Halitosis is an unpleasant odor discharged through the oral cavity with a prevalence as high as 30%–50% of the general population. Conventional diagnostic methods have been focused on mouth air analysis measuring the amount of sulfur compounds which does not directly reflect the cause of halitosis. Also, the possible role of halitosis as an indicator of general health status has been steadily suggested and inflammation has been constantly associated with aversive body odor. Therefore, this study aimed to search for inter-relationships between hematologic indicators, clinical characteristics, and halitosis measurement that can predict the presence of pathologic halitosis and its intensity. Furthermore, the tentative relationship between halitosis and the presence of systemic inflammation was investigated. A total of 125 patients were divided into 103 patients in the genuine halitosis group (value ⩾80 ppb) and 22 patients in the pseudo halitosis group (value <80 ppb) based on portable sulfide monitor measurements. Clinical examination and hematological indices including inflammatory prognostic factors and halitosis measurements including organoleptic testing, portable sulfide monitor, and gas chromatography were evaluated. The genuine halitosis group showed a significantly higher white blood cell (WBC) count (p< 0.01) compared to the pseudo halitosis group. Erythrocyte sedimentation rate (ESR, β = 0.341, p < 0.05) values and duration of halitosis (β= 0.353, p < 0.05) showed a significant association with halitosis intensity and neutrophil to lymphocyte ratio (NLR) values (β = 3.859, p < 0.05) were significantly related to genuine halitosis diagnosis. A new WBC cut-off value of 5575 μl−1 showed near to fair discriminative power in predicting genuine halitosis (area under the curve 0.661, p < 0.05). The results of this study showing an increased WBC count in genuine halitosis and its strong association with hematologic indices of subclinical inflammation including ESR and NLR suggest inflammatory hematologic markers as potential diagnostic tools in the diagnosis of genuine halitosis.
1.Introduction
Halitosis, characterized as an unpleasant odor discharged through the oral cavity, is a common disease with a prevalence as high as 30%–50% in the general population. There is still no overall consensus on the definition and subgrouping of halitosis. It is occasionally classified into genuine halitosis, pseudo halitosis, and halitophobia, and genuine halitosis is once again divided into physiological and pathological halitosis as shown in table 1 [1]. While physiological halitosis is a temporary symptom generally not subject to treatment that refers to bad breath related to physiological states such as morning breath caused by decreased salivation during sleep and halitosis during mensuration related to hormonal changes, genuine halitosis has been acknowledged as an independent disease entity. Halitosis does not directly threaten a patient's life but it is a condition that drastically lowers one's quality of life, especially in the modern ages in which complex human relationships and social interactions are crucial. Previous studies have demonstrated that the major cause of halitosis can be found in the oral cavity while 10% of all halitosis patients had an extra-oral cause [2]. Approximately 60% of the oral factors could be explained by gingivitis and periodontitis, and tongue biofilm-related issues accounted for the remaining 40%. Extra-oral factors include tonsillitis, sinusitis, gastrointestinal, and psychological problems [3–5]. If the patient's oral hygiene status is good, an accurate differential diagnosis is required to rule out the possibility of a serious systemic disease or a state simply due to the bacterial load on the surface of the tongue underlying bad breath.
Table 1.Classification of halitosis.
| Classification | Description |
|---|---|
| I. Genuine halitosis | Obvious perceived malodor |
| A. Physiological halitosis | Halitosis arises through putrefactive process within the oral cavity without any specific disease nor pathologic condition |
| B. Pathological halitosis | Halitosis caused by disease, pathologic condition or malfunction of oral tissues or extraoral tissue |
| II. Pseudo halitosis | 1. Obvious malodor is not perceived by others, although the patient complains of it |
| 2. Condition is improved by counselling and simple oral hygiene measures | |
| III. Halitophobia | 1. After treatment for genuine halitosis or pseudo-halitosis, the patient persistently believes that he/she has halitosis |
| 2. No evidence exists to suggest the presence of halitosis |
The classification followed Yaegaki and Coil [1].
The diagnosis of halitosis in many cases is based on empirical and subjective data such as sole organoleptic evaluation and self-reports from the patient and others. This causes delay in obtaining proper diagnosis and adequate treatment for the patient. Additionally, diagnostic methods have been focused on mouth air analysis measuring the amount of sulfur compounds which does not directly reflect the cause of halitosis [6]. The possible role of halitosis as an indicator of general health status has been steadily suggested and inflammation has been constantly associated with aversive body odor [7]. Disease diagnosis through breath analysis is in the spotlight due to its simplicity and non-invasiveness. However, related research is still at a basic level with investigators searching for a correlation between breath biomarkers and indices of systemic health that can be implemented in the diagnosis and prognosis prediction of halitosis. Hematological indices such as neutrophil to lymphocyte ratio (NLR), platelet to lymphocyte ratio (PLR), and lymphocyte to monocyte ratio (LMR) are well known indicators of systemic inflammation that show a close relationship with various systemic diseases including cancer [8] and rheumatic disorders [9].
Therefore, this study aimed to search for hematologic indicators that could be applied in differentiating patients with pathologic halitosis from those without by investigating their interrelationships with clinical characteristics and halitosis measurements. Furthermore, the presence of systemic inflammation was investigated as a possible cause of halitosis.
2.Materials and methods
2.1.Subjects and study design
This retrospective analysis evaluated consecutive patients who visited Seoul National University Dental Hospital complaining of halitosis from January, 2006 to December, 2014. Among these, adult patients (>18 years old) who had clinical, psychological, hematological evaluation and halitosis measurements including gas chromatography (GC), portable sulfide monitor, and organoleptic method were included. Exclusion criteria were subjects with a known active local and/or systemic inflammatory condition, uncontrolled diabetes and hypertension. Those with such a condition were referred to relevant experts based on subjective reports before the examination stage. During this period a total of 1636 patients visited with the chief complaint of halitosis and met the inclusion/exclusion criteria. Among this group 1345 and 166 patients who did not undergo GC and hematological tests independently were excluded from the study. The final analysis included 125 patients with clinical and hematologic examinations and halitosis measurement data. This study was approved by the Institutional Review Board of Seoul National University Dental Hospital (ERI20018). The review board authorized exemption of informed consent.
