The primary aim of the current study was to determine the association between halitosis detection (presence or absence) and periodontal status in non-smoking subjects, and also assess whether halitosis recordings were related to periodontal clinical parameters, tongue coating and quantities of two putative periodontal pathogens on the posterior region of the tongue determined by real-time PCR. Secondary, halitosis recordings were compared among subjects with chronic periodontitis, chronic generalized gingivitis and periodontal health.

A debate in the literature exists as to whether periodontitis is linked with oral malodour and this study attempts to understand the way in which periodontal disease is associated with oral malodour-related parameters.

Clinical protocol:

Seventy-eight systemically healthy non-smokers were screened for oral halitosis in the Department of Periodontology, Aristotle University of Thessaloniki, Greece. These were non-halitosis complaining subjects. After initial screening for suitability and obtaining a signed informed consent approved by the Ethics Committee of the Dental School, Aristotle University of Thessaloniki, a full-mouth periodontal charting was recorded including plaque index (PI), clinical probing depth (CPD), clinical attachment levels (CAL) and bleeding on probing (BOP) at six sites per tooth using a manual periodontal probe (Hu-Friedy XP-23/QW). The subjects that fulfilled the clinical inclusion criteria were classified as chronic periodontitis (N=28), chronic generalised gingivitis (N=23) or healthy individuals (27) and were further clinically and microbiologically monitored.

Assessment of tongue coating:

The Winkel Tongue Coating Index (Winkel et al., 2003) was visually determined by dividing the dorsum of the tongue into sextants. A score between zero and two was given to each sextant according to the amount of deposits and these scores were added giving a total ranging from zero to 12.

Collection of tongue sample:

A specimen was collected from the rear of the tongue dorsum for quantitative analysis of two putative periodontal pathogens, Porphyromonas gingivalis and Fusobacterium nucleatum, by real-time PCR.

In order to avoid interference of bleeding on probing and tongue sampling with the organoleptic assessments and primarily with the evaluation of the tongue odour, subjects were recalled on a second visit at the same time of the day and within a period of four days for oral malodour assessment. Within this short time interval, participants were asked not to change their routine dietary and oral hygiene habits except for 48 hours prior to the next visit when the halitosis assessments were carried out (Roldán et al., 2003; Donaldson et al., 2007). Written instructions were given to each participant regarding food and drink consumption and oral hygiene habits. In more detail, spicy food, onion, garlic, alcohol and mouthwashes were avoided 48 prior to the examination. On the morning of evaluation coffee, mints, chewing gum, oral hygiene practices, scented cosmetics or aftershave lotion were avoided. Consumption of water and oral rinses with water were allowed at least two hours prior to the odour assessment. In a similar manner, the judge who carried out the organoleptic assessments of oral malodour followed the same instructions as the participants for 24 hours before undertaking the examinations.

Oral malodour assessment:

• The Halimeter® (Halimeter® RH-17, Interscan Corporation, Chatsworth, CA, USA) was used to measure the concentration of volatile sulphide compounds (VSC) in parts per billion (p.p.b.) (Rosenberg et al., 1991). The assessments were always carried out between 10.00-12.00 at the same dental chair in the clinic to minimise daily variability in the room's temperature and humidity (Liu et al., 2006);
• Organoleptic assessments (ORG) ranging from zero to five (Rosenberg et al., 1991) assessed the odour of the following parameters:

Whole mouth air (ORG1) - subjects were asked to exhale gently into a 20cm long tube through a screen, to avoid eye contact with the odour judge, who evaluated breath odour at the other end of the tube (Murata et al., 2002); Anterior region of the dorsum of tongue (ORG2) - subjects were asked to lick their wrists and wait a couple of seconds to allow the area to dry out before the odour was assessed; Posterior region of tongue (ORG3) - material was harvested from the rear of the lingual dorsum in a standardized manner using a plastic spoon; Nose air - subjects were asked to exhale through one nostril after capping the opposite nostril with their finger in order to exclude extra-oral causes that may be related to bad breath.

A calibrated single "examiner" recorded the periodontal measurements and Halimeter® readings, whereas one calibrated "judge" carried out the organoleptic assessments of halitosis. All participants were assessed by the same examiner and judge. To assure unbiased data collection the odour judge was blinded regarding the Halimeter® measurements and the clinical status of the individual. Prior to and during this investigation the odour judge and an additional odour "evaluator" were repeatedly standardized against a wide range of n-butanol solutions with intensities from 25 to 6075 parts per million (Nachnani et al., 2005; Saad et al., 2005) and also on the mouth odour of patients attending the periodontal clinic. In order to assess the reliability of the judge's ratings, the judge and the evaluator examined the mouth odour of a number of study participants without being aware of one another's score. A statistically significant correlation was found between the scores (Spearman's rho=0.64 for the odour of whole mouth air, p<0.001) (S.A.C.M.O.T., 2002). It should be noted that only the judge's assessments were used in data analysis.

