabe T.
dontitis. J Periodontal Res 1992;27:233– tion of apoptotic cell death in peripheral Caspase activation is involved in chronic
238. blood mononuclear and polymorpho- periodontitis. FEBS Lett 2005;579:5559–
3. Rizzo A. Histologic and immunologic nuclear cells by an oral bacterium, Fuso- 5564.
evaluation of antigen penetration into oral bacterium nucleatum. Infect Immun 2000; 31. Bulut S, Uslu H, Ozdemir BH, Bulut OE.
tissue after topical application. J Perio- 68:1893–1898. Expression of caspase-3, p53, and Bcl-2 in
dontol 1970;41:210–212. 18. Gamonal J, Bascones A, Acevedo A, generalized aggressive periodontitis. Head
4. Ratkay LG, Waterfield JD, Tonzetich J. Blanco E, Silva A. Apoptosis in chronic Face Med 2006;2:17.
Stimulation of enzyme and cytokine pro- adult periodontitis analyzed by in situ 32. Ng W, Tonzetich J. Effect of hydrogen
duction by methyl mercaptan in human DNA breaks, electron microscopy, and sulfide and methyl mercaptan on the per-
gingival fibroblast and monocytes cell immunochemistry. J Periodontol 2001; meability of oral mucosa. J Dent Res
cultures. Arch Oral Biol 1995;40:337–344. 72:517–525. 1984;63:994–997.
5. Johnson P, Yaegaki K, Tonzetich J. Effect 19. Maurizio S, Cortellini D, Lang N. In situ 33. Johnson PW, Ng W, Tonzetich J. Modu-
of methyl mercaptan on synthesis and detection of apoptosis at sites of chronic lation of human gingival fibroblast cell
degradation of collagen. J Periodontal Res bacterially induced inflammation in human metabolism by methyl mercaptan. J Perio-
1996;31:323–329. gingiva. Infect Immun 1998;66:5190–5195. dontal Res 1992;27:476–483.
6. Johnson P, Yaegaki K, Tonzetich J. Effect 20. Yaegaki K, Qian W, Murata T et al. Oral 34. Setoguchi T, Machigashira M, Yamamoto
of volatile thiol compounds on protein malodorous compound causes apoptosis M et al. The effects of methyl mercaptan
metabolism by human gingival fibroblasts. and genomic DNA damage in human on epithelial cell growth and proliferation.
J Periodontal Res 1992;27:553–561. gingival fibroblasts. J Periodontal Res Int Dent J 2002;52(Suppl 3):241–246.
7. Setoguchi T, Machigashira M, Yamamoto 2008;43:391–9. 35. Takeuchi H, Setoguchi T, Machigashira
M et al. The effects of methyl mercaptan 21. Persson S. H2S and methyl mercaptan in M, Kanbara K, Izumi Y. Hydrogen sul-
on epithelial cell growth and proliferation. periodontal pockets. Oral Microbiol fide inhibits cell proliferation and induces
Int Dent J 2002;52:241–246. Immunol 1992;7:378–379. cell cycle arrest via an elevated p21 Cip1
8. Yaegaki K. Oral malodor and periodontal 22. van England M, Nieland LJ, Ramaekers level in Ca9-22 cells. J Periodontal Res
disease. In: Rosenberg M, ed. BAD FC, Schutte B, Reutelingsperger CP. 2008;43:90–95.
BREATH: Research Perspectives. Tel Annexin V-affinity assay: a review on an 36. van Cruchten S, van den Broeck W.
Aviv: Ramot Publishing-Tel Aviv Uni- apoptosis detection system based on Morphological and biochemical aspects of
versity, 1995:87–108. phosphatidylserine exposure. Cytometry apoptosis, oncosis and necrosis. Anat
9. Raff M. Cell suicide for beginners. Nature 1998;31:1–9. Histol Embryol 2002;31:214–223.
1998;396:119–122. 23. Debatin KM. Apoptosis pathways in 37. Green DR, Reed JC. Mitochondria and
10. Mariggio MA, Minunno V, Riccardi S, cancer and cancer therapy. Cancer Immu- apoptosis. Science 1998;281:1309–1312.
Santacroce R, De Rinaldis P, Fumarulo nol Immunother 2004;53:153–159. 38. Hill MM, Adrain C, Martin SJ. Portrait of
R. Sulfide enhancement of PMN apopto- 24. Friedlander RM. Apoptosis and caspases a killer: The mitochondrial apoptosome
sis. Immunopharmacol Immunotoxicol in neurodegenerative diseases. N Engl J emerges from the shadows. Mol Interv
1998;20:399–408. Med 2003;348:1365–1375. 2003;3:19–26.
11. Rinaldi L, Gobbi G, Pambianco M, 25. Sawa T, Nishimura F, Ohyama H, 39. Bossy-Wetzel E, Newmeyer DD, Green
Micheloni C, Mirandola P, Vitale M. H2S Takahashi K, Takashiba S, Murayama Y. DR. Mitochondrial cytochrome-c release
prevents apoptosis of human PMN via In vitro induction of activation-induced in apoptosis occurs upstream of
inhibition of p38 and caspase 3. Lab Invest cell death in lymphocytes from chronic DEVD-specific caspase activation and
2006;86:391–397. periodontal lesions by exogenous Fas independently of mitochondrial trans-
12. Sivarajah A, Collino M, Yasin M et al. ligand. Infect Immun 1999;67:1450–1454. membrane depolarization. EMBO J 1998;
Anti-apoptotic and anti-inflammatory 26. Koulouri O, Lappin DF, Radvar M, 17:37–49.
effects of hydrogen sulfide in a rat model Kinane DF. Cell division, synthetic 40. Porter AG, Janicke RU. Emerging roles of
of regional myocardial I/R. Shock capacity and apoptosis in periodontal caspase-3 in apoptosis. Cell Death Differ
2008;[Epub ahead of print]. lesions analyzed by in situ hybridization 1999;6:99–104.
13. Yang G, Sun X, Wang R. H2S induced- and immunohistochemistry. J Clin 41. Hengartner M. The biochemistry of
apoptosis of human aorta smooth muscle Periodontol 1990;26:552–559. apoptosis. Nature 2000;407:770–776.
 A. de Jongh, C. de Baat, M. Horstman, A.J. van Wijk Onderzoek en wetenschap

Subjectieve mondgeurperceptie en sociaal

functioneren
In een onderzoek werd de invloed van subjectieve mondgeurper- Wat weten we?
ceptie op het sociaal functioneren onderzocht. Een representatieve In het sociale verkeer tussen mensen bestaat een taboe op
steekproef van 1.082 personen van de Nederlandse bevolking van het bespreekbaar maken van halitose.
16 jaar en ouder werd hiervoor bevraagd. Gemiddeld beoordeelden
de deelnemers hun mondgeur op een schaal van 0-100 met een Wat is nieuw?
score van 66,8. Van alle deelnemers gaf 4,2% aan hun mondgeur Sommige mensen dichten zichzelf een frisse mondgeur
als ‘onfris’ te beschouwen. Ongeveer twee derde gaf aan rekening toe, maar bewaren in een gesprek desondanks afstand tot
te houden met het feit dat een ander tijdens een eerste ontmoe- een ander. Dit kan te maken hebben met een algemene
ting zijn of haar mondgeur zou kunnen ruiken. Deelnemers die onzekerheid in het sociaal functioneren.
hun eigen mondgeur als ‘onfris’ beoordeelden, bleken significant
meer afstand te houden tot een andere persoon dan deelnemers Praktijktoepassing
die hun eigen mondgeur als ‘fris’ beoordeelden. Opvallend was Mondzorgverleners kunnen een bijdrage leveren aan het
een subgroep van deelnemers die zichzelf een frisse mondgeur doorbreken van het taboe op halitose door het probleem
toedichtten, maar toch altijd afstand tot een ander bewaarden. De bespreekbaar te maken in hun mondzorgpraktijk.
subjectieve mondgeurperceptie bleek voor sommige deelnemers
een potentieel belemmerende factor tijdens sociaal functioneren.
chosociale gevolgen blijft, is recent aangetoond (De Jongh
Jongh A de, Baat C de, Horstman M, Wijk AJ van. Subjectieve mondgeurperceptie et al, 2012). Ofschoon minder sterk dan zweetlucht, wordt
en sociaal functioneren halitose door de Nederlandse bevolking in sociaal opzicht
Ned Tijdschr Tandheelkd 2013; 120: 194-198 als een van de grootste ‘afknappers’ beschouwd. Het is dus
doi: 10.5177/ntvt.2013.04.12198 niet ondenkbaar dat mensen uit onzekerheid of angst voor
halitose - en de mogelijke negatieve sociale consequenties
Inleiding ervan - hun sociaal functioneren hierop aanpassen. Of-
Halitose is een probleem waarvan betrekkelijk veel men- schoon deze redenering alleszins voorstelbaar is, ont-
sen last hebben en waarvan het voorkomen op 10-30% breekt echter informatie over de invloed van subjectieve
wordt geschat (Meskin, 1996; Tangerman, 2002; Liu et al, mondgeurperceptie op het sociaal functioneren.
2006; Van den Broek et al, 2007; Laine et al, 2011). Een Het doel van het onderhavige onderzoek was vast te
van de zaken die halitose compliceert is dat mensen er stellen welke subjectieve mondgeurperceptie Nederlan-
doorgaans zelf betrekkelijk weinig van merken. Er zijn wel ders in het algemeen hebben en in welke mate men hier-
testen om de eigen mondgeur te beoordelen, bijvoorbeeld mee tijdens het sociaal functioneren rekening houdt. Als
de zogeheten pols-liktest, waarbij men met een zo groot afgeleide hiervan werd de hypothese getoetst dat de sub-
mogelijk deel van de tong over de pols likt, het vocht even jectieve mondgeurperceptie samenhangt met de mate
laat drogen en vervolgens aan de pols ruikt. Dergelijke tes- waarin men hiermee rekening houdt tijdens het sociaal
ten komen echter normaal pas aan de orde als degene die functioneren.
het betreft zich van het probleem bewust is, bijvoorbeeld
omdat hij hierop door iemand is geattendeerd. Dat laatste Materiaal en methode
blijkt betrekkelijk weinig te gebeuren. Zo lijkt een taboe te Het onderzoek werd uitgevoerd via www.panelwizard. Uit
bestaan om een ander op halitose te wijzen. In een recent een bestand van bijna 20.000 panelleden van 16 jaar en
onderzoek, waarbij werd gebruikgemaakt van een repre- ouder werd een steekproef getrokken van 1.586 personen
sentatieve steekproef van de Nederlandse bevolking van die werden uitgenodigd aan het onderzoek deel te nemen.
16 jaar en ouder, werd gevonden dat naarmate de sociale Uiteindelijk participeerden 1.083 personen, een respons
afstand tussen personen toeneemt, de neiging om iemand van 68%. Alle onderzoeksvragen werden beantwoord door
op halitose te attenderen afneemt. Slechts 40% van de 1.082 personen. Deze steekproef was representatief voor
mensen is bijvoorbeeld bereid of geneigd een collega op de Nederlandse bevolking van 16 jaar en ouder ten aanzien
het werk op halitose aan te spreken. In geval van een onbe- van geslacht, leeftijd, opleidingsniveau, gezinssituatie en
kende heeft minder dan 7% deze bereidheid of neiging (De arbeidsparticipatie.
Jongh et al, 2012). De deelnemers kregen via internet vragen voorgelegd
Kortom, het complexe van halitose is dat men het zelf en kregen 4 dagen de tijd deze te beantwoorden. Bij de eer-
nauwelijks opmerkt, een ander er niets over zegt en beiden ste vraag konden zij op een lijnstuk van 0 (‘helemaal niet
er daarom last van blijven houden. Dat dit niet zonder psy- fris’) tot 100 (‘heel erg fris’) hun gemiddelde subjectieve

Nederlands Tijdschrift voor Tandheelkunde 194 120 | april 2013

 De Jongh e.a.: Mondgeurperceptie en sociaal functioneren

Achtergrondvariabele Geslacht Resultaten

Mannen (n = 561) Vrouwen (n = 522) Totaal De groep van 1.083 deelnemers be-
stond voor 48,2% uit vrouwen. In ta-
Leeftijd bel 1 wordt een overzicht gegeven van
16-29 jaar 19,8 20,7 20,2 hun leeftijd, opleidingsniveau, gezins-
30-39 jaar 20,1 16,1 18,2 situatie en arbeidsparticipatie uitge-
40-49 jaar 18,4 17,4 17,9 splitst voor mannen en vrouwen.
50-59 jaar 17,6 15,5 16,6 Mannen en vrouwen waren gemiddeld
60 jaar en ouder 24,1 30,3 27,1 ongeveer even oud (Student’s t-toets;
p = 0,15). Het opleidingsniveau van de
Gezinssituatie vrouwen bleek relatief vaker laag en
Alleenstaand 19,4 20,7 20,0 van de mannen relatief vaker hoog
Samenwonend/gehuwd zonder kinderen 50,1 43,9 47,1 (Mann-Whitney U test; p < 0,01). De
Samenwonend/gehuwd met kinderen 30,5 35,4 32,9 vrouwen werkten relatief gezien veel
minder vaak in een volledig dienstver-
Opleidingsniveau band en veel vaker in deeltijd of hele-
Laag 30,1 42,9 36,3 maal niet (chi-kwadraattest; p <
Middelbaar 40,1 35,4 37,9 0,001). Verder waren mannen en
Hoog 29,8 21,6 25,8 vrouwen gelijk verdeeld over de ver-
schillende leeftijdscategorieën (chi-
Arbeidsparticipatie kwadraattest; p = 0,79) en over de
Geen 33,5 57,5 45,0 verschillende gezinssituaties (chi-
Fulltime 57,6 11,5 35,4 kwadraattest; p = 0,16) (tab.1).
Parttime 8,9 31,0 19,6 Gemiddeld beoordeelden de deel-
nemers hun mondgeur met een score
Tabel 1. Achtergrondvariabelen van de 1.083 deelnemers uitgesplitst naar geslacht (%). van 66,8 ± 17,2. Van hen vond 4,2%
de eigen mondgeur onfris, terwijl ruim
mondgeurperceptie weergeven aan de hand van de volgen- de helft (56,3%) hun mondgeur fris vond. De hoogte van
de vraag: “Sommige mensen hebben een frisse mondgeur en deze score bleek niet statistisch significant samen te han-
anderen niet. Hoe denkt u in het algemeen over uw eigen gen met het geslacht (p > 0,05).
mondgeur?”. Om de reacties te kunnen interpreteren, werd Er werd een statistisch significant verschil in subjec-
een aantal arbitraire grenswaarden gehanteerd: een score tieve mondgeurperceptie gevonden tussen de deelnemers
tot en met 30 werd beschouwd als ‘onfris’ en een score van de 4 leeftijdsgroepen (ANOVA; p < 0,001). Deelnemers
vanaf 70 als ‘fris’. van 60 jaar of ouder beoordeelden hun subjectieve mond-
De tweede vraag had betrekking op het sociaal functio- geurperceptie gemiddeld beter dan de jongeren (71,5 ±
neren van de deelnemers: “Als u iemand voor de eerste keer 16,4), terwijl de andere leeftijdsgroepen op deze variabele
ontmoet, houdt u er dan rekening mee dat iemand uw mond- niet van elkaar verschilden (65,1 ± 17,2).
geur zou kunnen ruiken, bijvoorbeeld door de afstand die u tot Ook de gezinssituatie bleek samen te hangen met de
die persoon bewaart?” De 5 antwoordcategorieën waren: ik subjectieve mondgeurperceptie, waarbij deelnemers van
houd er helemaal geen/bijna geen/soms/vaak/altijd reke- gezinnen met kinderen relatief het slechtst scoorden
ning mee. (ANOVA; p = 0,03). Opleidingsniveau en arbeidsparticipa-
Van de deelnemers waren al diverse achtergrondvariabe- tie bleken niet samen te hangen met de subjectieve mond-
len bekend, waaronder geslacht, leeftijd, opleidingsniveau, geurperceptie (ANOVA; p > 0,05).
gezinssituatie en arbeidsparticipatie. Het opleidingsniveau, Ongeveer de helft van de deelnemers (50,2%) gaf aan
gedefinieerd als de hoogst gevolgde opleiding, werd onder- tijdens een eerste sociale ontmoeting soms tot vaak reke-
scheiden in laag (geen onderwijs, alleen basisonderwijs, ning te houden met het feit dat de ander zijn mondgeur
lbo/vbo/vmbo, mavo, of eerste 3 jaar havo of vwo), middel- zou kunnen ruiken, bijvoorbeeld door het bewaren van een
baar (mbo, havo, vwo en wo-propedeuse) en hoog (hbo/wo- zekere afstand. Ongeveer 15% zei zelfs hier altijd rekening
doctoraal of master). Als gezinssituatie werd onderscheid mee te houden. Vrouwen gaven in vergelijking met man-
gemaakt tussen alleenstaand en samenwonend/gehuwd nen (19% versus 12%) statistisch significant vaker aan
zonder of met kinderen. De arbeidsparticipatie werd onder- hiermee altijd rekening te houden (chi-kwadraattoets; p =
scheiden in: geen, volledig dienstverband en deeltijdaan- 0,001).
stelling. Het al dan niet rekening houden met subjectieve mond-
De onderzoeksgegevens werden ingevoerd en geanaly- geur tijdens een eerste sociale ontmoeting bleek ook af-
seerd in het computerprogramma Statistical Package for hankelijk te zijn van de leeftijd: deelnemers jonger dan 60
Social Sciences (SPSS) 18. Voor alle analyses werd een sig- jaar gaven aan hiermee vaker (‘soms tot altijd’) rekening te
nificantieniveau van 0,05 gehanteerd. houden dan de deelnemers van 60 jaar en ouder, namelijk

Nederlands Tijdschrift voor Tandheelkunde 195 120 | april 2013

De Jongh e.a.: Mondgeurperceptie en sociaal functioneren O n d e r zo e k e n w e t e n s c h a p

Aantal Gemiddelde ± *

1. Ik houd er helemaal geen rekening mee 155 73,8 ± 15,8 2, 3, 4, 5

2. Ik houd er bijna geen rekening mee 218 69,6 ± 15,4 1, 3, 4
3. Ik houd er soms rekening mee 305 64,3 ± 15,5 1, 2
4. Ik houd er vaak rekening mee 238 62,5 ± 18,6 1, 2, 5
5. Ik houd er altijd rekening mee 166 67,3 ± 19,2 1, 4
Totaal 1.082 66,8 ± 17,2

Tabel 2. Gemiddelden en standaarddeviaties (±) van de subjectieve mondgeurperceptie (0-100) ingedeeld in 5 groepen afhankelijk van de mate waarin de
deelnemers aangaven rekening te houden met hun mondgeur met daarbij de significante verschillen tussen de groepen onderling (*).