2.2.Clinical and hematological assessment
2.2.1.Subjective halitosis evaluation
A structured questionnaire was used to obtain data on subjective halitosis. The patient's degree of self-awareness of halitosis and disability in social activities and interpersonal relationships were questioned based on a visual analogue scale (0 [no halitosis/no disability] to ten [worst possible intensity of halitosis/impossible to conduct normal activates due to halitosis]). Also, data concerning the duration of the symptoms (months) was collected. Smoking history and information on oral hygiene habits including the frequency of tooth brushing, flossing, tongue cleaning and mouth-rinsing with chemical agents and dental check-ups were collected. Finally the patient was asked if the halitosis was recognized only by the patient or had ever been pointed out by another person, and also if the halitosis was continuously present or only perceived intermittently.
2.2.2.Periodontal evaluation
Periodontal evaluation was performed on six index teeth including the upper right first molar, upper left central incisor, upper left first premolar, lower left first molar, lower right central incisor, and lower right first premolar (Fédération Dentaire Internationale notation 16, 21, 24, 36, 41, 44). The bleeding index (BI) [10], gingival index (GI) [11] and probing pocket depth (PPD) were measured for each tooth. Each tooth was divided into four (mesial, buccal, distal, and lingual) and six (mesiobuccal, buccal, distobuccal, mesiolingual, lingual, and distolingual) sections when observed from the occlusal aspect for the measurement of GI and PPD, respectively. The mean BI and GI value were used for analysis. The deepest PPD was recorded for each tooth and the number of teeth with a PPD ⩾4 mm was also recorded for analysis.
2.2.3.Measurement of tongue coating
The dorsal surface of the tongue was repeatedly scraped using a disposable tongue blade in a motion starting from the root to the tip of the tongue until the motion did not yield any more visible tongue coating. Then the collected tongue coating was transferred to a gauze for measurement using a scale. Total tongue coating weight was calculated by subtracting the weight of the gauze. Unfortunately, microbial loading per se was not analyzed hence, the correlation between tongue coating and microbial loading is uncertain as stated in previous literature [12].
2.2.4.Salivary flow rate
Salivary flow rates, both unstimulated and stimulated, were measured by collecting whole saliva. The patient remained comfortably seated for 3 min before each measurement without talking. First, saliva was collected for 10 min by spitting method and then the patient chewed 1 g of gum base for 5 min while collecting saliva. The salivary flow rate was obtained as ml min−1. Also, the presence of subjective xerostomia was investigated by asking the patient if his/her mouth often felt dry.
2.2.5.Psychological evaluation
All patients completed Symptom Checklist-90-Revision (SCL-90R) [13]. Those with a global severity index value ⩾60 were also recorded.
2.2.6.Hematologic assessment
Venous blood was harvested from the antecubital vein of the patients included in the study on their first visit during 1–5 pm. Comprehensive laboratory assessments included complete blood cell with white blood cell (WBC) differential (cut-off value, 10 000 μl−1), red blood cell (RBC) indices (RBC, cut-off value 4 × 106 μl−1; hemoglobin [Hgb], cut-off value 12 g dl−1), and platelet count (cut-off value 130 × 103 μl−1). Blood chemistry including glucose (cut-off value, 110 mg dl−1), cholesterol (cut-off value, 240 mg dl−1), total protein (cut-off value, 8 g dl−1), aspartate aminotransferase (AST, cut-off value, 40 IU/L), alanine aminotransferase (ALT, cut-off value, 40 IU/L), blood urea nitrogen (BUN, cut-off value, 26 mg dl−1), and uric acid (cut-off value, 7 mg dl−1) levels were evaluated.
To assess systemic inflammation level NLR, derived neutrophil-to-lymphocyte ratio (d-NLR), PLR, and LMR were calculated from the hematologic test results to assess systemic inflammatory status. The cut-off value for NLR (3.193), d-NLR (2), PLR (218.006), and LMR (2.008) were based on reference values derived from Koreans [14]. Erythrocyte sedimentation rate (ESR, cut-off value, 20 mm hr−1) as a nonspecific inflammatory marker was also analyzed.
2.3.Measurement of halitosis
Halitosis assessment was performed through mouth air analysis. It was measured at three time points with three methods including GC, portable sulfide monitor, and organoleptic method which are widely recognized for the objective diagnosis of halitosis [15]. The first measurement was done when the patient arrived at the clinic between 9 and 11 am without eating or drinking anything and without conducting oral hygiene procedures to obtain a correct baseline value. The second measurement was done after tongue coating removal following the procedures described in section 2.2.3. The last measurement was done after full-mouth scaling and oral prophylaxis.
2.3.1.Organoleptic testing
All measurements were conducted by dentists trained in the same facility. Patients were instructed to sit in an upright position for 3 min with their mouths closed before collecting samples. A gastight syringe with a 6 cm long and 3 mm diameter polytetrafluoroethylene tube (Hamilton Co., Reno, NV, USA) was used with negative pressure method allowing passive flow of mouth air into the syringe for minimal loss of the volatile components. Organoleptic testing was performed by expelling the gathered mouth air into a disposable paper cup covering the examiner's nose following the Kim method [16]. The examiner evaluated the degree of halitosis on a 0–4 scale (0, no appreciable odor; 1, barely noticeable odor that is of low intensity and within acceptable limits; 2, slight to moderate odor that is clearly noticeable and slightly unpleasant; 3, moderate to high odor that is clearly noticeable, unpleasant, and of moderate intensity; and 4, offensive odor of strong intensity) [17].
2.3.2.Portable sulfide monitor
A Halimeter (Interscan Corp., Chatsworth, LA, USA) was applied to measure the amount of volatile sulfur compounds (VSCs) in parts per billion (ppb). The patient waited for 3 min without opening his/her mouth before each measurement. A disposable straw connected to the instrument was inserted into the oral cavity while the patient kept his/her mouth slightly open while breathing naturally through the nose to measure only the VSC generated in the oral cavity without being mixed with the exhalation from the lungs. The mean value of three measurements were used for analysis. A value of ⩾80 ppb was considered to indicate the presence of genuine halitosis and the final analysis was based on such grouping [18].