Subjects were designated as halitosis positive (+) when either the organoleptic score of the whole mouth air was two or more, and/or the readings of the Halimeter® exceeded 140 p.p.b.

Real-time polymerase chain reaction:

The DNA of P. gingivalis strain W50 and F. nucleatum strain ATCC 10953 was prepared at the Dental School, University of Glasgow following the standard protocol. Briefly, P. gingivalis and F. nucleatum were grown on fastidious anaerobe agar and harvested after three to seven days. The organisms from a single agar plate were harvested with swabs and dispersed into PBS containing 0.1mM disodium EDTA (PBSE). After one washing in PBS, they were resuspended in 1ml sterile deionised water and aliquoted into two 1.5ml screw top microcaps. DNA was extracted from the organisms by placing the tubes in a boiling water bath for ten minutes. The lysate was mixed with an equal volume of phenol: chloroform: isoamyl alcohol (25:24:1; v/v/v) for five minutes. After centrifugation (13000rpm, 10min) the aqueous upper phase from each preparation was transferred to a fresh tube and mixed with one volume of chloroform: isoamyl alcohol (24:1) for five minutes. After centrifugation (13000rpm, 10min) the aqueous upper phase from each preparation was transferred to a fresh tube and 0.1 volume of 4.5M sodium acetate solution (pH 5.4) and two volumes of ethanol were added. DNA was precipitated overnight at -20oC. The DNA was pelleted by centrifugation (13000rpm, 10min) washed sequentially in 1mL of 70% ethanol, 95% ethanol and allowed to almost dry. The DNA was then dissolved in 100µl water. The nucleic acid concentration and purity was measured at a wavelength of 260 nanometres and the ratio of the absorbance at 260 and 280 nanometres was measured using a sterile quartz cuvette in a spectrophotometer.

A real-time PCR assay using the TaqMan system (Yoshida et al., 2003) was used in the current study for the quantitative analysis of the 16S rRNA of F. nucleatum and P. gingivalis in tongue coatings. Lysates of samples were prepared by boiling the aliquot for 10 minutes after puncturing the cap with a fine sterile needle to prevent pressure build up. PCR was performed at a total volume of 20μl using the Light CyclerTM Fast Start DNA Master Hybridization Probes kit (Roche Diagnostics, Manheim, Germany), 500nM each of sense and antisense primers, 200nM probe and 5μl of lysed cells. The species-specific primers and TaqMan probes used are shown in Table 1. The specificity of the amplified products and therefore of the primers was confirmed previously (Kato et al., 2005; Yoshida et al., 2003). Amplification and detection were performed with the LightCycler Sequence Detection System (Roche Diagnostics, Manheim, Germany) with the following cycle conditions, identical for both assays: 95C for 10min, and then 45 cycles of 95C for 10sec, 58C for 25sec and a final elongation at 40C for 30sec. A standard curve was plotted for each assay by using serial dilutions of known quantities of purified genomic DNA of each test organism.

Statistical methods:

The statistical null hypothesis that there is no association between halitosis detection (presence or absence) and periodontal status in non-smoking subjects was tested by logistic regression. Since the size of each clinical group was above 20 and the minimum number of halitosis positive subjects per clinical group exceeded ten, with the exception of the healthy group (nine halitosis positive subjects), the sample size was deemed adequate for the application of this statistical model (Peduzzi et al. 1996).

The association between halitosis presence or absence and clinical indices (PI, CPD, CAL, BOP, sites > 5mm CPD, WTCI) was examined using logistic regression. Spearman's rho correlation coefficient estimated the association between scale variables and also the agreement between the scores of the two odour assessors. The assumption of normality for scale variables was tested by the Shapiro Wilk test. Between groups comparisons were performed using the Kruskal-Wallis test, while pair-wise comparisons were conducted by the Mann-Whitney test with Bonferroni's adjustment of type I error.

The analysis of data was performed using the SPSS version 16.0 software and the level of statistical significance was set at p<0.05.

Inclusion Criteria:

The clinical inclusion criteria were as follows:

(I) Chronic periodontitis patients had at least one site per quadrant with clinical probing depth (CPD) > 5mm and radiographic evidence of bone loss; (II) Generalised chronic gingivitis patients presented with bleeding on probing (BOP) at > 30% of sites, CPDs < 4mm and no evidence of bone loss; (III) Healthy subjects had < 10% sites with BOP and no sites with CPD > 3mm.

Exclusion Criteria:

Less than 20 teeth, smoking (during the previous year), presence of any systemic disease (e.g. diabetes mellitus, gastrointestinal disorders, etc.) including ENT complications, prescribed medication that can cause xerostomia based on the Greek National Formulary of drugs, full and/or partial dentures, heavily restored dentition, large carious cavities that can trap food remnants, pathology of the oral tissues (pericoronitis, dental / periodontal abscess, etc), antibiotic therapy within three months of recruitment, periodontal treatment within the last year, pregnancy or lactation.

Systemically healthy non-smokers were screened for oral halitosis in the Department of Periodontology, Aristotle University of Thessaloniki, Greece.