69,9% versus 53,8% (chi-kwadraattoets; p < 0,01). Ook het trouwbaarheidsinterval: 0,97-0,98). Kortom, naarmate
opleidingsniveau bleek hierbij een rol te spelen. Zo gaven men de subjectieve mondgeur als frisser omschreef, hield
deelnemers met een laag opleidingsniveau vaker aan altijd men er minder vaak geen tot bijna geen rekening mee. Al-
rekening te houden met hun mondgeur tijdens een eerste leen de variabele leeftijd kon als significante voorspeller
ontmoeting dan mensen met een hoog opleidingsniveau, aan het model worden toegevoegd, maar dit bleek geen
namelijk 18% versus 14% (chi-kwadraattoets; p = 0,018). verandering in odds ratio tussen subjectieve mondgeur en
Deelnemers werden ook vergeleken ten aanzien van de afstand houden tot gevolg te hebben (odds ratio = 0,98;
mate waarin zij aangaven rekening te houden met een 95% betrouwbaarheidsinterval: 0,97-0,99).
eventuele onfrisse mondgeur. Mensen die meenden een
onfrisse mondgeur te hebben, gaven aan hier vaker reke- Discussie
ning mee te houden dan mensen die meenden een frisse De resultaten van dit onderzoek lieten zien dat bij een sco-
mondgeur te hebben (Mann-Whitney U test; Z = - 2,53; p ringsmethode voor subjectieve mondgeurperceptie met
= 0,011). Er bleek een statistisch significant verschil in arbitraire grenzen van 30 of lager voor ‘onfris’ en 70 of ho-
subjectieve mondgeurperceptie te bestaan tussen de groe- ger voor ‘fris’ slechts een zeer klein deel van de deelnemers
pen deelnemers die in het sociaal functioneren ‘geen’, ‘bij- hun mondgeur als niet fris bestempelde. Daarbij beoor-
na geen’, ‘soms’, ‘vaak’, ‘altijd’ rekening hielden met het deelden vrouwen in vergelijking met mannen hun mond-
feit dat anderen hun mondgeur zouden kunnen ruiken geur als iets frisser. De gevonden prevalentie van
(ANOVA; p < 0,001). Kortom, naarmate mensen hun subjectieve mondgeurperceptie is beduidend lager dan die
mondgeur als slechter beoordeelden, had dat meer conse- van objectief vastgestelde halitose die in de literatuur
quenties voor hun gedrag. Zij gaven aan meer afstand tot wordt geschat op 10-30% (Meskin, 1996; Tangerman,
een persoon te nemen wanneer ze die voor het eerst ont- 2002; Liu et al, 2006; Van den Broek et al, 2007; Laine et
moeten dan mensen die meenden dat hun mondgeur fris al, 2011). Deze discrepantie is niet onlogisch gezien het
was (tab. 2). Een uitzondering op deze bevinding bleek de feit dat mensen nu eenmaal slecht in staat zijn hun eigen
groep die aangaf ‘altijd’ rekening te houden met mondgeur mondgeur te beoordelen. Er is weinig vergelijkbaar onder-
omdat deze 166 deelnemers (15,3%) meenden een goede zoek naar subjectieve mondgeurperceptie. In een onder-
mondgeur te hebben. Om deze uitkomst beter te kunnen zoek onder bewoners van de Zwitserse stad Bern (n = 419),
duiden, werd binnen de groep ‘altijd’ een frequentieanaly- waarbij werd gebruikgemaakt van een vierpuntenschaal
se van de subjectieve mondgeurperceptie uitgevoerd. Dit voor (on)frisse mondgeur, gaf 5% van de ondervraagden
leverde de volgende aanvullende informatie op. Binnen aan vaak last van halitose te hebben (Bornstein et al,
deze groep had 4,2% een score 30 of lager, 36,7% een 2009). Hoewel dit resultaat in de richting komt van de
score tussen 30 en 70 en 59% een score hoger dan 70. Dit 4,1% uit het onderhavige onderzoek, moet direct worden
betekent dat 59% van de deelnemers die aangaven bij het aangetekend dat dit bij een andere grenswaarde dan 30
sociaal functioneren altijd rekening te houden met hun voor ‘niet fris’tot een andere uitkomst zou hebben geleid.
mondgeur tegelijkertijd meenden dat hun mondgeur fris Veel deelnemers bleken rekening te houden met hun
was. Dit betrof 9,1% van alle deelnemers en daarvan was mondgeur. Ongeveer twee derde gaf aan tijdens een eerste
60,2% vrouw. ontmoeting rekening te houden met het feit dat de ander
Een statistisch significant verband werd gevonden tus- zijn of haar mondgeur zou kunnen ruiken, bijvoorbeeld
sen geslacht en in het sociaal functioneren rekening hou- door het bewaren van afstand tot die persoon. Ongeveer
den met mondgeur (chi-kwadraattoets; p = 0,04). Dit 15% zei zelfs hier altijd rekening mee te houden, waarbij
verband komt voort uit de aanwezigheid van onevenredig vrouwen en deelnemers jonger dan 60 jaar statistisch sig-
veel vrouwen in de groep die aangaf in het sociaal functio- nificant vaker aangaven hiermee rekening te houden dan
neren altijd rekening te houden met hun mondgeur. De mannen en deelnemers ouder dan 60 jaar.
resultaten van de logistische regressieanalyse lieten zien De belangrijkste uitkomst van dit onderzoek is dat de
dat subjectieve mondgeurperceptie een significante voor- hypothese wordt bevestigd dat samenhang bestaat tussen
speller is voor afstand houden (odds ratio = 0,98; 95% be- subjectieve mondgeurperceptie en de mate waarin men

Nederlands Tijdschrift voor Tandheelkunde 196 120 | april 2013

De Jongh e.a.: Mondgeurperceptie en sociaal functioneren O n d e r zo e k e n w e t e n s c h a p

Een beperking van het onderhavige onderzoek is dat hali-

tose niet objectief is vastgesteld, waardoor niet kon wor-
den beoordeeld welke deelnemers al dan niet terecht
meenden een frisse of onfrisse mondgeur te hebben. Een
gevolg hiervan is bovendien dat niet kon worden beoor-
deeld of deelnemers ‘passende’ in plaats van op de werke-
lijkheid berustende antwoorden hebben geven. Aan de
andere kant werden de vragen anoniem en online beant-
woord en daarmee zal men zich relatief vrij hebben gevoeld
de vragen eerlijk te beantwoorden.

Conclusie
In het sociale verkeer rust een taboe op het bespreekbaar
maken van halitose. Mensen vinden het lastig anderen op
halitose aan te spreken. Voor sommigen kan de subjectieve
mondgeurperceptie een bron van onzekerheid vormen.
Tweederde van de deelnemers in het onderhavige onder-
zoek gaf aan rekening te houden met hun mondgeur tij-
dens een eerste ontmoeting. Bovendien suggereren de
resultaten van het onderzoek dat de mate waarin mensen
Beeld: Shutterstock. afstand tot elkaar bewaren onder andere afhankelijk is van
de subjectieve mondgeurperceptie. Er is derhalve reden
hiermee rekening houdt tijdens het sociaal functioneren. aan te nemen dat angst omtrent eventuele halitose de soci-
Mensen die zichzelf een slechtere mondgeur toedichten, ale interactie beïnvloedt en daarmee dus letterlijk afstand
lijken dus in het algemeen geneigd te zijn meer afstand te tussen mensen creëert.
nemen tot een persoon wanneer ze die voor het eerst ont-
moeten dan mensen die menen dat hun mondgeur fris is. Literatuur
Terwijl dit gedragspatroon op zich niet een opzienbarend * Bornstein MM, Kislig K, Hoti BB, Seemann R, Lussi A. Prevalence of
gegeven is, werd een groep deelnemers geïdentificeerd die halitosis in the population of the city of Bern, Switzerland: a study
aangaven, ondanks dat zij zichzelf een frisse mondgeur comparing self-reported and clinical data. Eur J Oral Sci 2009; 117:
toedichtten, zich daar niet naar te gedragen en toch bijvoor- 261-267.
beeld afstand bewaarden tot een ander. Mogelijk heeft dit te * Broek AMWT van den, Feenstra L, Baat C de. A review of the current
maken met een algemene onzekerheid in het sociaal functi- literature on aetiology and measurement methods of halitosis. J Dent
oneren en is dit gedrag een manifestatie van sociale angst, 2007; 35: 627-635.
dat wil zeggen: angst te worden afgewezen, in dit geval van- * Jongh A de, Baat C de, Horstman M. Psychosociale aspecten van hali-
wege halitose. Bekend is dat sociale angst meer voorkomt tose. Ned Tijdschr Tandheelkd 2012; 119: 436-440.
bij vrouwen dan bij mannen (Furmark et al, 1999). Als * Furmark T, Tillfors M, Everz P, Marteinsdottir I, Gefvert O, Fredrikson M.
mensen obsessief ervan overtuigd zijn dat ze niet fris rui- Social phobia in the general population: prevalence and sociodemo-
ken, wordt dit in de psychologische literatuur aangeduid graphic profile. Soc Psych Psychiatr Epidemiol 1999; 34: 416-424.
met de term ‘olfactory reference syndrome’ (Lochner en * Laine ML, Slot DE, Danser MM. Halitose. Een probleem van velen. Ned
Stein, 2003). Dit syndroom is een ziektebeeld waarbij spra- Tijdschr Tandheelkd 2011; 118: 607-611.
ke is van een persisterende preoccupatie met lichaamsgeu- * Liu XN, Shinada K, Chen XC, Zhang BX, Yaegaki K, Kawaguchi Y. Oral
ren, zoals die van oksels, voeten, het genitale gebied en de malodor-related parameters in the Chinese general population. J Clin
mond. Opmerkelijk is dat in dit verband de term halitofobie Periodontol 2006: 33: 31-36.
wordt gebruikt. Men spreekt over halitofobie als ook na * Lochner C, Stein DJ. Olfactory reference syndrome: diagnostic criteria
subjectieve metingen met een negatief resultaat de patiënt and differential diagnosis. J Postgrad Med 2003; 49: 328-331.
nog steeds ervan overtuigd is halitose te hebben (Yaegaki * Meskin LH. A breath of fresh air. J Am Dent Assoc 1996; 127: 1282-
en Coil, 2000). Het verschil tussen mensen met een sociale 1286.
fobie enerzijds en ’olfactory reference syndrome’ of halito- * Stein DJ. Social anxiety disorder in the West and in the East. Ann Clin
fobie anderzijds is dat in het eerste geval mensen er niet per Psychiatr 2009; 21: 109-117.
se van zijn overtuigd daadwerkelijk halitose te hebben. De * Tangerman A. Halitosis in medicine: a review. Int Dent J 2002; 52:
deelnemers aan het huidige onderzoek die aangaven bij het 201-206.
sociaal functioneren ‘altijd’ rekening te houden met hun * Yaegaki K, Coil JM. Examination, classification, and treatment of
mondgeur, maar tevens meenden dat hun mondgeur fris halitosis; clinical perspectives. J Can Dent Assoc 2000: 5: 257-261.
was, leden dus niet aan halitofobie. Want als dit het geval
zou zijn geweest, zouden ze hun mondgeur niet als ‘fris’,
maar juist als ‘onfris’ hebben beoordeeld.

Nederlands Tijdschrift voor Tandheelkunde 197 120 | april 2013

De Jongh e.a.: Mondgeurperceptie en sociaal functioneren O n d e r zo e k e n w e t e n s c h a p

Summary

Self-perceived oral odour and social interaction

This study examined the influence of self-perceived oral odour on social inter-
action. A representative sample of 1,082 people from the Dutch population of
16 years and older, were surveyed. On average, the participants graded their
oral odour as 66.8 on a scale 0-100; 4.2% judged their oral odour as ‘not
fresh’ (score ≤ 30). Approximately 65% indicated that they took into account
the fact that, when meeting somebody for the first time, that person might
smell their oral odour. Participants judging their oral odour to be not fresh
were shown to keep significantly more distance when meeting somebody than
participants judging their oral odour as fresh. Noteworthy was a subgroup
of participants who judged their oral odour as fresh, but indicated that they
always kept a certain distance to other people. The results suggest that self-
perceived oral odour is a potential barrier in social interaction.

Bron
A. de Jongh1, C. de Baat2, M. Horstman3, A.J. van Wijk1
Uit 1de afdeling Conserverende en Preventieve Tandheelkunde van het
Academisch Centrum Tandheelkunde Amsterdam (ACTA), 2de vakgroep
Orale Functieleer van het Universitair Medisch Centrum St Radboud in
Nijmegen en 3het onderzoeksbureau Kien te Groningen
Datum van acceptatie: 28 december 2012
Adres: prof. dr. A. de Jongh, ACTA, Gustav Mahlerlaan 3004, 1081 LA
Amsterdam
a.de.jongh@acta.nl

Verantwoording
Het onderzoek werd uitgevoerd in opdracht van MedaPharma BV door
onderzoeksbureau Kien conform de richtlijnen van ISO 20252 (markt-
onderzoek) en ISO 26362 (access panels).

Nederlands Tijdschrift voor Tandheelkunde 198 120 | april 2013

 M.L. Laine, D.E. Slot, M.M. Danser Medisch

Halitose. Een probleem van velen

Halitose is een veelvoorkomend probleem waarvan de oorzaak Epidemiologie

veelal in de mond ligt. De uitdaging voor mondzorgverleners is Er is weinig bekend over de prevalentie van halitose en ver-
de juiste diagnose stellen en adequaat behandelen. Differentiële gelijking van verschillende onderzoeken is moeilijk van-
diagnostiek is van groot belang om het onderscheid te maken wege verschillende meetmethoden. De meeste onderzoeken
tussen halitose met een orale en een niet-orale oorsprong. Orale zijn zelfrapportage-onderzoeken en zijn dus gebaseerd op
halitose kan effectief worden behandeld, primair door een goede subjectieve waarnemingen. De meerderheid van de onder-
mondverzorging. Plaque-accumulatie op de tongrug is de meest zoeken geeft aan dat halitose voorkomt bij 20 tot 50% van
voorkomende oorzaak. Tongreiniging, eventueel in combinatie de volwassen bevolking.
met een specifieke mondspoelmiddel, wordt daarom als onderdeel Uit een epidemiologisch onderzoek in de Verenigde
van de mondverzorging geadviseerd. Andere mondgezondheids- Staten bleek dat 10 tot 30% van de Amerikaanse bevolking
problemen als parodontitis, cariës en gebitsprothesen met slechte regelmatig last heeft van een slechte adem (Meskin, 1996).
pasvorm dienen door adequate behandeling te worden uitgesloten In Japan was dat 24% en in Frankrijk 50 tot 60% (Miya-
als mogelijke oorzaak. zaki et al, 1995; Menningaud et al, 1999). In Nederland is
in 1966 een grootschalig onderzoek gedaan onder 11.625
Laine ML, Slot DE, Danser MM. Halitose. Een probleem van velen mensen. Hieruit bleek dat 25% van de personen ouder dan
Ned Tijdschr Tandheelkd 2011; 118: 607-611 60 jaar en 10% van de personen jonger dan 20 jaar halitose
doi: 10.5177/ntvt.2011.12.11160 had (De Wit, 1966). Dit zou kunnen betekenen dat de pre-
valentie van halitose toeneemt met de leeftijd. Een Brazili-
Inleiding aans onderzoek liet ook een 3 maal zo groot risico op
Halitose, ook wel bekend als slechte adem of foetor ex ore, halitose zien bij personen ouder dan 20 jaar als bij perso-
werd al duizenden jaren geleden beschreven in de litera- nen jonger dan 20 jaar (Nadanovsky et al, 2007). In een
tuur. Al in het Oude Testament staat geschreven dat Job Chinees onderzoek werd met behulp van organoleptische
betreurde: “Mijn vrouw walgt van mijn adem” (Job 19:17). scores en Halimeter®-scores bij respectievelijk 27% en
‘Halitose’ is afkomstig van het Latijnse woord halitus dat 20% van de onderzochte 2.000 personen halitose vastge-
adem betekent en het oud-Griekse osis, dat ‘abnormaal’ of steld (Liu et al, 2006). Uit een recent onderzoek in Zwitser-
‘ziek’ betekent. land bleek op basis van organoleptisch onderzoek dat 11%
Iedereen heeft wel eens een onaangename mondgeur, van de 419 onderzochte personen last had van halitose.
bijvoorbeeld ’s ochtends bij het wakker worden of na het Dezelfde onderzoekers kwamen op basis van een enquête
eten of drinken van bepaalde voedingsmiddelen. In derge- tot een score van 32% (Bornstein et al, 2008). Het verschil
lijke gevallen verdwijnt de slechte adem na een goede mond- tussen de 2 scores benadrukt het probleem voor een patiënt
verzorging en komt deze niet zonder meer terug. Maar om zelf halitose vast te stellen. Bij 16% van de patiënten
persisterende halitose is een probleem dat de hele dag of die het halitosespreekuur van de Universiteit van Leuven
langere perioden aanwezig is (Van den Broek et al, 2007). bezochten met klachten over een slechte adem kon geen
Halitose is een sociaal en psychologisch probleem dat in- halitose worden vastgesteld (Quirynen et al, 2009).
vloed heeft op het dagelijks leven van een grote deel van de
volwassen bevolking (Quirynen et al, 2009). Slechte adem Etiologie van orale halitose
kan zeer belemmerend zijn in communicatie en in contacten Verreweg het merendeel van de halitoseklachten (76%) had
met andere mensen. Door adaptatie van het eigen reuk- een orale oorzaak en van slechts 4% lag de oorsprong buiten
orgaan zijn de meeste mensen zich niet bewust van hun de mond, zoals in de keel, de neus of het maag-darmkanaal.
slechte adem, soms hun leven lang niet. Vaak komen men- Bij de overige 20% kon geen oorzaak worden gevonden of
sen erachter dat ze een slechte adem hebben door veelal was sprake van pseudohalitose (Quirynen et al, 2009). Elke
non-verbale reacties van hun omgeving: gesprekspartners plaats in de mond waarop een biofilm kan accumuleren,
draaien het hoofd weg of doen een stap naar achteren. biedt de mogelijkheid voor het ontstaan van halitose. Het
Halitose is in onze westerse maatschappij voor velen beslag op de tongrug is de voornaamste oorzaak van orale
bijna een taboe. Zelfs binnen de mondzorg blijft slechte halitose. Bij ongeveer 60% van de halitosepatiënten in
adem een moeilijke en gevoelige kwestie die leidt tot situa- Leuven werd tongbeslag gediagnosticeerd (Quirynen et al,
ties waarin zorgverleners niet weten hoe zij aan een patiënt 2009). Daarnaast speelden gingivitis en/of parodontitis bij
moeten of kunnen vertellen dat hij halitose heeft. Ook kan 30% van de patiënten een belangrijke rol. Ook factoren
halitose moeilijk te bespreken zijn door beperkte kennis van zoals hyposialie, medicijngebruik, mondademhaling en eet-
oorzaken en mogelijke behandelingen van halitose. en drinkgewoonten kunnen ademgeur beïnvloeden.