2.3.3.GC
A GC (HP 5890, Hewlett-Packard, Avondale, PA, USA) with a flame photometric detector specific for sulfur compounds and a 60/80-mesh size column (Chromosil 330, Supelco, Bellforte, PA, USA) was used to analyze the air collected from the oral cavity. The air was collected in the same way as for organoleptic testing and injected directly into the GC. Oven temperature protocol was 40 °C for 2 min, increased by 8 °C min−1 to 100 °C for 10 min. Carrier gas was nitrogen at 20 ml min−1 [19]. Personnel protective measures including gloves and masks were adopted during all collection procedures and all potentially contaminatable sampling devices and collected samples were immediately disposed following analysis. Sampling and analysis equipment was handled by trained personnel following safety procedures.
2.4.Statistical analysis
Kolmogorov-Smirnov test was used to test the normality of data. Differences in clinical and psychological characteristics between halitosis groups were analyzed by independent t-test, Mann-Whiney U test and chi-square test. Differences in hematologic characteristics between halitosis groups were analyzed by independent t-test and Mann-Whitney U test, and subgroups according to cut-off values were analyzed by Fisher's exact test. Repeated measures analysis of variance (ANOVA) were conducted to analyze changes in halitosis measurements with intervention. The relationship between halitosis measurements and hematologic inflammatory indices were analyzed by Pearson's and Spearman's correlation analysis. Multiple regression analyses were conducted to determine clinical and hematologic variables predicting halitosis intensity before treatment and logistic regression analysis was conducted to identify risk factors of having genuine halitosis. The receiver operating characteristic curve and area under the curve (AUC) was analyzed to obtain the cut-off value for inflammatory hematologic indices in evaluating genuine halitosis.
SPSS 25.0 software (IBM, Chicago, IL, USA) was used for all statistical analysis. The level of statistical significance was set at p < 0.05.
3.Results
3.1.Clinical and psychological characteristics of halitosis groups
The total 125 patients were divided into 103 patients in the genuine halitosis group (mean age 40.1 ± 12.9 years) and 22 patients in the pseudo halitosis group (mean age 38.2 ± 14.8 years) based on portable sulfide monitor values of the first measurement (⩾80 ppb).
As shown in table 2, there were no statistically significant differences between the two halitosis groups in most clinical characteristics. However genuine halitosis patients cleaned their tongue less frequently (p = 0.020) and more patients were taking systemic medication (p = 0.020) compared to the pseudo halitosis group. Among the 21 patients who were taking medication, ten were taking it for hypertension, six for cerebrovascular diseases, four for cardiovascular diseases and two for diabetes mellitus (one duplicate). There were no medications that were established to affect halitosis. Genuine halitosis patients had more tongue coating although the difference was not statistically significant. Pseudo halitosis patients were generally practicing more oral hygiene measures.
Table 2.Clinical and psychological characteristics of halitosis groups.
| Variables | Genuine halitosis (n = 103) | Pseudo halitosis (n = 22) | P-value |
|---|---|---|---|
| Age (years) a | 40.1 (12.9) | 38.2 (14.8) | 0.548 |
| Gender (M/F) c | 39/64 | 6/16 | 0.347 |
| Subjective halitosis intensity (VAS) b | 7.3 (5.0–10.0) | 7.4 (7.0–9.0) | 0.938 |
| VAS ⩾5 c | 58/63 (92.1%) | 14/15 (93.3%) | 0.868 |
| Subjective discomfort due to halitosis (VAS) b | 6.5 (5.0–9.0) | 6.0 (5.0–8.0) | 0.409 |
| VAS ⩾5 c | 56/63 (88.9%) | 13/15 (86.7%) | 0.809 |
| Duration of halitosis (months) b | 79.1 (12.0–120.0) | 72.5 (24.0–120.0) | 0.713 |
| Duration ⩾3 months c | 90/95 (94.7%) | 20/21 (95.2%) | 0.925 |
| Halitosis recognition-self c | 73/100 (73.0%) | 17/22 (77.3%) | 0.680 |
| Halitosis recognition-others c | 74/100 (74.0%) | 16/22 (72.7%) | 0.770 |
| Subjective persistent halitosis c | 49/100 (49.0%) | 14/22 (63.6%) | 0.214 |
| Gingival index b | 0.7 (0.3–1.1) | 0.4 (0.0–1.0) | 0.121 |
| Bleeding index b | 0.6 (0.2–0.8) | 0.4 (0.0–0.8) | 0.144 |
| Deepest pocket depth (mm) b | 3.8 (3.0–4.0) | 3.5 (3.0–4.0) | 0.336 |
| Number of teeth with pocket depth ⩾4 mm b | 1.3 (0.0–2.0) | 0.9 (0.0–2.0) | 0.285 |
| Tongue coating (mg) b | 169.0 (28.5–245.5) | 122.5 (12.0–173.5) | 0.506 |
| Unstimulated salivary flow rate (ml min−1) b | 0.3 (0.1–0.4) | 0.3 (0.1–0.4) | 0.918 |
| <0.1 ml min−1 c | 21/101 (20.8%) | 4/22 (18.2%) | 0.783 |
| Stimulated salivary flow rate (ml min−1) b | 1.0 (0.5–1.0) | 1.0 (0.3–1.3) | 0.391 |
| <0.5 ml min−1 c | 23/99 (23.2%) | 8/22 (36.4%) | 0.202 |
| Subjective dry mouth c | 56/97 (57.7%) | 11/22 (50.0%) | 0.509 |
| Oral hygiene care | |||
| Solution usage and flossing (per day) b | 1.1 (0.0–2.0) | 1.7 (0.0–3.0) | 0.069 |
| Use of solution or dental floss c | 33/63 (47.6%) | 11/15 (73.3%) | 0.249 |
| Tooth brushing (per day) b | 2.8 (2.0–3.0) | 2.9 (3.0–3.0) | 0.156 |
| Tongue cleaning (per day) b | 2.4 (2.0–3.0) | 3.0 (3.0–3.0) | 0.020* |
| Dental check-up (per year) b | 0.9 (0.0–1.0) | 0.9 (0.0–2.0) | 0.110 |
| Systemic medication intake c | 21/103 (20.4%) | 0/22 (0%) | 0.020* |
| Hypertension c | 11/103 (10.7%) | 0/22 (0%) | 0.108 |
| Diabetes mellitus c | 2/103 (1.9%) | 0/22 (0%) | 0.510 |
| Smoker c | 5/98 (5.1%) | 1/22 (4.5%) | 0.914 |
| Symptom checklist-90 revision b | |||
| Somatization | 45.6 (40.0–50.0) | 42.6 (38.0–47.0) | 0.080 |
| Obsessive-compulsive | 44.7 (38.0–49.0) | 40.4 (35.0–45.0) | 0.021* |
| Interpersonal-sensitivity | 45.6 (39.0–50.0) | 42.0 (38.0–46.0) | 0.059 |
| Depression | 44.1 (38.0–48.0) | 40.1 (36.0–45.0) | 0.079 |
| Anxiety | 44.6 (40.0–48.0) | 41.7 (38.0–45.0) | 0.045* |
| Hostility | 46.7 (40.0–51.0) | 42.4 (40.0–44.0) | 0.032* |
| Phobic anxiety | 46.2 (40.0–48.0) | 44.2 (42.0–45.0) | 0.722 |
| Paranoid ideation | 43.7 (38.0–48.0) | 41.0 (38.0–42.0) | 0.062 |
| Psychoticism | 44.7 (40.0–48.0) | 41.6 (39.0–43.0) | 0.026* |
| Global severity index | 44.3 (38.0–49.0) | 40.1 (36.0–42.0) | 0.013* |
| Global severity index ⩾60 c | 5/99 (5.1%) | 0/19 (0%) | 0.317 |
| Positive symptom distress index | 44.4 (39.0–48.0) | 41.1 (37.0–44.0) | 0.071 |
| Positive symptom total | 44.9 (36.0–52.0) | 39.2 (32.0–46.0) | 0.022* |
VAS, visual analog scale (0–10). a Independent t-test: mean (SD). b Mann-Whitney U test: median (lower quartile-upper quartile). cChi-square test: number of positive subjects. *Significant difference, P < 0.05.
