Halitosis affects a large proportion of the global population and may be the cause of a significant social or psychological handicap 1. The number of subjects who visit dental clinics complaining of halitosis has been increasing, and it is estimated that more than 50% of the population in North America suffer from halitosis 23. An epidemiological survey of the general population of Japan showed that 24% of the individuals examined complained about bad breath 4.

Much emphasis has been reported in the literature regarding unpleasant odors in the oral cavity. Volatile sulfur compounds (VSCs) generated by oral bacteria are the major causes of halitosis and are major components of intra-oral odors 56. On the other hand, nasal passages predominate among extra-oral etiologies of bad breath. Nasal malodor is indicative of a nasal infection, such as sinusitis, tonsillitis, and rhinitis 7. It has also been reported that patients with cleft lip and/or palate have a greater tendency to present nasal malodor compared to those without cleft lip and/or palate 8. Moreover, odorous gases emanating form the pharyngeal or bronchial region contaminate exhaled air from the lungs 9. Therefore, in order to manage patients with halitosis, an accurate and objective instrument for measuring malodor level in oral, exhaled, and nasal air is required. To date, however, only a few effective instruments like the halimeter exist that are capable of measuring malodor level originating from oral, exhaled and nasal air 7.

Currently, gas chromatography (GC) for determining the concentrations of VSCs is considered to be the most reliable method for measuring oral malodor 10. Unfortunately, since GC is not suitable in daily clinical practice because of its complexity, simpler devices for detecting VSCs have been developed and are widely used 111213. Actual biological gas, however, contains several odorous compounds other than VSCs, such as n-dodecanol, phenol, indole, and diamine 1415. Indeed, it is commonly observed that VSC levels do not always correspond to actual odor levels perceived by humans. Therefore, organoleptic measurement, the subjective measurement of halitosis, is considered to be the most reliable, due to the fact that the human sense of smell can simultaneously detect several kinds of compounds, including VSCs and other odorous gases 316.

Recently, the B/B Checker^®, a new portable gas detector for measuring the malodor level from human oral, exhaled, and nasal air has been developed. As a single unit, this device is capable of detection of several kinds of gases, and also measuring of the malodor level in oral, exhaled, and nasal gases independently in a short period of time. The purpose of the present study was to examine the effectiveness of this new device in measuring malodor level in oral, exhaled and nasal air.

The mean level ± standard deviation of each gas measured using different procedures are shown in Table 1. Since methyl mercaptan and dimethyl sulfide were not detected in most of subjects, the measurement using GC was expressed as hydrogen sulfide level and total concentration of VSCs. All measurement procedures determined that exhaled air was the most odorous, followed by oral air and nasal air. All odor levels, i.e., B/B Checker^® measurements, OT scores, and GC analyses for nasal air were significantly lower than those for oral and exhaled air. There was no significant difference of malodor level between oral air and exhaled air.

The correlation between the B/B Checker^® measurements and total concentration of VSCs is shown in Figure 2. In oral and nasal air, total concentration of VSCs measured by GC significantly correlated to the malodor level detected using the B/B Checker^®. However, there was no significant relationship between the two measurements in exhaled air.

Since the organoleptic scoring method is still considered to be the gold standard for determining oral malodor, the correlation coefficients of OT scores with the B/B Checker^® and total concentration of VSCs measured by GC was calculated (Table 2). In all three types of air, a significant correlation was observed between the OT scores and B/B Checker^® measurements (oral air: 0.892, p < 0.001; exhaled air: 0.748, p < 0.001; and nasal air: 0.534, p < 0.001). On the other hand, the correlation between the OT scores and total concentration of VSCs was significant only for oral air (0.790, p < 0.001) or nasal air (0.431, p = 0.002); not for exhaled (0.310, p = 0.096).

As shown in Table 1 the value of malodor level was relatively low. In order to examine the usability of B/B Checker^® in halitosis patients, those with OT scores (≥ 3) were selected (n = 20). In oral air, a significant correlation was observed between concentration of hydrogen sulfide and B/B Checker^® measurements (0.790, p < 0.001).

The appropriate level of the B/B Checker^® for screening subjects with malodor was determined by calculating sensitivity and specificity. The subjects with OT scores (≥ 3) were defined as those with clearly noticeable malodor (Figure 3). When the screening level of the B/B Checker^® was set to 50 for oral air, the sensitivity and specificity was 1.00 and 0.90, respectively. On the other hand, when the screening level of the B/B Checker^® was set to 60 for exhaled air, the sensitivity and specificity was 0.82 and 1.00, respectively. Since there were few subjects with OT scores (≥ 3) in nasal air, this particular analysis was not performed.

The correlation coefficients of OT scores with the B/B Checker^® measurements were significant for the oral, exhaled, and nasal air. It is now commonly accepted that the organoleptic scoring method is a gold standard for the study of malodor. Moreover, the B/B Checker^® has relatively high sensitivity and specificity for screening subjects with noticeable malodor. Ueno et al. (2008) developed a new portable sulfide monitor, and reported that sensitivity is 0.8-0.9 and specificity is 0.6-0.8 12. The study by Ueno et al. employed 475 subjects, whereas only 30 participated in the present study. Although the high sensitivity (0.82-1.00) and specificity (0.90-1.00) of the B/B Checker^® must be re-evaluated in a larger population, it may be useful for objective evaluation of malodor in oral, exhaled and nasal air and for screening subjects with halitosis.