Nederlands Tijdschrift voor Tandheelkunde 607 118 | december 2011

Laine e.a.: Halitose M edis ch

Vluchtige zwavelverbindingen, vooral waterstofsulfide en

methylmercaptaan en in mindere mate dimethylsulfide,
veroorzaken een onaangename ademgeur en zijn bij zeer
lage concentraties al waarneembaar (Tangerman, 2002).
Waterstofsulfide ruikt als rotte eieren, methylmercaptaan
heeft een penetrante geur van rotte kool en dimethylsulfide
heeft een onaangenaam zoetige geur. Deze vluchtige zwa-
velverbindingen worden geproduceerd door voornamelijk
Gram-negatieve anaerobe bacteriën zoals Porphymonas
gingivalis, Prevotella intermedia, Fusobacterium nucleatum
en Treponema denticola (Persson et al, 1990). Deze bacte-
riën zijn in staat aminozuren uit voedselresten, cellen,
speeksel, creviculaire vloeistof of bloed om te zetten in de
vluchtige zwavelverbindingen. Onderzoeken met nieuwe
DNA-detectiemethoden hebben recent laten zien dat de
tong in vergelijking met andere orale niches een specifieke
en de meest diverse bacteriële compositie heeft (Zaura et
al, 2009; Preza et al, 2009).

Niet-orale halitose Afb. 1. Tong met tongbeslag.

Naast orale oorzaken zijn er ook andere oorzaken voor ha-
litose zoals keel- en neusaandoeningen, gastro-intestinale
ziekten, sommige stofwisselingsziekten en carcinomen.
Deze kunnen resulteren in slechte geur uit de mond en de
neus (Van den Broek et al, 2007).
Bij kinderen kan een infectie die ontstaat als gevolg van
het inbrengen van vreemde voorwerpen zoals klein speel-
goed in de neus een oorzaak van halitose zijn. Bij volwas-
senen kunnen infecties in de luchtwegen, zoals chronische
sinusitis, tonsillitis en bronchitis, of nasofaryngeale abces-
sen of tumoren verantwoordelijk zijn voor slechte adem. De
aanwezigheid van tonsillolieten verhoogde de kans op een
verhoogde hoeveelheid vluchtige zwavelverbindingen in de
adem 10 keer (Rio et al, 2008).
Halitose kan ook een symptoom zijn van gastro-oeso-
fageale reflux (Van den Broek et al, 2007). Helicobacter
pylori is ook gesuggereerd als een oorzaak van halitose
omdat de bacterie vluchtige zwavelverbindingen kan pro-
duceren (Lee et al, 2006). Bovendien kunnen bij een aantal
stofwisselingsziekten onaangenaam ruikende stoffen in de Afb. 2. Losgeschraapt tongbeslag
bloedbaan circuleren en worden uitgeademd door alveo-
laire gasuitwisseling. Dimethylsulfide is de belangrijkste verdunningsreeksen, door te oefenen op patiënten en met
vluchtige zwavelverbinding die bijdraagt aan niet-orale behulp van een gestandaardiseerde geurtest (Sensonics®,
halitose (Tangerman et al, 2007). Diabetes mellitus kan www.sensonics.com). De gebruiker moet 40 uiteenlopen-
ketongeur in de adem veroorzaken. Trimethylaminuria, de geuren identificeren en zo wordt het reukvermogen
het visgeursyndroom, is een zeldzame genetische stof- getest. Deze test biedt tevens de mogelijkheid een eventu-
wisselingsziekte waarbij trimethylamine niet goed wordt eel gebrek aan reukvermogen (anosmie) vast te stellen.
afgebroken. Trimethylamine veroorzaakt een lichaamsgeur Voor het stellen van een diagnose zijn verschillende
die lijkt op de geur van rotte vis. Deze lucht komt vrij via het onderzoeksmethoden beschikbaar. De meestgebruikte
zweet, de urine en de adem. Ten slotte kan het gebruik van methode is een organoleptische test en deze wordt als
een aantal medicijnen bijdragen aan een slechte adem. gouden standaard gehanteerd bij onderzoek naar halitose.
Vrijwel altijd wordt de schaal gebruikt zoals die is ontwik-
Diagnostiek keld door Rosenberg et al en later is aangepast door Green-
Een mondzorgverlener moet zelf goed geuren kunnen ruiken man et al (Rosenberg et al, 1991; Greenman et al, 2004).
en onderscheiden om patiënten te kunnen attenderen op De mogelijke scores zijn: 0 = geen geur, 1 = nauwelijks
een slechte adem en om de ademgeur te kunnen beoorde- merkbare geur, 2 = lichte geur, 3 = matige geur, 4 = sterke
len. Dat is te bereiken met een training met chemische geur en 5 = hele sterke (overweldigende) geur. De test kan

Nederlands Tijdschrift voor Tandheelkunde 608 118 | december 2011

Laine e.a.: Halitose M edis ch

taan en dimethylsulfide apart weer (www.abilit-medical-

and-environmental.jp). De resultaten van de Halimeter® en
de organoleptische testen bleken met elkaar te correleren
(Rosenberg et al, 1991). Het aantal micro-organismen in
de biofilm op de tong was sterker gecorreleerd aan de orga-
noleptische testen en de uitslagen van de Halimeter® dan
de hoeveelheid tongbeslag (Saad en Greenman, 2008). De
correlatie tussen organoleptische testen en metingen met
de OralChroma™ is nog niet vaak onderzocht maar deze
lijkt positief (Saad et al, 2011).
Het is belangrijk niet alleen te meten maar ook naar het
Afb. 3. Halimeter®. verhaal van de patiënt te luisteren en door te vragen en
vooral de klachten serieus te nemen. Hierbij moet worden
aangetekend dat de metingen een momentopname zijn.
Veel patiënten met een gerichte klacht proberen bovendien
met allerlei hulpmiddelen die voor de consument beschik-
baar zijn de onaangename geur te verdoezelen. Eten, roken
en het gebruik van cosmetica voorafgaande aan de metin-
gen kan een sterke invloed op de metingen hebben. Het is
belangrijk de patiënt er op te wijzen wat hij wel of niet kan
doen vóór het onderzoek.
Veel mensen hebben ’s ochtends een minder frisse
mondgeur door de sterk verminderde speekselsecretie-
snelheid tijdens een nacht slapen. Door het nuttigen van
een ontbijt en een goede mondverzorging kan dit verdwij-
nen en dan is er sprake van alleen ochtendhalitose. Als de
halitose daarmee niet verdwijnt, wordt het persisterende
Afb. 4. OralChroma™-apparatuur en bijbehorende software. halitose genoemd en die diagnose wordt op basis van
meerdere onderzoeken al dan niet bevestigd. Als bij een
op verschillende wijzen worden uitgevoerd. De waarnemer patiënt na de metingen geen halitose kan worden vastge-
kan recht tegenover of haaks op de patiënt plaatsnemen. steld, had de patiënt tot op dat moment pseudo-halitose.
De patiënt kan al dan niet actief uitademen of spreken. Het Als een patiënt er niet van kan worden overtuigd dat hij
is ook mogelijk de geur in de mond te ruiken met de neus geen halitose heeft en halitose blijft ervaren, is de diagnose
van de waarnemer vlak voor de mond van de patiënt. De ‘halitofobie’ en kan het wenselijk zijn de patiënt te verwijzen
persoon in kwestie kan ook door een buisje uitademen, naar een psycholoog.
terwijl de waarnemer aan de andere kant van het buisje
ruikt (Wigger-Alberti et al, 2010). Naast de mondgeur kan Behandeling
ook het tongbeslag en de neusadem organoleptisch worden Behandeling van halitose richt zich primair op het dagelijks
gescoord. effectief verwijderen van micro-organismen en de hoeveel-
Van het tongbeslag (afb. 1 en 2) wordt ook de dikte ge- heid voedingstoffen voor micro-organismen. Afhankelijk
scoord (0 = geen beslag, 1 = een beetje beslag en 2 = dik van de locatie van de accumulerende biofilm in de mond
beslag), en de kleur. De uitgebreidheid van het tongbeslag dient een behandelplan te worden opgesteld. Het beslag op
kan worden gemeten door de tong in 9 of 6 compartimen- de tongrug, de meest voorkomende oorzaak van orale hali-
ten te verdelen (Mantilla Gomez et al, 2001; Winkel et al tose, kan mechanisch en/of chemisch worden verwijderd
2003). Iedereen heeft een biofilm op de tong, deze is echter (Quirynen et al, 2009). Als er sprake is van cariëslaesies,
niet bij iedereen visueel waarneembaar. De consistentie gebitsprothesen met een slechte pasvorm, parodontitis of
van de biofilm blijkt van groter belang dan de visuele waar- gingivitis worden die problemen ook behandeld.
neembaarheid (Saad en Greenman, 2008). De meeste onderzoeken hebben behandelingen en
Op de markt zijn diverse apparaten beschikbaar om de producten getest bij personen met ochtendhalitose en de
mondgeur objectief vast te stellen. De bekendste zijn Hali- resultaten daarvan zijn dus niet zonder meer extrapoleer-
meter® en OralChromaTM (afb. 3 en 4). Het zijn gemakkelijk baar voor persisterende halitose (Haas et al, 2007).
verplaatsbare instrumenten die in de praktijk kunnen wor-
den gebruikt voor het stellen van de diagnose en voor de Mechanische middelen
evaluatie van de behandeling van halitose. De Halimeter® Het beslag op de tongrug kan met een tongschraper, een
is een apparaat dat de totale hoeveelheid vluchtige zwavel- tandenborstel of een ander instrument worden verwijderd.
verbindingen meet (www.halimeter.com). De OralChroma™ Dit kan lastig zijn doordat veel mensen daarbij last krijgen
geeft de concentraties van waterstofsulfide, methylmercap- van de natuurlijke kokhalsreflex.

Nederlands Tijdschrift voor Tandheelkunde 609 118 | december 2011

Laine e.a.: Halitose M edis ch

Het is gebleken dat dagelijkse mechanische tongreiniging Literatuur

de hoeveelheid tongbeslag verminderde (Van der Sleen et * Bordas A, McNab R, Staples AM, Bowman J, Kanapka J, Bosma MP.
al, 2010). De invloed daarvan op de hoeveelheid micro- Impact of different tongue cleaning methods on the bacterial load of
organismen op de tongrug is niet duidelijk. Het ene onder- the tongue dorsum. Arch Oral Biol 2008; 53 (Suppl. 1): S13-S18.
zoek liet zien dat tongreiniging eventueel de hoeveelheid * Bornstein MM, Kislig K, Hoti BB, Seemann R, Lussi A. Prevalence of hali-
bacteriën kan verminderen, maar uit een ander onderzoek tosis in the population of the city of Bern, Switzerland: a study comparing
bleek geen invloed op het aantal bacteriën (Quirynen et al, self-reported and clinical data. Eur J Oral Sci 2009; 117: 261-257.
2004; Bordas et al, 2008). De mate van halitose hangt af * Bosy A, Kulkarni GV, Rosenberg M, McCulloch CAG. Relationship of
van de productie van vluchtige zwavelverbindingen. Een oral malodor to periodontitis: evidence of independence in discrete
aantal onderzoeken heeft een relatie aangetoond tussen subpopulations. J Periodontol 1994; 65: 37-46.
mechanische verwijdering van tongbeslag en verlaging van * Broek AMWT van den, Feenstra L, Baat C de. A review of the current
organoleptische scores en vermindering van de hoeveel- literature on aetiology and measurement methods of halitosis. J Dent
heid vluchtige zwavelverbindingen (Bosy et al, 1994; See- 2007; 35: 627-635.
mann et al, 2001). * Broek AMWT van den, Feenstra L, Baat C de. A review of the current
literature on management of halitosis. Oral Dis 2008; 14: 30-39.
Chemische middelen * Greenman J, Duffield J, Spencer P, et al. Study on the organoleptic in-
Mondspoelmiddelen, tandpasta’s of tonggels kunnen op 2 tensity scale for measuring oral malodor. J Dent Res 2004; 83: 81-85.
manieren een slechte adem verminderen: door de bacteri- * Haas AN, Silveira EM, Rösing CK. Effect of tongue cleansing on morning
en te elimineren en door de hoeveelheid geproduceerde oral malodour in periodontally healthy individuals. Oral Health Prev
gassen te reduceren. Dent 2007; 5: 89-94.
De actieve componenten in mondspoelmiddelen zijn * Lee H, Kho HS, Chung JW, Chung SC, Kim YK. Volatile sulfur com-
meestal antimicrobiële middelen zoals chloorhexidine, ce- pounds produced by Helicobacter pylori. J Clin Gastroenterol 2006:
tylpyridiniumchloride, etherische oliën (Listerine®), chloor- 40; 421-426.
dioxide, metaalionen (zinklactaat en zinkchloride), triclosan * Liu XN, Shinada K, Chen XC, Zhang BX, Yaegaki K, Kawaguchi Y. Oral
en waterstofperoxide. Al deze middelen kunnen tijdelijk malodor-related parameters in the Chinese general population. J Clin
halitose verminderen. De meeste middelen zijn relatief korte Periodontol 2006; 33: 31-36.
tijd (3-4 uur) effectief. Waterstofperoxide, cetylpyridinium- * Loesche WJ, Kazor C. Microbiology and treatment of halitosis. Periodon-
chloride, een combinatie van chloorhexidine, cetylpyridini- tol 2000 2002; 28: 256-279.
umchloride en zinklactaat (Halita®), een combinatie van * Mantilla Gómez S, Danser MM, Sipos PM, et al. Tongue coating and
triclosan en zink en een combinatie van amino- en tinfluori- salivary bacterial counts in healthy/gingivitis subjects and periodonti-
de en zinklactaat (Meridol Halitosis®) lijken langere tijd ef- tis patients. J Clin Periodontol 2001; 28: 970-978.
fectief te zijn (Rosenberg et al, 1992; Raven et al, 1996; * Menningaud JP, Bado F, Favre E, Bertrand JC, Guilbert F. Halitosis in
Winkel et al, 2003; Wigger-Alberti et al, 2010). Voor een 1999. Rev Stomatol Chir Maxillofac 1999; 100: 240–244.
optimaal effect is het belangrijk dat de patiënt gorgelt, bij * Meskin LH. A breath of fresh air. J Am Dent Assoc 1996; 127: 1282-
voorkeur met een uitgestoken tong zodat het achterste 1286.
gedeelte van de tongrug wordt bereikt. Het beste resultaat * Miyazaki H, Sakao S, Katoh Y, Takehara T. Correlation between volatile
wordt behaald met tongschrapen, eventueel in combinatie sulphur compounds and certain oral health measurements in the
met gebruik van mondspoelmiddelen, naast dagelijkse inter- general population. J Periodontol 1995; 66: 679-684.
dentaal reinigen en tandenpoetsen (Roldan et al, 2005). * Nadanovsky P, Carvalho LB, Ponce de Leon A. Oral malodour and its
Een aantal onderzoeken heeft laten zien dat borstelen association with age and sex in a general population in Brazil. Oral Dis
van het achterste deel van de tongrug met tandpasta 2007; 13: 105-109.
effectief kan zijn. Vooral tandpasta’s met natriumbicarbo- * Persson S, Edlund MB, Claesson R, Carlsson J. The formation of hydro-
naat, triclosan met zinkcitraat en tinfluoride lijken effectief gen sulfide and methyl mercaptan by oral bacteria. Oral Microbiol
te zijn (Van den Broek et al, 2008). Tongreiniging met Immunol 1990; 5:195-201.
tandpasta leek wel minder effectief dan een mondspoel- * Preza D, Olsen I, Willumsen T, Grinde B, Paster BJ. Diversity and site-
middel (Loesche en Kazor, 2002). specificity of the oral microflora in the elderly. Eur J Clin Microbiol
Producten zoals pepermuntjes en kauwgom kunnen Infect Dis 2009; 28: 1033-1040.
tijdelijk de slechte adem maskeren door stimulering van * Quirynen M, Avontroodt P, Soers C, Zhao H, Pauwels M, Steenberghe D
de speekselsecretiesnelheid en het verspreiden van de van. Impact of tongue cleansers on microbial load and taste. J Clin
mintgeur. Periodontol 2004; 31: 506-510.

* Quirynen M, Dadamio J, Velde S van den, et al. Characteristics of 2000

Ten slotte patients who visited a halitosis clinic. J Clin Periodontol 2009; 36:
Omdat het voor een patiënt niet mogelijk is zijn eigen 970-975.
ademgeur te ruiken, is het aan te raden dat de patiënt een * Raven SJ, Matheson JR, Huntington E, Tonzetich J. The efficacy of
vertrouwenspersoon zoekt die regelmatig zijn of haar a combined zinc and triclosan system in the prevention of oral
adem kan ruiken en commentaar kan geven over de behan- malodour. In: Steenbergen D van, Rosenberg M (red.). Bad breath: a
delresultaten. multidisciplinary approach. Leuven: Leuven University Press, 1996.