Psychological evaluations showed significantly higher somatization (p = 0.021), anxiety (p = 0.045), hostility (p = 0.032), psychoticism (p = 0.026), global severity index (p = 0.013), and positive symptom total (p = 0.022) scores in the genuine halitosis group.
3.2.Hematologic characteristics between halitosis groups
Table 3 showed the differences in hematologic variables between halitosis groups. The genuine halitosis group showed a significantly higher WBC count (p = 0.004), and lower cholesterol (p = 0.045) level than the pseudo halitosis group. There was no statistically significant difference between groups in other inflammatory hematologic variables.
Table 3.Hematologic characteristics of halitosis groups.
| Variables | Genuine halitosis (n = 103) | Pseudo halitosis (n = 22) | P-value |
|---|---|---|---|
| WBC a | 6.3 (1.6) | 5.5 (1.1) | 0.004* |
| Abnormal (>10 000 μl−1) c | 3/103 (2.9%) | 0/22 (0%) | 1.000 |
| RBC a | 4.6 (0.5) | 4.5 (0.3) | 0.138 |
| Abnormal (<4 × 106 μl−1) c | 6/103 (5.8%) | 1/22 (4.5%) | 1.000 |
| Hgb a | 14.0 (1.4) | 14.5 (1.1) | 0.141 |
| Abnormal (<12 g dl−1) c | 6/103 (5.8%) | 1/22 (4.5%) | 1.000 |
| Platelet count a | 242.0 (54.3) | 243.3 (70.3) | 0.923 |
| Abnormal (<130 × 103 μl−1) c | 2/103 (1.9%) | 0/22 (0%) | 1.000 |
| ESR b | 8.0 (2.0–11.0) | 6.8 (2.8–9.3) | 0.929 |
| Abnormal (>20 mm h−1) c | 7/102 (6.9%) | 0/22 (0%) | 0.351 |
| NLR b | 1.9 (1.3–2.3) | 1.6 (1.2–1.9) | 0.143 |
| Abnormal (>3.193) c | 9/103 (8.7%) | 0/22 (0%) | 0.359 |
| dNLR b | 0.6 (0.4–0.6) | 0.6 (0.5–0.7) | 0.223 |
| Abnormal (>2) c | 2/103 (1.9%) | 0/22 (0%) | 1.000 |
| PLR a | 123.9 (41.4) | 127.0 (36.4) | 0.748 |
| Abnormal (>218.006) c | 3/103 (2.9%) | 0/22 (0%) | 1.000 |
| LMR b | 5.1 (3.7–6.3) | 4.8 (3.8–5.8) | 0.930 |
| Abnormal (<2.008) | 0/103 (0%) | 0/22 (0%) | — |
| Glucose b | 93.0 (84.8–99.3) | 95.0 (85.0–99.0) | 0.930 |
| Abnormal (>110 mg dl−1) c | 9/103 (8.8%) | 2/22 (9.1%) | 1.000 |
| Cholesterol b | 175.1 (148.8–195.5) | 193.0 (163.0–224.8) | 0.045* |
| Abnormal (>240 mg dl−1) c | 4/103 (3.9%) | 3/22 (13.6%) | 0.106 |
| Total protein b | 7.5 (7.2–7.8) | 7.1 (7.2–7.6) | 0.239 |
| Abnormal (>8 g dl−1) c | 12/102 (11.8%) | 1/22 (4.5%) | 0.461 |
| AST b | 20.7 (16.8–23.0) | 20.0 (17.0–22.0) | 0.844 |
| Abnormal (>40 IU/L) c | 1/103 (1.0%) | 0/22 (0%) | 1.000 |
| ALT b | 21.4 (13.0–26.0) | 16.4 (13.0–20.3) | 0.154 |
| Abnormal (>40 IU/L) c | 7/103 (6.8%) | 0/22 (0%) | 0.352 |
| BUN b | 12.0 (9.9–14.0) | 11.2 (10.0–12.3) | 0.808 |
| Abnormal (>26 mg dl−1) c | 1/103 (1.0%) | 0/22 (0%) | 1.000 |
| Uric acid b | 4.7 (3.8–5.6) | 4.4 (3.8–4.9) | 0.428 |
| Abnormal (>7 mg dl−1) c | 5/103 (4.9%) | 0/22 (0%) | 0.585 |
WBC, white blood cell; RBC, red blood cell; Hgb, hemoglobin; ESR, erythrocyte sedimentation rate; NLR, neutrophil to lympohcyte ratio; dNLR, derived neutrophil to lymphocyte ratio; PLR, platelet to lymphocyte ratio; LMR, lymphocyte to monocyte ratio; AST, aspartate aminotransferase; ALT, alanine aminotransferase; BUN, blood urea nitrogen. a Independent t-test: mean (SD). b Mann-Whitney U test: median (lower quartile-upper quartile). c Fisher's exact test: number of positive subjects. *Significant difference, P < 0.05.