Significant association between OT scores and total concentration of VSCs by GC was observed only for oral air, and not for exhaled or nasal air. On the other hand, OT scores and B/B Checker^® measurements were significantly correlated for three kinds of air. In addition, there was no significant association between the total concentration of VSCs and B/B Checker^® measurements in exhaled air. While the reason remains unclear, the GC focuses only on VSCs; hydrogen sulfide, methyl mercaptan and dimethyl sulfide. Therefore, some of the major components of oral air are VSCs, substances that both the B/B Checker^® and GC are capable of detecting. It is likely that exhaled air, probably or nasal air, contain other major components that both the B/B Checker^® and human nose, and not GC, could detect. In fact, it has been hypothesized that indole, skatole, putrescine, and cadaverine, and some organic acids, like acetic and butyric acid, also contribute to malodor 15. Recently, Van den Velde et al. (2009) reported that di- and trimethyl sulfide are associated with malodor 21. Based on these findings, devices sensitive to several kinds of gasses are required for malodor study.

Tangeman and Winkel (2007) proposed to clearly differentiate the origin of halitosis, intra- and extra-orally. The most predominant sources of halitosis (80-90%) are present within the oral cavity and include bacterial reservoirs such as the dorsum of the tongue, saliva and periodontal pockets 22. This corresponds to the malodor in oral air in the present study. However, less attention has been paid to extra-oral halitosis 222324. One of the origins of extra-oral halitosis is malodor in nasal air, which is considered to be related to serious systemic disease. Furthermore, we suspected that the air that passes through the oral cavity from the lung after expiring oral air that has stagnated in the oral cavity might be another origin of extra-oral halitosis. As a result, all three kinds of measurements, i.e., B/B Checker^® measurements, total concentration of VSCs, and OT scores, showed that the malodor level was greatest in the exhaled air, followed by oral air and nasal air. Since healthy subjects, without any medical disorders, were employed in the present study, the lowest scores in the nasal air were deemed reasonable. Moreover, the present study indicates that future studies should focus on exhaled air for halitosis management.

Many articles have discussed the generation of bad breath in the oral cavity. However, to date, few devices have been developed for effectively measuring malodor in nasal or exhaled air and few studies of nasal malodor and its causes have been published 17. It has been reported that patients with cleft lip and/or palate have a greater tendency to present higher nasal VSC measurements, indicating the usefulness of VSC measurements for evaluation of quality of life (QOL) in these patients. The B/B Checker^® has a special adaptor for the nasal cavity, allowing the sensor to be placed directly into the nasal cavity without gas contamination. Accurate assessment and testing are important to determine which point is an actual problem. Therefore, the B/B Checker^® is applicable for QOL determination for cleft patients. The present study also examined the ability of conducting measurements of exhaled air in a clinical situation. Moreover, the B/B Checker^® showed adequate sensitivity for detecting malodor and a fair specificity for identifying non-malodor cases. The results of the present study suggest that the B/B Checker^® could be used as an adjunct instrument to a GC system and organoleptic methods currently in use at clinical situation for determining malodor in oral, exhaled and nasal air.

The malodor levels of exhaled air were higher than of oral air, however, no significant differences were observed in the levels of OT score, B/B value, hydrogen sulfide and VSCs. The sources of hydrogen sulfide may be multiple, including the bacterial flora 25, as well as endogenous sulfide-generating enzymes that exist in many cells and tissues of the human body 26. Oral air was thought to be relatively low in subjects at the present study because periodontally healthy subjects were analyzed. If we have measured in malodor levels at periodontal patients, oral air would be higher than exhaled breath.

This study has some limitations. At first, the subjects who participated in the present study had relatively mild halitosis. The mean VSC levels ranged from 40 to 110 ppb of the total concentration of VSCs. It remains unclear whether the B/B Checker^® measurements correlate to GC or organoleptic ratings in subjects with severe halitosis. Since it is apparent that VSCs cause severe halitosis and that the B/B Checker^® is capable of detecting several gases, the B/B Checker^® might be less sensitive to severely malodorous air in 5 subjects complaining of oral malodor. Secondly, the present study is only a cross sectional study. Therefore, the effect of malodor management in clinic should be monitored using the B/B Checker^®. Then, the usefulness of this device can be actually provided.

The results from the present study show that the correlation coefficients of OT scores with measurements from the B/B Checker^® were significant for oral, exhaled and nasal air. In oral and nasal air, the total concentration of VSCs measured by GC significantly correlated to the malodor level determined using the B/B Checker^®. The B/B Checker^® may be effective for objective evaluation of malodor in oral, exhaled and nasal air and for screening subjects with halitosis in the clinical setting. Further clinical studies are needed to investigate the capability of measuring the malodor of each type of air (i.e., oral, exhaled, nasal) in different sample groups with systemic disease or periodontal disease.

One of the author (T. Toshikawa) is employees of Taiyo Instrument Inc., which has been involved in the B/B Checker^® development.

NT has made substantial contribution to the study conception and design. KK, ME, TT¹, DE, and SH implemented this study and participated in the acquisition, analysis and interpretation of data. TT³ gave technical advices using the B/B Checker^®. MM has been intimately involved in drafting and editing the manuscript. All authors read and approved the final manuscript.