Nederlands Tijdschrift voor Tandheelkunde 610 118 | december 2011

Laine e.a.: Halitose M edis ch

* Rio AC, Franchi-Teixeira AR, Nicola EM. Relationship between the Bron
presence of tonsilloliths and halitosis in patients with chronic caseous M.L. Laine1, 2, D.E. Slot1, M.m. Danser1
tonsillitis. Br Dent J 2008: 204; E4. Uit 1de sectie Parodontologie en 2het Halitose spreekuur, afdeling Conser-

* Roldán S, Herrera D, O’Connor A, González I, Sanz M. A combined verende en Preventieve Tandheelkunde, van het Academisch Centrum voor
therapeutic approach to manage oral halitosis: a 3-month prospective Tandheelkunde Amsterdam (ACTA)
case series. J Periodontol 2005; 76:1025-1033. Datum van acceptatie: 25 augustus 2011
* Rosenberg M, Kulkarni GV, Bosy A, McCulloch CA. Reproducibility and Adres: mw. dr. M.L. Laine, ACTA, Gustav Mahlerlaan 3004,
sensitivity of oral malodor measurements with a portable sulphide 1081 LA Amsterdam
monitor. J Dent Res 1991; 70:1436-1440. m.laine@acta.nl
* Rosenberg M, Gelernter I, Barki M, Bar-Ness R. Day-long reduction of
oral malodor by a two-phase oil:water mouthrinse as compared to Dankwoord
chlorhexidine and placebo rinses. J Periodontol 1992; 63: 39-43. De auteurs danken Halimeter en Abimedical Corporation voor de toestem-
* Saad S, Greenman J. Tongue biofilm areal density and tong coating ming voor het gebruik van de foto’s van de Halimeter® en de OralChroma™.
index. J Breath Res 2008; 2: 017008.

* Saad S, Greenman J, Shaw H. Comparative effects of various commer- Belangenconflict en bron van financiering
cially available mouthrinse formulations on oral malodour. Oral Dis De auteurs verklaren geen financiële banden te hebben met industriële
2011; 17: 180-186. producten die gerelateerd zijn aan het onderwerp van dit artikel. Dit artikel
* Seemann R, Kison A, Bizhang M, Zimmer S. Effectiveness of mechanical werd mogelijk gemaakt door financiering uit eigen middelen en de instel-
tongue cleaning on oral levels of volatile sulfur compounds. J Am Dent ling waarbij de auteurs werken.
Assoc 2001; 132: 1263-1267.

* Sleen MI van der, Slot DE, Trijffel E van, Winkel EG, Weijden GA van
der. Effectiveness of mechanical tongue cleaning on breath odour and
tongue coating: a systematic review. Int J Dent Hyg 2010; 8: 258-268.

* Tangerman A. Halitosis in medicine: a review. Int Dent J 2002; 52:

201–206.

* Tangerman A, Winkel EG. Intra- and extra-oral halitosis: finding of

a new form of extra-oral blood-borne halitosis caused by dimethyl
sulphide. J Clin Periodontol 2007; 34: 748-755.

* Wigger-Alberti W, Gysen K, Axmann E-M, Wilhelm K-P. Efficacy of a

new mouthrinse formulation on the reduction of oral malodour in
vivo. A randomized, double-blind, placebo-controlled, 3 week clinical
study. J Breath Res 2009; 4: 017102.

* Winkel EG, Roldán S, Winkelhoff AJ van, Herrera D, Sanz M. Clinical

effects of a new mouthrinse containing chlorhexidine, cetylpyridinium
chloride and zinclactate on oral halitosis. A dual-center, double-blind
placebo-controlled study. J Clin Periodontol 2003; 30: 300-306.

* Wit G de. Foetor ex ore. Ned Tijdschr Geneeskd 1966; 10: 1689-1692.
* Zaura E, Keijser BJ, Huse SM, Crielaard W. Defining the healthy “core
microbiome” of oral microbial communities. BMC Microbiol 2009;
15: 259.

Summary
Halitosis. A common problem
Halitosis is a frequently occurring problem, the cause of which is generally
to be found in the mouth. The challenge for oral health care providers is to
diagnose it correctly and treat it effectively. Differential diagnosis is of great
importance in making a distinction between halitosis which originates in
the mouth and which does not originate in the mouth. Oral halitosis can
be treated effectively by good oral health care. Plaque accumulation on the
tongue is the most common cause of oral halitosis. Tongue cleansing, possibly
in combination with a specific mouth wash, is consequently recommended
as an element of oral hygiene care. Other oral health problems, such as
periodontal disease, caries and ill-fitting removable dentures should be
treated adequately to eliminate these problems as potential causes of halitosis.

Nederlands Tijdschrift voor Tandheelkunde 611 118 | december 2011

 Vol. 113 No. 1 January 2012

Comparison between self-perceived and clinical oral malodor

Thuy A.V. Pham, DDS, PhD,a,b Masayuki Ueno, DDS, MPH, PhD,a Kayoko Shinada, DDS, PhD,c and
Yoko Kawaguchi, DDS, PhD,a Tokyo, Japan, and Ho Chi Minh City, Vietnam
TOKYO MEDICAL AND DENTAL UNIVERSITY AND NATIONAL HOSPITAL OF ODONTO-STOMATOLOGY

Objectives. The aims were to compare self-perceived with clinical oral malodor and to examine risk factors of oral malodor.
Study Design. The study was performed on 565 dental patients. Information on sociodemographics, dental health behavior,
and self-perceived oral malodor was collected. Clinical oral malodor, oral health status, and the proteolytic activity of the
N-benzoyl-DL-arginine-2-napthilamide (BANA) test in tongue coating were assessed.
Results. The sensitivity and specificity of self-perceived oral malodor were 47.2% and 59.2%, respectively. Risk factors for
self-perceived oral malodor were smoking habit and alcohol consumption, whereas those for clinical oral malodor were level
of education, dental visit frequency, tongue-brushing frequency, mouth rinse use, deep periodontal pockets, gingivitis, tongue
coating, and a high BANA test score.
Conclusions. Self-perception was considered an invalid method of judging one’s own oral malodor. Factors related to self-
perceived oral malodor were different from those found in clinical oral malodor. (Oral Surg Oral Med Oral Pathol Oral
Radiol 2012;113:70-80)

Halitosis, also called oral malodor, is a condition that Clinically, oral malodor can be measured by the organ-
causes problems in many aspects of daily life.1,2 It is oleptic test or with specific devices. The organoleptic test,
one of the strong negative factors affecting social com- which uses a human’s nose to smell and rank the intensity
munication. Increasing numbers of patients present at of the odors emanating from the mouth, is a gold standard
dental offices complaining of oral malodor.3 In western for measurement of oral malodor.8,11 It is easy to perform,
countries, personal discomfort and social embarrass- requires no extra apparatus, and reflects the presence of an
ment are the main reasons people seek oral malodor objective malodor detected by an examiner. Thus, the
treatment by professionals. organoleptic test is believed to be the most reliable and
Oral malodor originates from a variety of products of practical method for clinical oral malodor diagnosis.11-15
bacterial metabolism of amino acids, desquamated cells, Specific devices, such as gas chromatographs and sulfide
and leukocytes. These metabolites include volatile sulfur monitors (e.g., gas chromatography, Oral Chroma, Hali-
compounds (VSCs), indole, skatole, amine, and ammonia. meter, Breathtron) have been used to estimate oral mal-
Of these compounds, VSCs with principal components of odor, based on levels of VSCs within the oral cavity.
hydrogen sulfide (H2S) and methyl mercaptan (CH3SH) Recent research has reported good agreement between the
are the main causes of oral malodor.4-6 Several studies Oral Chroma and gas chromatography.16
have reported that periodontal disease–associated bacteria, Information on self-perceived oral malodor, history,
such as Porphyromonas gingivalis, Treponema denticola, and the symptoms of oral malodor, obtained from a
and Tannerella forsythiae are mainly responsible for VSC questionnaire, is very important for the diagnosis of
production.7-9 The presence of these bacteria was detected psychological halitosis, such as pseudohalitosis and
based on their ability to hydrolyze the synthetic trypsin halitophobia. Attempts to estimate oral malodor only
substrate, namely by the N-benzoyl-DL-arginine-2- by self-perception sometimes are not reliable because
napthilamide10 (BANA) test. they often reflect subjective opinion and do not corre-
spond to the objective presence of halitosis.1 Although
most persons who complain of halitosis have oral mal-
a
Department of Oral Health Promotion, Graduate School of Medical odor clinically, some do not, and they often suffer for
and Dental Sciences, Tokyo Medical and Dental University, Tokyo, psychological reasons.3
Japan. Because of the important health and social implica-
b
National Hospital of Odonto-Stomatology, Ho Chi Minh City,
Vietnam.
tions of halitosis, many epidemiologic and etiologic
c
Department of Oral Health Care Promotion, School of Oral Health studies have been conducted to estimate the prevalence
Care Sciences, Faculty of Dentistry, Tokyo Medical and Dental of halitosis, to identify the factors associated with the
University, Tokyo, Japan. condition, and to evaluate various halitosis assessment
Received for publication Jun 9, 2011; returned for revision Jul 26, methods. Some studies have used self-perception or
2011; accepted for publication Aug 6, 2011.
© 2012 Elsevier Inc. All rights reserved.
self-reporting to estimate the prevalence of oral mal-
2212-4403/$ - see front matter odor, however.17-19 A previous study demonstrated that
doi:10.1016/j.tripleo.2011.08.012 sniff magnitude measures were altered by a number of

70
OOOO ORIGINAL ARTICLE
Volume 113, Number 1 Pham et al. 71

factors, including the type of ordorant used in testing, (yes or no), dental visit frequency (often, sometimes,
and suggested that they were modulated by judgment never, rarely or when having dental problems), tooth-
and prior experience.20 Patients visting the dental clin- brushing frequency (number of times per day), daily
ic/hospital are considered a special group with higher tongue cleaning (yes or no), and mouth rinse use at
dental needs and the self-perception of their oral health least 3 times per week (yes or no). The last section
status may be more sensitive than the general popula- asked about self-perceived oral malodor: “Do you think
tion. In reviewing the literature, oral malodor and its you presently have oral malodor?” If subjects answered
related factors in different age group populations, such “yes,” they were asked about “the duration of your oral
as children,21 adolescents,22 or elderly23; the effect of malodor” and “who complained about your oral mal-
salivary protein24 or probiotic25 or mouthwash26,27 on odor.”
genuine oral malodor; and relationship between halito-
sis and psychological condition28 in dental patients Oral malodor measurement
have been well reported. There are few studies, how- Oral malodor measurements were performed in the
ever, that have evaluated the validity of self-perceived morning between 7:30 AM and 11:30 AM. Subjects were
oral malodor by comparing it with clinical oral malodor asked to refrain from eating, drinking, smoking, or oral
based on an organoleptic test and VSC measurements, hygiene at least 2 hours before the measurement. Oral
and that have provided information regarding sociode- malodor was measured by an organoleptic test and with
mographics, oral health behavior, oral health condition the Oral Chroma (Abimedical Corporation, Osaka,
characteristics, and BANA-positive bacteria as associ- Japan).
ated or controlled factors of oral malodor in a popula- One examiner, a well-trained and experienced dentist
tion of dental patients. Hence, the purpose of this study from the Fresh Breath Clinic of Dental Hospital, Tokyo
was to evaluate the validity of self-perceived oral mal- Medical and Dental University, performed the organo-
odor and examine the relationship between oral mal- leptic test for all study subjects. Subjects were asked to
odor and sociodemographics, dental health behaviors, close their lips tightly for 3 minutes sitting upright in a
and oral health conditions in dental patients. dental chair and then exhale briefly and softly from the
mouth through a paper tube. Subjects were at a distance
MATERIAL AND METHODS of about 10 cm from the examiner who was seated
Subjects behind a privacy screen (50 ⫻ 70 cm), separating the
A total of 603 patients visited the Diagnosis Depart- examiner and the subject. The breath was evaluated
ment of the National Hospital of Odonto-Stomatology with a 6-point scale, where 0 ⫽ no odor, 1 ⫽ question-
in Ho Chi Minh City, Vietnam, in 2009, and agreed to able odor, 2 ⫽ slight but clearly noticeable odor, 3 ⫽
participate in the study. All subjects signed a written moderate odor, 4 ⫽ strong odor, and 5 ⫽ extremely
informed consent form. Thirty-eight participants who foul odor.11 The organoleptic test was used to diagnose
suffered from systemic diseases (e.g., diabetes mellitus, clinical oral malodor in this study and subjects were
gastrointestinal disorder, respiratory dysfunction, neo- diagnosed as having oral malodor when their organo-
plasia, various carcinoma), who took medication or anti- leptic score was 2 or greater.13
bacterial substances, and who were pregnant or lactating, For the Oral Chroma measurement, a disposable
conditions known as causes of extraoral halitosis, were 1-mL-capacity syringe was inserted into the subject’s
excluded. Finally, 565 patients (287 men and 278 mouth, positioned above the posterior part of the dor-
women), aged 18 to 60 years (mean age 43.6 ⫾ 11.1) sum of the tongue without touching the oral mucosa or
were recruited for this study. The study protocol was the tongue. Subjects were asked to close their mouth
approved by the National Hospital of Odonto-Stomatol- and keep breathing gently through their nose for 1
ogy in Ho Chi Minh City, Vietnam, and by the Tokyo minute. A 0.5-mL aspirated air sample was then in-
Medical and Dental University in Japan. jected into the inlet of the measurement device. The
VSC gases were analyzed automatically in 8 minutes
Questionnaire and their concentration values were displayed in ng/10
The subjects completed a self-administered question- mL. A previous study suggested that the threshold
naire before their oral malodor measurement and oral levels of oral malodor by VSC concentrations were H2S
examination. The first section of the questionnaire in- of 1.5 ng/10 mL or higher and CH3SH of 0.5 ng/10 mL
quired about sociodemographic information, including or higher.6
age, gender, and level of education. The second section
comprised questions about smoking habits (nonsmoker/ Oral examination
past smoker or current smoker at least 4 cigarettes per Following the oral malodor measurement, all subjects
day), daily alcoholic beverage or alcohol consumption underwent a standard oral examination. The examina-
ORAL MEDICINE OOOO
72 Pham et al. January 2012

tion included dental caries experience (Decayed, Miss- SPSS 17.0 (SPSS Japan, Tokyo, Japan IL) and signif-
ing, Filled Teeth), periodontal conditions, tongue-coat- icance was set at P less than .05.
ing status, and bacterial levels on the tongue dorsum.
Measurements of the dental and periodontal conditions RESULTS
were made on all teeth, excluding the third molars. Sociodemographic and health behavioral
Dental caries assessment was based on World Health characteristics
Organization criteria. Pocket depths were assessed us- Men accounted for about half of the sample (50.8%).
ing a 1-mm scaled Williams periodontal probe and Mean ages of men and women (42.9 ⫾ 11.0 years and
measured at 6 sites on each tooth. The deepest pocket 44.3 ⫾ 11.1 years, respectively) did not differ signifi-
depth was recorded for the tooth. Gum bleeding was cantly (P ⫽ .117). The sociodemographic and health
recorded as present or absent within 30 seconds after behavioral characteristics of the subjects are shown in
the pocket depth measurement.29 Dental plaque and the Table I. About two-thirds of the sample had completed
gingival index by Löe criteria were also recorded.30 at least a high school education (65.5%). About 46% of
The area of tongue coating was recorded on a scale subjects were current smokers, and a significantly
from 0 to 3 by visual inspection as 0 ⫽ no coating, 1 ⫽ higher percentage of men (76.0%) were smokers than
less than one-third of tongue dorsum with a coating, 2 ⫽ women (14.4%) (P ⬍ .001). Almost one-fourth of
one-third to two-thirds of tongue dorsum with a coat- subjects consumed alcohol and 68.5% of them were
ing, and 3 ⫽ more than two-thirds of tongue dorsum men (P ⬍ .001). Fewer than half of subjects reported
with a coating.3 that they never or rarely had a dental checkup, or
A tongue-coating sample was collected using a cotton- visited the dentist when they had dental problems. The
tip swab. The sample was applied to the lower matrix of percentage of women who often or sometimes had a
the BANA test (BANAMet LLC, Ann Arbor, MI) dental checkup (66.5%) was significantly higher than
immediately after the collection, and the upper matrix that of men (47.0%) (P ⬍ .001). All subjects reported
was moistened with distilled water. Then, the lower brushing their teeth every day, and about 70% of those
matrix was folded back to make contact with the upper did so at least twice a day. More than one-third of
matrix and inserted into an incubator at 35°C for 5 subjects had a daily tongue cleaning, and about two-
minutes.31 Results of the BANA test were scored as 0 thirds of subjects used a mouth rinse. No significant
(negative reaction) when no blue color was visible, 1 differences in oral hygiene behaviors, such as tooth
(weak positive reaction) when a faint blue color was brushing (P ⫽ .786), tongue cleaning (P ⫽ .427), and
detected, and 2 (positive reaction) when distinct blue mouth rinse use (P ⫽ .379) were detected by gender.
color appeared in the area in contact with the sample.
Oral health status
All study subjects were dentate, and their oral health
Statistical analysis status is presented in Table II. The mean numbers of
Chi-square tests were used to examine distributional teeth present and decayed were 25.4 ⫾ 3.7 and 3.1 ⫾
differences by gender and to determine the relationship 3.2, respectively. No significant differences in dental
between oral malodor and factors, such as sociodemo- conditions were detected by gender. About one-third of
graphics and behavioral characteristics. Independent- the subjects (30.6%) had at least 1 gingival pocket of 5
sample t tests were used to detect the mean difference mm or more, and almost 70% of subjects (67.3%) had
in age by gender and to examine the differences of the at least 1 tooth with bleeding sites. Mean values of the
oral health parameters and H2S and CH3SH parameters plaque index and the gingival index were 2.6 ⫾ 1.5 and
between the subjects with or without oral malodor. 1.2 ⫾ 1.0, respectively. There were no statistically
Multiple logistic regression analysis was performed significant differences in periodontal conditions by
using perceived oral malodor (0 ⫽ “without perceived gender, except for dental plaque being significantly
oral malodor” and 1 ⫽ “with perceived oral malodor”) higher in men than women.
or clinical oral malodor (0 ⫽ “without clinical oral A low percentage of subjects had no tongue coating
malodor” and 1 ⫽ “with clinical oral malodor”) as a (2.7%). Almost one-fourth of subjects had a score of 1
dependent variable and the sociodemographics, behav- (23.4%). The highest percentage (43.2%) was observed
ioral characteristics, and those oral health factors that in subjects with a score of 2, and more than 30% of
showed statistically significant associations in the chi- subjects (30.8%) had a score of 3. Men had signifi-
square tests and independent t tests as independent cantly more tongue coating than women (P ⫽ .048).
variables. Adjusted odds ratios and corresponding 95% Almost one-fourth of subjects (22.3%) had a BANA
confidence intervals were generated for all significant test score of 0. About one-half of the subjects (48.3%)
variables. Statistical analysis was carried out using had a score of 1 and 29.5% had a score of 2. No
OOOO ORIGINAL ARTICLE
Volume 113, Number 1 Pham et al. 73

Table I. Sociodemographic and health behavioral characteristics of the subjects

Variables Male Female Total P value
Level of education
Primary and secondary school 94 (32.8) 101 (36.3) 195 (34.5) .378
High school or higher 193 (67.2) 177 (63.7) 370 (65.5)
Smoking habit
Non- and past-smoker 69 (24.0) 238 (85.6) 307 (54.3) ⬍.001
Current smoker 218 (76.0) 40 (14.4) 258 (45.7)
Alcohol consumption
Yes 87 (30.3) 40 (14.4) 127 (22.5) ⬍.001
No 200 (69.7) 238 (85.6) 438 (75.5)
Dental visit frequency
Often and sometime 135 (47.0) 185 (66.5) 320 (56.6) ⬍.001
Never, rarely, only when having a problem 152 (53.0) 93 (33.5) 245 (43.4)
Tooth brushing frequency
Once/d 89 (31.0) 90 (32.4) 179 (31.7) .786
ⱖ2 times/d 198 (69.0) 188 (67.6) 386 (68.3)
Tongue-cleaning frequency
Yes 94 (32.8) 100 (36.0) 194 (34.3) .427
No 193 (67.2) 178 (64.0) 371 (65.7)
Mouth rinse use
Yes 191 (66.6) 175 (62.9) 366 (64.8) .379
No 96 (33.4) 103 (37.1) 199 (35.2)
Data are presented as number (%).