3.3.Change in halitosis with intervention
As shown in table 4, halitosis measurements with all three methods showed significant reduction after mechanical intervention in both groups. There was a significant difference in all halitosis measurements at T1 and T3 between groups, with the genuine halitosis group showing a more significant reduction than the pseudo halitosis group in organoleptic score (60.0% vs 16.7% reduction, p < 0.001), portable sulfide monitor (65.5% vs 31.8% reduction, p < 0.001), and GC (80.1% vs 59.5% reduction, p = 0.029) after tongue coating removal, full mouth scaling and oral prophylaxis.
Table 4.Change in halitosis measurements with intervention of halitosis groups.
| Variables | Group | T1 | T2 | T3 | Repeated-measures ANOVA (F) | ||
|---|---|---|---|---|---|---|---|
| Time | Group | Time x group | |||||
| Organoleptic score | Genuine (n = 62) | 2.0 (0.8) | 1.8 (0.9) | 0.8 (0.6) | 24.999*** | 29.226*** | 18.131*** |
| Pseudo (n = 14) | 0.6 (0.6) | 0.5 (0.5) | 0.5 (0.5) | ||||
| p-value | 0.000*** | 0.000*** | 0.053 | ||||
| Portable sulfide monitor (ppb) | Genuine (n = 63) | 305.4 (273.2) | 297.7 (221.5) | 105.2 (72.7) | 10.236*** | 15.622*** | 6.155** |
| Pseudo (n = 14) | 58.8 (32.7) | 72.1 (42.9) | 40.1 (30.0) | ||||
| p-value | 0.000*** | 0.000*** | 0.002** | ||||
| Gas chromatography (ppm) | Genuine (n = 59) | 0.472 (0.673) | 0.338 (0.588) | 0.094 (0.259) | 3.557* | 4.652* | 2.077 |
| Pseudo (n = 14) | 0.079 (0.119) | 0.081 (0.118) | 0.032 (0.091) | ||||
| p-value | 0.000*** | 0.002** | 0.359 | ||||
T1, before tongue coating removal (without brushing); T2, after tongue coating removal; T3, after full mouth scaling and oral prophylaxis; Organoleptic score: 0, no appreciable odor; 1, barely noticeable odor that is of low intensity and within acceptable limits; 2, slight to moderate odor that is clearly noticeable and slightly unpleasant; 3, moderate to high odor that is clearly noticeable, unpleasant, and of moderate intensity; and 4, offensive odor of strong intensity; Portable sulfide monitor, sum of amount of hydrogen sulfide and methyl mercaptan; ppb, parts per billion; Gas chromatography, sum of amount of hydrogen sulfide, methyl mercaptan and dimethyl sulfide; ppm, parts per million.Results were obtained from repeated measures ANOVA and independent t-test: mean (SD). * Significant difference, P< 0.05. ** Significant difference, P< 0.01. *** Significant difference, P < 0.001.
3.4.Relationship among halitosis measurements and hematologic inflammatory indices
As shown in table 5, there was a significant correlation between WBC (r = 0.238, p = 0.036), NLR (r = 0.309, p = 0.006) and dNLR (r = −0.316, p = 0.005) with halitosis measurement values using a portable sulfide monitor after tongue coating removal.
Table 5.Correlations among halitosis measurements and hematologic inflammatory indices.
| Variables | Organoleptic score | Portable sulfide monitor (ppb) | Gas chromatography (ppm) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| T1 | T2 | T3 | T1 | T2 | T3 | T1 | T2 | T3 | |
| WBC a | .138 | .138 | .059 | .131 | .238* | .099 | .168 | .211 | .058 |
| ESR b | .073 | .076 | −.087 | −.023 | .059 | −.024 | .003 | .038 | −.027 |
| NLR b | .021 | .149 | .142 | .153 | .309** | .163 | −.005 | .025 | .008 |
| dNLR b | −.018 | −.147 | −.153 | −.157 | −.316** | −.140 | .010 | −.054 | −.022 |
| PLR a | −.058 | −.075 | −.103 | .025 | .082 | .012 | −.151 | −.117 | −.081 |
| LMR b | .003 | −.196 | −.207 | −.047 | −.214 | −.106 | .101 | −.027 | .098 |
WBC, white blood cell; ESR, erythrocyte sedimentation rate; NLR, neutrophil to lymphocyte ratio; dNLR, derived neutrophil to lymphocyte ratio; PLR, platelet to lymphocyte ratio; LMR, lymphocyte to monocyte ratio. a The appearing values are correlation coefficients of Pearson's correlation analysis. b The appearing values are correlation coefficients of Spearman's correlation analysis. *Significant difference: p < 0.05, **Significant difference: p < 0.01.
3.5.Regression analysis of variables predicting halitosis
Table 6 shows clinical and hematologic variables predicting halitosis intensity before intervention through multiple regression analysis. There was a significant effect of duration of halitosis (β = 0.382, p = 0.038), ESR (β = 0.555, p = 0.011), glucose level (β = −0.476, p = 0.049), total protein (β = −0.658, p = 0.031) values.
Table 6.Clinical and hematologic variables predicting halitosis intensity before intervention.
| Predictor variables | Standardized β | Standard error | t | P-value |
|---|---|---|---|---|
| Age (years) | −.191 | 5.262 | −.677 | .505 |
| Duration of halitosis (months) | .382 | .547 | 2.193 | .038* |
| Gingival index | .082 | 146.107 | .266 | .793 |
| Bleeding index | −.210 | 166.502 | −.639 | .529 |
| Tongue coating amount (mg) | −.040 | .310 | −.188 | .853 |
| Unstimulated salivary flow rate (ml min−1) | .128 | 243.300 | .557 | .583 |
| Stimulated salivary flow rate (ml min−1) | −.373 | 141.618 | −1.599 | .123 |
| WBC (×103 μl−1) | .024 | 37.104 | .103 | .919 |
| RBC (×106 μl−1) | .220 | 189.709 | .607 | .550 |
| Hgb (g dl−1) | .114 | 52.354 | .368 | .716 |
| Platelet count (×103 μl−1) | −.059 | 1.096 | −.309 | .760 |
| ESR (mm hr−1) | .555 | 7.841 | 2.756 | .011* |
| NLR | .583 | 69.840 | 1.859 | .075 |
| LMR | .306 | 41.477 | 1.069 | .296 |
| Glucose (mg dl−1) | −.476 | 4.603 | −2.072 | .049* |
| Cholesterol (mg dl−1) | −.171 | 1.525 | −.880 | .388 |
| Total protein (g dl−1) | −.658 | 139.694 | −2.297 | .031* |
| AST (IU/L) | .175 | 15.463 | .671 | .509 |
| ALT (IU/L) | .244 | 8.000 | .794 | .435 |
| BUN (mg dl−1) | .130 | 16.877 | .459 | .651 |
| Uric acid (mg dl−1) | −.052 | 38.427 | −.251 | .804 |
WBC, white blood cell; RBC, red blood cell; Hgb, hemoglobin; ESR, erythrocyte sedimentation rate; NLR, neutrophil to lympohcyte ratio; LMR, lymphocyte to monocyte ratio; AST, aspartate aminotransferase; ALT, alanine aminotransferase; BUN, blood urea nitrogen.Results were obtained from multiple regression analysis. *Significant difference: p < 0.05.
As shown in table 7, logistic regression analysis results showed that NLR values (β = 3.859, p = 0.021) were significantly related to genuine halitosis diagnosis.
Table 7.Risk factors of genuine halitosis.
| Predictor variables | Standardized β | Standard error | Odds ratio | 95% CI | P-value |
|---|---|---|---|---|---|
| Age (years) | −.028 | .032 | .972 | .914–1.034 | .366 |
| Duration of halitosis (months) | .010 | .007 | 1.010 | .996–1.025 | .147 |
| Gingival index | .873 | 1.332 | 2.393 | .176–32.563 | .512 |
| Bleeding index | .901 | 1.641 | 2.462 | .099–61.352 | .583 |
| Unstimulated salivary flow rate (ml min−1) | −1.680 | 2.109 | .186 | .003–11.632 | .426 |
| Stimulated salivary flow rate (ml min−1) | −.382 | .545 | .683 | .235–1.986 | .483 |
| WBC (×103 μl−1) | −.098 | .408 | .907 | .408–2.015 | .810 |
| Platelet count (×103 μl−1) | −.012 | .007 | .988 | .976–1.001 | .070 |
| ESR (mm hr−1) | .124 | .096 | 1.132 | .937–1.367 | .199 |
| NLR | 3.859 | 1.672 | 47.398 | 1.788–1256.465 | .021* |
| LMR | .728 | .470 | 2.070 | .824–5.199 | .121 |
CI, confidence interval; WBC, white blood cell; ESR, erythrocyte sedimentation rate; NLR, neutrophil to lympohcyte ratio; LMR, lymphocyte to monocyte ratio.Results were obtained from logistic regression analysis. *Significant difference: p < 0.05.
3.6.Effectiveness of hematologic inflammatory indices in predicting genuine halitosis
Table 8 and figure 1 shows that WBC with a cut-off value of 5575 μl−1 leads to an AUC of 0.661 for genuine halitosis (p = 0.018) indicating fair discriminative power. Other evaluated indices did not show sufficient discriminative power for genuine halitosis. As shown in table 9, when the patients were grouped according to this new cut-off value of WBC, both groups showed a significant improvement with intervention with all halitosis measurement methods (p < 0.001). The abnormal WBC count group showed significantly higher potable sulfide monitor values after tongue cleaning compared to the normal group and all measurement values were consistently higher at all evaluation points.
Table 8.Sensitivity, specificity, PPV, NPV, and error rate of hematologic inflammatory indices in evaluating genuine halitosis.
| Cut-off value | Halitosis group | AUC | Sensitivity (%) [95% CI] | Specificity (%) [95% CI] | PPV (%) [95% CI] | NPV (%) [95% CI] | Error rate (%) | P-value | ||
|---|---|---|---|---|---|---|---|---|---|---|
| Genuine | Pseudo | |||||||||
| WBC | ⩾5575 μl−1 | 67 | 10 | 0.661 | 65.1 [55.0, 74.2] | 54.6 [32.2, 75.6] | 87.0 [80.6, 91.5] | 25.0 [17.3, 34.6] | 36.8 | 0.018* |
| <5575 μl−1 | 36 | 12 | ||||||||
| ESR | ⩾4.5 mm h−1 | 57 | 13 | 0.506 | 55.3 [45.2, 65.1] | 40.9 [20.7, 63.7] | 81.4 [74.8, 86.6] | 16.4 [10.2, 25.3] | 47.8 | 0.930 |
| <4.5 mm h−1 | 46 | 9 | ||||||||
| NLR | ⩾1.320 | 76 | 13 | 0.605 | 73.8 [64.2, 82.0] | 40.9 [20.7, 63.7] | 85.4 [80.2, 89.4] | 25.0 [15.5, 37.7] | 32.0 | 0.123 |
| <1.320 | 27 | 9 | ||||||||
| dNLR | ⩾0.463 | 67 | 15 | 0.412 | 65.1 [55.0, 74.2] | 31.8 [13.8, 54.9] | 81.7 [76.5, 86.0] | 16.3 [9.1, 27.5] | 40.8 | 0.199 |
| <0.463 | 36 | 7 | ||||||||
| PLR | ⩾108.3 | 67 | 12 | 0.497 | 65.0 [55.0, 74.2] | 45.5 [24.4, 67.8] | 84.8 [78.8, 89.4] | 21.7 [14.1, 32.0] | 38.4 | 0.969 |
| <108.3 | 36 | 10 | ||||||||
| LMR | ⩾4.18 | 65 | 15 | 0.510 | 63.1 [53.0, 72.4] | 31.8 [13.9, 54.9] | 81.3 [75.9, 85.7] | 15.6 [8.7, 26.3] | 42.6 | 0.883 |
| <4.18 | 38 | 7 | ||||||||
AUC, area under the curve; CI, confidence internal; PPV, positive predictive value; NPV, negative predictive value; WBC, white blood cell; ESR, erythrocyte sedimentation rate; NLR, neutrophil to lympohcyte ratio; dNLR, derived neutrophil to lymphocyte ratio; PLR, platelet to lymphocyte ratio; LMR, lymphocyte to monocyte ratio.Sensitivity was obtained from TP/(TP + FN) × 100, Specificity was obtained from TN/(TN + FP) × 100, PPV was obtained from TP/(TP + FP) × 100, NPV was obtained from TN/(TN + FN) × 100, Error rate was obtained from (FN + FP)/(TN + TP + FN + FP).