Table II. Oral health parameters of the subjects

Oral health parameters Male Female Total P value
No. of decayed teeth 3.12 ⫾ 3.31 3.02 ⫾ 3.08 3.07 ⫾ 3.20 .719
No. of teeth with 5 mm or greater pockets 2.29 ⫾ 3.93 1.97 ⫾ 3.89 2.13 ⫾ 3.91 .341
No. of teeth with bleeding sites 4.94 ⫾ 5.39 5.21 ⫾ 5.26 5.07 ⫾ 5.32 .551
Plaque Index 2.76 ⫾ 1.49 2.47 ⫾ 1.50 2.61 ⫾ 1.50 .021
Gingival Index 1.28 ⫾ 1.03 1.13 ⫾ 1.01 1.20 ⫾ 1.02 .091
Tongue-coating score 2.09 ⫾ 0.84 1.95 ⫾ 0.77 2.02 ⫾ 0.81 .048
BANA test score 1.11 ⫾ 0.74 1.03 ⫾ 0.69 1.07 ⫾ 0.72 .211
Data are presented as mean ⫾ SD.
BANA, N-benzoyl-DL-arginine-2-napthilamide.

significant difference in BANA test score was observed (57.9%). Subjects with clinical oral malodor were sig-
by gender. nificantly older (44.6 ⫾ 9.7 years) than those without
oral malodor (42.2 ⫾ 12.6 years), but no significant
Oral malodor status distributional difference in oral malodor status was
According to the questionnaire, 44.6% of subjects re- found by gender.
ported currently having oral malodor. A significantly According to Oral Chroma measurement, 321 sub-
higher prevalence of self-perceived oral malodor was jects (56.8%) were diagnosed with oral malodor by H2S
found in men (54.0%) than in women (34.9%). Of the and 313 subjects (55.4%) by CH3SH. Concentrations of
patients, 19.4% had complaints of oral malodor for less H2S and CH3SH in those with and without self-per-
than 1 year, 44.0% had complaints for 1 to 5 years, and ceived oral malodor are shown in Figure 1. The H2S
36.6% had complaints for more than 5 years. Also, and CH3SH concentrations of subjects who did not
47.2% of subjects perceived their oral malodor them- perceive their oral malodor were 4.2 ⫾ 4.0 ng/10 mL
selves and 52.8% had the condition pointed out by and 3.0 ⫾ 3.7 ng/10 mL, respectively, and of those who
others. perceived their oral malodor were 4.6 ⫾ 4.1 ng/10 mL
According to the organoleptic test, 1.1% of subjects and 3.6 ⫾ 4.3 ng/10 mL, respectively. These concen-
had an organoleptic score of 0, 41.1% had a score of 1, trations were all higher than the threshold levels of
14.9% had a score of 2, 37.7% had a score of 3, and VSCs for oral malodor but did not differ significantly
5.3% had score of 4. No subject had a score of 5. between the 2 groups. Concentrations of H2S and
Subjects who had an organoleptic score of 2 or more CH3SH of those with and without clinical oral malodor
were diagnosed as having clinical oral malodor are shown in Figure 2. The mean H2S and CH3SH
ORAL MEDICINE OOOO
74 Pham et al. January 2012

Fig. 1. Concentrations of H2S (left) and CH3SH (right) by self-perceived oral malodor.

Fig. 2. Concentrations of H2S (left) and CH3SH (right) by clinical oral malodor.

concentrations of subjects without clinical oral malodor perceived oral malodor were 0.615 and 0.450, respec-
were below the threshold levels of VSCs for oral mal- tively, indicating that 61.5% of subjects who perceived
odor (1.0 ⫾ 0.6 ng/10 mL and 0.4 ⫾ 0.3 ng/10 mL, their oral malodor actually had clinical oral malodor,
respectively). Corresponding concentrations for subjects and 45.0% of subjects who did not perceive their oral
with clinical oral malodor were higher than the threshold malodor actually did not have clinical oral malodor.
levels of VSCs (6.9 ⫾ 3.7 ng/10 mL and 5.4 ⫾ 4.1
ng/10 mL, respectively), and significantly higher than
those without clinical oral malodor. Bivariate relationship of oral malodor status
The relationships of self-perceived and clinical oral
Agreement between self-perceived and clinical malodor with sociodemographics, health behavioral
oral malodor characteristics, and oral health status of the subjects are
When clinical oral malodor was diagnosed by the or- shown in Table III. Smoking habit, alcohol consump-
ganoleptic test, the sensitivity and specificity of self- tion, and dental visit frequency were significant factors
perceived malodor were 0.474 and 0.592, respectively, for self-perceived oral malodor. Subjects who per-
meaning that 47.4% of subjects correctly perceived that ceived their oral malodor were more likely to be current
they had oral malodor, whereas 59.2% correctly per- smokers or alcohol drinkers or persons who often vis-
ceived that they did not. Conversely, 40.8% of subjects ited a dentist for a checkup. On the other hand, level of
thought they suffered from oral malodor when it was education, smoking habit, dental visit frequency, and
not clinically diagnosed (false positive) and 52.6% did tongue-cleaning frequency were significant factors for
not recognize their actual oral malodor (false negative). clinical oral malodor. Significantly higher percentages
The positive and negative predictive values of self- of subjects who had lower education, were current
OOOO ORIGINAL ARTICLE
Volume 113, Number 1 Pham et al. 75

Table III. Bivariate relationships of oral malodor with sociodemographic, health behavioral characteristics and oral
health status
Self-perceived oral malodor Clinical oral malodor
(⫹) (⫺) (⫹) (⫺)
Variables No. (%)/Mean (SD) No. (%)/Mean (SD) P value No. (%)/Mean (SD) No. (%)/Mean (SD) P value
Sociodemography and health
behavior
Level of education
Primary and secondary school 95 (48.7) 100 (51.3) .156 149 (76.4) 46 (23.6) ⬍.001
High school or higher 157 (42.4) 213 (57.6) 178 (48.1) 192 (51.9)
Smoking habit
Non- and past-smoker 95 (30.9) 212 (69.1) ⬍.001 150 (48.9) 157 (51.1) ⬍.001
Current smoker 157 (60.9) 101 (39.1) 177 (68.6) 81 (31.4)
Alcohol consumption
Yes 85 (66.9) 42 (33.1) ⬍.001 82 (64.6) 45 (35.4) .102
No 167 (38.1) 271 (61.9) 245 (55.9) 193 (44.1)
Dental visit frequency
Often and sometime 124 (38.8) 196 (61.2) .002 126 (39.4) 194 (60.6) ⬍.001
Never, rarely, only when having 128 (52.2) 117 (47.8) 201 (82.0) 44 (18.0)
a problem
Tooth brushing frequency
Once/d 80 (44.7) 99 (55.3) .999 101 (56.4) 78 (43.6) .648
ⱖ2 times/d 172 (44.6) 214 (55.4) 226 (58.5) 160 (41.5)
Tongue-cleaning frequency
Yes 88 (45.4) 106 (54.6) .859 51 (26.3) 143 (73.7) ⬍.001
No 164 (44.2) 207 (55.8) 276 (74.4) 95 (25.6)
Mouth rinse use
Yes 168 (45.9) 198 (54.1) .426 196 (53.6) 170 (46.4) .006
No 84 (42.2) 115 (57.8) 131 (65.8) 68 (34.2)
Oral health status
No. of decayed teeth 2.82 (2.83) 3.27 (3.46) .091 3.24 (3.39) 2.84 (2.89) .144
No. of teeth with 5 mm or greater 2.44 (4.03) 1.89 (3.80) .095 3.57 (4.59) 0.17 (0.83) ⬍.001
pockets
No. of teeth with bleeding sites 5.42 (5.37) 4.79 (5.28) .159 7.17 (5.61) 2.18 (3.12) ⬍.001
Plaque Index 2.75 (1.51) 2.50 (1.48) .051 3.01 (1.49) 2.07 (1.34) ⬍.001
Gingival Index 1.28 (1.08) 1.15 (0.97) .132 1.55 (1.02) 0.73 (0.82) ⬍.001
Tongue coating score 2.02 (0.78) 2.03 (0.82) .887 2.47 (0.62) 1.41 (0.61) ⬍.001
BANA test score 1.03 (0.67) 1.10 (0.75) .245 1.37 (0.60) 0.66 (0.65) ⬍.001
(⫹) or (⫺) was expressed as “with” or “without” in each oral malodor status.
BANA, N-benzoyl-DL-arginine-2-napthilamide.

smokers, did not visit dental offices for dental checkups justed for age and gender. The resulting factors signif-
regularly, did not clean their tongue daily, and did not icantly associated with self-perceived oral malodor
use a mouth rinse were in the group with clinical oral were smoking habit (odds ratio [OR] ⫽ 2.7) and alco-
malodor. hol consumption (OR ⫽ 2.4). Subjects who were
There were no significant relationships between self- current smokers or drank alcohol daily were more
perceived oral malodor and any of the oral health likely to perceive oral malodor; however, dental visit
parameters. On the other hand, the group with clinical behavior was not significantly associated with self-
oral malodor had a significantly higher number of teeth perceived oral malodor.
with 5 mm or greater pockets, number of teeth with The significantly related factors for clinical oral mal-
bleeding sites, plaque index, gingival index, tongue- odor in the bivariate analysis were also entered into a
coating score, and BANA test score. The number of logistic regression model that was adjusted for age and
decayed teeth was not a significant factor of clinical gender. Factors significantly associated with clinical
oral malodor. oral malodor were level of education (OR ⫽ 3.7),
dental visit frequency (OR ⫽ 4.1), tongue-cleaning
Logistic regression analysis of oral malodor status frequency (OR ⫽ 3.0), mouth rinse use (OR ⫽ 2.6),
Factors significantly related to self-perceived oral deep periodontal pockets (OR ⫽ 5.4), gingivitis (OR ⫽
malodor in the bivariate analyses were further en- 3.2), a tongue coating (OR ⫽ 9.0), and a high BANA
tered into a logistic regression model that was ad- test score (OR ⫽ 2.8 and 7.9). Smoking habit, bleeding
ORAL MEDICINE OOOO
76 Pham et al. January 2012

Table IV. Logistic regression analyses of self-perceived and clinical oral malodor
Variables No. (%) OR (95% CI) P value
Self-perceived oral malodor
Smoking habit
Non- and past-smoker 307 (54.3) 1.0
Current smoker 258 (45.7) 2.7 (1.7-4.4) ⬍.001
Alcohol consumption
No 438 (77.5) 1.0
Yes 127 (22.5) 2.4 (1.5-3.9) ⬍.001
Dental visit frequency
Often and sometimes 320 (56.6) 1.0
Never, rarely, only when having a problem 245 (43.4) 1.1 (0.8-1.7) .491
Clinical oral malodor
Level of education
High school or higher 370 (65.5) 1.0
Primary and secondary school 195 (34.5) 3.7 (2.0-6.7) ⬍.001
Smoking habit
Non- and past-smoker 307 (54.3) 1.0
Current smoker 258 (45.7) 1.3 (0.6-2.7) .482
Dental visit frequency
Often and sometimes 320 (56.6) 1.0
Never, rarely, only when having a problem 245 (43.4) 4.1 (2.3-7.4) ⬍.001
Tongue-cleaning frequency
Yes 194 (34.3) 1.0
No 371 (65.7) 3.0 (1.7-5.3) ⬍.001
Mouth rinse use
Yes 366 (64.8) 1.0
No 199 (35.2) 2.6 (1.5-4.6) .001
Tooth with 5 mm or greater pocket
0 392 (69.4) 1.0
ⱖ1 173 (30.6) 5.4 (2.5-11.5) ⬍.001
Tooth with bleeding sites
0 185 (32.7) 1.0
ⱖ1 380 (67.3) 0.5 (0.2-1.1) .096
Plaque Index
Median 0–2.5 288 (51.0) 1.0
Median 3–5 277 (49.0) 1.4 (0.7-2.7) .313
Gingival Index
Median 0–1 373 (66.0) 1.0
Median 1.5–3 192 (34.0) 3.2 (1.6-6.6) .001
Tongue-coating score
0 and 1 161 (28.5) 1.0
2 and 3 404 (71.5) 9.0 (4.5-18.1) ⬍.001
BANA test score
0 126 (22.3) 1.0
1 273 (48.3) 2.8 (1.4-5.9) .005
2 166 (29.4) 7.9 (3.2-19.8) ⬍.001
Adjusted by age and gender.
OR, adjusted odd ratio; CI, confidence interval; BANA, N-benzoyl-DL-arginine-2-napthilamide.

gums, and having dental plaque were not significantly ior and dental caries did not show any relationship with
associated with clinical oral malodor (Table IV). either self-perceived or clinical oral malodor.
There have been many epidemiologic studies on oral
DISCUSSION malodor, but comparison of the results is often difficult
Prevalence of oral malodor because each researcher used different criteria of oral
Current findings showed that self-perception is not a malodor and different study populations. The overall
valid method of judging one’s own oral malodor. Self- prevalence of clinical oral malodor was 57.9% in the
perceived oral malodor was associated with smoking current sample, a higher prevalence than in previous
and alcohol consumption, factors different from those studies conducted in the general population.32,33 This
found in clinical oral malodor. Tooth-brushing behav- may be because our study subjects were patients visit-
OOOO ORIGINAL ARTICLE
Volume 113, Number 1 Pham et al. 77

ing the dental hospital and a high percentage of them own malodor. Further, the perception of halitosis is
had periodontal diseases or poor oral hygiene, which reported to differ among culturally different popula-
are major causes of oral malodor. tions.40 Therefore, findings on self-perceived oral mal-
The organoleptic test and VSC concentrations were odor may be highly study specific.
measured in this study. All organoleptic measurements In this study, more than 50% of subjects were not
were conducted by the same examiner, and there was aware of their actual oral malodor: they might be un-
good agreement between the organoleptic test and mea- concerned about their breath or lack enough knowledge
sured H2S and CH3SH concentrations. The H2S and about oral malodor to be aware of it. People who are
CH3SH concentrations of subjects without clinical oral concerned about their own oral malodor tend to ask
malodor (organoleptic score 0 or 1) were below the others if it is noticeable. In contrast, most people who
threshold levels for oral malodor, whereas those with have oral malodor are unlikely to have others point it
an organoleptic score of 2 or more were higher than the out to them.41 Because many current subjects did not
threshold levels. However, self-perception of oral mal- have regular dental checkups, they may also have had
odor did not show good agreement with concentrations no opportunity to receive information on halitosis from
of these gases. dental health professionals that would make them
In this study, the prevalence of self-perceived oral aware of their oral malodor.
malodor was 44.6%, which is higher than in studies by
Al-Ansari et al.,18 Struch et al.,19 Nadanovsky et al.,17 Clinical oral malodor and related factors
or Settineri et al.,34 but is similar to the findings of
Age and gender were not significantly associated with
Bornstein et al.35 and Iwanicka-Grzegorek et al.36 In
either clinically diagnosed or self-perceived oral mal-
the current study, a high percentage of subjects reported
odor in this study. These findings were similar to those
that they had oral malodor; moreover, most subjects
in a Japanese sample33 but different from those of
complaining of oral malodor suffered from this condi- university students in Brazil17 and Kuwaiti patients.18
tion for more than 1 year, and more than 50% of Level of education and dental visit frequency were
subjects with self-perceived oral malodor stated that the linked to clinical oral malodor in the current study.
condition had been reported by others. Thus, these Personal oral hygiene behavior is known to differ by
findings suggest the possibility that perceived oral mal- level of education.42 A higher level of education is
odor might have a negative impact on daily life. related to less oral malodor because these subjects may
have better oral health and be more concerned about
Comparison between self-perceived and clinical professional oral health care and oral hygiene practices.
oral malodor This study indicated that subjects making regular dental
Iwakura et al.37 reported that most patients with com- visits were less likely to have oral malodor. A regular
plaints of halitosis at a dental clinic did not actually dental visit for a checkup might lead to less severe
have halitosis, but suffered from an imaginary halitosis conditions of oral diseases and more prevention-ori-
because of presumptions based on others’ attitudes. ented dental care.
Further, recent research by Bornstein et al. showed a The association between mouth rinse use and clinical
weak correlation35 or no correlation38 between self- oral malodor has also been demonstrated in previous
reported halitosis and organoleptic or VSC measure- studies.26,43 Although we did not explore the kind of
ments. Similarly, in this study we found self-perceived mouth rinse, it was plausible that an antibacterial mouth
oral malodor to have a low sensitivity and specificity; rinse could decrease the production of VSCs.
however, these findings did not agree with those of It is well documented that VSCs are produced in
Iwanicka-Grzegorek et al.,36 who reported that pa- periodontal pockets of individuals with periodontal dis-
tients’ subjective opinions correlated well with objec- ease.33,15,44,45 Our study also confirmed an association
tive evaluation of halitosis, resulting in an estimated between periodontal disease and clinical oral malodor.
sensitivity and specificity for self-perceived oral mal- Patients with gingivitis or at least 1 tooth with a 5-mm
odor of 0.89 and 0.61, respectively. Rosenberg et al.39 or deeper periodontal pocket were more likely to have
suggested that corresponding values were 0.65 and oral malodor. Recent research has also increasingly
0.68, also higher than ours. The apparently contradic- supported the idea that VSCs contribute to periodontal
tory results may be because of a wide range of study disease.46
sample sizes, criteria for sample recruitment, and the The dorsum of the tongue is considered to be a major
procedures for oral malodor measurement. In addition, site of oral malodor, and tongue coatings provide an
self-perception of oral malodor may be influenced by ideal environment for VSC generation5,8,33,45 owing to
various factors, such as level of education, occupation, the association of oral malodor with the load of gram-
knowledge about halitosis, and the sensitivity to one’s negative anaerobes in tongue coatings.47,48 In addition,
ORAL MEDICINE OOOO
78 Pham et al. January 2012