Table 9.Change in halitosis measurements with intervention of groups according to new cut-off value of white blood cell count (⩾5575 μl−1).
| Variables | Group | T1 | T2 | T3 | Repeated-measures ANOVA (F) | ||
|---|---|---|---|---|---|---|---|
| Time | Group | Time x group | |||||
| Organoleptic score | Normal (n = 30) | 1.7 (0.9) | 1.5 (1.0) | 0.7 (0.6) | 75.291** | 0.435 | 0.027 |
| Abnormal (n = 46) | 1.8 (1.0) | 1.6 (1.0) | 0.8 (0.7) | ||||
| p-value | 0.342 | 0.527 | 0.607 | ||||
| Portable sulfide monitor (ppb) | Normal (n = 31) | 198.7 (137.3) | 199.6 (165.7) | 78.6 (47.6) | 31.797** | 3.800 | 1.801 |
| Abnormal (n = 46) | 302.2 (318.8) | 295.1 (242.9) | 103.3 (82.8) | ||||
| p-value | 0.116 | 0.033* | 0.138 | ||||
| Gas chromatography (ppm) | Normal (n = 28) | 0.291 (0.390) | 0.237 (0.405) | 0.053 (0.097) | 10.492** | 1.233 | 0.458 |
| Abnormal (n = 45) | 0.463 (0.732) | 0.322 (0.611) | 0.100 (0.292) | ||||
| p-value | 0.244 | 0.426 | 0.469 | ||||
T1, before tongue coating removal (without brushing); T2, after tongue coating removal; T2, after oral prophylaxis; WBC, white blood cell.Results were obtained from repeated measures ANOVA and independent t-test: mean (SD). *Significant difference, P < 0.05. **Significant difference, P < 0.001.
4.Discussion
The results of this study showed for the first time in literature that certain inflammatory hematologic indices are closely related to the presence of genuine halitosis and suggested a new WBC count cut-off value of ⩾5575 μl−1 that may be applied in differentiating those with genuine halitosis from pseudo halitosis.
4.1.WBC count in halitosis
A high WBC count is known to indicate the presence of low-grade systemic inflammation. Numerous previous studies showed a close relationship with increased WBC count and cardiovascular disease [20, 21], metabolic syndrome [22], insulin resistance [23] and posttraumatic stress disorder [24] establishing it as a risk factor for each disease entity and also increased mortality [25]. WBC count is known to be affected by gender, Hgb level, and smoking and all factors did not show a significant difference in the two halitosis groups of this study supporting the accuracy of its results [26]. A previous large-scale cross-sectional study reported 5600 μl−1 for men and 5800 μl−1 for women as optimal cutoff values for predicting the presence of metabolic syndrome [27]. Another study on older adults reported leukocyte counts of 7514 mm−3 for men and 5626 mm−3 for women as meaningful parameters associated with metabolic syndrome [28]. A higher cutoff value of 12 × 109 l−1 has been suggested for serious inflammation accompanied by sepsis [29]. The cutoff values provided by studies on metabolic syndrome are close to that found with our halitosis patients implicating that subclinical levels of inflammation may be involved in the pathophysiology of both diseases. Unfortunately there is no previous study showing the prevalence of halitosis in metabolic syndrome patients to further support their coexistence. Diabetes is known to be associated with halitosis [30] while, halitosis prevalence did not differ between before and after bariatric surgery in relation to obesity [31]. Information on other major systemic conditions are majorly lacking. So, the first step towards further verification would be to understand the accurate prevalence of halitosis in various systemic conditions while controlling confounders.
The association between low-grade inflammation and halitosis may be explained in part by pro-inflammatory cytokine release by activated immune cells in systemic inflammatory status. This may contribute to endothelial dysfunction, vascular damage, and further aggravation of systemic inflammation both directly and indirectly leading to worsening of halitosis. Elastase released from activated neutrophils leads to endothelial damage causing abnormal blood flow [32]. WBC count is also known to show a long-term positive correlation with smoking [33] and cigarette smoking is a well-established risk factor of halitosis [34]. Literature on WBC count and inflammatory cytokines following environmental exposures showed that systemic cardiopulmonary and cardiovascular effects could be mediated by inflammatory mechanisms [35, 36]. For now we can only speculate the mechanism underlying genuine halitosis in patients with high WBC counts since, the role of inflammation in halitosis has not been investigated at a systemic level and all literature is focused on inflammation due to local infectious diseases including tonsillar pathology [37] and periodontal disease [34]. A correlation between periodontal pocket inflammation and hydrogen sulfide concentration was identified which may in turn cause further inflammation in the surrounding periodontium [38] and the presence of local inflammation is associated with more salivary sediments including leucocytes and epithelial cells which are a source for bacterial putrefaction [39]. The mechanism through which systemic and local inflammation result in halitosis may differ considerably so the assessment of systemic inflammatory status should be considered in the diagnostic process of halitosis in addition to the evaluation for local inflammation in periodontal tissue and tongue which is regarded as routine practice for now. WBC count measurements are well standardized and is commonly assessed as routine hematologic evaluation. Also, it is relatively affordable allowing it to be easily applied in clinical settings. Since the value itself is not indicative of apparent systemic inflammation, repetitive long-term evaluation for changes in WBC count is not essential in evaluating disease status and treatment response. This could be another factor that increases its accessibility as a possible diagnostic test for halitosis. Due to its invasiveness hematologic evaluation should not replace mouth air analysis but rather should be considered when further verification of the etiology is called upon. The results from our study showed that according to the general cutoff value of 10 000 μl−1 for abnormal WBC count only 2.4% of the total subjects would be differentiated as abnormal and this suggests the need for a unique value to be applied in diagnosing each specific disease. The WBC count value of ⩾5575 μl−1 derived from our objectively defined genuine halitosis patients could act as a tentative criterion for such use with further validation based on other halitosis groups.