tongue cleaning has an important effect on oral malodor This study had some limitations. We used a cross-
reduction.15,47 In accordance with these previous stud- sectional design and convenience sample of dental pa-
ies, our study indicated that tongue-coating accumula- tients. Hence, the sample may not be representative of
tion, levels of BANA-positive bacteria in the tongue the general population. Future studies should be con-
coating, and daily tongue-cleaning behavior were sig- ducted with a randomly drawn sample at a population
nificant factors affecting clinical oral malodor. level; however, the findings from the current study
Smoking and alcohol ingestion can result in transient revealed a poor agreement between self-perceived and
oral malodor because some substances in tobacco cause clinical oral malodor and a clear difference between the
xerostomia and some alcoholic beverages produce vol- factors related to each malodor status in a large group
atile compounds.49,50 Cigarette smoke contains a vola- of dental patients. Such a study population and design
tile sulfur compound that could be detected using the has not been reported previously. The current study
Halimeter.51,52 Alcohol may also dry out the mouth, would provide further evidence that might help dental
contributing to oral malodor formation.41 A correlation professionals to diagnose and treat patients with psy-
between clinical oral malodor and smoking behavior33 chological halitosis. Our study also demonstrated that
or alcohol consumption39 has been demonstrated; how- people are incapable of precisely perceiving their own
ever, smoking habit and alcohol consumption were not oral malodor in an objective fashion or have oral mal-
associated with clinical oral malodor in this study. The odor self-image, as in a previous report.54 The results
conflicting results might be a consequence of differ- also suggested that patients’ knowledge and behavior
ences in measurement methods or cultural differences concerning oral health, as well as oral pathologic and
in the habit and consumption style, including type and psychological halitosis, needs improvement. The role
frequency of smoking or alcohol drinking. of dental professionals in maintaining good oral health
and preventing oral malodor should be emphasized to
dental patients. Further, community programs and oral
Self-perceived oral malodor and related factors health promotion activities should also be strengthened
Identification of the factors potentially relating to a with provision of adequate information on halitosis and
patient’s complaint is essential for a proper diagnosis its risk factors.
and management in psychological halitosis. We found
that self-perceived oral malodor was not associated CONCLUSIONS
with any of the oral health parameters in this study. Current findings indicated that self-perception is not a
Sociodemographics and health behavioral characteris- valid method to judge one’s own oral malodor. Self-
tics were also not linked with self-perceived oral mal- perceived oral malodor was associated with a smoking
odor. In this and previous studies, smoking habit18 and habit and alcohol consumption, factors different from
alcohol consumption34 were identified as factors related those found in clinical oral malodor.
only to self-perceived oral malodor. This may have
been because the examiner excluded the smoking and REFERENCES
alcohol smell from the criteria of oral malodor diagno- 1. Rosenberg M, Kozlovsky A, Gelernter I, Cherniak O, Gabbay J,
sis. In addition, smoking or alcohol smells might not Baht R, et al. Self-estimation of oral malodor. J Dent Res
1995;74:1577-82.
have been detected by the organoleptic test at the time
2. Bosy A. Oral malodor: philosophical and practical aspects. J Can
of measurement because the concentration of detectable Dent Assoc 1997;63:196-201.
smell depends on the time elapsed since the last tobacco 3. Oho T, Yoshida Y, Shimazaki Y, Yamashita Y, Koga T. Char-
was smoked or the last alcohol was consumed. acteristics of patients complaining of halitosis and the usefulness
A high percentage of subjects were current smokers of gas chromatography for diagnosing halitosis. Oral Surg Oral
Med Oral Pathol Oral Radiol Endod 2001;91:531-4.
(76% in men) in this study. This finding was in agree- 4. Tonzetich J, McBride BC. Characterization of volatile sulphur
ment with a previous report,53 which indicated that production by pathogenic and non-pathogenic strains of oral
Vietnam had the highest rate of male smoking in the Bacteroides. Arch Oral Biol 1981;26:963-9.
world, with more than 72% of men who smoke. This 5. Lee CH, Kho HS, Chung SC, Lee SW, Kim YK. The relationship
between volatile sulfur compounds and major halitosis-inducing
situation, however, might be a possible confounder of
factors. J Periodontol 2003;74:32-7.
oral malodor. Clinical oral malodor was not associated 6. Tonzetich J. Production and origin of oral malodor: a review of
with smoking habit or alcohol consumption. It is plau- mechanisms and methods of analysis. J Periodontol 1977;48:
sible that current study subjects might perceive their 13-20.
smoking and alcohol smells as oral malodor. Such an 7. Persson S, Edlund MB, Claesson R, Carlsson J. The formation of
hydrogen sulfide and methyl mercaptan by oral bacteria. Oral
explanation is supported by the fact that a high percent- Microbiol Immunol 1990;5:195-201.
age of men, who smoked or drank more than women, 8. Loesche WJ, Kazor C. Microbiology and treatment of halitosis.
reported perceiving their own oral malodor. Periodontol 2000 2002;28:256-79.
OOOO ORIGINAL ARTICLE
Volume 113, Number 1 Pham et al. 79

9. Awano S, Gohara K, Kurihara E, Ansai T, Takehara T. The 28. Suzuki N, Yoneda M, Naito T, Iwamoto T, Hirofuji T. Relation-
relationship between the presence of periodontopathogenic ship between halitosis and psychologic status. Oral Surg Oral
bacteria in saliva and halitosis. Int Dent J 2002;52(Suppl Med Oral Pathol Oral Radiol Endod 2008;106:542-7.
3):212-6. 29. Van der Weijden GA, Timmerman MF, Nijboer A, Reijerse E,
10. Loesche WJ, Bretz WA, Kerschensteiner D, Stoll J, Socransky SS, Van der Velden U. Comparison of different approaches to assess
Hujoel P, et al. Development of a diagnostic test for anaerobic bleeding on probing as indicators of gingivitis. J Clin Periodontol
periodontal infections based on plaque hydrolysis of benzoyl-DL- 1994;21:589-94.
arginine-naphthylamide. J Clin Microbiol 1990;28:1551-9. 30. Löe H. The Gingival Index, the Plaque Index and the Retention
11. Rosenberg M, McCulloch CA. Measurement of oral malodor: Index systems. J Periodontol 1967;38(Suppl):610-6.
current methods and future prospects. J Periodontol 1992; 31. Loesche WJ, Kazor CE, Taylor GW. The optimization of the
63:776-82. BANA test as a screening instrument for gingivitis among sub-
12. Donaldson AC, Riggio MP, Rolph HJ, Bagg J, Hodge PJ. Clin- jects seeking dental treatment. J Clin Periodontol 1997;24:
ical examination of subjects with halitosis. Oral Dis 718-26.
2007;13:63-70. 32. Liu XN, Shinada K, Chen XC, Zhang BX, Yaegaki K, Kawa-
13. Murata T, Yamaga T, Iida T, Miyazaki H, Yaegaki K. Classifi- guchi Y. Oral malodor-related parameters in the Chinese general
cation and examination of halitosis. Int Dent J 2002;52(Suppl population. J Clin Periodontol 2006;33:31-6.
3):181-6. 33. Miyazaki H, Sakao S, Katoh Y, Takehara T. Correlation between
14. van den Broek AM, Feenstra L, de Baat C. A review of the volatile sulphur compounds and certain oral health measure-
current literature on aetiology and measurement methods of ments in the general population. J Periodontol 1995;66:679-84.
halitosis. J Dent 2007;35:627-35. 34. Settineri S, Mento C, Gugliotta SC, Saitta A, Terranova A,
15. Pham TAV, Ueno M, Zaitsu T, Takehara S, Shinada K, Lam PH, Trimarchi G, et al. Self-reported halitosis and emotional state:
et al. Clinical trial of oral malodor treatment in patients with impact on oral conditions and treatments. Health Qual Life
periodontal dideases. J Periodontal Res 2011;46(6):722-9. Outcomes 2010;8:34.
16. Murata T, Rahardjo A, Fujiyama Y, Yamaga T, Hanada M, 35. Bornstein MM, Kislig K, Hoti BB, Seemann R, Lussi A. Prev-
Yaegaki K, et al. Development of a compact and simple gas alence of halitosis in the population of the city of Bern, Swit-
chromatography for oral malodor measurement. J Periodontol zerland: a study comparing self-reported and clinical data. Eur
J Oral Sci 2009;117:261-7.
2006;77:1142-7.
36. Iwanicka-Grzegorek E, Michalik J, Kepa J, Wierzbicka M, Alek-
17. Nadanovsky P, Carvalho LB, Ponce de Leon A. Oral malodour
sinski M, Pierzynowska E. Subjective patients’ opinion and
and its association with age and sex in a general population in
evaluation of halitosis using halimeter and organoleptic scores.
Brazil. Oral Dis 2007;13:105-9.
Oral Dis 2005;11(Suppl 1):86-8.
18. Al-Ansari JM, Boodai H, Al-Sumait N, Al-Khabbaz AK, Al-
37. Iwakura M, Yasuno Y, Shimura M, Sakamoto S. Clinical char-
Shammari KF, Salako N. Factors associated with self-reported
acteristics of halitosis: differences in two patient groups with
halitosis in Kuwaiti patients. J Dent 2006;34:444-9.
primary and secondary complaints of halitosis. J Dent Res
19. Struch F, Schwahn C, Wallaschofski H, Grabe HJ, Völzke H,
1994;73:1568-74.
Lerch MM, et al. Self-reported halitosis and gastro-esophageal
38. Bornstein MM, Stocker BL, Seemann R, Bürgin WB, Lussi A.
reflux disease in the general population. J Gen Intern Med
Prevalence of halitosis in young male adults: a study in Swiss
2008;23:260-6.
army recruits comparing self-reported and clinical data. J Peri-
20. Tourbier IA, Doty RL. Sniff magnitude test: relationship to odor
odontol 2009;80:24-31.
identification, detection, and memory tests in a clinic population. 39. Rosenberg M, Knaan T, Cohen D. Association among bad
Chem Sens 2007;32:515-23. breath, body mass index, and alcohol intake. J Dent Res
21. Nalçaci R, Sönmez IS. Evaluation of oral malodor in children. 2007;86:997-1000.
Oral Surg Oral Med Oral Pathol Oral Radiol Endod 40. Rayman S, Almas K. Halitosis among racially diverse popula-
2008;106:384-8. tions: an update. Int J Dent Hyg 2008;6:2-7.
22. Kida IA, Manyori C, Masalu JR. Prevalence and correlates of 41. Rosenberg M. The science of bad breath. Sci Am 2002;286:72-9.
perceived oral malodor among adolescents in Temeke district, 42. Davidson PL, Rams TE, Andersen RM. Socio-behavioral deter-
Dar es Salaam. East Afr J Public Health 2010;7(1):49-53. minants of oral hygiene practices among USA ethnic and age
23. Nalcaci R, Baran I. Oral malodor and removable complete den- groups. Adv Dent Res 1997;11:245-53.
tures in the elderly. Oral Surg Oral Med Oral Pathol Oral Radiol 43. Quirynen M, Mongardini C, van Steenberghe D. The effect of a
Endod 2008;105:e5-9. 1-stage full-mouth disinfection on oral malodor and microbial
24. Sopapornamorn P, Ueno M, Shinada K, Yanagishita M, Kawa- colonization of the tongue in periodontitis. A pilot study. J
guchi Y. Relationship between total salivary protein content and Periodontol 1998;69:374-82.
volatile sulfur compounds levels in malodor patients. Oral Surg 44. Coli JM, Tonzetich J. Characterization of volatile sulphur com-
Oral Med Oral Pathol Oral Radiol Endod 2007;103:655-60. pounds production at individual gingival crevicular sites in hu-
25. Iwamoto T, Suzuki N, Tanabe K, Takeshita T, Hirofuji T. Effects mans. J Clin Dent 1992;3:97-103.
of probiotic Lactobacillus salivarius WB21 on halitosis and oral 45. Yaegaki K, Sanada K. Volatile sulfur compounds in mouth air
health: an open-label pilot trial. Oral Surg Oral Med Oral Pathol from clinically healthy subjects and patients with periodontal
Oral Radiol Endod 2010;110:201-8. disease. J Periodont Res 1992;27:233-8.
26. Shinada K, Ueno M, Konishi C, Takehara S, Yokoyama S, 46. Yaegaki K, Qian W, Murata T, Imai T, Sato T, Tanaka T, et al.
Kawaguchi Y. A randomized double blind crossover placebo- Oral malodorous compound causes apoptosis and genomic DNA
controlled clinical trial to assess the effects of a mouthwash damage in human gingival fibroblasts. J Periodont Res
containing chlorine dioxide on oral malodor. Trials 2008;9:71. 2008;43:391-9.
27. Polat HB, Ozdemir H, Ay S. Effect of different mouth rinses on 47. Quirynen M, Avontroodt P, Soers C, Zhao H, Pauwels M, van
third molar surgery-related oral malodor. Oral Surg Oral Med Steenberghe D. Impact of tongue cleansers on microbial load and
Oral Pathol Oral Radiol Endod 2008;105:e1-8. taste. J Clin Periodontol 2004;31:506-10.
ORAL MEDICINE OOOO
80 Pham et al. January 2012

48. De Boever EH, Loesche WJ. Assessing the contribution of of the transnational tobacco corporations. JAMA 1997;277:
anaerobic microflora of the tongue to oral malodor. J Am Dent 1726-31.
Assoc 1995;126:1384-93. 54. Eli I, Baht R, Koriat H, Rosenberg M. Self-perception of breath
49. Hughes FJ, McNab R. Oral malodour—a review. Arch Oral Biol odor. J Am Dent Assoc 2001;132:621-6.
2008;53(Suppl 1):S1-7.
50. Dal Rio AC, Nicola EM, Teixeira AR. Halitosis—an assessment Reprint requests:
protocol proposal. Braz J Otorhinolaryngol 2007;73:835-42. Masayuki Ueno, DDS, MPH, PhD
51. Rosenberg M. Clinical assessment of bad breath: current con- Tokyo Medical and Dental University
cepts. J Am Dent Assoc 1996;127:475-82. Department of Oral Health Promotion
52. Christen AG. The impact of tobacco use and cessation on oral 1-5-45 Yushima
and dental diseases and conditions. Am J Med 1992;93:25S-31S. Bunkyo-ku
53. Jenkins CN, Dai PX, Ngoc DH, Kinh HV, Hoang TT, Bales S, et Tokyo 113– 8549, Japan
al. Tobacco use in Vietnam. Prevalence, predictors, and the role ueno.ohp@tmd.ac.jp
Eur J Oral Sci 2015; 123: 72–79 Ó 2015 Eur J Oral Sci
DOI: 10.1111/eos.12169 European Journal of
Printed in Singapore. All rights reserved
Oral Sciences

o G. Soares1, Grazyna Jonski2,

Le
Short-term effect of strontium- and Eduardo M. B. Tinoco1, Alix Young3
1
Department of Periodontology, Universidade
zinc-containing toothpastes and Estadual do Rio, Rio de Janeiro-RJ, Brazil;
2
Clinical Research Laboratory, Faculty of
Dentistry, University of Oslo, Oslo, Norway;
mouthrinses on volatile sulphur 3
Department of Cariology and Gerodontology,
Institute of Clinical Dentistry, Faculty of
Dentistry, University of Oslo, Oslo, Norway
compounds in morning breath: a
randomized, double-blind, cross-over
clinical study
Soares LG, Jonski G, Tinoco EMB, Young A. Short-term effect of strontium- and
zinc-containing toothpastes and mouthrinses on volatile sulphur compounds in morning
breath: a randomized, double-blind, cross-over clinical study.
Eur J Oral Sci 2015; 123: 72–79. © 2015 Eur J Oral Sci
Zinc (Zn) reduces the formation of volatile sulphur compounds (VSCs) associated
with oral malodour. Although strontium (Sr) is included in some products for
reducing dental hypersensitivity, it may also have anti-halitosis properties. This ran-
domized, double-blind, cross-over clinical study compared the anti-VSC effect of
brushing with commercial toothpastes and rinses containing Zn and Sr. The volun-
teers (n = 30) either brushed/rinsed with/without tongue brushing using Zn-contain-
ing toothpaste/rinse, Sr-containing toothpaste/rinse, or placebo (control). Volatile
sulphur compounds [hydrogen sulphide (H2S) and methyl mercaptan (CH3SH)]
were measured, in morning breath, using gas chromatography. The anti-VSC effects
of the test toothpastes and test rinses were significantly better than the anti-VSC
effects of the respective controls. Toothbrushing with test toothpastes gave median
reductions, compared with the control, of 70% for H2S and 55–57% for CH3SH.
o G. Soares, Rua Alfredo da Costa Mattos
Le
Rinsing with the Sr- and Zn-containing solutions had the same anti-VSC effect as Jr, 75/2 - Paraı́ba do Sul, RJ – RJ, CEP
toothbrushing and tooth- and tongue brushing with the Sr- and Zn-containing 25850-000, Brazil.
toothpastes. Zinc-containing rinse resulted in a significantly higher median salivary
E-mail: dr_leog@hotmail.com
level of Zn compared with brushing with Zn-containing toothpaste, although this
effect did not correlate with the anti-VSC effect. It can be concluded that the
Sr- and Zn-containing toothpastes and the Zn- and Sr-containing rinses, when used Key words: gas chromatography; halitosis;
in the evening, are equally effective in reducing morning-breath VSCs the following strontium; volatile sulphur compound; zinc
day. Accepted for publication December 2014

Oral malodour, or bad breath, is a common problem coating, and periodontal disease, are the main oral
that affects individuals of different ages, and it has sources of malodour (4–7). When oral bacteria catabo-
become a topic of increasing interest in the scientific lize available sulphur-containing substrates in the oral
community (1). A recent study reported that among cavity, many species produce volatile sulphur com-
patients attending a halitosis clinic, 83% were diag- pounds (VSCs) (4).
nosed with true halitosis, of which 96% had an oral Hydrogen sulphide (H2S), methyl mercaptan (MM;
cause (2). For most people it is not uncommon to expe- CH3SH), and dimethyl sulphide [DMS; (CH3SH)2S] are
rience some degree of bad breath (also called morning- the main oral VSCs (8, 9). These compounds have
breath odour) upon awakening after a night’s sleep. unpleasant odours, even when present at very low levels,
This is caused, in part, by a physiological reduction in and the concentration of compound is strongly corre-
saliva secretion during sleep (3) that results in the build lated to the degree of oral malodour (7, 10). In addition
up of sulphur-containing gasses in the oral cavity (4). to using organoleptic methods for detection and diagno-
Oral malodour is usually a symptom of local oral con- sis of oral malodour, several portable devices are
ditions or infections, whereby accumulations of oral available for analysis and quantification of VSCs. How-
bacteria, often associated with gingivitis, tongue ever, for research purposes, gas chromatography (GC)
 Effect of zinc and strontium on oral VSCs 73

provides the most objective and accurate quantification study protocol was approved by the Committee for Medical
of the different VSC components (9, 11). Research Ethics (REK sør-øst: 2012/548D). The experiment
In many cases the treatment of oral malodour will was performed at the Clinical Research Laboratory, Faculty
involve physical measures to resolve gingivitis and/or of Dentistry, University of Oslo (Oslo, Norway) from June
systematic treatment of periodontal disease (5, 12, 13). 2012 to November 2012.
Removal of the tongue coating is also considered to be
an important measure (10, 14, 15). In some cases, a Study participants
chemical approach may also be appropriate for short
The number of participants recruited to the study was
periods of time, and various oral hygiene products
determined using a standardized sample size calculation
(such as mouthrinses or toothpastes containing metal (SigmaPlot statistical software; Systat Software, San Jose,
salts and various antibacterial substances) are targeted CA, USA). The calculation was based on the number of
for reducing oral malodour (16–18). The anti-VSC effi- treatment groups, the cross-over study design, a statistical
cacy of metal ions relates to a combination of their power of 80%, and a significance level of 5%. A true dif-
antibacterial activity and their affinity for sulphur (S) ference in the percentage reduction of VSCs between test
(8, 18–21). Zinc (Zn) has known antibacterial effects as treatments and controls was suggested to be 2.4%, with a
well as good affinity for S ions. When Zn ions bind to common standard deviation of the error of 2%. The esti-
S ions, the resulting ZnS compounds have low solubil- mated number of individuals needed for the chosen study
ity, and the odiferous gases are, in this way, eliminated design was therefore 28 (2). Interested persons responded
to an advertisement mounted on information boards at
from the oral cavity, thus explaining the well-docu-
university student accommodation. In order to allow
mented anti-VSC effects of ZnS (22, 23). for some dropout, the first 30 persons who fulfilled the
In vitro studies have shown that other metals, such inclusion criteria were recruited and received written infor-
as tin and copper, are also effective in inhibiting VSCs mation about the study. Respondents recruited were
(24). Strontium (Sr) is an alkaline, soft, highly reactive non-smokers, 18–50 yr of age, with no self-reported his-
metal that forms neutral aqueous salt solutions. Stron- tory of systemic diseases or periodontal disease, no
tium chloride (SrCl) is used in cosmetics (skin condi- complaints of xerostomia, and no regular use of medicines.
tioning and soothing) and is considered to have Before starting the study, informed written consent
relatively low toxicity (25). Strontium has been shown was obtained from all participants: 17 male subjects and
to adsorb strongly to calcified tissues, including den- 13 female subjects, with a mean  SD (range) age of
28  7 (20–45) yr. In a short prestudy questionnaire,
tine. It has been widely tested for its effect in reducing
participants were asked about their current oral-hygiene
dentine hypersensitivity, where the beneficial effects routines for brushing, as well as interdental and tongue-
have been attributed to blockage of the organic matrix hygiene measures and their usual diet.
on tooth root surfaces (26, 27). Strontium salts are
included in several commercially available toothpastes
developed for patients suffering from dentine hypersen- Test toothpastes and solutions
sitivity (26, 28, 29).
Details of the Zn- and Sr-containing toothpastes and solu-
Dentine hypersensitivity is often a problem in tions tested in the study are given in Table 1. When pre-
patients with periodontal disease. Following mechani- paring the rinsing solutions, the concentration of Zn and
cal-debridement procedures, tooth-root surfaces often Sr was calculated to match the concentration of the metal
become exposed and patients can experience sensitivity. salts in the respective toothpastes. The preparation and
Although Sr is expected to have S-binding properties, coding of the test agents was performed by a different
no studies have been reported on the anti-VSC effect of researcher from the one in contact with the participants
Sr-containing toothpaste. It was therefore of interest to and who also performed the VSC analyses.
examine whether a commercial Sr-containing tooth-
paste for reducing dentine hypersensitivity can also
Experimental phase
reduce the levels of VSCs in morning breath.
The aim of the present study was to test the 12-h The participants received standard fluoride toothpaste for
anti-VSC effects of toothbrushing (with and without use in a 7- to 10-d washout period before starting the
tongue brushing) with two commercial toothpastes – study. The same toothpaste was used during the study per-
one containing Sr and one containing Zn. Aqueous iod, except on evenings before test days. Based on the
solutions of Zn and Sr were also tested in order to results of pilot studies, a washout period of a minimum of
3 d between different treatments was used in order to
compare the anti-VSC effects of these metal ions when
avoid any carry-over/interference between the test treat-
present in mouthwashes. The null hypothesis tested was ments. The participants were randomly assigned to test the
that there is no difference between the anti-VSC effect different treatments in different orders, and were blinded
of the Sr- and Zn-containing toothpastes and rinses, to the treatments.
using morning breath as a study model system.

Test toothpastes

Material and methods Participants were instructed to brush their teeth for 2 min
with 1 g of preweighed samples of the test and control
This study had a randomized, double-blind, cross-over toothpastes in the evening before their allocated test days.
design and tested 12-h morning breath in volunteers. The They were provided with new disposable toothbrushes for
74 Soares et al.

Table 1
Details of the test methods and test agents

Code Method Test agent Fluoride (p.p.m. F as NaF)

C-t 2 min of toothbrushing Control toothpaste containing no Zn or Sr 1500

C-tt 2 min of toothbrushing + tongue
brushing
Zn-t 2 min of toothbrushing Toothpaste containing 0.003 g of Zn/g of toothpaste 1450
Zn-tt 2 min of toothbrushing + tongue
brushing
Sr-t 2 min of toothbrushing Toothpaste containing 0.08 g of Sr/g of toothpaste 1040
Sr-tt 2 min of toothbrushing + tongue
brushing
C-r 1 min of mouth rinsing Deionized water control – no Zn or Sr –
Zn-r 1 min of mouth rinsing Aqueous solution containing 0.003 g of Zn/10 ml solution –
Sr-r 1 min of mouth rinsing Aqueous solution containing 0.08 g of Sr/10 ml solution –

F, fluorine; NaF, sodium fluoride; Sr, strontium; Zn, zinc.

each toothpaste test and were instructed to use their usual were sealed, coded, and immediately frozen until required
brushing method. It was emphasized to the volunteers that for analysis. Saliva samples collected on test days when
they must use the same technique for each of the test days the participants had brushed with the toothpaste without
involving toothbrushing. When the test involved both Zn or Sr, and when they had rinsed with water, functioned
tooth- and tongue brushing, participants were instructed as the controls for metal concentrations in the saliva for
to brush their tongue using three strokes over the dorsum each participant.
of the tongue after they had brushed their teeth. Before analysis, the saliva was thawed at room tempera-
ture and centrifuged at 3619 g for 20 s at 20°C. Three-mil-
lilitre aliquots of the saliva samples were mixed with 1 ml
Test solutions of 65% nitric acid (HNO3) and 0.5 ml of 5% lanthanum
Participants were instructed to rinse with 10 ml of the test/ oxide (La2O3). The samples were analysed blind for either
control solution for 1 min in the evening before each allo- Zn or Sr ions using atomic absorption spectrometry
cated rinse test day. No toothbrushing or dental flossing (AAS) (Model 3300; Perkin Elmer Analytical Instrument,
was performed on the evenings when the participants rinsed Norwalk, CT, USA) with a wavelength of 422.7 nm and
with the test or control solutions. The participants were an air-acetylene flame. Solutions containing 5 p.p.m. Sr
instructed to refrain from eating/drinking and from other and Zn and 0.5 ml of 5% La2O3 were used as the stan-
oral hygiene practices following the evening treatments and dards for calibration.
until after they were tested, 12 h later. Test participants
met up at the research laboratory in the morning of their Statistical analysis
allocated test days and times, and single morning-breath
samples were immediately collected and analysed for VSCs. For the GC analysis of VSCs, the raw data consisted of
the area under the chromatogram curve (AUC) measure-
ments for each VSC in each breath analysis. The data
VSC measurements were not normally distributed according to the Shapiro–
Morning breath levels of H2S and MM were measured Wilk test, and the Friedman test and post hoc Tukey test
using a gas chromatograph specifically calibrated for sul- were performed to detect differences between treatments
phur detection (Shimadzu, Kyoto, Japan). The volunteers for both H2S and CH3SH. The effects of each treatment
were instructed to keep their mouth closed for 90 s, before method were also calculated as a percentage of the con-
samples of mouth air were aspirated using a mouthpiece trol. These data were also not normally distributed and
and a 10-ml syringe. Samples were injected into a 6-ml the Wilcoxon test was performed to test for differences
sample loop connected to the auto injector of the gas between the treatment effects for the different VSCs.
chromatograph. Analysis was performed directly by sepa- For the salivary metal analysis, the raw data consisted
ration using a Teflon column (3.66 m 9 0.32 cm, tempera- of metal ion concentrations (p.p.m.) in the saliva samples,
ture 70°C, nitrogen gas flow 32 ml min 1, hydrogen gas as measured by AAS. These data were not normally dis-
flow rate 125 ml min 1 and airflow rate 43 ml min 1) tributed and the same statistical analyses were performed
packed with polyphenol ether (5%) and phosphoric acid as for the VSC data.
(0.05%) on 40/60 mesh Chromosorb T and a flame photo- Statistical analyses were performed using SIGMAPLOT
metric detector. statistical software. All differences were considered signifi-
cant at P < 0.05.

Saliva collection and analysis

Unstimulated whole saliva samples were collected from the Results
participants immediately after collection of breath samples.
Participants were instructed to spit into a 15-ml test tube All participants completed all stages of the study and
until at least 3 ml of saliva had been collected. Samples there were no complaints or reactions associated with
 Effect of zinc and strontium on oral VSCs 75

brushing with the toothpastes or rinsing with the were all significantly more effective against MM than
solutions. The entire study was performed over a per- were their respective controls (P < 0.001). Toothbrush-
iod of 18 wk. The results of the prestudy questionnaire ing alone gave a median reduction in MM of 55–57%
showed that 15 (50%) volunteers reported using dental for the test toothpastes compared with the control
floss at least four times per week and 17 (56%) volun- toothpaste (Table 2). The results for tooth- and tongue
teers brushed their tongue regularly. brushing were not significantly different from the
results for tongue brushing only, although the Sr-con-
taining toothpaste tended to have a lower anti-MM
GC analyses of VSCs
effect (median 34% reduction) when the tongue was
Hydrogen sulphide: The area under the chromatogram also brushed. Rinsing with the test solutions resulted in
curve (AUC) for H2S for each treatment method is a median MM reduction, compared with the rinse con-
shown in Fig. 1. Toothbrushing with the Zn- and Sr- trol, of 55% for Zn and 65% for Sr (Table 2). This dif-
containing toothpastes (with and without additional ference was not significant, and the anti-MM effect of
tongue brushing), and rinsing with the Zn- and Sr-ace- the test rinses was not significantly different from that
tate-containing solutions were all significantly more of the test toothpastes.
effective against H2S than their respective controls
(P < 0.001). Toothbrushing alone gave a median reduc-
Analyses of salivary Zn and Sr
tion in H2S of about 70% compared with the control
toothpaste (Table 2). Although there were no signifi- The results of the salivary analyses are shown in Fig. 3.
cant differences in anti-VSC effects between tooth- and There was a significantly higher median 12-h salivary
tongue brushing and toothbrushing alone, the Zn-con- metal concentration (Zn) after rinsing with the Zn-ace-
taining toothpaste tended to have a better anti-H2S tate-containing solution than after the other treatments.
effect (median 85% reduction) when the tongue was Statistical analyses did not reveal any significant rela-
also brushed. Rinsing with the test solutions resulted in tionships between the salivary Zn or Sr levels and the
median H2S reductions, compared with the control respective VSC levels for the different treatment
rinse, of 82% for Zn and 77% for Sr (Table 2). methods.
Although this was slightly higher than for toothbrush-
ing alone for Zn, this difference was not significant.

Methyl mercaptan: The area under the chromatogram

Discussion
curve for MM for each treatment method is shown in The results of this double-blind, cross-over clinical
Fig. 2. Toothbrushing with the Zn- and Sr-containing study on morning breath showed that there was no sig-
toothpastes (with and without tongue brushing), and nificant difference between the anti-VSC effect of the
rinsing with the Zn- and Sr-acetate-containing solutions Zn- and Sr-containing toothpastes tested. Furthermore,
when solutions containing the same concentration of
metal ions as in the respective toothpastes were tested
for their anti-VSC effect, no difference could be seen
between the anti-VSC effects of rinsing with the Zn-
and Sr-acetate solutions. The null hypothesis was there-
fore proven to be correct. Furthermore, the study also
showed that brief tongue brushing in addition to tooth-
brushing had no additional anti-VSC effect.
The protocol for the present study was in accordance
with Standardization of clinical protocols in oral mal-
odour research, including a cross-over study design to
avoid the effects of individual variations, and the use of
GC (11). The randomized cross-over design reduces
variability and enables different products to be tested
on the same subjects, and GC is the preferred method
used at halitosis centres to measure the different VSC
components.
Oral hygiene products contain various active ingredi-
Fig. 1. Box plots of hydrogen sulphide levels (H2S, ents aimed at consumers looking for specific needs.
AUC = area under the chromatogram curve) recorded in Zinc is one agent that has commonly been included in
morning breath of test participants, 12 h after brushing with products for oral malodour, and the efficacy of Zn-con-
the toothpastes or rinsing with the solutions (n = 30 test sub- taining agents against oral malodour has been con-
jects). See Table 1 for an explanation of the codes. The box
plots show the median and upper/lower quartiles. The whis-
firmed. Several studies that have tested the effect of
kers represent the upper 90th and lower 10th percentiles. Out- toothpaste formulations containing various Zn salts
liers are shown as separate points. (Friedman test, P < 0.05, (Zn-chloride, -sulphate, -citrate, and -gluconate) have
¥, Δ, #, *: groups with the same signs are not statistically sig- demonstrated good efficacy against oral malodour (22,
nificantly different from each other). 23). Zinc salts are also demonstrated as efficacious in
76 Soares et al.