4.2.Inflammatory hematologic indices in halitosis
ESR also showed a significant relationship with halitosis intensity. Similar to WBC count ESR values are also closely associated with systemic levels of inflammation and shows significant relationships with increased risk of cardiovascular disease [40]. The index has also been used as a clinical predictor of all-cause mortality [41]. As new indices that combine the patient's systemic inflammation level and disease burden, inflammation-based factors including NLR, PLR, and LMR are attracting interest in multiple conditions [8, 9]. The results of our study also showed that NLR was strongly associated with the presence of genuine halitosis. In a study based on rheumatoid arthritis a value of NLR 1.4 was suggested to classify patients in deep remission which was similar to the value of 1.32 from our study [42]. The reference values for NLR, PLR, and LMR in healthy south Korean adults was reported as 1.65, 132.40, and 5.31, respectively [14]. The mean value of NLR was higher and the LMR value was lower from the genuine halitosis group our study compared to such values. Additional investigations are essential to verify the magnitude of association between such inflammatory prognostic factors and halitosis and establish a unique reference value with sufficient predictive power as they may differ according to halitosis severity and origin. The fact that previous studies based on different disease entities provide widely diverse results related to such inflammatory indices underlines the need. A composite formula of several inflammatory indices may provide more power in accurately diagnosing genuine halitosis and further studies are necessary to explore possible combinations. Since factors known to influence hematologic indices levels including systemic conditions such as diabetes and hypertension, periodontal, and smoking status did not differ according to the presence of genuine halitosis, the results gained from this study could be interpreted as more precise.
4.3.Inflammation in halitosis and systemic health
The results of this study showing that halitosis may be associated with the presence of a low-grade systemic inflammation suggest the possibility of evaluating systemic health conditions through breath analysis and verification of genuine halitosis. Attempts of analyzing exhaled breath for diagnostic purposes are actively being made and results related to cancer [43], gastrointestinal disorders [44], and tuberculosis [45] propose the possibility and benefits of such an approach while, emphasizing the need of standardization of breath collection methods and analysis procedures. Since breath analysis is noninvasive and can be done at the point of care, long-term data collection to evaluate overall health of an individual based on such an approach would be an attractive alternative to the conventional methods of blood and urine analysis. Since there is no consensus on which component of breath more accurately reflects systemic health conditions, commonly used measurements of halitosis may also be considered and its diagnostic value should be verified. The results of this study showed that genuine halitosis may have a close relationship with systemic subclinical inflammation which in turn is known to lead to disability and mortality [46]. Previous literature on the analysis of exhaled breath condensate also showed increased levels of inflammatory markers such as cytokines and leukotrienes suggesting novel methods to evaluate systemic health. However, multiple factors including food, medication, environmental exposure, injury, and circadian rhythm may cause intra-subject variability and influence results. Such effects can identically be applied to this study and may in part explain the modest results from predictive power analysis of inflammatory indices [47]. Another interesting result was the lower cholesterol level of the genuine halitosis group. It is difficult to speculate based solely on the results of this study however, evidence shows that higher cholesterol accumulation in immune cells which is related to less cholesterol efflux is directly associated with the promotion of inflammatory response [48].
4.4.Study limitations
There are several limitations of this study to be discussed. Due to its retrospective nature the results cannot be considered to support a direct causal relationship between systemic inflammation and genuine halitosis since factors that may have influenced the results could have been overlooked. In addition, the high exclusion rate due to lack of halitosis measurement data based on GC and hematological evaluation may have affected the final results. Future studies of a prospective longitudinal design are needed to fully elucidate the role of inflammation in halitosis and the value of inflammatory markers in its diagnosis. Another limitation is the relatively small sample size of pseudo halitosis patients that may have weakened the statistical power of the results. The size of the study may have caused the insignificant results from statistical analysis related to several inflammatory indices. Also the study population consisted of Korean adults only limiting the application of the results to other ethnic and age groups since factors such as NLR are known to differ according to ethnic populations [49]. Analysis of this study did not consider possible gender effects which should be another point to expand on in future studies. Finally, the lack of a well-established VSC cutoff value based on diverse ethnic populations to differentiate those with halitosis and not may act as a limitation since a certain value was used in this study and the results may differ according to different criteria. In spite of such limitations the results of this longitudinal study based on the comprehensive evaluation of various aspects of halitosis with well-standardized tests provide in-depth knowledge into the clinical characteristics of halitosis and its possible relationship with the presence of systemic low-grade inflammation.
5.Conclusions
The results of this study showing an increased WBC count in genuine halitosis patients and a strong association between halitosis and inflammatory indices including ESR and NLR suggest subclinical inflammation as a possible etiology of halitosis and underlines the need to further investigate perturbed hematological indicators that may be linked to oral malodor in otherwise healthy individuals. Hematologic testing reflecting subclinical inflammation may become a useful tool when there is a need to further investigate the cause of halitosis and provide essential information on one's systemic health status in addition to mouth air analysis and other non-invasive diagnostic approaches. The clinician should further investigate possible causes of persistent systemic inflammation when intraoral factors are not significant and raised titers of related indices are identified to provide definitive treatment of halitosis and lessen the burden of unnecessary treatment.
Acknowledgment
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Data availability statement
The data generated and/or analysed during the current study are not publicly available for legal/ethical reasons but are available from the corresponding author on reasonable request.
Authors' contributions
S E P and J H J participated in acquisition of data and interpretation of data. And also in drafting and revising the manuscript critically for important intellectual content. Y K K participated in acquisition of data and interpretation of data. J W P initiated the study project and participated in acquisition and interpretation of data. All authors reviewed and revised the manuscript. All authors read and approved the final manuscript.
Conflict of interest
The authors state that there are no conflicts of interest to report in connection with this article.
Ethical statement
The research was conducted in accordance with the principles embodied in the Declaration of Helsinki and in accordance with local statutory requirements. This was a retrospective chart review study based on medical records so the Institutional Review Board (IRB) of Seoul National University Dental Hospital authorized exemption of informed consent (ERI20018).