Table 2
Percentage reductions, compared with controls, in morning-breath volatile sulphur compounds (VSCs) [hydrogen sulphide (H2S) and
methyl mercaptan (CH3SH)]

Method
Toothbrushing Tooth- and tongue brushing Mouth rinsing
H2S CH3SH H2S CH3SH H2S CH3SH
Test person Sr-t Zn-t Sr-t Zn-t Sr-tt Zn-tt Sr-tt Zn-tt Sr-r Zn-r Sr-r Zn-r

1 97.58 55.93 52.82 59.52 79.72 58.52 38.19 66.53 0 53.28 0 0

2 99.90 94.57 99.36 97.94 81.64 33.12 0 0 0 38.89 0 12.80
3 98.61 97.36 96.41 92.83 97.96 96.83 90.10 85.51 55.19 70.41 84.63 52.13
4 63.93 67.81 46.07 92.56 94.07 81.75 94.66 94.21 89.54 0 70.53 0
5 99.55 0 85.89 0 93.59 0 21.40 63.04 0 0 0 0
6 57.67 0 36.01 21.27 0 0 0 0 0 0 0 0
7 54.75 86.28 52.20 74.01 67.82 0 0 0 39.62 27.95 49.41 52.11
8 5.06 63.54 43.46 93.09 26.18 27.30 43.49 28.00 91.63 87.92 62.51 58.15
9 65.93 22.33 14.08 23.49 0 0 59.47 0 0 0 0 0
10 72.74 72.21 84.65 33.59 32.17 63.22 14.31 54.78 95.97 98.00 88.98 70.90
11 100 88.05 81.36 82.35 55.02 97.70 11.65 88.42 66.38 59.92 72.14 71.35
12 43.86 0 58.80 50.03 99.48 0 86.89 0 90.66 92.86 69.92 51.03
13 43.39 0 21.78 0 97.72 0 77.54 0 76.85 95.11 82.93 97.67
14 81.36 79.58 76.53 17.64 66.24 53.81 30.96 0 99.96 96.79 89.08 51.80
15 0 0 0 0 43.92 0 15.09 0 48.17 0 53.87 0
16 57.30 96.58 14.83 82.27 37.61 96.25 0 86.28 10.20 87.26 36.18 73.25
17 43.42 82.17 30.70 54.25 91.95 96.01 37.08 68.44 87.71 94.11 66.49 85.65
18 54.48 8.99 0 14.42 56.77 86.57 62.63 84.65 72.17 64.66 8.54 73.49
19 68.46 11.68 35.64 0 45.61 84.40 0 47.38 80.85 65.92 69.14 16.62
20 95.43 97.81 77.89 98.22 82.12 93.48 95.63 81.33 98.58 98.26 96.81 86.46
21 93.59 99.60 85.26 95.46 43.37 94.21 0 0 99.79 99.53 16.85 0
22 97.45 97.33 89.52 83.11 73.68 96.15 0 0 77.72 77.15 68.34 100
23 96.10 100 81.49 88.11 92.19 93.23 40.91 26.62 77.16 96.76 42.37 81.02
24 77.59 7.91 84.41 0 62.21 99.25 12.70 89.07 26.58 68.91 0 32.29
25 85.24 99.67 0 78.07 97.53 86.77 79.45 98.08 89.57 97.78 63.63 79.64
26 6.56 31.65 57.24 26.88 81.66 93.63 86.12 91.00 74.76 0 54.92 74.62
27 76.78 99.65 70.52 90.01 6.92 94.50 0 79.39 95.10 97.16 84.94 93.57
28 46.36 0 19.83 15.60 82.12 74.83 51.55 50.66 80.90 99.99 99.69 99.92
29 0 56.28 0 50.47 69.97 98.24 0 89.00 97.38 97.87 94.99 84.13
30 98.08 90.02 85.59 82.44 92.51 99.44 60.06 90.97 99.42 97.87 72.58 45.09
Median 70.60 70.01 55.03 56.88 71.82 85.48 34.02 58.91 77.44 82.21 65.06 55.14

Values are given as percentage reductions in morning-breath VSCs. The median percentage reduction in morning-breath VSCs, following
application of the different test methods, is given for all test participants at the foot of the table.
‘0’, no reduction in VSCs, compared with controls, in morning breath following the treatments.
See Table 1 for details of treatment codes.

mouthrinse formulations (17, 20, 30, 31, 32). The when both the teeth and the tongue were brushed,
present study confirmed the good efficacy of Zn in both compared with when only the teeth were brushed. Fur-
toothpaste and as a rinse against morning-breath odour. thermore, the Sr-containing toothpaste tended to have
The other metal tested, Sr, has not been commonly a poorer effect on MM when both the teeth and the
included in oral hygiene products. Strontium has a doc- tongue were brushed, compared with when only the
umented effect in reducing dental hypersensitivity and teeth were brushed. It would be interesting to perform
has been included in toothpastes for this reason (28). further studies in a more homogeneous test population
Like Zn ions, divalent Sr ions could also be expected to to determine whether these trends indicate true effects.
bind S ions and to have some anti-VSC effect. How- Such a population could be a group of patients with
ever, there are no other studies reporting on the effect periodontitis who have higher background levels of
of Sr-containing products for oral malodour, and the VSCs in morning breath.
Sr-containing toothpaste tested in the present study is The levels of Zn and Sr in saliva were measured, 12 h
not marketed as having an anti-VSC effect. The results after brushing/rinsing with the test toothpastes/solutions,
of the present study indicate, for the first time, that Sr- in order to examine the possible relationship between
containing toothpaste and solutions have an anti-VSC them and their anti-VSC effects. Elevated salivary levels
effect in morning breath. of Zn, 3–4 h after brushing with a Zn-citrate dentifrice,
In the present study, it may be worth noting that has been reported previously (33). Oral retention of Zn
although not statistically significant, the Zn-containing following rinsing with aqueous Zn salts has been mea-
toothpaste tended to have a better effect against H2S sured previously, and the salivary concentration of Zn
 Effect of zinc and strontium on oral VSCs 77

Fig. 3. Box plots showing the salivary concentration (p.p.m.)

Fig. 2. Box plots of methyl mercaptan levels (MM, of zinc (Zn) or strontium (Sr), 12 h after using the tooth-
AUC = area under the chromatogram curve) recorded in pastes and solutions (n = 30 test subjects). See Table 1 for an
morning breath of test participants, 12 h after brushing with explanation of the codes. The box plots show the median and
the toothpastes or rinsing with the solutions (n = 30 test sub- the upper/lower quartiles. The whiskers represent the upper
jects). See Table 1 for an explanation of the codes. The box 90th and lower 10th percentiles. Outliers are shown as sepa-
plots show the median and the upper/lower quartiles. The whis- rate points. (Friedman test, P < 0.05; #, *: groups with the
kers represent the upper 90th and lower 10th percentiles. Out- same signs are not statistically significantly different from each
liers are shown as separate points. (Friedman test, P < 0.05; ¥, other). NB: Zn and Sr were not detectable in the saliva after
Δ, #, *: groups with the same signs are not statistically signifi- brushing/rinsing with control toothpaste and solutions.
cantly different from each other).

Physiological inter- and intra-individual variations in

was shown to decrease rapidly to about 15% of the rinse VSC levels are reported (38, 39) and provide a chal-
concentration during the first 4 h (34). Normal mean lenge for experimental protocols. In the present study,
whole salivary Zn concentrations have previously been the large variation in VSC levels (both control levels
reported in the range of 0.460–0.478 p.p.m. (35, 36). In and treatment response levels) may partly be a result of
the present study, the mean salivary Zn concentration, the heterogeneous nature of the group of volunteers.
12 h after rinsing with Zn-acetate, was about three times Approximately half of the test participants reported
that level (1.32 mg l 1), clearly suggesting the oral reten- using dental floss four times per week, suggesting a rel-
tion of Zn ions. atively high level of oral hygiene. Although it has been
Regarding Sr, there are few studies reporting normal demonstrated that the low frequency of toothbrushing
salivary Sr concentrations. In a study on 14-yr-old sub- is correlated with halitosis, one study found that there
jects, a mean salivary Sr concentration of 0.035 mg l 1 was no association between the reported use of dental
(range 0.02–2.93 mg l 1) was reported (37). In the pres- floss and halitosis (40). Furthermore, somewhat more
ent study, 12 h after rinsing with Sr-acetate, the mean than half of the test participants in the present study
salivary Sr concentration was 3.89 mg l 1, indicating brushed their tongue regularly.
retention of Sr. The median retention of metal ions Two clinical epidemiological studies found that tongue
was, however, significantly greater following rinsing coating was the main factor associated with high VSC
with the Zn-acetate solution than for the other treat- levels (41, 42). In a group of periodontally healthy dental
ments. This finding may indicate an important differ- students, tongue scraping appeared to be the most
ence in oral substantivity between Zn and Sr ions, at important hygiene procedure for reducing morning hali-
least for these metals in aqueous form, although this tosis (43). For patients with periodontitis, periodontal
difference was not seen with the two metals in the treatment was shown to play an important role in reduc-
toothpastes. ing halitosis, while tongue cleaning contributed to a les-
Toothpastes include many different components, and ser extent; in the same study, tongue cleaning alone was
the rationale for including rinsing with the Zn- and Sr- also shown to be the best approach to reduce halitosis in
acetate solutions in the present study was to evaluate patients with gingivitis (44). A systematic review on the
the effect of the metal ion alone. However, the present effectiveness of mechanical tongue cleaning and tongue
study design and protocol could not demonstrate that coating concluded that tongue cleaning has the potential
rinsing with the Zn- and Sr-containing solutions was to reduce both tongue coating and breath odour (15).
more efficacious than brushing with Zn- and Sr-con- Despite this, the authors concluded that the evidence was
taining toothpastes. Furthermore, no significant corre- not yet convincing regarding the effectiveness of tongue
lations were observed between the salivary metal cleaning on halitosis. In the present study, no significant
concentrations and the anti-VSC effect, 12 h after the difference in the effect on VSCs could be shown when
different treatments. The latter finding may be a func- tooth- and tongue brushing were compared with tooth-
tion of the long time between the evening treatment brushing only. This result may be partly explained by the
regimes and the time of saliva collection the next day. fact that half of the volunteers may not have had much
78 Soares et al.

tongue coating to remove and therefore received no addi- 9. VAN DEN VELDE S, VAN STEENBERGHE D, VAN HEE P, QUIRY-
NEN M. Detection of odorous compounds in breath. J Dent
tional effect of tongue brushing in this study.
Res 2009; 88: 285–289.
It can be concluded, from this study, that there was 10. SEEMAN R, KISON A, BIZHANG M, ZIMMER S. Effectiveness of
no significant difference between the Sr- and Zn-con- mechanical tongue cleaning on oral levels of volatile sulfur
taining toothpastes tested, and their anti-VSC effect on compounds. J Am Dent Assoc 2001; 132: 1263–1267.
morning breath. Furthermore, a similar anti-VSC effect 11. YAEGAKI K, BRUNETTE DM, TANGERMAN A, CHOE Y, WINKEL
was achieved by rinsing with aqueous solutions of Sr EG, ITO S, KITANO T, II H, CALENIC B, ISHKITIEV N, IMAI T.
Standardization of clinical protocols in oral malodor research.
and Zn as for toothbrushing with the test toothpastes. J Breath Res 2012; 6: 1–10.
The null hypothesis was therefore proven to be correct. 12. TONZETICH J, NG SK. Reduction of malodor by oral cleansing
The fact that toothpastes containing Sr, which are procedures. Oral Surg Oral Med Oral Pathol 1976; 42: 172–
aimed at reducing dentine sensitivity, may also have a 181.
13. QUIRYNEN M. Management of oral malodour. J Clin Period-
clinical anti-VSC effect, can be important for some ontol 2003; 30: S17–S28.
patients. Patients with periodontitis usually undergo 14. QUIRYNEN M, AVONTROODT P, SOERS C, ZHAO H, PAUWELS M,
systematic mechanical debridement of their teeth and VAN STEENBERGHE D. Impact of tongue cleansers on microbial
this often results in exposed tooth-root surfaces. Peri- load and taste. J Clin Periodontol 2004; 31: 506–510.
15. VAN DER SLEEN MI, SLOT DE, VAN TRIJFFEL E, WINKEL EG,
odontitis is also very often associated with halitosis.
VAN DER WEIJDEN G. Effectiveness of mechanical tongue
Periodontal patients will, in many cases, experience cleaning on breath odour and tongue coating: a systematic
dental hypersensitivity following treatment, and tooth- review. Int J Dent Hyg 2010; 8: 258–268.
paste that helps with this problem, and at the same 16. SOARES LG, FALABELLA MEV, TEIXEIRA HGC, TINOCO EMB,
time reduces VSCs, may therefore be very useful. SILVA DG. The use of buccal mouthrinses in the reductions of
halitosis: review. Perionews 2009; 3: 48–52.
Interestingly, although a significantly higher median 17. BLOM T, SLOT DE, QUIRYNEN M, VAN DER WEIJDEN GA. The
salivary metal ion level was observed 12 h following effect of mouthrinses on oral malodor: a systematic review.
rinsing with the Zn-containing rinse compared with the Int J Dent Hyg 2012; 10: 209–222.
other treatments, this was not significantly correlated 18. DADAMIO J, LALEMAN I, QUIRYNEN M. The role of toothpastes
in oral malodour management. Monogr Oral Sci 2013; 23:
with the respective anti-VSC effect of the toothpastes
45–60.
and solutions. Furthermore, as noted by other 19. SCHEIE AA. Modes of action of currently known chemical
researchers, no additional effect of tongue brushing anti-plaque agents other than chlorhexidine. J Dent Res 1989;
was seen in the volunteers in the study, who generally 68: 1609–1616.
had a good level of tongue hygiene before starting the 20. YAEGAKI K, SUETAKA T. The effect of zinc chloride mouth-
wash on the production of oral malodour, the degradations
study. of salivary cellular elements, and proteins. J Public Health
Dent 1989; 39: 377–386.
Acknowledgements – The authors would like to thank our volun- € 
21. YOUNG A, JONSKI G, ROLLA G, WALER SM. Effects of metal
teers, the foreign students in Oslo, Norway, for their excellent
salts on the oral production of volatile sulfur-containing com-
cooperation in this study. No external funding was used for this
pounds (VSC). J Clin Periodontol 2001; 28: 776–781.
research apart from the support of the authors’ institution. L
eo
22. NAVADA R, KUMARI H, LE S, ZHANG J. Oral malodor reduc-
G. Soares had a scholarship from CAPES/Brazil – Process no.
tion from a zinc-containing toothpaste. J Clin Dent 2008; 19:
2438-11-9.
69–73.
23. NEWBY EE, HICKLING JM, HUGHES FJ, PROSKIN HM, BOSMA
Conflicts of interest – The authors declare that there are no con- MP. Control of oral malodour by dentifrices measured by gas
flicts of interest for this study. chromatography. Arch Oral Biol 2008; 53: S19–S25.
24. JONSKI G, YOUNG A, WALER  SM, ROLLA€ G. Insoluble zinc,
cupric and tin pyrophosphates inhibit the formation of vola-
tile sulphur compounds. Eur J Oral Sci 2004; 112: 429–432.
References 25. ROGERS CH, SOINE TO, WILSON CO. A textbook of inorganic
pharmaceutical chemistry, 1st edn. Philadelphia, PA: Lea and
1. YAEGAKI K. Advances in breath odor research: re-evaluation Febiger, 1952; 449–450.
and newly-arising sciences. J Breath Res 2012; 6: 1–3. 26. PARKINSON CR, WILLSON RJ. An in vitro investigation of two
€
2. ZURCHER A, FILIPPI A. Findings, diagnoses and results of a currently marketed dentin tubule occlusion dentifrices. J Clin
halitosis clinic over a seven year period. Schweiz Monatsschr Dent 2011; 21: 6–10.
Zahnmed 2012; 122: 205–216. 27. OBERG C, POCHAPSKI MT, FARAGO PV, GRANADO CJ, PILATTI
3. DAWES C. Circadian rhythms in human salivary flow rate and GL, SANTOS FA. Evaluation of desensitizing agents on dentin
composition. J Physiol 1972; 220: 529–545. permeability and dentinal tubule occlusion: an in vitro study.
4. TONZETICH J. Production and origin of oral malodor: a review Gen Dent 2009; 57: 496–501.
of mechanisms and methods of analysis. J Periodontol 1977; 28. LAYER T, HUGHES N. Evidence for the efficacy of an 8%
48: 13–20. strontium acetate dentifrice for instant and lasting relief of
5. SCULLY C, EL-MAAYTAH M, PORTER SR, GREENMAN J. Breath dentin hypersensitivity. J Clin Dent 2010; 21: 56–58.
odor: etiopathogenesis, assessment and management. Eur J 29. PATEL R, CHOPRA S, VANDEVEN M, CUMMINS D. Comparison
Oral Sci 1997; 105: 287–293. of the effects on dentin permeability of two commercially
6. GREENMAN J. Microbial aetiology of halitosis. In: NEWMAN available sensitivity relief dentifrices. J Clin Dent 2011; 22:
HN, WILSON M, eds. Dental plaque revisited: oral biofilms in 108–112.
health and disease. Cardiff: Bioline Publications, 1999; 419– 30. YOUNG A, JONSKI G. Effect of a single brushing with two Zn-
442. containing toothpastes on VSC in morning breath: a 12h,
7. ROSENBERG M. Bad breath and periodontal disease: how randomized, double-blind, cross-over clinical study. J Breath
related are they? J Clin Periodontol 2006; 33: 29–30. Res 2011; 5: 1–8.
8. TONZETICH J. Direct gas chromatographic analysis of sulphur 31. THRANE PS, JONSKI G, YOUNG A. Comparative effects of vari-
compounds in mouth air in man. Arch Oral Biol 1971; 16: ous commercially available mouth-rinse formulations on hali-
587–597. tosis. Dent Health 2010; 49: 6–10.
 Effect of zinc and strontium on oral VSCs 79

32. SAAD S, GREENMAN J, SHAW H. Comparative effects of vari- morning breath of human volunteers. Arch Oral Biol 2011; 56:
ous commercially available mouth rinse formulations on oral 29–34.
malodor. Oral Dis 2010; 17: 180–186. 40. YOUNGNAK-PIBOONRATANAKIT P, VACHIRAROJPISAN T. Preva-
33. SAXTON CA, HARRAP GJ, LLOYD AM. The effect of dentifrices lence of self-perceived oral malodor in a group of Thai dental
containing zinc citrate on plaque growth and oral zinc levels. patients. J Dent 2010; 7: 196–204.
J Clin Periodontol 1986; 13: 301–306. 41. BORNSTEN MM, KISLIG K, HOTI BB, SEEMAN R, LUSSI A.
34. AFSETH J, HELGELAND K, BONESVOLL P. Retention of Cu and Prevalence of halitosis in the population of the city of Bern,
Zn in the oral cavity following rinsing with aqueous solutions Switzerland: a study comparing self-reported and clinical
of copper and zinc salts. Scand J Dent Res 1983; 91: 42–45. data. Eur J Oral Sci 2009; 117: 261–267.
35. MATHUR A, WALLENIUS K, ABDULLA M. Relation between 42. BORNSTEN MM, STOCKER BL, SEEMANN R, BURGIN WB, LUSSI
zinc content in saliva and blood in healthy human adults. A. Prevalence of halitosis in young male adults: a study in
Scand J Clin Lab Invest 1977; 37: 469–472. swiss army recruits comparing self-reported and clinical data.
36. GREGER JL, SICKLES VS. Saliva zinc levels: potential indica- J Periodontol 2009; 80: 24–31.
tors of zinc status. Am J Clin Nutr 1979; 32: 1859–1866. 43. FAVERI M, HAYACIBAR MF, PUPIO GC, CURY JA, TSUZUKI
37. CURZON MEJ. Strontium concentrations in whole human sal- CO, HAYACIBAR RM. A cross-over study on the effect of vari-
iva. Arch Oral Biol 1984; 29: 211–214. ous therapeutic approaches to morning breath odour. J Clin
38. YOUNG A, JONSKI G, RØLLA G. Variation in oral volatile sul- Periodontol 2006; 33: 555–560.
phur compound formation. Acta Odontol Scand 2002; 60: 44. PHAM TA, UENO M, ZAITSU T, TAKEHARA S, SHINADA K, LAM
321–324. PH, KAWAGUCHI Y. Clinical trial of oral malodor treatment
39. SNEL J, BURGERING M, SMIT B, NOORDMAN W, TANGERMAN A, in patients with periodontal diseases. J Periodont Res 2011;
WINKEL EG, KLEEREBEZEM M. Volatile sulphur compounds in 46: 722–729.